INCINERATION-AT-SEA

          RESEARCH STRATEGY
         FEBRUARY 19, 1985
U.S. Environmental Protection Agency
          Office of Water

-------
                         TABLE OF CONTENTS
1.0  INTRODUCTION
     1.1  Previous Research
     1.2  Relationship of This Strategy to Other Agency
          Activities
     1.3  Development of This Research Strategy and Future
          Plans  .......................      3

2.0  OVERALL STRATEGY  .....................    4

     2.1  Summary of the Research Strategy ...........    4
     2.2  Area 1:  Methods Development for Aquatic Toxicity
          Tests and Emissions Sampling .............   12
     2.3  Area 2:  Monitoring of Incineration-at-Sea
          Operations ......................   15

          A. Emission Characterization .............   15
          B. Exposure and Effects  ............. [     15

     2.4  Area 3:  Additional Tests  ..............   19

3.0  RATIONALE FOR SPECIFIC PLAN COMPONENTS  ..........   20

     3.1  Emission Characterization Considerations .......   20

          3.1.1  Overview  ...................   20
          3.1.2  Development of Emissions Sampling
                 Methods  ....................   22
          3.1.3  Determination of the Composition of
                 Incineration Emissions  ............   24

                 3.1.3.1   Fuel Oil  Control Test  ........   24
                 3.1.3.2   Emissions Sampling  ..........   24
                 3.1.3.3   Continuous Monitoring  ........   25
                 3.1.3.4   Chemical  Analysis Methods   ......   26
     3.2   Exposure  Considerations
                                                                   26
          3.2.1   Overview	    26
          3.2.2   Site  Selection  and  Environmental
                 Monitoring	    26

-------
3.2.2.1 Site Selection • • . . . . . . . . . . 27
3.2.2.2 Environmental Control Study . . . . . 29
3.2.2.3 Environmental Sampling During
Research Burn . . . . . . . . . . . . 30
3.2.3 Model Evaluation and Exposure
Assessment . . . . . . . . . . . . . . . . . . 30
3.2.3.1 Atmospheric Model Evaluation . . . . . 31
3.2.3.2 Aquatic Model Evaluation . . . . . . . 33
3.3 Biological Assessment Consideration . . . . . . . . . 36
3.3.1 Overview . . . . . . . . . . . . . . . . . . . 36
3.3.2 Direct Biological Effects . . . . . . . . . . . 37
3.3.3 Effects on the Surface Microlayer . . . . . . . 40
3.3.4 Comparative Environmental Risk
Assessment . . . . . . . . . . . . . . . . . . 41

-------
1.0 INTRODUCTION
1.1 Previous Research
The Environmental Protection Agency (EPA) has been involved
in ocean incineration research for over 10 years. .Starting in
1974, a series of 4 research burns have been conducted under EPA
permits to gather scientific information about the Incineration of
liquid hazardous wastes at sea and to evaluate ocean incineration
as an alternative to various land based disposal options. These
research burns were conducted under the authority of the Marine
Protection, Research, and Sanctuaries Act of 1972, as amended,
and the Convention on the Prevention of Marine Pollution by Dumping
of Wastes and Other Matter (London Dumping Convention).
The major objective of these research burns was to determine
if the incinerator could be operated efficiently and could attain
the minimum acceptable destruction efficiently (DE) required for
the particular waste being incinerated. This determination was
necessary because the permits were defined by incinerator operating
conditions rather than being specific to the toxicities of the
emissions. Environmental monitoring has been conducted and the
emissions have been analyzed for specific products of incomplete
combustion. Plumes have been followed and sampled by ships and
planes in an effort to trace the dispersion of the emissions over
the ocean, and marine water samples have been collected for analysis
of possible emission constituents. Caged marine organisms have
been exposed in the incinerator area to determine if any effects
could be observed. No products of incomplete combustion (dioxins
or furans) were identified in the emission samples and no potential
toxic emission products were observed at elevated levels in the
environment.
During these past ten years, the scientific community has
developed several different methods for sampling incinerator emis-
sions for DE. These basic procedures have been used in the ocean
incineration research burns. The complexities of sampling at sea
and the peculiarities of ocean incinerators has led to the use of
modifications of the accepted land based methods. The research
burn results indicated that incineration at sea could be a viable
technology for destroying hazardous wastes. However, the previous
studies did not address a number of questions and issues which have
subsequently emerged. This research strategy is proposed to address
these questions.
1.2 Relationship of This Strategy to Other Agency Activities
On May 22, 1984, the Agency decided not to grant special (i.e.
operational.) incineration—at—sea permits as requested by Chemical
Waste Management and Ocean Combustion Services. That decision was
based on:
I

-------
(a) deficiencies in available information considered by the Agency
in determining the need for incineration at sea;
(b) the lack of specific EPA regulations for incineration at
sea; and
(c) the fact that the research permits recommended by EPA’s
Hearing Officer after a public hearing in Brownsville, Texas
had never been the subject of a tentative determination, of
a public hearing, or of public comment.
While the requested special permits were denied, future use of
incineration at sea was not ruled out. Rather, EPA deferred permit
issuance until completion of a more deliberative approach for ocean
incineration. To that end, EPA began development of a research
strategy that would respond to the needs of the program and to
propose specific regulations for ocean incineration.
EPA is currently developing specific regulations by applying
the Agency’s experience with land—based incineration facilities and
ocean incineration vessels. In addition, as a Contracting Party to
the London Dumping Convention (LDC), EPA is bound by the LDC
“Regulations for the Control of Incineration of Waste and Other
Matter at Sea” (Regulations) and is required “to take full account
of” the “Technical Guidelines” implementing the Regulations to the
extent the requirements of the Marine Protection, Research, and
Sanctuaries Act (MPRSA) are not relaxed. The Agency also considered
extensive public comments on the “Tentative Determination To Issue
Special and Research Permits to Chemical Waste Management” (48 FR
48986, October 21, 1983) in preparing the proposed regulations.
In recognition of the questions regarding: the environmental
efficacy of incinerating liquid wastes at sea; the environmental
and human health risks involved in the process; the risks involved
in transportation and loading activities that support incineration
at sea, particularly the risk of a catastrophic spill; the need for
incineration at sea; and public acceptance of incineration at sea
to support this hazardous waste disposal option, the Agency has
also initiated an “Incineration Study” in addition to this research
strategy.
The Agency’s Incineration Study is to provide an analysis of
the technical, economic, and environmental benefits of, and problems
with, incineration as a means of disposing of combustible liquid
hazardous wastes. Both land—based and at—sea incineration are
being assessed with respect to capability, risks and economics in
2

-------
order to make comparisons between the two technologies. Currently
available data is being used in this effort. Other technologies
with potential for treating or disposing of combustible liquid
hazardous wastes such as pyrolysis and wet oxidation, are also
being assessed.
The analysis included in the Incineration Study will discuss
the risks of land—based and at—sea incineration, the efficacy of
the technology, and the capacity of existing incineration facilities
to meet present and future demand for incineration under different
regulatory scenarios. The regulatory scenarios that will be examined
include potential restrictions on land disposal and deep—well
injection of hazardous wastes, the use of hazardous wastes in
industrial boilers, and the elimination of the small hazardous waste
generator exemption. The risk of spills from ocean—going vessels
will also be addressed.
The Incineration Study should put into perspective the risks
and uncertainties of ocean incineration as compared to the available
land—baed alternatives using the technical information currently
available. This research strategy is being designed to allow EPA
to obtain more detailed data which will be needed to more accurately
evaluate these risks.
1.3 Development of This Research Strategy and Future Plans
The research strategy is designed to address the issues of
concern and provide a strategy for organizing the research necessary
to determine the human health and environmental risks of at sea
incineration.
Legal, policy, regulatory and procedural issues are not within
the scope of this strategy. Only the risks of incinerator operation
are included because the risks related to issues such as spills,
fugitive emissions, etc., are being evaluated in the Agency’s
overall “Incineration Study.” Long—term environmental monitoring
is also not addressed in this research strategy but is the subject
of another document being developed by EPA.
In July 1984, a scientific working group which included
scientists from EPA’S Office of Research and Development who haie
been deeply involved in the Agency’s incineration program, met to
prepare a strategy for gathering the information to address the
major scientific issues of concern.
3

-------
The scientific work group was asked to develop a strategy to
meet the following three objectives:
1. Define the major technical and operational issues related
to incineration at sea permits and of conc rn to EPA,
the public and the scientific community;
2. Provide the incineration at sea industry with a scienti-
fically valid program design for gathering information
to address these issues; and
3. Provide the Agency with a basis for making future decisions
on ocean incineration permit applications.
Major scientific/technical issues generated by the public
regarding ocean—based hazardous waste incineration were organized
under four headings. These four headings serve as an organizational
framework for the research strategy (Table 1).
The working group was directed to develop a research strategy
which would address the information needs of as many of the issues
as possible. The work group developed a strategy with each member
contributing in his own area of expertise. The research strategy
does not address long—term environmental monitoring or many of the
issues covered in the Nincineration study.” Only research is
addressed.
The draft strategy was circulated and was the focus of a public
meeting held in Washington, D.C., on November 13, 1984. The comments
received from the public on the draft strategy were then the focus
of another EPA Task Force meeting on December 19, where the draft
strategy was modified and the current document developed.
EPA plans to have this document and the actual subsequent
research plans for the strategy reviewed by qualified scientists
prior to implementation. The reports resulting from this work will
also be reviewed by a qualified scientific group.
2.0 OVERALL STRATEGY
2.1 Summary of the Research Strategy
The major focus for the research strategy is the development
of a rational, scientifically defensible, methodology for an updated
environmental risk assessment of hazardous waste incineration at
sea. The alternative approach of monitoring an area around the
site before, during, and after incineration to measure environmental
concentration of pollutants and ecological effects will not gi nerate
an acceptable environmental risk assessment. Risk assessment is a
confidence—bounded estimate of the probability of unacceptably
altering the aquatic environment as a result of hazardous waste
incineration at sea. The estimate is derived from a comparison
of the predicted organism exposure concentrations of pollutants
with concentrations that cause adverse biological effects.
I

-------
TABLE 1
The specific issues that were developed from public comments are
grouped under the few major headings of emissions, exposure, effects
and site selection.
1.0 Emissions
1.1 General Issues
1.1.1 What are the emissions from ocean incineration?
1.1.2 What is the relationship between emissions and
incinerator specifications, operating conditions
and waste feedstock characteristics?
1.1.3 What are acceptable parameters for continuous
monitoring which are related to emissions?
1.1.4 What effect does the addition of control techno-
logies have on emissions?
1.2 Specific Issues
1.2.1 What are the types and quantities of products of
incomplete combustion (PIC5) generated by the
types of wastes and incinerator operating charac-
teristics proposed for operational, and research
burns at sea?
1.2.2 What are alternative parameters (e.g., total
unburned hydrocarbons or total organochiorides)
for monitoring total PlC emissions?
1.2.3 Is it necessary to identify and measure all
PICs as well as principal organic hazardous con-
stituents (POHCs)?
1.2.4 What happens to PlC production during upset
(clogged burner, etc.) Conditions?
1.2.5 What are the sampling protocols, detection
limits and quality assurance requirements for
monitoring ocean incineration emissions?
1.2.6 What sampling procedure should be used to obtain
a representative sample of the entire stack?
5

-------
1.2.7 Within the regulatory window, what is the
relationship between destruction and combustion
efficiencies over a range of conditions of Ca 2 ,
CO, etc.?
1.2.8 Should destruction efficiency (DE) be measured
continuously, throughout all burns?
1.2.9 Should combustion efficiency (CE) be based on
total carbon in waste, and not just the oxidized
emissions?
1.2.10 What is the relationship between PlC production
and destruction/combustion efficiency?
1.2.11 What performance standards and specifications
should be met during ocean incineration burns?
1.2.12 What are the organic compounds found in the plume
when it reaches ambient temperatures and what
are their concentration levels?
1.2.13 What are the emission levels of metals and chloride
during ocean incineration?
1.2.14 Should research protocols be tested with non—toxic
Substances prior to burns involving liquid organic
hazardous wastes?
2.0 Exposure
2.1 General Issues
2.1.1 How do we measure and model transport and dis-
persion of the plume generated during ocean incine-
ration?
2.1.2 How do we verify model predictions?
2.1.3 To what extent do these substances in the plume
persist in the atmosphere and enter the micro—
layer and water column?
2.1.4 To what extent is the biota exposed to these
substances?
S

-------
2.2 Specific Issues
2.2.1 What is the potential for uptake, concentration
and transport by each of the followii g: eddies,
slicks (microlayers), particles, bubbles, winds,
waves, thermoclines, currents, and precipitation
(wet deposition)?
2.2.2 How is the rate of uptake and residence time of
emissions components in the microlayer measured?
2.2.3 Are plume spiking (e.g., with perfluorocarbons)
and tracking needed?
3.0 Effects
3.1 General Issues
3.1.1 What biological and ecological effects result
from exposure to ocean incineration emissions?
3.1.2 What human health effects result from ocean
incineration?
3.2 Specific Issues
3.2.1 To what extent do emission components which
accumulate in the surface microlayer cause adverse
impacts in the microlayer?
3.2.2 To what extent do emission components bioaccumu—
late in food chain organisms exposed to the micro—
layer and water column?
3.2.3 What are the aquatic organism acute and chronic
toxicities of specific emission components and
byproducts and how do they compare to the toxi—
cities of the dispersed and the undispersed plumes?
3.2.4 What levels of emissions cause behavioral or
sublethal population and ecosystem effects?
3.2.5 What species can be used as indicators of adverse
impacts (e.g., the Mussel—watch Program) at incine-
ration sites?
3.2.6 How can current models for predicting human and
ecological effects from ocean incineration be
improved?
7

-------
4.0 Siting
4.1 General Issues
4.1.1 What are characteristics of a site which make
it suitable for incineration?
4.1.2 How can the long—term assimilative capacity
of a site for incinerator emissions be assessed?
4.2 Specific Issues
4.2.1 What are criteria to identify areas with “redun-
dant” ecosystems as candidate sites for incinera-
tion?
4.2.2 What parameters should be measured to determine
the capacity of a site to handle additional
incineration without causing adverse impacts?
4.2.3 What additional information is needed at unde—
signated sites?
8

-------
Given the scientific concerns regarding previous research,
EPA has developed this research strategy based upon what appears
to be the three most important issues that need to be addressed
(area 1 and 2 below) to develop an adequate risk assessment for
incineration at sea, and tests that may need to follow (area 3).
These areas are:
Area 1. — Methods development for aquatic toxicology tests and
emissions sampling
Methods need to be developed for determining the aquatic
toxicity of incinerator emissions. This will involve the development
of a system for removing a known volume of emissions from an
incinerator with subsequent incorporation of these emissions into
seawater which would then be subjected to various bioassays to
determine its toxicity to various standard laboratory aquatic test
organisms. If possible, preliminary toxicity tests will be conducted
using emissions from a land—based hazardous waste incinerator to
field test the system and obtain information describing the possible
toxic effects of real incinerator emissions (range finding).
Tests also need to be conducted at either a land—based or at—
sea incinerator to determine the need for stack traversal while
collecting high temperature, low particulate emissions for subse-
quent chemical characterization or toxicity testing.
Area 2 — Monitoring of incineration—at—sea operations
During incineration—at—sea operations, emissions need to be
collected for chemical characterization and toxicity testing. Envi-
ronmental sampling during these operations is also needed to
determine if emissions can be detected in the environment near the
incinerator and if environmental effects can be detected. This
sampling is also needed to modify/verify emission transport models.
Area 3 — Additional research
Studies will be needed to evaluate the environmental impacts
of alternative incineration—at—sea technologies and wastes. Other
studies which may be conducted include detailed studies of transport
and fate of emissions in the air and water column, toxicity studies
using organisms indigenous to burn sites, and long range chronic
studies of emission toxicity and bioaccumulation.
There is a substantial difference between a risk assessment
based on a specific hazardous substance and the planned risk assess-
ment based on hazardous waste emissions from an incinerator. The
9

-------
TABLE 2
RESEAROI STRATEGY
Area 1 Area 2 Area 3
Methods Developuent I. e of New Methods Mditional Tests and Research
and At-Sea to Update Risk Assessment
Range Finding ( Research Burn) ( Yearly )
Develop nissions ° Baseline and environ— &nissions transport testing
sampling technique mental nonitoring. in lab situation.
for toxicity testing.
o F)uissions sampling for ° lbxicity testing with other
o Develop appropriate chemical analysis. organisms and various methods
toxicity tests, of exposure and exposure times.
o fluissions sampling for
o Determine appropriate toxicity tests. ° Area 2 tests using other vessels
a einis ions sampling in research or trial burns.
procedure for sub— ° Transport nodel develop-
sequent chemical ment and verification. ° Area 2 tests using other wastes.
characterization
and toxicity testing. ° Risk asses nt. ° Other tests as required.
o Obtain toxicity data
fr a land—based
incinerator if
possible.
TIME

-------
planned assessment is much more complex because the emissions are
difficult to sample at high temperatures, may be in 2 phases
(gas and scrubber water), and only poorly characterized. The
emitted material is potentially a complex mixture composed of
small amounts of the original waste material, partly destroyed
original material, and new compounds generated during the incine-
ration phase or during the cooling of the emitted gases in air
or in scrubber waters.
Even if the emissions can be fully quantified and characteri-
zed, this will not yield the needed toxicity and environmental
exposure information. Incinerator emissions are complex materials
with potentially additive, synergistic, or subtractive interac-
tions of toxicological properties. The approach of defining the
composition of the emissions and developing a toxicity data base
for each constituent would require extensive chemical analysis
and long, expensive toxicity testing. Instead, EPA plans to
chemically analyze the emissions in an effort to identify the
most abundant substances present and to treat the emissions as a
complex mixture to conduct a risk assessment based upon the
toxicity of the mixture itself.
To obtain the needed toxicity information, methodology will
be developed to trap the emission and introduce it in a controlled
manner into laboratory aquatic systems to create a known dilution
of emissions in sea water. The emissions will be directly intro-
duced into a water media which will subsequently be used to dose
aquatic laboratory test systems.
When this methodology has been developed, some preliminary
tests on standard test animals, which have documented tolerances
to a wide range of toxic substances, will be conducted at an on-
shore facility. Additional tests at a land—based hazardous waste
incinerator may be conducted as a preliminary assessment of the
aquatic toxicity of emissions which may be expected during at—sea
incineration operations.
An at—sea research burn would then be conducted where the
newly developed emission toxicity sampling and bioassay procedures
would be used. Emission samples would also be collected for
chemical analysis, and environmental samples would be collected
for plume fate and transport model verification.
The field measurement program is also designed to respond to
EPA concerns and issues raised by the public and scientific
community on incinerator destruction efficiency variability during
normal operation of the incinerator, the characterization of
the emissions, and the fate of the emissions leaving the stack.
Simultaneously, a control area of the site region will be moni-
tored for background information.
It

-------
Because very small amounts of unburned waste material may
pass through the incinerator, EPA plans to burn polychlorjnated
biphenyls (PCBs) as the first waste material to be subjected to
a research burn. This will enable EPA to use the large existing
toxicity data base for PCBs for comparison to the toxicity data
obtained from the proposed tests. The PCB waste material, can also
be more easily characterized than a complex waste composed of
several toxic constituents.
By combining the results of the emission bioassay dose—
response tests and the plume transport and exposure models, EPA
intends to estimate the potential impact of the emissions on the
environment (risk assessment). Field samples collected during
the research burn from control and plume areas will be used in an
attempt to verify the existence of impacts in the plume touchdown
area.
Risk assessments will be conducted at several stages in the
strategy at a series of exposure conditions. This will allow
derivation of a risk assessment for operation of incinerator
vessels at some prescribed level of efficiency. The risk assess-
ment process will be tiered as appropriate to examine acute,
chronic, and genotoxic effects of a range of spatial and temporal
environmental concentrations.
Data and risk assessments generated from implementation of
this strategy will be useful for decision making on incineration—
at—sea permits and for incineration site management.
Specific activities anticipated to occur in each of these
three areas of this strategy are briefly outlined below.
2.2 Area 1 — Method development for aquatic toxicity tests and
emissions sampling
Toxicity Test Sampling System Development
This will require the design, development and fabrication of
an emission sampling system which can draw hot emissions from an
incinerator system and incorporate the emissions to a volume of
synthetic sea water.
This system must be capable of removing organics from the
emissions and will subsequently collect HC1 and water vapor. Its
development will therefore require consideration of the extra
volume of water which will be collected and the need for pH and
oxygen adjustments that may be needed prior to use in bioassays.
12

-------
o Sampling System Performance Testing
There will be an initial testing of the system using an
easily accessible combustion chamber such as a small—scale incine-
rator or fuel oil burner.
Emissions will be collected from the exhaust duct using
the newly developed sampling system. The water from the system
will then be subjected to a series of lab tests to evaluate the
usefulness of the overall sampling system. Emissions may need
to be spiked with water vapor, HC1, and organic substances
to approximate real incinerator emissions.
Tests will be performed to evaluate the effect of the
collection procedure on the quality of the sample water. These
tests will include:
— the effect of, and necessary counter—measures for, the
decreased salinity due to collection of water
vapors
— the effect of, and necessary counter—measures for, the
the decreased pH due to HC1 collection
— the effect of temperature during sampling
— the effect of particulates collected during sampling
— the effect of, and necessary counter—measures for, changes
in buffering capacity due to the collection procedures
and substances collected
Tests shall also be performed to determine the appropriate
species and bioassay endpoints which will become the major focus
of future bioassay tests. The species and endpoints to be
considered will include:
— fish embryo/larvae tests for hatch viability and growth.
Test durations of 48 and 96 hours will be considered
— molluscan larvae 48—hour growth tests and shell
development success
— crustacean larvae (mysids and grass shrimp) survival
and growth at 48 and 96 hours
— zooplankton (nauplii or juvenile copepods) survival at
48 and 96 hours
— phytoplankton growth for 48 and 96 hours
13

-------
Additional consideration will be given to the necessary
handling procedures of the sample bioassay water between collection
and use. These will include evaluation of the need for immediate
use versus acceptable storage and transport procedures.
When the sampling and bioassay procedures are developed,
preliminary field testing may be conducted at a land—based
hazardous waste incinerator if available. Samplers at this type
of facility will present some of the same sampling conditions as
an at—sea incinerator and the emissions themselves should be
similar. Emissions collection can be conducted at a land—based
incinerator and the methods for exposing test animals to the
emissions can be tested. This approach will allow range finding
tests to be conducted and will permit the field testing of the
overall procedure while avoiding the complications of conducting
an at—sea research burn.
Fuel oil will be burned during part of these tests for
blank” information since many PCB wastes contain fuel oil, and
samples will also be collected while PCBs are being incinerated.
° Incinerator Emission Sampling Point/Points Determination
A series of tests will be conducted to examine the need for
traversing a section of the incinerator stack while collecting
emissions for subsequent chemical characterization or toxicity
testing. These tests may be conducted during an at—sea research
burn, on shore at a land_based hazardous waste incinerator, during
a fuel oil dockside burn of an at—sea incinerator, or at a
combination of the three locations based upon the availability of
a suitable location for the tests.
The incinerator stack or duct will be sampled for determina-
tion of combustion efficiency using a traverse across the stack
or duct. Velocity, Ca, C0 2 , and 02 will be measured using EPA
Methods 1 through 3. Three full traverses will be conducted with
measurements taken at each cross—sectional sampling point.
Additional tests at the land—based hazardous waste incinerator
or during an at—sea research burn when PCBs are being destroyed,
will again include sampling the single point and traverse sampling
points, but this time a modified Method 5 procedure will be used
for organics collection. The samples will be analyzed for PCBs
(as will the PCB fuel), and destruction efficiency will be calcu—
lated. Particulate samples will be collected at each individual
point in the traverse and weighed for determination of the varia-
bility in particulate loadings across the stack.
‘4

-------
The results of the particulate CE and DE single—point and
traverse analyses will be compared to determine the need for
traversing stacks in future studies. If CE should show a con-
stantly low value at a specific sampling point, this point will
be used for the single—point DE determination. If this point
shows equivalent or lower DE than the traverse sampled, then it
may be used by EPA in future tests as a worst case. If the single
point and traverse CE’S and DE’s are not statistically different,
traversal may not be required in future tests.
The oxygen concentrations in the gas at the stack sampling
points of an at—sea incinerator will also be useful in determi—
fling if ambient air is intruding down into the stack to the
sampling level. The analysis will allow EPA to determine if the
sampling ports in the stack should be lowered, moving them farther
from the exit plane of the stack.
2.3 Area 2 — Monitoring of incineration—at—sea operations
A — Emission Characterization
° Emissions will be collected for subsequent chemical charac-
terization. Samples will be collected using either the
EPA Modified Method 5 (MM5) Sampling System or Source
Assessment Sampling System (SASS) for semivolatile organic
analyses, and the Volatile Organic Sampling Train (VOST)
for volatile organic analyses. Sampling points in the
stack will be determined based upon the results of the
Area 1” studies.
Emission samples will be analyzed using gas chromato-
graphy/mass spectroscopy (GC/MS), gas chromatography!
electron capture (CC/EC) and other methods to identify
and quantify the most abundant compounds observed in
CC/MS scan chromatograms (including PCBs, dioxins, furans
and other chlorinated organics).
° Emissions samples will be collected for subsequent total
organohalogen (TOX) analysis using a GC/Hall Detector
technique.
° Total unburned hydrocarbons, plus CO, CO 2 and 02 will
be monitored continuously.
° Emissions will be collected for determination of
particulate levels at various points across the stack.
15

-------
o Sampling will occur during normal operation of the incine—
rator (including upsets if they occur).
B — Exposure and Effects
o Incineration Research Site Selection
Prior to the implementation of the ocean research burn of
hazardous waste, an environmentally acceptable site will be
selected specifically for the burn.
Information will be collected on the geographic, physical,
chemical, and biological characteristics of the research site.
Physical information will include data on currents, hydrology,
and meteorology.
o Environmental Control Sampling
Immediately prior to an at—sea research burn (on the order
of days) and during the burn, an environmental control study will
be conducted at the site.
— Chemical concentration data will be collected
on samples of biota living in surface waters at the
site (neuston), as well as in the air and water to
establish background concentrations of any emission
related contaminants.
— Air samples will be collected and analyzed for several
types of substances. Air will be sampled for organic
substances using a high volume filtering system which
uses polyurethane foam plugs to trap organics. HC1
in air will be measured using a real—time HC1 analyzer,
such as a Geomet HC1 analyzer, which has the needed
sensitivity.
— Water samples will be collected using several proce-
dures. One procedure will be a high volume polyurethane
foam plug filtering system for organics collection
and the other will be a syringe sampling method for
sulfurhexafluoride (SF 6 ) or perfluorinated hydro-
carbon (PFH) determination. Water bottles will be
used to collect water for pH, alkalinity, tempera-
ture, salinity, chlorophyll, and ATP determinations.
16

-------
— Neuston will be collected using standard neuston nets,
and the catch will be taxonomically identified into
major groups and analyzed for organic residues.
— Background SF 6 or PFH levels (plume tracers) will
be measured in air using either a continuous monitor or
a grab sample — GC/EC procedures.
— Chemical analysis of the samples will be conducted
using the previously described methods for waste and
emissions samples and will be conducted after the
emissions samples have been analyzed.
o Environmental Sampling in the Research Burn Area
Samples will be collected during the research burn in
the vicinity of the incinerator vessel.
— Samples of air, water and biota will be collected
in the plume area.
— Chemical analysis of the environmental samples will
be the same as for the control samples and will not
be conducted until the substances present in the
emissions samples are identified. The environmental
samples will then he analyzed specifically for these
substances.
— Air samples for HC1 and SF 6 or PFH analysis will
be collected in the plume.
o Transport Model Development
Appropriate atmospheric and aquatic transport models will
be selected to predict transport of the emissions over
and in the water to the point where they are indistin-
guishable from the background.
• Transport Model Verification 1
The results of the environmental sampling during the
research burn cruise will be used to verify the transport
models.
o The atmospheric transport models will be empirically
calibrated with the HC1 in the plume and emission plume
markers such as SF 6 and PFH.
17

-------
o In addition, limiting atmospheric conditions for incine-
ration will be identified, and best— and worst—case
scenarios of environmental exposure determined.
o Biological Effects Assessment
The impact of emissions on biological processes in
the marine environment will be assessed from laboratory
studies conducted with stack emissions that have been
collected during the ocean research burn. These emis-
sions will be collected using the procedures developed
in Area 1 of this strategy.
— Test organisms will be obtained for bioassays and
will include the standard test species for which we
have extensive comparative toxicological data.
— Test organisms will be exposed to concentrations
of the emissions including the expected/measured range
of water column concentrations.
— To insure that the toxicological data developed for
this plan are quality—assured and meet the require-
ments of good laboratory practice, the following
standard test protocols will be used:
a. ASTM Standard Practice for Conduct of Acute
Toxicity Tests with Fishes and Invertebrates.
b. ASTM Standard Practice for the Conduct of
Acute and Chronic Tests with the Mysid
Shrimp, Mysidopsis bahia .
c. ASTM Standard Practice for the Conduct of
Early Life Stage Tests with Fishes.
d. OPTS Interlaboratory Calibration Studies
with Oyster Larvae.
e. OPTS Interlaboratory Calibration Studies
with Penaeid Shrimp.
f. OPTS Interlaboratory Calibration Studies with
Marine Algae.
18

-------
— Estimates of the potential bioaccumu]atjon of toxic
pollutants in the emissions will be predicted from
estimated trophic level productivity structure acti-
vity and the log of the octanol—water equilibrium
partition coefficient for the substances identifjed
in the emissions and, depending on these results,
direct measures of tissue residues will be attempted
on chronically—exposed species as dictated by tissue
sample size and analytical detection levels.
2.4 Area 3 — Additional tests
o Tests can be conducted to determine the transport pro-
perties of emissions through various media including
air, microlayer, surface water, deep water and sediments.
o Emissions toxicity tests can be conducted using indigenous
organisms from incineration sites, and more complicated
toxicity tests could be developed and used.
o Tests for bioaccumulation in indigenous species and
food chains can be investigated.
o Areas 2 and 3 type tests can be conducted using various
waste materials.
o Areas 2 and 3 type tests can be conducted using different
incineration vessels or other incinerator designs.
o Tests can be conducted to determine if cooling of emissions
in the plume or scrubber changes their chemical composi-
tion.
o Tests can be conducted to determine the effects of possible
incinerator upset conditions on DE, CE, and emission
composition.
o Tests can be conducted to determine the physical form
and characteristics of substances in the emissions (i.e.,
particulate, droplet, and gaseous form).
o Tests can be conducted for additional verification of
the applicability of current emissions sampling methods
(particularly the Modified Method 5 procedure for organic
substance collection) when sampling at high temperature
and acid conditions.
19

-------
3.0 RATIONALE FOR SPECIFIC PLAN COMPONENTS
3.1 Emission Characterization Considerations
3.1.1 Overview
One of the essential elements to an understanding of the
impacts from ocean incineration is an accurate determination of
the types and quantities of substances emitted from the incin-
eration system. These include unburned waste materials and
substances formed inside the incinerator as products of incom-
plete combustion.
Another essential element is an understanding of the com-
bustion process to the extent that easily monitored substan-
ces/parameters can be realistically used as surrogates for
continuously measuring all the substances being emitted from
the incinerator.
EPA has been increasing its knowledge and understanding
of both these issues for many years regarding on—shore and ocean
incineration. Methods have been developed for collecting!
analyzing/monitoring samples for constituents in addition to
the several emission gas constituents currently being used
to continuously monitor incineration to ensure proper operation.
EPA ’S emissions analysis and monitoring techniques are
continuously evolving and have been developed and used largely
at land-based incinerators or other types of on—shore combustion
facilities.
Questions have been raised as to whether the newest, state—
of—the—art techniques for emissions sampling, analysis, and
monitoring have been used on ocean incinerators, and whether
these procedures are actually applicable or necessary to ocean
incinerators due to the high stack gas temperatures, high HC1
concentrations, cyclonic flow in the short stacks, and low
particulate levels.
Several ocean incineration research burns have been comple-
ted in the past. During eact burn, testing for destruction
efficiency of the major waste components was done using the
best technology available at that time. Certain issues have
not been fully addressed during previous incineration burns
and the associated analytical studies. In retrospect, there
are three main technical areas which are now viewed as deficient.
These areas are described separately below.
20

-------
A. Rigorous Testing of Performance According to Best Available
EPA Methods :
EPA has not fully applied the Agency’s most rigorous stack
sampling methods such as stack traversing with is.okinetic
collection of emissions during past research burns. Also,
the previous evaluations were only designed o test for
compliance of earlier incineration—at—sea requirements
for 99.9 or 99.99 percent destruction efficiency (DE),
and some were insufficient in sampling time duration
to show whether the land—based regulatory requirements
for DRE of 99.9 to 99.9999 percent were being met, or
whether any significant quantification of particulate
matter emission was accomplished. Although the previous
work which EPA has reviewed is accepted as valid and
accurate, public comments and recent or proposed revi-
sions to Agency policy now dictate that the above defi-
ciencies in this area should be addressed through addi-
tional research and testing. The additional research
tests would include longer (e.g., 8 hours versus 2
hours) sampling, fully traversing the stack(s) isokine—
tically (or providing suitable justification that traver-
sing is not necessary), and quantifying the particulate
matter.
B. A Complete Chemical Characterization of All Incinerator
E niss ions :
Previous work with incinerator ships concentrated on des-
truction efficiencies but failed to quantify fully the
amount of all the other substances in the stack emissions.
Although a similar deficiency exists in most of EPA’S
land—based incinerator performance data, recent comments
request that the Agency close this data gap in any
future incineration tests whether for land or at—sea
incinerator performance.
EPA recognizes that identification and quantification of
“all stack effluents” rather than just the major PCB and
selected emission substances may be impractical in terms
of the cost and time. ! owever, the analytical search
should be expanded to the greatest practical extent.
C. Potential for Changes in Stack Effluent Chemical Composi-
tion Upon Leaving the Stack :
Previous at—sea incinerator performance evaluations did
not evaluate the potential for further chemical changes
21

-------
(e.g., chemical reformation or recombination of mole-
cules, etc.) in scrubbers or within the stack effluent
gases upon leaving the stack exit plane and entering
the cooler atmosphere immediately above or downstream
of the ship. Such studies have also not been conducted
for land—based incinerators; however, EPA agrees that
this is an area which merits investigation, particularly
for those incinerators which discharge hot effluents
directly into the atmosphere. Although it is likely
that hazardous compounds passing through the incinerator
hot zones become disassociated, oxidized, and separated
such that they will not further change or recombine, a
test of the “permanency” of separation and its “stability”
would be a significant piece of information to demon-
strate incineration effectiveness.
In response to the data needs described above and else-
where, a series of at—sea incineration research burns should
be conducted to provide sufficient amounts of emission conta-
minants for various tests to (1) demonstrate the destruction
efficiency of designated organic compounds, (2) demonstrate
particulate collection and quantification in stack emissions,
(3) demonstrate the production and survival of other chemical
compounds during the incineration process, (4) demonstrate
whether reactive products are produced in the plume as cooling
occurs, (5) aid in plume model validation studies and biolo-
gical testing, and (6) demonstrate the effects of liquid
scrubbers on stack emissions. The following sections address
these data needs.
3.1.2 Development of Emissions Sampling Methods
In order to characterize adequately the emissions
from at—sea incineration of hazardous waste, a test burn
should first be conducted to determine the optimum sampling
and analysis procedures.
EPA’s land—based RCRA and TSCA trial burn procedures
and stack sampling procedures require that the incinerator
effluents be sampled isokinetica]].y at a number of individual
points (typically 12 points) across each of two perpendicular
diameters of the stack. While this elaborate and costly
procedure is not absolutely known to be necessary (versus
single—point readings), it nevertheless has been required on
all land—based incinerator trial burn testing.
22

-------
Recently, EPA has reviewed the need for traversing by examining
the typical turbulence/mixing/Reynolds number levels in most
incinerator stacks. They found that, in most cases, the flow
patterns are extremely turbulent and rarely laminar. If good
mixing and uniformity across any diameter of stack effluents
can be verified, then traversing should not be necessary. Be-
cause of the large stack diameters, the extreme heat, and
the necessity for personnel to be in unsafe locations during
some of the traverse procedures, traversing during waste
burning at sea is both hazardous to personnel and costly.
This strategy suggests conducting traverse/single point com-
parison tests on at—sea incinerators with the alternate
use of dockside nonhazardous fuel (diesel oil) stack emis-
sions testing if these tests cannot be conducted at sea.
This at—sea or dockside emissions testing will provide
meaningful information on the variation across stack diameters
of velocity, CO, C0 2 , O , THC, NO, etc. It will also provide
information showing the relationship of the particulates
obtained from single—point samples at all the traverse points.
These parameters will help to determine if variations are sig-
nificant and therefore justify requiring traversing during
sampling. Alternatively, should these tests identify a point
of Nworst emissions, the option for placing a single—point
probe at these points may be considered during waste burns.
Traverse vs. single point DE tests may be conducted at a
land—based hazardous waste incinerator for the collection
of data to supplement the dockside data.
EPA has reviewed data submitted by a permit applicant
which shows clearly the combustion products for an oil burn
would be similar to an actual hazardous waste burn, except
for the chlorine content of a chlorinated waste. The chlorine
content may be in the 3 percent (by weight and volume) range.
Therefore, it appears that data from an oil burn would be
representative, in a fluid dynamics/flow pattern sense, of
the flow of hazardous waste emissions from an incinerator
stack under the same velocity and net flow rate conditions,
etc.
The recognized EPA methods for collecting organic sub-
stances for DE determination include the Modified Method 5
(MM5) train for organic sampling, the Source Assessment
Sampling System (SASS) train, which is a larger version of
the MM5 capable of collecting five times the organic sample
as the MM5, and the Volatile Organic Sampling Train (yOST).
Both the MM5 and the SASS trains collect organic compounds
which boil over 100°C, while the VOST collects organic corn—
23

-------
pounds in the 30—100°C boiling range. EPA Methods 1 through
4 will be used to obtain gas flow information from the stack.
3.1.3 Determination of the Composition of Incineration Emissions
The results of the Area 1 tests will be used to design
the procedure for collecting the emissions from the incinera-
tion of hazardous waste at sea. This will require use of
either the single—point or traverse procedure, whichever is
appropriate, for collection of high volumes of stack gas.
3.1.3.1 Fuel Oil Control Test
When the incinerator vessel is at the incineration site,
a one—day test will be conducted using fuel oil for comparison
to the dockside fuel oil test if it is conducted. The results
will be used to determine if being at sea has any effects on
either the incinerator combustion efficiency or the sample
collection efficiency.
3.1.3.2 Emissions Sampling
Emissions samples will be collected with the MM5 or SASS
train for subsequent trace chemical analysis. Ambient air
will also be collected during the hazardous waste burn at
sea as a background control using the same collection system
as is used in the stack. Waste samples will also be collected
for analysis. In the future (Area 3 studies), systems inclu-
ding seawater scrubbers will either be sampled upstream of
the scrubbing device or require the development of adequate
gaseous and scrubber water sampling procedures.
The recognized EPA methods for the collection of parti—
culates in stack emissions through the use of traversing
will be used only if the results of the previous tests demon-
strate that one—point sampling will not provide a representa-
tive or even a worst case collection point.
EPA wishes to determine if the substances detected in the
standard MM5 or SASS train samples (collected from inside the
stack) are the same as those that actually enter the environ-
ment in the cooled off plume from the incinerator. During
future tests, EPA will use a new sampling system which collects
gas from the stack much like the MM5 and SASS trains, but
cools and dilutes the gas before the particulates and organics
are collected on the filter and resin. It is anticipated
that the cooling of diluted gas before sample collection
will more closely imitate the real world environment of the
plume from a non—scrubbed stack.
24

-------
This dilution sampling system has been used by EPA on
industrial boilers but must be reconstructed of glass before it
can be used in the acid environment of the ocean incineration
plumes. This system will collect samples for organics analysis
and may provide a better representation of the substances
actually entering the environment from the incinerator.
3.1.3.3 Continuous Monitoring
During the entire research burn, several parameters
will be continuously measured. The amount of oxygen, carbon
monoxide, and carbon dioxide will be determined by continuous,
rapid response monitors. The temperature of the incinerator
wall at several points will be monitored. The measurement
of these parameters is required by LDC regulations. The
ratio of CO to CO 2 is defined as combustion efficiency (CE =
(CO 2 — CO)/C0 2 x 100). CE is a generally accepted guideline
to the proper operation of an incinerator. In addition, a
total hydrocarbon (THC) continuous monitor will be required.
The use of a THC monitor will note if the organic portion of
the stack emissions were to increase during an incineration
burn indicating a decrease in optimum incineration operation.
The monitoring of the oxygen provides assurance that the
required air is present for proper oxidation of the organic
compounds.
To date, incinerator experts from EPA and the rest of
the hazardous waste technical community both here and in
Europe have not found any reliable theoretical or empirical
relationships among such performance parameters as destruction
efficiency (DE), destruction removal efficiency (DRE), CE, C0 2 ,
02, THC, or any mathematical ratio or combination thereof.
From a theoretical sense, finding such relation-
ships is a very difficult task due to the complexity of the
high temperature flame processes which simultaneously occur
within an incinerator. EPA’S existing incineration data base
is quite large; however, it is composed of performance data
from dozens of different hardware designs, which adds yet
another variable or dimension to complicate finding any
of these relationships. The only possibilities for finding
such relationships among DE, CE, etc., might be on an empiri-
cal basis, or as derived from extensive parametric (systema-
tically varying all parameters during operation) on individual
fixed designs of incinerators in a research laboratory. EPA
is currently embarking on such studies on laboratory and
pilot scale incinerators; however, the results of these
studies are likely to take many years.
25

-------
Meanwhile, EPA has gained from its broad incineration
data base some reliable and useful engineering judgment
rules or guidelines; for example, what levels of CO, C0 2 ,
02, temperatures, and other design and operating parameters
usually result in acceptable DE or DRE levels overall.
3.1.3.4 Chemical Analysis Methods
The stack emission samples will be analyzed for compounds
present by methods capable of measuring in the nanogram per
sample range. In general, Gas Chromatography/Mass Spectro—
scopy (GC/MS) will be the analytical finish; however, other
techniques will be used as necessary. All stack emission
samples will be analyzed for all major compounds identified
in the feedstock plus the most abundant compounds that can
be detected. This will allow the identification of products
of incomplete combustion and/or recombined compounds. Samples
will be analyzed for total organohalogen (TOX) using a GC/Hall
technique. If any unexpected compounds are found in the
emission samples, the waste will be specifically re—analyzed
for those compounds to determine if they are uncombusted
compounds from the waste.
3.2 Exposure Considerations
3.2.1 Overview
In addition to knowing the identity and quantity of the
emissions from ocean incineration, it is necessary to know
where the emissions go after leaving the stack and at what
concentrations they exist in the environment. This strategy
includes the use of plume transport models to estimate where
gaseous emissions will go and the use of ambient monitoring
to validate the models. It is set up in a tiered approach
where less sophisticated exposure estimates are used in
worst—case scenarios first and, if necessary, sophisticated
exposure estimates (models, etc.) are used later. This
approach is designed to be used in a tiered risk assessment
as described in section 3.3.4.
3.2.2 Site Selection and Environmental Monitoring
Before any incineration can take place in the ocean, a
suitable site must be located which meets the requirements of
the Marine Protection, Research, and Sanctuaries Act (MPRSA)
and which will be appropriate for model validation studies.
26

-------
3.2.2.1 Site Selection
Research burns must be conducted at sites believed to be
environmentally acceptable, but need not be conducted in a
location which has been officially designated by EPA as an
ocean incineration site. The MPRSA allows rese’arch at
non—designated sites, and the low volume of wastes to be
incinerated indicates that the evaluation of an acceptable
short—term site need not consider high—volume/long_term con-
tamination to the degree needed to fully designate sites for
operating permits.
This type of site may be useful for research into detailed
plume modeling, environmental effects, and detection and quan-
tification of trace substances in incinerator emissions from
a source which is believed to operate efficiently. This
type of site may not be appropriate for long—term use.
Sites, as described in this document, will require additional
evaluation before being considered for long—term use.
Sites used for incineration research must meet the general
criteria requirements as defined in the ocean dumping regula-
tions. Sites must be in areas of minimal interference with
fishery or shelifishery areas and regions of heavy commercial
or recreational activity. Sites must be in areas of low
biological activity as exemplified by plankton density and
the absence of sensitive biota, and should not have prevailing
winds or currents toward sensitive areas or shore, In order
to be useful for model validation, the site should be in an
area of relatively consistent wind and predictable current
velocities, or the models will not be capable of explaining
environmental concentrations of the emissions.
In addition to the general criteria, 12 factors should
be considered in site approval. These factors are the speci-
fic criteria described in EPA’s ocean dumping regulations
(40 CFR 228) with the addition of a twelfth factor suggesting
that the occurrence of endangered and threatened species be
considered. The twelve criteria are presented below.
(a) Geographical position, depth of water, bottom topo-
graphy, and distance from coast or special marine resource
areas. The site should not be near shore where the plume
could potentially contact land prior to dispersal.
27

-------
(b) Location in r elation to breeding, spawning, nursery,
feeding, or passage areas of living resources in adult or
juvenile phases. Sites should not be located near these
areas.
Cc) Location in relation to beaches, marine sanctuaries,
and other amenity areas. Sites should not be located in areas
where winds or currents are likely to transport emission pro-
ducts toward these areas.
Cd) Types and quantities of wastes proposed to be
incinerated, and details of the proposed methods of incinera-
tion.
Types and quantities of waste need to be evaluated in
order to determine the rates at which incineration can take
place at a site without endangering local biota. The inciner-
ator designs, including the use of sea water scrubbers, need
to be evaluated in conjunction with the waste type in order
to estimate how emissions will enter the environment (i.e.,
directly into the water as scrubber discharge, or to the
atmosphere in hot gases). Because the site will be used for
research, the total quantities of waste incinerated will be
low.
Ce) Feasibility of surveillance and monitoring.
(f) Seasonal circulation patterns, dispersal, horizontal
transport, and vertical mixing characteristics of the area,
including prevailing water current and wind direction and
speed, air temperature, atmospheric stability, frequency of
inversions and fog, precipitation amounts, and relative humi-
dity.
These parameters of the site area must be known in order
for research to be conducted at a time when air and water cir-
culation patterns will not carry emissions into unsuitable
areas, such as beaches, reefs, or fishing areas, and will
assist in model validation.
(g) Existence and effects of current and previous
discharges from incineration or other disposal activities in
the area.
The existence of previous dumping activities at a site
would be important in evaluating environmental data from base-
line or trend assessment monitoring. The cumulative effect of
ocean incineration and other discharges in the site must be
considered in evaluating a site.
28

-------
(h) Interference with shipping, fishing, recreation,
mineral extraction, desalination, fish and shellfish culture,
areas of special scientific importance, and other legitimate
uses of the ocean.
Ci) The existing water quality and ecology of the site
as determined by available data or baseline surveys. Environ-
mental characteristics of the marine ecosystem at the site
should be described to identify sensitive organisms which
may occur at the site during the test and the background
levels of combustion emission related substances that are
present at the site.
(j) Potential for the recruitment of nuisance species
in the site.
(k) Existence at or in close proximity to the site of
any significant natural or cultural features of historical
importance.
(1) Occurrence of endangered or threatened species at
or near the site. Seasonal migration patterns of endangered
species should be described prior to site use. Seasonal migra-
tion may lead to specific seasonal use of the site so as not
to interfere with endangered species.
3.2.2.2 Environmental Control Study
When a site has been selected and a research burn is
about to begin, environmental control sampling should begin at
the site. Environmental background data must be collected at
the research burn site immediately prior to and during the
research burn. This background data collection will be
oriented toward determining background levels of organic
substances in the area of the site during the research burn
for comparison to levels determined during and after the
burn. Other background data for sulfur hexafluoride (SF 6 )
or perfluorinated hydrocarbons (PFH) and HC1 will also be
collected for use in model validation studies. Indigenous
organisms may also be collected and appropriate species
selected for future laboratory bioassays. Sampling will
begin a day or two before the incineration vessel arrives at
the site to ensure EPA that the rest of the control data taken
during the burn is in fact not contaminated by the incinerator,
29

-------
Environmental baseline samples to be collected at the
site will include the surface water, air, and neuston.
These types of samples will be taken because any substances
exiting the incinerator will likely be at their most’ concen-
trated levels in these media near the incinerator during the
research burn. These baseline concentrations in the near
field area are essential if the environment is to be sampled
during a burn for detection of elevated levels of emission—
related substances. These near field samples will undergo
the same chemical analysis as that described for emission
samples and will also be analyzed for background levels of
HC1 (air only) and SF 6 or PFH (air and surface water only).
3.2.2.3 Environmental Sampling During Research Burn
During a research burn, the same types of samples as
described for controls will be collected in the plume area and
analyzed after the emission samples are analyzed so that the
specific substances observed in the emission samples can be
searched for in the environmental samples.
HC1 and SF 6 or PFH concentrations in the air will be
vigorously tested during the burn to locate and describe the
plume. SF 6 or PFH will be sought in surface water to determine
areas of plume touchdown.
3.2.3 Model Evaluation and Exposure Assessment
Prior to the implementation of a research burn, various
atmospheric and water transport models will be evaluated.
The reason for studying air quality plume dispersion mo-
dels is that they will be used in the future at other sites
to predict the location and extent of the incinerator emis-
sions’ sea surface impact zone and the amounts of material
deposited in this zone based on stack emission rates. This
information will also be used to determine where future air
samples will be taken during operational burns. The study
must include control data, modified overland plume dispersion
models to accommodate overwater meteorology (air movements,
stability, etc.) and model verification to demonstrate that
these predictions are reasonable. The background data collec-
tion plan is designed to obtain data which will estimate
background levels of organics, SF 6 or PFH tracer, and HCI.
These data will be compared to the emission data collected
during the research burns and used to estimate the extent
and diffusion of the incineration plume. These comparisons
will also be used to verify the use of the modified overland
plume dispersion models for predicting the dispersion of
plumes from incineration at sea.
30

-------
3.2.3.1 Atmospheric Model Evaluation
Models exist for predicting the movement and impact of
air masses, and the EPA has a series of air quality disper-
sion models referred to as UNAMAP that can be used. However,
these models were designed for overland dispersion and not
explicitly for overwater air quality modeling. Xn previous
applications over water, poor results have been obtained
mainly because the assumptions made in the models were based
on wind speed and net radiation levels generally found over
land and not generally found over water. Modification of
the models and additional measurements at the site during
the environmental testing will be used to overcome some of
these deficiencies.
Two approaches are proposed for tracking the plume for
model verification during the implementation of this strategy.
First, the high concentration of HC1 (which is produced
during the combustion of chlorinated organic substances)
released from the burning of PCB will provide an airborne
tracer, the HC1. However, this cannot be used to determine
the actual interaction of the plume with the sea surface
since HCJ is neutralized once it contacts seawater. The
second approach will be to add a material such as perfluori—
nated hydrocarbons (PFH) or sulfur hexafluoride (SF 6 ) that
can be tracked in the plume and in the water. Sulfur hexa—
fluoride is one good candidate for use as a tracer, since it
has been used as a plume tracer in many studies; analytical
techniques, including continuous monitors capable of detecting
as little as iO 2 parts SF 6 in air, are available. Several
studies with SF 6 as the plume tracer have been carried out
over the ocean. Tests have also shown that SF 6 is not
destroyed by seawater and can be detected in low concentra-
tions in seawater. SF 6 is not generally found in seawater;
thus, a very low background and good analytical techniques
will permit low detection limits. Tests of the validity of
these tracers as surrogates of emissions will be required.
Since the PFH’s are organic compounds, they should be the
best surrogates for tracking the transport of the emission
organics in the air and water. These compounds are also
very stable in water and should not be present in the ocean.
Detection limits are also very low for PFH’s. The use of an
organic tracer such as one of the PFH’s as a surrogate for
tracking the transport of the organic substances in the
emissions is the more desirable approach. However, cost or
technical problems may make this approach impossible, in
which case SF 6 will be used as the organic emissions plume
31

-------
tracer even though it is not organic. Generally, when tracers
such as PFH and SF 6 are used, a second material or plume
property such as plume visibility is used to help track the
plume. In this study, HC1 will be used as the supplementary
tracer, and, as with PFH and SF 6 very good, highly sensitive
detection methods are available for tracking the HC1.. Ammo-
nia can also be added to the plume to make it visible.
Several releases of PFH or SF 6 are planned to cover a
variety of meteorological conditions. The tracer will be re-
leased in the plume just above the stack at a point where
the stack gases have cooled below the decomposition tempera-
ture of the tracers. Tracking by ships and planes is recom-
mended. If the plume follows the predicted course, a limited
use of planes will be needed. If the plume rises, extensive
airplane sampling may be needed. Seawater samples and air
samples will be taken in the site area during the baseline
cruise for background data, and a wide range of samples will
be taken during the burn in the areas of plume touchdown.
These will include but not be limited to air, sea slicks,
and surface water samples.
The needed meteorological measurements will be made from
the incinerator ship and along the plume path. The standard
wind speed, wind direction, temperature, and relative humidity
measurements logged by the incinerator ship and sampling ships
can be supplemented with pilot balloon soundings and airplane
measurements for wind profiles and boundary—layer conditions.
When the plume has been identified using PFH or SF 6 and
HC1, the environmental samples will be collected under the
plume to be chemically analyzed for emissions—related sub-
stances for use in model verification.
The task of statistically identifying background levels
of the potential organic substances in the air, water, and
neuston of a pelagic site is quite different from defining
SF 6 or PFH background levels. An approach for sampling
and statistically describing the background levels of organic
substances will be developed. The numbers, kinds, and sizes
of the environmental samples will be an important considera-
tion. The approach selected will be used for the samples
taken before, during, and after the burns.
The on—site measurements of the tracer, HC1, and the
incinerator—emitted organics in the plume and at the surface
impact zone will be used to verify predicted values from
the modified overland atmospheric dispersion models. Addi—
32

-------
tional modeling and laboratory studies need to be conducted,
however, to determine how emissions—related substances are
incorporated into the surface microlayer and subsequently
transported into the surface water and biota.
3.2.3.2 Aquatic Model Evaluation
Assuming a small level of incomplete combustion plus
possible formation of organic compounds during the combus-
tion process, release to the atmosphere of some amount of
organic material is possible. The emissions during actual
operations may also include small amounts of metals that were
present in the original waste material. The most probable
point of impact of incineration stack emissions, and scrubber
water if applicable, will be the ocean surface and consequen-
tly the organisms directly or indirectly associated with
this area of the marine ecosystem. Chemical characterization
of emission exhaust will determine the types and levels of
potential emissions including organics and HC1. The predic-
tive modeling and tracer studies used in the research burn
will determine the area and points of plume impact. Models
will also be used to predict the levels and composition of
emissions that ultimately reach the ocean surface and subsur-
face waters and associated organisms. Prevailing currents,
wave action, and other hydrological disturbances will play a
major role in final fate and exposure concentrations. In
addition, biodegradation rates of emissions, potential for
bioaccuznulation, and residence time of Nd in surface micro—
layers are important components of an exposure assessment in
the aquatic environment.
In Area 3 of this strategy, studies are planned to deter-
mine if, how much, and for how long emission components
reside in the water surface microlayers. Laboratory ecosys-
tems (microcosms) with simulated lipid surface microlayers
will be sprayed with appropriate organic compounds similar
to expected emissions as determined by previous trial burns.
Periodic measurements will be made, using Nucleopore membranes,
to determine:
1) the rate of organic compound uptake;
2) residence time in the microlayer; and
3) disruption of microlayer integrity by Nd.
Chemical constituents of the surface films can affect these
rates and will be varied to determine their impact. Constitu-
ents such as protein, lipid, and carbohydrate content will
be varied.
33

-------
To determine exposure concentrations in marine surface
rnicrolayers, the microbiological fate of emissions will be
investigated. These studies can be conducted in the labora-
tory with real micro]ayer materials and associated microflora
or simulated microlayers as described above. Studies should
be conducted in sterile and non—sterile systems and every
effort made to account for mass balance of substances added
to the system(s). Microbes associated with the surface film
(i.e., microlayer) will be determined by chemical constitu-
ents of the film. Variation of these chemical constituents,
as discussed above, will be correlated to changes in the
microflora and assess their abilities to degrade the organic
emissions.
Field samples (surface microlayers) may be taken during
research burn events to determine concentrations of organic
and acid emission accumulation and to verify data from labora-
tory ecosystem studies. Background samples will be taken
prior to the burn to subtract levels of pollutants resident
in the surface microlayers and associated organisms.
Various sampling techniques are available to retrieve
environmental surface microlayers. These include screen tech-
niques, a teflon disk technique, a Nucleopore or Millipore
membrane technique, and a rollerdam technique. Each of
these techniques have benefits for retrieving certain consti-
tuents or components of films. Most are limited in volume,
thickness of film, or area which can be retrieved. One or
more of these methods would need to be used in a statistically
sound design in order to obtain valid information on film
components prior to and during the research burn. At the
present time, the statistical variability from sample to
sample using any of these procedures is unknown and would
need to be determined before a meaningful sampling program
could be designed.
Because of the intensive decomposer—herbivore (detriti—
vor) food chain in the niicrolayer, it is highly probably
that organic chemicals from stack emissions, if not decomposed
in the microlayer (biodegradation/photodecomposition), will
be transferred to higher organisms via the food chain (i.e.,
dinoflagellates, ciliates, and fish). Heavy surface slicks
on inland waterways and estuaries have been observed to
contain an extensive food chain with small fish feeding di-
rectly on the slimy and filamentous organic layer. Therefore,
if area 1 and 2 studies show that significant levels of
organics are being emitted from the incineration stacks
studies should be performed to determine if these substances
accumulate in the surface microlayer of the ocean and if so,
3A

-------
what is the potential toxicity of these chemicals to represen-
tative marine organisms 1hich may live in or feed on the micro—
layer. These organisms may feed during the night and then
migrate to deeper water during the day, thus dispensing the
emission constituents. The microlayer might therefore not
accumulate emissions at all but serve as a distribution
source.
Since surface microlayers contain a diverse community of
bacteria, it is highly probable that small quantities of
organic chemicals can be biodegraded to microbial biomass or
harmless organic or inorganic products. The densities of the
surface films, associated bacterial populations, and their
catabolic capabilities will vary with the quantity of nutrients
available, both organic and inorganic, as well as the type
of substrates. For example, highly proteinous films will
contain a higher proportion of peptide—utilizing bacteria
than a film containing a high concentration of lipids.
Since surface films collect and concentrate hydrophobic
chemicals across the ocean surface, these chemicals are exposed
to a higher ultraviolet (LW) radiation than if they were dis-
persed in the water column and adsorbed to particulates.
This enhanced liv exposure adds to the probability of photo—
decomposition of these compounds, particularly for photosen-
sitive chemicals. Photodecomposition may or may not reduce
the chemicals to harmless products;or render a normally
non—biodegradable chemical more susceptible to microbial
attack.
The ultimate fate of any incinerator emissions and there-
fore their exposure concentration to organisms of the marine
ecosystem will be a function of biodegradation, photodecom—
position, potential for bioaccumulation, and physical activi-
ties such as waves 1 currents, etc.
The microlayer and its associated organisms may be
exposed to both organics and acid emissions from organo—
chlorine combustion. While it is understood that relatively
alkaline seawater has the capacity to neutralize significant
quantities of acid (HC1), research may be done in Area 3 to
determine the fate and potential impact of fairly large
quantities of Nd on the important microflora associated
with this microlayer. Laboratory studies could be conducted
to determine the residence time of the acid precipitation in
the surface microlayers. If the acid is rapidly neutralized
by the alkaline seawater, no effect on the resident microbial
populations would be expected. If, on the other hand, the
acid precipitation significantly changes the pH in the surface
microlayer more than transiently, coincident changes in micro-
bial species composition and function could result.
35

-------
Field verifications of predicted acid residence time in
the surface water is needed. Surface water samples taken
during burns will help to verify laboratory data and assist
in monitoring potential depositions whether expected or unex-
pected.
The transport and exposure considerations addressed in
this research strategy will therefore try to describe the
transport and fate of any substances emitted from ocean
incineration and estimate the levels of these substances
that could exist in the environment based upon the results
of stack emissions analyses. Best— and worst—case scenarios
will be run through the models. If the worst-case scenario
shows an extremely low probability of exceeding background
ambient levels of emissions—related substances, the decision
could be made that additional toxicity—related studies may
not be necessary.
3.3 Biological Assessment Consideration
3.3.1 Overview
Information obtained from studies conducted to determine
the types and quantities of substances emitted from ocean
incinerators and the exposure concentrations of these sub-
stances in surface microlayers, subsurface waters, and air
above background levels, will help determine if toxicological
effects due to incineration stack emissions can be expected.
If these substances accumulate in the marine environment and
are not biologically degraded or photodecomposed, they may
have the potential to cause adverse effects or bioaccumulate
through the food chain. Laboratory studies will be conducted
to determine the concentration of emissions that will produce
significant changes in a suite of biological responses ranging
from acute mortality to growth, reproduction, or ecosystem
process interruption. If evidence does not exist which
indicates that elevated levels of substances will appear in
the environment (i.e., worst—case scenario), extensive labora-
tory studies of biological effects may not be needed. The
effects testing proposed in this plan is therefore tiered
with the least sophisticated and least complicated tests first,
with more sophisticated test following——if necessary. This
tiered approach will parallel the exposure—tiered approach and
be used in the tiered Risk Assessment described in section
3.3.4.
36

-------
3.3.2 Direct Biological Effects
The objective of the aquatic toxicology part of this
strategy is to determine from laboratory studies the concen-
tration of emissions that produce significant chan es in a
suite of biological responses ranging from acute mortality
to growth and reproduction. These biological responses to
various concentrations of emission substances can then be
compared to the concentrations estimated to occur in the
environment, and decisions can be made regarding the accepta-
bility of ocean incineration (see hypothetical example in
Figure 1).
Before any biological testing can be initiated using
emissions from burning hazardous wastes at sea, the testing
methods themselves must be developed and proven effective.
Preliminary studies will therefore be conducted where an
operational, land—based combustion facility will be used as
a source of emissions, and various emissions collection
methods will be tested. These methods will include passing
emission gas through a tank of seawater which will act directly
as an adsorbent. This bubble—through method will require
water, pH, and temperature adjustment before being used in
bioassays on selected organisms.
This testing will also include a determination of the
appropriate bioassay procedures, standard test organisms,
and end points which should be used in at—sea studies.
After sampling methods and toxicity test procedures are
developed, the approach is to collect emission products
during an at—sea research burn and to use emissions to gene-
rate exposure response curves for the most sensitive life
stages of standard test species using intercalibrated test
protocols. A test of these procedures may be conducted at a
land—based hazardous waste incinerator for toxicity range
finding prior to a research burn at sea. Additional test
species will subsequently be selected to include representa-
tives of the indigenous neuston population and sensitive
life stages of commercially important crustaceans, molluscs,
and fishes. The chronic toxicity tests will follow the
recommendations of the Water Quality Guidelines Committee
(1984) as much as possible, and the specific toxicity test
protocols for conducting the test will follow ASTM standard
practices whenever applicable.
31

-------
100.0 -
10.0 -
1.0
V. £
FIGURE 1
I
p
.
I
I
_____Worst Case Concen rat
-
tion
. QsPers ion
SUPE Sc ARIO
0.00001
EFFECT CONCENTRATIONS AND WORST CASE EXPOSURE ESTIMATE
(HYPOTHETICAL)
Acute
a
CONC.
Sub—lethal
I
Cenotoxic
I
I
I
I
0.01
0.001
Chronic
0.0001
I
S
Bioaccuinulatjon
5
10
15
20
TIME
25
30
38

-------
Because f the potential limitation of the test material,
a hierarchical testing strategy will be employed in first
attempts to define the relationship between acute and chronic
responses for standard test species and then determine the
range of acute toxicities for indigenous species. ‘ The mean
of the acute: chronic ratios from the standard test species will
then be applied to the acute toxicity values of the indigenous
species to estimate the chronic values of the most sensitive
species.
Specific test species cannot be recommended with certainty
at this time. However, the strategy proposed in previous sec-
tions will logically suggest the selection of certain species
types. For example, in order to optimize the predictive nature
of the effects testing, species such as those for which we
have standard protocols and extensive comparative toxicological
data bases will be given priority. Next, we need to look at
indigenous species to address the question of whether the
standard species are representative of the sensitivity of the
indigenous species. This will allow us to place some confidence
limits on predictions made from standard test species. The
precise approach to these types of tests will require further
investigation.
The types of biological responses that should be measured
range from survival to integrative responses such as growth
and reproduction. For selected standard test species, methods
have been developed for measuring the latter integrative respon-
ses from short—exposure durations and should be applied to
this problem since the duration of the tests will be strongly
influenced by the availability of the emission condensate.
Estimates of the potential for bioaccumulatiori of emission
products will be conducted using predictive methodologies inclu-
ding the application of log—octanol water partition coefficients
to those emission products that are identified in the emissions.
Direct measures of bioaccumulation will be difficult due to
the limited amount of emission product but will be attempted
on those chronic studies where calculations indicate that the
tissue residues could conceivably reach levels of analytical
detection given the exposure concentrations, bioaccumulatjon
factors, and available sample size.
Because of the nature of the original feedstock Contami-
nant composition and the combustion processes, there is the
possibility of genetic toxicants being present in the emissions.
To test this hypothesis, a hierarchy of genotoxicity tests
ranging from the Ames Test for mutagenicity, to tests for
genetic damage (Sister Chromatid Exchange), to tissue culture
tests designed to address the potential for tumor promotion
are included in Area 3” studies. However, the need to conduct
these higher level tests will depend on initial data on the
chemical characterization of the emissions and the availability
of material.
39

-------
3.3.3 Effects on the Surface Microlayer
Because surface microlayers can serve as food for higher
organisms, the potential exists for food chain bioaccumula—
tion of organic substances accumulated in the microlayer.
If laboratory tests are indicated, studies should be conducted
to determine the potential for bioaccumulation of these
chemicals where appropriate.
EPA already possesses information showing that signifi-
cant quantities of acid (HC1) will be emitted from combustion
of organochiorine waste. If these acid emissions contact
the ocean surface too rapidly and are not neutralized fast
enough, they could adversely affect the organism structure or
functions of the microlayer. Tests should be undertaken to
determine the potential impact of acid emissions on surface
film organism structure and function.
Since microorganisms accumulate in the surface microlay—
ers (dinoflagellates, ci]iates, and bacteria), acid precipi-
tation from plume fallout could affect their life cycles.
The potential damage to the ecosystem from such changes is
only specul on. Certain microbes, however, can only func-
tion within narrow pH ranges, and speciation of certain
metal ions necessary for growth and metabolic reaction are
regulated by small pH changes.
Endpoints of concern other than death and bioaccumulation
can be addressed in a laboratory situation using bioassays to
determine potential impacts of incineration acid fallout.
These bioassays include: ATP (total viable microorganisms
predominantly algae), phospholipid (total bacterial content),
and 14C—glucose metabolism (heterotrophic activity levels).
Other biochemical characterizations of microbiological and
macrobiologica]. communities can be utilized to investigate
changes in community structure and function.
Disruption of the community structure may or may not
result in changes in the community function when changing
the environmental pH. It has been suggested that acid rain
deposition on terrestrial ecosystems can change the nitrogen
recycling to some extent. Investigations of critical cycles
in marine surface films can be addressed. However, initial
studies will be needed to determine what function these
films play in the current cycling in the open waters. It is
believed that these films are vastly important to detrital
and fecal degradation since the decomposer populations are
10 — 10,000 times higher in this area than in subsurface
water.
40

-------
3.3.4 Comparative Environmental Risk Assessment
Ecosystem impacts are typically inferred from tests on
individual species from which forecasts are made to the popu-
lation and community levels of biological organization. This
can be accomplished by interfacing the source inp 1ts (e.g.,
pollutant type and loading rate) to the biological effects
through measurements of environmental exposure. Pollutant input
rates are then used as source terms for transportation and
fate models which are used to predict pollutant concentration
isopleths on defined spatial and temporal scales. This form
of output is directly compatible with biological effects measure-
ments in the laboratory that are typically described as func-
tions of pollutant concentration and exposure duration. The
conduct of a risk assessment occurs when the probability of
environmental impact is estimated from a comparison of the
environmental contaminant exposure with contaminant concentra-
tions producing biological effects.
One of the principal features of a risk assessment stra-
tegy is that the effects and exposure components are tiered.
The ordering of the tiers implies increasing degrees of com-
plexity, resolution, and predictive confidence. An additional
assumption is that there are clearly defined criteria for
decisions at each level of the hierarchy which then trigger
the need for increased data acquisition at the higher, more
complex tiers. A prediction of hazard is made by comparing
the predicted environmental exposure concentration of the mate-
rial (exposure assessment) and the concentration producing
biological effects in the laboratory studies (effects asses-
sment) at each tier. When properly synthesized, these data
provide a confidence—bounded estimate of the probability (risk)
of unacceptably altering the aquatic environment as the result
of the disposal of the waste. If either the confidence of the
prediction or the probability of risk is unacceptable, then
additional levels of testing can be conducted. This strategy
places the biological tests in a hierarchy of complexity which
reflects a continuum from detection to assessment methods with
increasing predictive confidence, and directly relates the
contaminant concentrations causing effects to the predict d
environmental concentrations determined at the same hierar-
chical level. This approach permits the development of a
series of testable hypotheses that is amenable to both labora-
tory and field verification.
41

-------