v>EPA
United States
Environmental Protection
Agency
Office of Solid Waste
& Emergency Response
Washington, D.C. 20460
EPA/530-SW-84-011
August 1984
Solid Waste
A Risk Assessment
of Waste Oil Burning
in Boilers & Space Heaters
-------
A RISK ASSESSMENT OF WASTE OIL BURNING
IN BOILERS AND SPACE HEATERS
FINAL REPORT
This report was prepared for
the Office of Solid Waste
under contract no. 68-02-3173
U.S Environmental Protection Agency
Region V, Library
230 South Dearborn Street
Illinois 60604
U.S. ENVIRONMENTAL PROTECTION AGENCY
January 1984
-------
This report was prepared by PEDCo Environmental, Inc.,
Cincinnati, Ohio, under contract no. 68-02-3173, and is reproduced
as received from the contractor.
Publication does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental Protection
Agency, nor does mention of commercial products constitute endorse-
ment by the U.S. Government.
-------
CONTENTS
Figures iv
Tables v
Acknowledgment x
1. Summary 1-1
1.1 Introduction 1-1
1.2 Approach 1-2
1.3 Results 1-7
2. Introduction 2-1
References for Section 2 2-8
3. Source Characterization: Facilities that Burn
Waste Oil 3-1
3.1 Boiler characterization 3-2
3.2 Oil space heaters 3-8
3.3 Modeled population 3-13
3.4 Emissions from boilers and space heaters 3-16
References for Section 3 3-26
4. Air Dispersion Modeling 4-1
4.1 Introduction : - 4-1
4.2 Point-source analysis with ISC model 4-4
4.3 Urban-scale analysis with the Hanna-Gifford
model 4-44
References for Section 4 4-60
5. Environmental Impact and Health Risks from Waste
Oil Burning 5-1
5.1 Environmental impact and health consequence
of threshold contaminants 5-1
5.2 Environmental impact and health consequence
of nonthreshold contaminants 5-5
References for Section 5 5-9
11
-------
6. Chlorinated Dibenzodioxins and Dibenzofurans 6-1
6.1 Health effects 6-1
6.2 Potential for dioxin and dibenzofuran formation
from the combustion of waste oil in boilers 6-6
6.3 Air dispersion modeling of dioxins 6-12
6.4 Risk assessment 6-20
References for Section 6 6-28
Appendix A - Oil-Fired Boiler Characterization A-l
Appendix B - Air Quality Modeling Techniques and
Assumptions B-l
Appendix C - Estimates of Waste Oil Burned in Urban Study
Area C-l
Appendix D - Comparison of Contaminant Emission Rates
from the Combustion of Waste Oil with
those from Other Combustion Sources D-l
Appendix E - Health Effects Assessment Method E-l
111
-------
FIGURES
Number Page
2-1 Quantities of Waste Oil Generated Annually and
Their Uses and Disposal 2-2
2-2 Plow Diagram of Project Approach 2-5
3-1 Occurrence of Various Size Boilers by Type and
Use 3-4
3-2 Combustion Principles of the Low-Pressure Atomiz-
ing and Vaporizing Pot Burners 3-11
4-1 Wind Rose for JFK International Airport, 1974-
1978 4-5
4-2 Isopleths of Lead Concentration over a 30-Day
Averaging Period for a 19 Liter/h Small Boiler 4-12
4-3 Isopleths of Lead Concentration over a 30-Day
Averaging Period for a 246 Liter/h Small Boiler 4-13
4-4 Isopleths of Lead Concentration over a 30-Day
Averaging Period for a 1330 Liter/h Medium
Boiler 4-34
4-5 Isopleths of Lead Concentration over a 30-Day •
Averaging Period for a 2660 Liter/h Medium
Boiler 4-35
4-6 Urban Area and Associated Study Grid Used in the
Hanna-Gifford Model Analyses 4-48
IV
-------
TABLES
Number Page
1-1 Summary of Waste Oil Concentration at the 50th,
75th, and 90th Percentile 1-4
1-2 Threshold and Nonthreshold Contaminants Con-
sidered in Dispersion Modeling Analyses 1-5
1-3 Results of Risk Analysis 1-7
2-1 Waste Oil Contaminants Considered in Dispersion
Modeling Analyses 2-3
3-1 Boilers with Potential for Burning Waste Oil 3-6
3-2 An Order of Magnitude Estimate of Boilers
Burning Waste Oil 3-7
3-3 Stack Parameters for Boilers with Capacities of
Less than 3 Million Btu/h 3-9
3-4 Stack Parameters for Boilers with Capacities of
Between 3 and 15 Million Btu/h 3-10
3-5 Oil Space Heater Parameters 3-13
3-6 Properties of Waste Oil and Residual Fuel Oil 3-18
3-7 Summary of Waste Oil Concentration at the 50th,
75th, and 90th Percentile 3-19
3-8 Calculated Emitted Portion of Total Input of
Selected Metals 3-20
3-9 Particle Size Distribution of Some Major Contam-
inants in Emissions from Waste Oil Combustion 3-23
3-10 Particulate Size Distribution of Uncontrolled
Emissions from an Oil-Fired Boiler 3-23
3-11 Half-Lives of Selected Waste Oil Organic Emis-
sions in Air 3-24
3-12 Assumed Emissions of Metals and Organics Used in
Air Dispersion Modeling 3-25
v
-------
TABLES (continued)
Number
4-1 Model Selection
4-2 Source Characteristics for Single and Multiple
Source Analysis: Small Boilers and Space
Heaters 4-7
4-3 Contaminant Emission Rates Used in ISC Modeling
Analysis: Small Boilers and Space Heaters 4-9
4-4 Ambient Air Impact from Combustion of Waste Oil
Containing Contaminants with a Threshold
Response in Individual Boilers and Space
Heaters Smaller than 15 x 10 But/h 4-10
4-5 Ambient Air Impact from the Combustion of Waste
Oil Containing Various Input Concentrations of
Barium in Individual Boilers,and Space
Heaters Smaller than 15 x 10 But/h 4-15
4-6 Ambient Air Impact from the Combustion of Waste
Oil Containing Various Input Concentrations of
Lead in Individual Boilers and Space Heaters
Smaller than 15 x 10 But/h 4-16
4-7 Ambient Air Impact of HC1 from the Combustion of
Waste Oil Containing Various Input Concentra-
tions of Chlorinated Organics in Individual
Boilers and Space Heaters Smaller than 15 x
10 But/h 4-17
4-8 Ambient Air Impacts from the Combustion of Waste
Oil Containing Various Input Rates of Metal
Contaminants in Space Heaters 4-18
4-9 Ambient Air Impact from the Combustion of Waste
Oil Containing Various Input Concentrations of
Barium in Individual Space Heaters 4-19
4-10 Ambient Air Impact from the Combustion of Waste
Oil Containing Various Input Concentrations of
Lead in Individual Space Heaters 4-20
4-11 Ambient Air Impacts from Multiple-Source Com-
bustion of Waste Oil Containing Contaminants
with a Threshold Response 4-24
4-12 Ambient Air Impacts from Multiple-Source Com-
bustion of Waste Oil Containing Barium at
Various Input Concentrations 4-25
vi
-------
TABLES (continued)
Number Page
4-13 Ambient Air Impacts from Multiple-Source Com-
bustion of Waste Oil Containing Lead at Various
Input Concentrations in Four Clustered Sources 4-26
4-14 Ambient Air Impacts from Multiple-Source Com-
bustion of Waste Oil Containing Chlorinated
Organics at Various Input Concentrations 4-27
4-15 Ambient Air Impacts from Sixteen Clustered
Sources Burning Waste Oil Containing Lead at
Various Input Concentrations 4-29
4-16 Source Characteristics for Single Source
Analysis: Medium-Size Boilers 4-30
4-17 Contaminant Emissions Used in Single Source ISC
Modeling Analysis: Medium-Size Boilers 4-31
4-18 Ambient Air Impacts from the Combustion of Waste
Oil Containing All Contaminants with a
Threshold Response in Medium-Size Boilers 4-33
4-19 Ambient Air Impacts from the Combustion of Waste
Oil Containing Various Input Concentrations
of Barium in Medium-Size Boilers 4-36
4-20 Ambient Air Impacts from the Combustion of Waste
Oil Containing Various Input Concentrations
of Lead in Medium-Size Boilers 4-37
4-21 Ambient Air Impacts of HC1 from the Combustion
of Waste Oil Containing Various Input Concen-
trations of Chlorinated Organics in Medium-
Size Boilers 4-38
4-22 Ambient Air Lead Impacts at Elevated Receptors
Resulting from the Combustion of Waste Oil
Containing Lead 4-41
4-23 Maximum 30-Day Concentrations Assuming 1 Percent
Levels of Two Organics in Waste Oil and 97
Percent Destruction Removal Efficiencies 4-45
4-24 Areas of Each Study Subarea 4-49
4-25 Estimated Contaminant Emission Rates per Liter
of Waste Oil Burned 4-50
vii
-------
TABLES (continued)
Number Page
4-26 Waste Oil Burning Characteristics Derived for
the Urban-Scale Dispersion Modeling Analysis 4-51
4-27 Area Source January Emissions from the Combustion
of Waste Oil Containing Contaminants with a
Threshold Response 4-52
4-28 Area Source Annual Emissions from the Combustion
of Waste Oil for Nonthreshold Contaminants 4-53
4-29 Ambient Air Impacts from the Combustion of Waste
Oil Containing All Contaminants with a
Threshold Response in the Urban Study Area 4-55
4-30 Ambient Air Impacts from the Combustion of Waste
Oil Containing Chlorinated Organics at Various
Plant Input Concentrations in the Urban Study
Area 4-56
4-31 Concentrations of Nonthreshold Contaminants from
Waste Oil Combustion in the Urban Area, Based
on 97 Percent Destruction Removal Efficiencies
of Organics and Emission of 75 Percent of the
Metals in the Oil 4-58
4-32 Concentrations of Nonthreshold Contaminants in
the Urban Area, Based on 99 Percent Destruction
Removal Efficiencies and Emission of 50 Percent
of the Metals in the Oil 4-59
5-1 A comparison of Estimated Maximum Exposure
Concentrations from Waste Oil Burning with
Environmental Exposure Limits 5-2
5-2 Lifetime Cancer Risk Associated with Waste Oil
Burning in a High-Density Urban Study Area,
Based on 97 Percent Destruction Removal Effi-
ciency for Organics and the Emission of 75
Percent of the Metals in the Oil 5-6
5-3 Lifetime Cancer Risk Associated with Waste Oil
Burning in a High-Density Urban Study Area,
Based on 99 Percent Destruction Removal Effi-
ciency of Organics and the Emission of 50
Percent of the Metals in the Oil 5-7
5-4 Waste Oil Constituents Presenting a Potentially
Unacceptable Cancer Risk 5-8
Vlll
-------
TABLES (continued)
Number Page
6-1 Contaminant Emissions Used in Single Source ISC
Modeling Analysis: Small Boilers and Space
Heaters 6-17
6-2 Ambient Air Impact from Waste Oil Combustion in
Individual Boilers and Space Heaters for
Variable TCDD Concentrations 6-18
6-3 Dioxin Emissions Used in Single-Source ISC
Modeling Analysis: Medium-Size Boilers 6-19
6-4 Maximum 30-Day Dioxin Concentrations in Ambient
Air as a Result of Waste Oil Combustion in
Medium-Size Boilers 6-19
6-5 Dioxin Emission Rates Per Waste Oil Burned
Used in Urban Scale Modeling 6-21
6-6 Area Source 30-Day-Averaged Emissions for TCDD
in Waste Oil (January) 6-21
6-7 Area Source Annual-Averaged Emissions for TCDD
in Waste Oil 6-21
6-8 Average Annual TCDD Concentrations for Waste
Oil Burning in the Urban Area 6-22
6-9 Estimated Risk from Dioxin Emissions 6-25
IX
-------
ACKNOWLEDGMENT
This report was prepared under Contract No. 68-02-3173 for
the U.S. Environmental Protection Agency's Office of Solid Waste.
The EPA Project Officer was Michael J. Petruska. PEDCo Envi-
ronmental, Inc., was the prime contractor; Franklin Associates
Limited was the major subcontractor. Paul Moskowitz of Brook-
haven National Laboratories was also a subcontractor. The PEDCo
Project Director was Robert S. Amick; the PEDCo Project Manager
was Catherine E. Jarvis. Major authors were Catherine Jarvis,
Diane Albrinck, George Schewe, and Les Ungers, all from PEDCo
Environmental, and Paul Moskowitz of Brookhaven Labs.
PEDCo gratefully acknowledges contributions from the fol-
lowing in the form of review comments and suggestions as the
study was being conducted: EPA Office of Air Quality Planning
and Standards of North Carolina, and Michael J. Petruska, Project
Officer.
-------
SECTION 1
SUMMARY
1.1 INTRODUCTION
The regulations promulgated under the Resource Conservation
and Recovery Act (RCRA) in 1980 (and the additions/revisions of
1981 and 1982) covered hazardous wastes, but left used oil large-
ly unregulated. In line with sections that allow exemptions for
materials that are reused, recycled, or reclaimed, used (waste)
oils have been used as fuels and dust suppressants without being
manifested or tested. Also, waste oils sometimes serve as car-
riers for other hazardous wastes, such as chlorinated and non-
chlorinated organic solvents. As a result of this lack of regu-
lation, waste oils containing heavy metals, organic solvents, and
other contaminants (e.g., polychlorinated biphenyls) are dispersed
into the environment with little knowledge of the potential
health impact and resulting risks to exposed populations. The
purpose of this study is to analyze the risks associated with the
use of waste oil as a fuel.
Risk assessment is an estimate of the probability and severity
of harm to human health or to the environment as a result of some
occurrence. This study is a quantitative assessment of the risks
associated with waste oil burning. The sources were characterized
in Section 3 and in Appendix A by examining waste oil burning
1-1
-------
practices and estimating emissions from oil space heaters, small
waste oil boilers (defined as those with capacities of less than
15 million Btu/h), and medium-size oil boilers (defined as those
with capacities of 15 to 150 million Btu/h). This effort was
followed by air dispersion modeling to estimate the ground-level
concentrations of waste oil contaminants in the emissions. (See
Section 3 and Appendices B, C, and D.) The resulting concentra-
tions in air were then used to estimate doses to exposed popula-
tions and to determine the response to the population in the form
of threshold toxic effects or excess cancers. (See Section 5 and
Appendix E.)
1.2 APPROACH
The numbers used to estimate risk to exposed populations
were derived from Threshold Limit Values (TLV's) or from carcino-
genic potency factors developed by the U.S. Environmental Protec-
tion Agency's (EPA's) Cancer Assessment Group. For those waste
oil contaminants that have a threshold response (i.e., a thresh-
old level below which no adverse effects are observed), the TLV's
were modified in two ways: 1) a factor was added to account for
lifetime vs. workweek exposure, and 2) another factor was added
to account for exposure to the most susceptible portion of the
population vs. the typical adult male worker. Risk was estimated
by comparing the modified TLV's—referred to as Environmental Ex-
posure Limits (EEL's)—with the concentrations calculated by the
dispersion models. For those waste oil contaminants classed as
carcinogens, the current theory is that no safe threshold exists.
1-2
-------
For these contaminants, the carcinogenic potency factors modified
for airborne exposure (referred to as reference concentrations)
were used to calculate excess cancers. The ambient air concentra-
tion estimated from the dispersion modeling is compared with the
reference concentrations to calculate the number of cancers that
would occur from exposure to the ambient air concentration. Risk
is determined by stating the number of cancers per 10,000, 100,000,
—4 —5
or 1,000,000 people, referred to as risk levels of 10 ,10 , or
10~ , respectively. Risk was also stated as the risk of cancer
to an individual (one chance in 300,000, etc.). Appendix B
explains the method for assessing health effects in more detail.
Waste oil varies widely in its composition, depending on the
type of oil (e.g., industrial oils such as hydraulic oil vs.
crankcase oil from automobiles and diesel engines), the extent of
its previous use, and the addition of other wastes (such as
degreasing solvents) to the oil. Because its composition varies,
waste oil cannot be easily characterized. The approach taken in
this study was to compile available waste oil composition data
and to look at the distribution of concentrations of contami-
nants. The median, 75th percentile, and 90th percentile concen-
trations were calculated for each contaminant, as shown in Table
1-1. The waste oil composition used for the modeling represented
the 90th percentile concentrations so that risk estimation would
err on the side of overestimating rather than underestimating the
risk. Table 1-2 identifies the modeled contaminants that exhibit
threshold responses (referred to as threshold contaminants) and
those that do not exhibit threshold responses (carcinogens, which
1-3
-------
TABLE 1-1. SUMMARY OF WASTE OIL CONCENTRATION AT THE
50TH, 75TH, AND 90TH PERCENTILE
lumber
of
samples
Median
concentration
at 50th
percent! le,
ppm
Concentration
at 75th .
percent-He,
ppm
Concentration
at 90th
percentile,
ppm
Metals
Arsenic
Barium
Cadmium
Chromium
Leadd
Zinc
17
159
189
273
227
232
11
50
1.1
10
220
469
14
200
1.3
12
420
890
16
485
4.0
28
1,000
1,150
Chlorinated solvents
Dichlorodifluoromethane
Trichlorotrifluoroethane
1,1, 1-tri chl oroethane
Trichloroethylene
Tetrachloroethylene
Total chlorine
78
26
123
126
100
62
20
<1
270
60
120
1,400
210
33
590
490
370
2,600
860
130
1,300
1,049
1,200
6,150
Other organics
Benzene
Toluene
Xylene
Benzo(a)anthracene
Benzo(a)pyrene
Naphthalene
PCB's
32
30
22
17
19
15
264
46
190
36
16
9
290
9
77
490
270
26
12
490
41
160
1,200
570
35
33
580
50
At the median, 50 percent of the analyzed waste oil samples had contaminant
concentrations below the given value.
Seventy-five percent of the analyzed waste oil samples had contaminant con-
centrations below the given value.
c Ninety percent of the analyzed waste oil samples had contaminant concentra-
tions below the given value.
d Values for lead were taken from 1979-1983 data only.
1-4
-------
are referred to as nonthreshold contaminants). Hydrogen chloride,
listed as a threshold contaminant, is assumed to be formed from
chlorinated compounds during the combustion of waste oil.
In addition to the waste oil contaminants listed in Table
1-2 and modeled in Section 4, dioxins are covered in a separate
section. These were treated separately for two reasons:
1. They are not contaminants that are usually found in
waste oil, but they may be formed during its combustion
as a result of precursors present in the oil.
2) Some dioxins are extremely toxic and their health
impacts are potentially significant; thus, more atten-
tion was given to the health effects description.
TABLE 1-2. THRESHOLD AND NONTHRESHOLD CONTAMINANTS
CONSIDERED IN DISPERSION MODELING ANALYSES
Threshold
Barium
Cadmium
Chromium
Lead
Zinc
1,1,1-Trichloroethane
Toluene
Hydrogen chloride
Nonthreshold
Arsenic
Chromium
1,1,2-Trichloroethane
Tetrachloroethy1ene
Trichloroethylene
Carbon tetrachloride
PCB's
Dioxins
Two air dispersion models were used to estimate ground-level
concentrations of contaminants from sources burning waste oil:
the Industrial Source Complex (ISC) model and the Hanna-Gifford
model. The ISC model, a point-source model used to estimate max-
imum concentrations around a single point source or several point
sources, was used to estimate ground-level concentrations of
threshold contaminants. The Hanna-Gifford model, an area-source
model used to estimate ground-level concentrations from many
1-5
-------
sources throughout an entire area (such as a city and its sub-
urbs) , was used as the "urban model" to determine concentrations
of both threshold and nonthreshold contaminants from the wide-
spread burning of waste oil. Air dispersion modeling results are
presented in Section 4.
For proper interpretation of the results, it is important to
know several of the assumptions used in the modeling. Emissions
of metals from waste oil burning can vary from 20 percent (or
less) to 100 percent, depending on boiler operation and condi-
tion. This range is very wide, but a review of data in the
literature indicated that the emission of about 50 percent of the
metals in the waste oil is common. In the dispersion modeling,
it was assumed that 75 percent of the metals are emitted. Al-
though this assumption is realistic, it could result in overesti-
mating risk in some cases. A limited amount of modeling was also
done based on an assumed emission of 50 percent of the metals in
the oil.
Destruction removal efficiencies (DRE's) of organics in
waste oil boilers were assumed to be 97 percent. Some additional
modeling was done based on a 99 percent ORE. Some recently com-
pleted waste oil test burns done as part of a current study for
the EPA confirm that the usual DRE exceeds 99 percent (occasion-
ally, slightly less). Again, the DRE assumption used in the
modeling is realistic, but may err slightly on the side of over-
estimating risk.
1-6
-------
1.3 RESULTS
The concentrations calculated from the air dispersion model-
ing were compared with the EEL's and the reference concentrations
for threshold and nonthreshold contaminants (see Section 5 for
detailed results). The threshold contaminants that appear to
present a potentially significant risk are barium, hydrogen chlo-
ride, and lead (see Table 1-3). Concentrations of each of these
substances from sources burning waste oil could have a signifi-
cant impact on air quality.' The other threshold pollutants (cad-
mium, chromium, zinc, naphthalene, toluene, and 1,1,1-trichloro-
ethane) do not appear to have a serious impact on air quality or
pose a significant health risk.
TABLE 1-3. RESULTS OF RISK ANALYSIS
Threshold substances posing a significant risk:
Barium
Hydrogen chloride
Lead
Nonthreshold substances posing given cancer risk levels:
Risk level Nonthreshold substances
-4
10 Chromium
10" Chromium
Arsenic
Dioxins
10" Chromium
Arsenic
Cadmium
Dioxins
Potential cancer risk estimates are also summarized in Table
1-3. At a risk level of approximately one cancer in 10,000 or
1-7
-------
-4
10 , chromium is a contaminant of concern. At a risk level of
10" , arsenic and, in some cases, dioxins become additional
contaminants of concern. At a risk level of 10~6, or one excess
cancer in a million people, cadmium also becomes a contaminant of
concern.
Other waste oil nonthreshold contaminants pose lesser cancer
risk levels of 10~ (carbon tetrachloride, PCB's, tetrachloro-
ethylene, and 1,1,2-trichloroethane) or 10~8 (benzene and tri-
chloroethylene).
1-8
-------
SECTION 2
INTRODUCTION
Of the approximately 4.3 billion liters (1.1 billion gal-
lons) of waste oil generated per year in the United States, a
major portion, about 1900 to 2500 million liters (500 to 660
million gallons), is burned in boilers, kilns, diesel engines,
and waste oil heaters. This is the largest single use of waste
oil; the remainder is re-refined, used as dust suppressants,
landfilled, or dumped. (See Figure 2-1.)
Waste oils contain many contaminants, either because of
their uses or because they are mixed with other chemical wastes.
Some of the contaminants found in waste oil include heavy metals,
particularly lead; organic solvents such as benzene, xylene, and
toluene; and chlorinated organics such as trichloroethane, tri-
chloroethylene, and polychlorinated biphenyls (PCB's). Many
contaminants commonly found in waste oil are either toxins or
carcinogens and are therefore potentially hazardous.
Table 2-1 shows the waste oil contaminants that were modeled
for this study. The metals and the organics listed were chosen
because they are commonly found in waste oil and because all can
be measured in emissions. This latter point is important be-
cause stack emissions from waste oil-fired boilers are being
measured in concurrent studies (by other firms), and results
2-1
-------
10
GENERATION
LITERS GALLONS PERCENT OF TOTAL
4350 1150 100
USE AND DISPOSAL
LITERS GALLONS PERCENT OF TOTAL
TOTAL 3200 850 74
BURNING 1900-2500 500-660 43-57
ROAD OILING 190-380 50-100 4-9
LAND DISPOSAL/ 950-1100 250-300 22-26
DUMPING
LUBE OIL 340-380 90-100 8-9
OTHER 190-380 50-100 4-9
NOT RECEIVED
LITERS GALLONS PERCENT OF TOTAL
1100 300 26
ALL VALUES IN MILLIONS
OF LITERS AND MILLIONS
OF GALLONS
Figure 2-1. Quantities of waste oil generated annually and their uses and disposal.
-------
from these studies will be used to modify assumptions used in
the modeling if the data show that adjustments are necessary.
TABLE 2-1. WASTE OIL CONTAMINANTS CONSIDERED IN
DISPERSION MODELING ANALYSES
Response
Threshold
Barium
Cadmium
Chromium
Lead
Zinc
1,1,1-Tri chloroethane
Toluene
Hydrogen chloride
Nonthreshold
Arsenic
Cadmium
Chromium
1,1,2-Trichloroethane
Tetrachloroethy1ene
Trichloroethylene
Benzene
Carbon tetrachloride
Dioxins
PCB's
a Assumed to be formed during combustion rather than
present as a waste oil contaminant.
Contaminants in Table 2-1 are also divided into those that
elicit a toxic response above a threshold level, and those that
are classed as nonthreshold substances (e.g., those for which no
safe threshold is known). This distinction is important in the
models chosen to estimate ambient concentrations and in the
interpretation of the results.
To date, the environmental impacts of burning waste oil
have not been fully assessed. Several studies have measured
2—4
lead emissions, and some have used modeling to predict
ambient concentrations of lead from a source burning waste
2 5—fl
oil. ' Lead emissions from a source burning waste oil are
significantly greater than those from a source burning fuel oil.
2-3
-------
The impact of waste oil burning on ambient air depends on many
factors: number of sources, emission control equipment, stack
heights, meteorological conditions, background levels, and the
waste oil itself. Ground-level concentrations can approach or
exceed the National Ambient Air Quality Standard for lead of 1.5
ug/m . ' ' ' Thus, the data indicate that the burning of waste
oil containing lead may pose a human health hazard, at least
under some conditions.
Data on the environmental impact from other metals and
organic contaminants as a result of burning waste oil are far
9
more limited. Recent media coverage has focused attention on
the potential for adverse effects from chemical wastes in waste
oil, and a limited number of studies have concluded that there
is a potential for adverse effects from both organic contami-
3 4
nants and from trace metals other than lead. '
The U.S. Environmental Protection Agency's Office of Solid
Waste is funding a study to assess the environmental impact of
various waste oil practices, including its use as a fuel, its
use as a dust suppressant, and its storage. Three separate
reports (this and two others) characterize each of the practices
and analyze the associated risks.
This report describes the practices of waste oil burning,
estimates its impact on air quality, and assesses the resulting
risks to human health. Figure 2-2 is a flow diagram of the
approach to this study, which consists of three main parts: a
source characterization, air dispersion modeling, and risk
assessment. The source characterization in Section 3 describes
2-4
-------
SOURCE CHARACTERIZATION
0 SMALL BOILERS
0 MEDIUM BOILERS
0 SPACE HEATERS
KEY:
MAJOR TASK
TASK OUTPUT
0 STACK PARAMETERS
0 EMISSIONS
AIR DISPERSION MODELING
0 POINT SOURCE MODEL
0 URBAN AREA MODEL
AMBIENT CONCENTRATIONS OF CONTAMINANTS
0 POINT SOURCE - THRESHOLD
0 URBAN AREA - THRESHOLD
AND NONTHRESHOLD
RISK ANALYSIS
0 HEALTH EFFECTS
0 ENVIRONMENTAL EXPOSURE LIMITS - THRESHOLD CONTAMINANTS
0 REFERENCE CONCENTRATIONS - NONTHRESHOLD CONTAMINANTS
0 COMPARISON OF EEL's AND AMBIENT CONCENTRATIONS OF
THRESHOLD CONTAMINANTS
0 COMPARISON OF REFERENCE CONCENTRATIONS AND AMBIENT
CONCENTRATIONS OF NONTHRESHOLD CONTAMINANTS:
CALCULATION OF CANCER RISK TO AN INDIVIDUAL
0 CONCLUSIONS
Figure 2-2. Flow diagram of project approach.
2-5
-------
the boilers and space heaters modeled. This analysis focuses on
burning of waste oil in small boilers (defined as those with a
capacity of less than 15 million Btu/h) and in waste oil space
heaters (commonly used in service stations). Some modeling and
discussion of medium-sized boilers (defined as those between 15
and 150 million Btu/h) is also included. The stack parameters
chosen for each modeled source and the emissions calculated for
each source are used in the air dispersion modeling. •
Section 4 presents a description of the air dispersion
models used to estimate ground-level concentrations of contami-
nants emitted by waste oil boilers and space heaters. The two
types of air dispersion modeling used were 1) a point source
model to estimate emission dispersion from an individual source,
and 2) an urban area model to estimate emission dispersion from
many sources spread over a densely populated area. Section 4
presents the results of the two air dispersion models. The
output of point source modeling is ambient concentrations of
waste oil contaminants eliciting a threshold response. The
urban area modeling results include concentrations of both
threshold and nonthreshold contaminants.
The preliminary work described in Sections 2 through 4
provides the background data for the risk analysis in Section 5.
The risk analysis is a quantitative assessment of the hazard
that waste oil burning poses to human health. It is divided
into an evaluation of the risks associated with exposure to
the threshold and nonthreshold contaminants in the oil. For
threshold substances, "Environmental Exposure Limits" (EEL's)
2-6
-------
were derived, based on modifications of Threshold Limit Values
for occupational exposure. For nonthreshold contaminants (car-
cinogens) , reference concentrations were derived as a means of
calculating the cancer risk to an individual from exposure to
contaminants released into the air from waste oil burning.
Appendix C presents the derivation of the EEL's and reference
concentrations; Section 5 focuses on the air dispersion modeling
results and includes a discussion of the associated risks. Sec-
tion 6 addresses dioxins, their potential for formation during
combustion of waste oil, potential health effects, dispersion
modeling, and risk assessment.
2-7
-------
REFERENCES FOR SECTION 2
Franklin Associates Limited and PEDCo Environmental, Inc.
Survey of the Waste Oil Industry and Waste Oil Composition.
(Draft Report.) April 1983.
Waite, D. A., et al. Waste Oil Combustion: An Environmen-
tal Case Study. APCA Paper No. 82-5.1. Presented at Annual
Meeting of Air Pollution Control Association, New Orleans,
June 20-25, 1982. p. 1-15.
Brinkman, D. W., P. Fennelly, and N. Suprenant. The Fate of
Hazardous Wastes in Used Oil Recycling. GCA Corporation.
July 1983.
Walker, W. B. Pollution of the Environment by Burning of
Waste Oils. Environmental Pollution Management, 11(3):
80-82, May/June 1981.
Recon Systems, Inc., and ETA Engineering, Inc. Used Oil
Burned as a Fuel. Vols. 1 and 2. October 1980.
Devitt, T., et al. Population and Characteristics of Indus-
trial/Commercial Boilers in the U.S. EPA-600/7-79-178a,
August 1979.
New Jersey State Department of Environmental Protection.
Memo from J. Held to G. Pierce on calculations of metal
contaminants from waste oil burning. October 6, 1982.
New York State Department of Environmental Conservation
Division of Air. Proposed regulations Subpart 225-2 (Fuel
Oil and Waste Fuel) of 6NYCRR 225 (Fuel Composition and
Use). September 1982.
American Broadcasting System. ABC News "20/20" Transcript
on Use of Waste Oil as Fuel. December 17, 1981.
2-8
-------
SECTION 3
SOURCE CHARACTERIZATION:
FACILITIES THAT BURN WASTE OIL
Waste oil can be burned in virtually any facility designed
to burn No. 6 fuel oil, and in most facilities designed to burn
No. 4 and No. 5 fuel oils, although modifications may be neces-
sary in systems designed for the lighter fuels. The available
literature indicates that waste oil is currently burned in the
following facilities or devices: boilers, oil heaters, incin-
erators, asphalt plants, cement kilns, and diesel engines. '
Although no currently available survey data show the exact amount
of waste oil burned in each of these units/devices, a study by
Development Planning and Research Associates, Inc., reported the
proportional distribution to various combustion units or devices
as follows :
Combustion unit/device
Cement kilns
Diesel engines
Oil space heaters
Oil-fired boilers
Relative distribution of
waste oil by source, %
2
2
3
92
These figures indicate that oil-fired boilers consume the
largest portion of waste oil; therefore, these are the sources of
concern in this study.
3-1
-------
3.1 BOILER CHARACTERIZATION
3.1.1 Types of Boilers
The three major types of boilers are water-tube, fire-tube,
and cast iron. Water-tube boilers are generally the largest, and
they also have the greatest range of capacity. Cast iron boilers
are usually very small.
A water-tube boiler is one in which the hot combustion gases
resulting from combustion of fuel contact the outside of the heat
transfer tubes, whereas the boiler water and steam contact the
inside of the tubes. These boilers, which generate high-pressure,
high-temperature steam, are available in many sizes, generally in
the range of 15 x 106 to 1500 x 106 Btu/h.
In fire-tube boilers, the hot combustion gases flow from the
combustion chamber, which is usually enclosed in a boiler shell,
through a tube that is surrounded by a water basin that absorbs
heat through the shell and tubes. These units, which are usually
small (<20 x 10 Btu/h), are used where loads are relatively
constant.
Cast iron boilers are similar to fire-tube boilers and are
sometimes classified as such. In these very small boilers (0.003
x 10 Btu/h to 14 x 10 Btu/h), the combustion chamber is sur-
rounded by a water basin that is lanced with flues, or in commer-
cial units, a maze of tubes. The flues and tubes allow the
combustion gases to transfer heat to the water and escape from
the chamber. These boilers require little maintenance and can
handle overloading or demand surges.
3-2
-------
3.1.2 Boiler Sizes
Figure 3-1 classifies boiler sizes by type and by use. Both
residual and distillate fuel oil can be burned in boilers of all
sizes. Commercial/institutional use includes space heating, wa-
ter heating, and cooking at nonmanufacturing establishments, such
as apartment buildings, motels, restaurants, wholesale and retail
businesses, health and educational institutions, and government
office buildings. Domestic or residential use is defined as
space heating, water heating, cooking, and other household opera-
tions at private households, including farmhouses. The small
boilers of most interest in this study are fire-tube and cast
iron boilers used for commercial/institutional and residential
purposes. The medium-sized boilers modeled in the study are
usually water-tube boilers used for space heating.
The following listing of typical applications provides some
perspective to the boiler sizes discussed in this report:
Boiler size Typical application
60,000 Btu/h 1,000 ft* apartment
100,000 Btu/h 2,000 ft* house
300,000 Btu/h 6,000 ft2 - 5 small apartments
2,000,000 Btu/h 40,000 ft2 - large church and school
7,000,000 Btu/h 140,000 ft2 - large 7-story office building
15,000,000 Btu/h 280,000 ft2 - large 10-story hospital
3.1.3 Boilers Most Likely to be Fueled by Waste Oil
Despite the lack of data on waste oil users, some assump-
tions can be made as to the kinds of boilers most amenable to
waste oil burning and those for which waste oil is not a likely
fuel.
3-3
-------
Ul
I
Parameter
Boiler type
Water-tube
Fire-tube
Cast iron
Usage
Utility
Industrial
(process)
Industrial
(space heat)
Commercial
Domestic
Capacity range, 106 Btu/h
0.4 1.0 10 25 100 500 1,500
Figure 3-1. Occurrence of various size boilers by type and use.
-------
Waste oil can be burned in any facility designed for No. 6
fuel oil, and in most facilities designed for No. 4 and No. 5
fuel oils (with some possible modifications); therefore, it is
assumed that waste oil is more likely to be burned at facilities
that currently burn residual fuel oil rather than at those that
burn distillate fuel oil.
Because of the potential tube and furnace fouling associated
with water-tube boilers, waste oil burning in these boilers may
require some design modifications, e.g., the installation of fuel
filters, air- and steam-assisted burners, "dirty" tanks, soot
blowers (on larger units), and air pollution control equipment.
Fire-tube and cast iron boilers lend themselves more readily to
"dirty" oil firing than do water-tube boilers. For this reason,
waste oil is believed to be fired in many boilers of this type.
Table 3-1 presents the potential population of waste-oil-
burning boilers. As shown in Table 3-2, an estimated 50,000 to
60,000 of these boilers now burn waste oil. The assumptions
necessary for arriving at this estimate are as follows:
0 Waste oil that is burned in boilers is blended with
virgin oil: 25 percent waste oil and 75 percent virgin
oil.
0 Waste oil market penetration is higher in larger boilers
and residual oil boilers and lower in the smaller units
and distillate oil boilers.
3.1.4 Boiler Stack Parameters
PEDCo compiled a list of stack parameters for small boilers
(defined as those less than 15 x 10 Btu/h), based on a 1982
3-5
-------
TABLE 3-1. BOILERS WITH POTENTIAL FOR BURNING WASTE OIL
Boiler type
Water-tube
Fire-tube
Fire-tube
Cast iron
Cast iron
Total
Fuel
Residual oil
Residual oil
Distillate oil
Residual oil
Distillate oil
Size, 106 Btu/h
<25
25-250
>250
0.4-25
25-50
0.4-25
25-50
<0.4
0.4-10
<0.4
0.4-10
Total number
7,119
8,464
370
72,850
833
47,598
543
203,569
95,899
127,833
60,224
625,302
Total
capacity,
10fe Btu/h
62,300
527,000
166,400
253,400
31,200
165,800
20,400
51,300
132,200
32,200
82,900
1,525,100
Load
factor3
0.106
0.106
0.106
0.295
0.295
0.469
0.469
0.295
0.295
0.469
0.469
Annual fuel
consumption,
1015 Btu
0.058
0.489
0.155
0.655
0.081
0.681
0.084
0.133
0.342
0.132
0.341
3.151
CJ
I
a\
Based on Table 3-1 of Reference 1. Load factor refers to the fraction of total boiler capacity actually
utilized.
-------
TABLE 3-2. AN ORDER-OF-MAGNITUDE ESTIMATE OF BOILERS BURNING WASTE OIL"
Boiler type
Water-tube, residual" "oil
Fire- tube/cast iron,
residual oil
Fire-tube/cast iron,
distillate oil
Total
Boiler
size (106 Btu/h)
Medium (100-500)
Large (500-1500)
Power plant (1500)
Small (10-50)
Very small (0.4-10)
Small (10-50)
Very small (0.4-10)
Estimated number
of boilers
using waste oil
615
15
1
2,362
45,175
927
3,144
52,239
Source: Reference 1.
3-7
-------
retrieval from the Ohio EPA boiler inventory. Tables 3-3 and 3-4
show the numbers obtained from the inventory for boilers with
capacities of less than 3 million Btu/h and those with capacities
between 3 and 15 million Btu/h.
3.2 OIL SPACE HEATERS
For purposes of this discussion, oil heaters are defined as
combustion units with capacities of 0.4 million Btu/h or less and
that used to heat either air or water. The principal users of
space heaters are automobile service stations. Other users
include automobile and truck dealerships; truck, automobile, and
taxicab fleet operators; automotive repair shops; and farm opera-
tors.
3.2.1 Types and Number of Oil Space Heaters
The firing methods of the two principal types of oil heaters
(vaporizers and atomizers) are different. Figure 3-2 illustrates
these two different combustion principles. The vaporizing unit
burns heated oil vapor, whereas atomizing burners use mechanical
force to atomize the fuel before it is ignited. The major dif-
ferences between these two heating systems are 1) the vaporizing
unit leaves a significant amount of residue compared with that
left by the atomizing unit, and 2) the atomizing unit yields
higher gas-phase concentrations of most inorganic species than
4 5
the vaporizing unit does. ' Because the vaporizing-pot combus-
tion systems trap metal contaminants in the residue, they emit
fewer metal pollutants to the air. Atomizers, on the other hand.
3-8
-------
TABLE 3-3. STACK PARAMETERS FOR BOILERSWITH CAPACITIES OF
LESS THAN 3 MILLION Btu/h*
Boiler size,
106 Btu/h
1
3
3
1
3
2
1
2
2
1
Average
1.9
Stack
height,
m
5.5
18.9
12.5
18.3
7.3
9.1
10,7
12.2
19.8
7.3
12.2
Stack
diameter,
m
0.46
0.67
0.40
0.46
0.52
0.88
0.43
0.85
1.04
0.40
0.61
Maximum
temperature,
°C
10
77
107
232
204
93
177
177
177
204
146
Feed rate,
liters/h
26
83
83
30
114
53
38
53
53
26
56
Gas flow ,
m3/min
9.91
31.18
31.18
11.4
42.54
19.9
14.2
19.9
19.9
9.91
21.0
Exit
velocity,
m/s
1.0
1.5
4.21
1.16
3.35
0.55
1.6
0.58
0.40
1.3
1.56
Source: Ohio EPA Boiler Inventory (1982 retrieval).
Assuming "22.48 m3/liter of oil burned. Calculated by use of North American
Combustion Handbook, 1st edition, 1957 (average value for fuel oil).
Exit velocity (m/s) =
46
60 * dz
where G = gas flow, m3/min
d = stack diameter, m
3-9
-------
TABLE 3-4. STACK PARAMETERS FOR BOILERS WITH CAPACITIES OF
BETWEEN 3 AND 15 MILLION Btu/ha
Boiler size,
106 Btu/h
9
4
12
13
4
9
4
Average
7.86
Stack
height,
m
12.2
13.7
20.4
9.14
7.62
7.62
7.32
11.14
Stack
diameter,
m
0.40
0.55
0.76
0.73
0.76
0.46
0.52
0.61
Maximum
temperature,
°C
149
107
316
288
399
288
204
250
Feed rate,
liters/h
261
114
348
329
95
227
110
212
Gas flowb,
m3/nrin
97.85
42.54
130.5
123.4
35.46
85.10
41.12
79.41
Exit
velocity,
m/s
13.23
2.99
4.79
4.91
1.31
8.66
3.26
5.59
. Source: Ohio EPA Boiler Inventory (1982 retrieval).
Assuming ~22.48 m3/liter of oil burned. Calculated by use of North American
Combustion Handbook, 1st edition, 1957 (average value for fuel oil).
Exit velocity (m/s)
where G = gas flow, m3/min
d = stack diameter, m
4G
3-10
-------
u»
I
DAMPER—\
\
STACK
*—OUTLET EMISSIONS
DAMPER
I
\
\
STACK
OUTLET EMISSIONS
-VAPORS
OIL
POT RESIDUE
LOW-PRESSURE AIR
ATOMIZAT ION BURNER
VAPORIZING POT BURNER
Figure 3-2. Combustion principles of the low-pressure atomizing and vaporizing pot burners.
-------
release to the air nearly all of the metal pollutants contained
in the fuel.
Compared with most burning facilities, oil heaters consume
relatively low volumes of waste oil; therefore, the overall abso-
lute quantity of emitted metals is low. Because of the heaters'
low stack heights and probable low gas emission velocities,
however, little dispersion of the metallic contaminants occurs,
so ground-level concentrations (particularly for atomizing burn-
ers) are relatively high.
Oil space heaters are easily adaptable to waste oil burning
and are often designed exclusively for that purpose, particularly
those used at facilities that generate waste oil (e.g., gas
stations). If waste oil is burned, it is usually the only fuel;
it is rarely blended with fuel oil.
A recently completed study of the waste oil space heater
industry estimates that there are 10 manufacturers of waste-oil
space heaters in the United States and that two brands are made
4
in Europe and sold here. Nearly 34,000 space heaters were sold
in this country in the past 3 years; about 90 percent of these
were vaporizing units and the other 10 percent were atomizing
units. Average life expectancy on such units is believed to be
about 10 years.
3.2.2 Oil Space Heater Parameters
Table 3-5 summarizes the parameters of oil heaters. Stack
heights vary with the location of the heater. Because they are
usually vented to the roof, stack heights can vary from 2 to 10
3-12
-------
meters. Stack diameters are generally in the range of 15 to 25
centimeters. Stack temperatures vary with the unit: most are
between 300° and 400°C, but some are as low as 165°C. Fuel feed
rates vary from 0.4 to about 15 liters (0.1 to 4 gallons) per
hour.
TABLE 3-5. OIL SPACE HEATER PARAMETERS
Stack temperature, °C
Stack diameter, cm-
Stack height, m
Exit velocity, m/min
Volumetric flow rate at STP
(20°C, 1 atm), m3/s
Fuel feed rate, liters/h
165-420
15-25
2-10
75-130
0.014-0.025
0.4-15
Source: References 5 through 13.
3.3 MODELED POPULATION
3.3.1 Population Density and Distribution of Boilers
Most small residual-oil-fired boilers can be classified as
industrial, commercial/institutional, or residential. Some in-
formation on the distribution of industrial and commercial/insti-
tutional boilers was obtained for this study and is presented in
detail in Appendix B, along with some data on utility boilers.
Only a brief summary is included here.
Nearly 90 percent of the U.S. industrial residual-oil-fired
boilers have a heat input capacity of less than 10 million Btu/h;
however, this group represents only 20 percent of the total
industrial boiler capacity. Fifty percent of the capacity is
generated by the larger boilers (in the 25 to 250 million Btu/h
range). (See Appendix B for details.)
3-13
-------
Another area of interest was the location of industrial
boilers in densely populated urban areas versus nonurban areas.
With the number of production employees used as an indicator of
the number of boilers and data from 22 states, the following
statistics were compiled:
0 In the 22 states investigated, approximately 44 percent
of all production workers are located in cities with
populations of 25,000 or more and at least 450 manufac-
turing employees; 56 percent are located in nonurban
areas. If this is used as an indicator of industrial
boiler distribution, the same percentages would apply,
which means that 44 percent of the industrial boilers
are located in cities and 56 percent in nonurban areas.
0 Based on a broader definition of urban and nonurban
areas, about 78 percent of all production workers are
located either within cities (same definition as above)
or in Standard Metropolitan Statistical Areas (SMSA's),
and the remaining 22 percent are located in nonurban
areas. Again, if industrial boiler distribution corre-
lates with production workers, then more than three-
quarters of the industrial boilers are in urban areas,
and a little less than one-fourth are in urban areas.
(See Appendix A for details.)
Commercial/institutional oil-fired boilers are those used in
hospitals, greenhouses, shopping malls, and similar applications.
Because the nature of this type of boiler makes their distribution
correlate well with population, no attempt was made to locate
actual commercial/institutional boilers and categorize them as
being in an urban or nonurban location.
Unlike industrial boilers, most commercial/institutional
boilers are small: 88 percent (by number) have heat input capac-
ities of less than 10 million Btu/h. Much of this heat input
capacity (i.e., about 25 percent) is found in the range of 1.5 to
10 million Btu/h range. About half of the total generated capac-
ity is represented by boilers with heat input capacities of less
3-14
-------
than 10 million Btu/h. Tables A-9 through A-14 in Appendix A
provide more detailed data on the numbers and capacities of
various sized commercial/institutional boilers by type of boiler
(i.e., water-tube, fire-tube, etc.).
Although it was not an objective of this study to survey
small boilers, it is useful to provide some indication of the
relative potential use of waste oil across the nation. Two
approaches to obtaining an estimate of the geographical distri-
bution of small boilers were considered: 1) to base the distri-
•
bution on residual fuel oil deliveries to the commercial sector
in each state, and 2) to obtain data from a survey of states and
waste oil collectors/processors. These approaches are discussed
in detail in Appendix A. Based on the latter method of estimating
the amount of waste oil generated and burned in each state (as
presented in a survey by Franklin Associates, Ltd. ), the states
with the greatest potential number of small boilers that burn
waste oil are New York, New Jersey, Pennsylvania, Illinois,
Michigan, Texas, and California. This is a reasonable estimate,
but its accuracy is questionable. (See Appendix A.)
3.3.2 Population Density and Distribution of Oil Space Heaters
Actual data are not available on the population density and
distribution of oil space heaters. Users of space heaters in-
clude service stations, automobile dealerships, trucklines, and
farmers. The distribution of the population of about 34,000
space heaters is assumed to be similar to that of the general
population except in very warm climates, where the need is limited.
3-15
-------
3.4 EMISSIONS FROM BOILERS AND SPACE HEATERS
Emissions resulting from burning waste oil depend on the
design of the combustion unit and control equipment, unit capac-
ity, burner type, operating parameters, waste oil contaminants,
and feed rate. Large water-tube boilers with firebox residence
times of at least 1 second and firebox exit temperatures of high-
er than 1500°F provide more complete combustion than the smaller
fire-tube boilers. Many of the factors that affect the quality
of combustion (e.g., excess air, fuel homogeneity, firebox^heat
release rate, and on/off cycling) are not as carefully controlled
14
in small boilers.
In an oil-fired boiler, the four basic atomization methods
are air, steam, mechanical, and rotary-cup atomization. The
trend is toward mechanical atomization in small units, and toward
air atomization in larger units. Generally, rotary-cup burners
are being phased out because of their high maintenance require-
ments; however, these burners are most adaptable to waste oil
firing and continue to be used for that purpose. Small water-
tube, fire-tube, and cast iron boilers that are fired by the same
type of single burners have similar combustion characteristics
and, consequently, similar emission factors. Small commercial/
institutional boilers generally do not have air pollution equip-
ment. When such boilers burn waste oil as fuel, however, the
lack of controls becomes a concern because of the relatively high
levels of lead and other heavy metals that can be emitted as
particulates. The potential emissions of most concern are metals,
halide acids, and organics (solvents and chlorinated organics).
3-16
-------
3.4.1 Metals Emissions
Table 3-6 presents average physical and chemical properties
of waste oil and residual fuel oil. Table 3-7 presents the aver-
age contaminant concentrations of nearly 400 waste oil samples.
The numbers represent median concentrations, 75th percentile
concentrations, and 90th percentile concentrations. The metals
of particular concern and interest are arsenic, barium, cadmium,
chromium, lead, and zinc.
Some test data are available on emissions of metals from
waste oil boilers, and calculations of a value for percent metals
emitted was based on these data (total concentration in the feed,
feed rate, and measured emission rates). Table 3-8 presents
these values for the metals of interest. (Lead and zinc values
represent an average of five data points; values of all other
metals represent averages of two data points.) The average
emission rates of these metals is 31 to 75 percent; individual
emission rates ranged from 20 to 100 percent. As a maximum,
worst-case conditions would be based on the assumption that all
metals in the fuel are quantitatively emitted. A reasonably
conservative, but more realistic, estimate would be 75 percent.
Thus, if a facility were burning 100 percent waste oil at a feed
rate of approximately 132 liters/h (35 gal/h), the estimated
potential lead emission from this source would be 99 g/h or 0.028
g/s, calculated as follows:
3-17
-------
TABLE 3-6. PROPERTIES OF WASTE OIL AND RESIDUAL FUEL OIL
Parameter
Heat value, Btu/lb
Btu/gal
API gravity at 60°F
Density at 15°C, g/ml
Viscosity at 40°C, cs
Water, %
Ash, %
Carbon, %
Hydrogen, %
Nitrogen, %
Sulfur, %
Chlorine, %
Arsenic, yg/g
Barium, ug/g
Cadmium, ug/g
Chromium, ug/g
Lead, ug/g
Zinc, ug/g
Estimated
typical value
of waste oil
18,500!)
138,000a
26a
0.8960s
66. 6a
2.4a
1.16a
83. 9a
13. 2a
0.093a
0.5a
1.4C
llc
50C
l.lc
10C
220c'e
469C
Estimated
typical value of
residual fuel oil
18,500a
149,000a
13-16b
350b
0.08b
0.2-0.3b
86. 4b
12. 9b
0.3b
0.2-5b
0.2-0.7d
0.3-5.0d
0.003-ld
0.7-4d
l-4d
0.4-2.0d
Reference 17.
Reference 18.
c Reference 16.
Reference 19.
A
This is an average of mixed oils containing both crankcase oil (high in lead)
with industrial oils (no lead content).
3-18
-------
TABLE 3-7. SUMMARY OF WASTE OIL CONCENTRATION AT THE
50TH, 75TH, AND 90TH PERCENTILE
Number
of
samples
Median
concentration
at 50th
percentile,
ppm
Concentration
at 75th b
percentile,
ppm
Concentration
at 90th
percentile,
ppm
Metals
Arsenic
Barium
Cadmium
Chromium
Leadd
Zinc
17
159
189
273
227
232
11
50
1.1
10
220
469
14
200
1.3
12
420
890
16
485
4.0
28
1,000
1,150
Chlorinated solvents
Di chl orodi f 1 uoromethane
Tri chl orotri f 1 uoroethane
1,1,1-trichloroethane
Trichloroethylene
Tetrachloroethylene
Total chlorine
78
26
123
126
100
62
20
<1
270
60
120
1,400
210
33
590
490
370
2,600
860
130
1,300
1,049
1,200
6,150
Other organics
Benzene
Toluene
Xylene
Benzo(a)anthracene
Benzo(a)pyrene
Naphthalene
PCB's
32
30
22
17
19
15
264
46
190
36
16
9
290
9
77
490
270
26
12
490
41
160
1,200
570
35
33
580
50
At the median, 50 percent of the analyzed waste oil samples had contaminant
concentrations below the given value.
Seventy-five percent of the analyzed waste oil samples had contaminant con-
centrations below the given value.
Ninety percent of the analyzed waste oil samples had contaminant concentra-
tions below the given value.
Values for lead were taken from 1979-1983 data only.
3-19
-------
Q - (A) (F) (0.75) (grams/103 ug)
where Q = lead emission rate in g/h
A = lead concentration in feed in ppm (assumed to be
1000 mg/liter)
F = feed rate in liters/h
0.75 = estimated percent of lead emitted to the air
g/10 ug = conversion factor
TABLE 3-8. CALCULATED EMITTED PORTION OF
TOTAL INPUT OF SELECTED METALS
Metal
Arsenic
Barium
Cadmium
Chromium
Lead
Zinc
Percent
emitted
70
58
40
31
64
75
Source: References 17 and 20.
3.4.2 Organics Emissions
Because organics such as polynuclear aromatics (PNA's) and
polycyclic organic matter (POM's) may be found in all heavy fos-
sil fuels, they do not represent a contaminant resulting from the
use of waste oil. Polychlorinated biphenyls (PCB's), on the
other hand, are not present in virgin fuel oil and therefore are
a contaminant in waste oil. Other organics that may be present
in waste oils are gasoline, glycol (from antifreeze) , pesticides,
and solvents. Nonhalide solvents, glycols, and gasoline are
normally readily combustible. Table 3-7 shows concentrations of
some organic contaminants found in waste oil.
Commercial/institutional boilers consume less fuel than the
residential, industrial, and utility boilers and therefore are
3-20
-------
not major contributors to the total national emissions. Never-
theless, their environmental impact may be significant because of
their high seasonal fuel consumption, their relative abundance
and proximity to population centers, their almost total lack of
pollution control equipment, and their release of emissions rela-
tively close to ground level. The same reasoning applies to
waste oil space heaters, which have a significant potential im-
pact because of their low stack height and proximity to popula-
tions.
In addition to the potential emissions resulting from con-
taminants in the waste oil, emissions could also arise from in-
complete combustion (carbon monoxide, hydrocarbons, carbonaceous
particles, and possibly other'chemical species, such as dioxin).
Very few emission test results are available, and it is difficult
to make a quantitative estimate of these emissions, which often
vary with operating conditions (i.e., temperature, residence
time, excess air, etc.).
The limited test data available on organic emissions from
burning waste oil indicate mostly nondetectable levels. For the
purpose of the air dispersion modeling in this study, emissions
of specific organic compounds were estimated based on the follow-
ing assumptions:
0 Organic compounds present in waste oil as potential
hazardous air emissions are 1,1,1-trichloroethane,
tetrachloroethylene, toluene, trichloroethylene, carbon
tetrachloride, and polychlorinated biphenyl compounds.
0 Each organic compound is present in waste oil at the
concentration shown in Table 3-7 as the 90th percentile
concentration. Some additional runs of one of the air
3-21
-------
dispersion models were based on 1 percent levels of
some organics in waste oil (10,000 ppm) .
0 Destruction removal efficiency for these specific
organic compounds is assumed to be 97 or 99 percent for
all sources modeled (space heaters, small boilers, and
medium-sized boilers).
Calculated emissions used for the modeling are presented in
Section 4.
3.4.3 Particle Size Distribution
Some available test data on,.particle siz-e, distribution of.
emissions from fuel oil and waste oil burning are presented in
Tables 3-9 and 3-10. Table 3-9 gives the distribution,on a
weight percent basis, for some major contaminants resulting from
burning 100 percent waste oil in a small boiler. As the table
shows, in most cases 90 percent or more of the particulate emis-
sions are 10 ym or smaller in size. This is the respirable
range, which is the most detrimental to human health. Table 3-10
gives the distribution, on a weight percent basis, for particulate
from the combustion of No. 6 fuel oil in a 10-MW (30 x 10 Btu/h)
industrial boiler. Here, 95 percent of the particulate was in
the respirable range of 10 um or smaller.
3.4.4 Atmospheric Half-Life
The organic waste oil contaminants have a finite life in the
atmosphere. After a time, they transform into simpler compounds
as a result of sunlight or because of chemicals in the atmo-
sphere. Table 3-11 shows the approximate half-lives of the
organic compounds of interest. Because carbon tetrachloride,
toluene, and trichloroethylene are simpler compounds, their
3-22
-------
TABLE 3-9. PARTICLE SIZE DISTRIBUTION OF SOME
MAJOR CONTAMINANTS IN EMISSIONS FROM WASTE OIL COMBUSTION6
(wt. percent of the contaminant falling within
the indicated particle size range)
Particle size
<1 micrometer
1-10 micrometers
>10 micrometers
Lead
76-79
16-21
2.7-4.4
Calcium
10-19
71-74
10-15
Phos-
phorus
23-42
49-66
8.9-10
Zinc
56-73
23-39
3.4-5.0
Iron
2.7-36
51-80
13-80
Barium
3.3-51
40-79
8.9-18
a Source: Reference 21.
TABLE 3-10. PARTICULATE SIZE DISTRIBUTION OF,
UNCONTROLLED EMISSIONS FROM AN OIL-FIRED BOILER6
Aerodynamic diameter
size range, pm
1-3
3-10
Weight %
20
1
74
5
Source: Reference 22.
3-23
-------
half-lives are 5 to 12 hours. Tetrachloroethylene and 1,1,2-
trichloroethane remain in the atmosphere between 20 and 40 hours
before half of their original concentration is transformed;
1,1,1-trichloroethane, which is very persistent in the atmo-
sphere, has a half-life in excess of 1700 hours (nearly 71 days),
TABLE 3-11. HALF-LIVES OF SELECTED WASTE OIL
ORGANIC EMISSIONS IN AIR
Organic substance
Carbon tetrachloride
Tetrachloroethylene
Toluene
Trichloroethane
Tri chloroethylene
Ha If-life,
hours
<8
20-40
5; 7
(1,1,1) »1700
(1,1,2) 20-40
5-12
3.4.5 Summary of Assumptions Used to Model Emissions
The assumed emissions of metals and organics used in the air
dispersion modeling are shown in Table 3-12. For boilers, it is
assumed that 75 percent of the metals in the waste oil are emit-
ted. Additional analyses were conducted based on 50 percent of
the metals being emitted. For space heaters, a range of emission
rates of 5 to 95 percent is used. For both boilers and space
heaters, destruction removal efficiencies of 97 and 99 percent
are used, which result in respective emissions of 3 percent and 1
percent of the organics in the waste oil. Sensitivity analyses
(described in Section 4) showed that particle deposition and
half-life had negligible effects on dispersion modeling results.
3-24
-------
Therefore, they are not included in the point source and urban
area modeling.
TABLE 3-12. ASSUMED EMISSIONS OF METALS AND ORGANICS
USED IN AIR DISPERSION MODELING
Source
Space heaters
Boilers (small and medium)
Type of
contaminant
Metals
Organics
Metals
Organics
Emissions,3 %
5, 10, 15, 25, 50,
75, 85, 95
3, lb
75, 50
3, lb
Expressed as percent of contaminant in waste oil that is emitted.
Corresponds to destruction removal efficiencies of 97 percent and
99 percent, respectively.
3-25
-------
REFERENCES FOR SECTION 3
1. Recon Systems, Inc., and ETA Engineering, Inc. Used Oil
Burned as a Fuel. Vols. 1 and 2. U.S. Environmental
Protection Agency, Washington, D.C. October 1980.
2. Development Planning and Research Associates, Inc. Risk/
Cost Analysis of Regulatory Options for the Waste Oil Man-
agement System. Vol. 1. U.S. Environmental Protection
Agency, Washington, D.C. January 1982.
3. Devitt, T., et al. Population and Characteristics of Indus-
trial/Commercial Boilers in the U.S. EPA-600/7-79-178a,
August 1979.
4. Development Planning and Research Associates. Selected
Characteristics of the Waste Oil Space Heater Industry.
July 1983.
5. Barbour, R. L., and W. M. Cooke. Chemical Analysis of Waste
Crankcase Oil Combustion Samples. U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina. May
1982.
6. Waite, D. A., et al. Waste Oil Combustion: An Environmen-
tal Case Study. APCA Paper No. 82-5.1. Presented at the
Annual Meeting, Air Pollution Control Association, New
Orleans, June 20-25, 1982.
7. Thome-Kozmiensky, K. J. Report on Article 1, No. 7. of the
First Regulation for the Implementation of the German Fed-
eral Emission Control Act of September 22, 1978. Berlin
Technical University. December 15, 1980.
8. Brookes, B. I. Emissions From the Combustion of Waste Oil
in a Space Heater. (Unpublished report) Glasgow, Strath-
clyde Regional Council, re. August 1981.
9. Interpoll, Inc. Results of the November 20 and 21, 1979,
particulate emission compliance tests of the Kutrieb Corpo-
ration Chetek 500 and F.A.I. 70 incinerators. (Unpublished
report) Chetek, Wisconsin, Kutrieb Corporation, December 17,
1979.
3-26
-------
10. Personal communication from Timothy A. Vanderver, Jr.,
Patton, Boggs, and Blow, to Docket Clerk (Docket No. 3001),
Office of Solid Waste, U.S. Environmental Protection Agency,
August 14, 1980. Comments on EPA's proposal to designate
waste oil as a hazardous waste and to regulate certain
reuses of waste oil.
11. Rourke, J. Emission Test Results - Lanair Waste Oil Heater.
Tests performed October 15, 1980.
12. Thome-Kozmiensky, K. J. Comparison of the Environmental
Effects of Different Waste Oil Recycling Methods. Berlin
Technical University. December 1980.
13, Dravo/Hastings, Inc.- P--50 Paraflow Waste Oil Heater Emis-
sion Test Report Project No. 78-152. Test date February
6-7, 1979.
14. Fred C. Hart Associates, Inc. Impact of Burning of Hazard-
ous Waste in Boilers. (Unpublished report) SCA Chemical
Services, Inc., Boston. August 1982.
15. Surprenant, N. F. Emissions Assessment of Conventional
Stationary Combustion Systems, Vol. 4. National Technical
Information Service, EPA-600/7-81-003b, January 1981.
16. Franklin Associates Limited and PEDCo Environmental, Inc.
Survey of the Waste Oil Industry and Waste Oil Composition.
Draft Report. April 1983.
17. American Petroleum Institute. Energy From Used Lubricating
Oils. Prepared by the Task Force on Utilization of Waste
Lubricating Oils. Publication No. 1588. October 1975.
18. Schmidt, P. F. Fuel Oil Manual. Industrial Press, Inc.
New York. 1969.
19. Electric Power Research Institute. Study of Electrostatic
Precipitators Installed on Oil-Fired Boilers. EPRI FP-792,
Project 413-1, Task 1.4, Volume 2, Final Report. June 1978.
20. Hall, R., et al. Test Plan and Quality Assurance Plan for
Emissions Characterization of Waste Oil Combustion. U.S.
Environmental Protection Agency, Cincinnati. June 1982.
21. U.S. Environmental Protection Agency. Report to Congress;
Waste Oil Study. Accession No. PB-257693. April 1974.
22. U.S. Environmental Protection Agency. Environmental Assess-
ment of Coal- and Oil-Firing in a Controlled Industrial
Boiler. Volume III. EPA-600/7-78-164c, August 1978.
3-27
-------
SECTION 4
AIR DISPERSION MODELING
4.1 INTRODUCTION
The objective of this part of the analysis was to quantify
the impacts of hazardous emissions from the combustion of waste
oil on ambient air quality. The dispersion models provided a
mechanism for assessing the air quality impact of a large number
of sources under various scenarios and for predicting the con-
centrations of waste oil contaminants in ambient air.
Of primary concern was the combustion of waste oil in small
commercial/industrial and residential boilers and space heaters
(<15 x 10 Btu/h). Secondary emphasis was the use of this oil
in medium-sized boilers (50 to 100 x 10 Btu/h). Thus three
source types were modeled: small boilers, medium-sized boilers,
and space heaters. Preliminary investigations revealed several
factors concerning the population of small boilers and space
heaters:
0 Emissions from these boilers are generally uncon-
trolled.
0 The distribution of these boilers is much the same as
population distribution.
0 Stack heights are generally less than 30 meters.
The most reasonable analysis entails modeling the impacts
of waste oil burning based on specific assumptions concerning
4-1
-------
the source types and the distribution of sources and emissions.
Because little is known about the population of small boilers
and space heaters that burn waste oil, individual and urbanwide
emissions were modeled to identify problem areas. Although the
air quality impact of an individual boiler may be minimal, when
combined with that of many other small boilers in the same
geographical area, the total impact may become significant.
Such impact depends on boiler density, emissions, and source
characteristics.
Two dispersion models were selected to estimate the air
quality impacts of contaminant emissions from waste oil burning.
The Industrial Source Complex Model (which is recommended by
2
the Guideline on Air Quality Models ) was chosen to model indi-
vidual sources and small groups of multiple sources at distances
up to 10 km. Ambient air concentrations due to individual boil-
ers were estimated only for those waste oil contaminants that
elicit a threshold response.
For a spatial scale of up to 50 km, the Hanna-Gifford
Model (a simplified, empirical model for areawide sources) was
used. Waste oil burning over an urban area was estimated for a
worst-case month and for the year. Total emissions were com-
bined with derived constants and climatologically averaged
meteorological conditions to determine monthly and annual pollu-
tant concentrations. The urban modeling included estimates for
both threshold and nonthreshold contaminants.
4-2
-------
These two models are easily adaptable, they are accepted by
the modeling community, and their uses and limitations are well
documented. The simplified and generic application of these
models is appropriate because of the hypothetical nature of the
air quality analyses performed.
Table 4-1 describes the spatial and temporal scales used
for the analysis. (Further details for each model are given in
Appendix B.) The specific applications of these models are as
follows:
0 The short-range modeling concentrated on producing
worst-case estimates resulting from individual or
multiple point sources, assuming each source burned
100 percent waste oil; concentrations of contaminants
eliciting a threshold response were of particular
concern.
0 The urban-wide modeling focused on the burning of all
available waste oil in a given area; urban-wide im-
pacts of contaminants eliciting a threshold response
were of considerable concern; impacts of nonthreshold
substances across the selected urban area were also
studied, but these offer limited utility for assessing
risk in other urban areas.
TABLE 4-1. MODEL SELECTION
Model
Industrial Source
Complex (ISC)
Hanna-Gifford
Spatial scale
0-5 km
Individual or multi-
ple-point sources
0-50 km
Urban-wide area
sources
Temporal scale
30-day
30-day, annual
Pollutants
Threshold3
Threshold and.
nonthreshold
Threshold pollutants are those eliciting a threshold limit response.
Nonthreshold pollutants are those causing excess cancers per a specified
population over a lifetime.
4-3
-------
4.2 POINT-SOURCE ANALYSIS WITH ISC MODEL
4.2.1 Overview
All modeling discussed in this section was done with an ISC
model. Individual or small group sources were modeled to estimate
the air quality impact provided by a threshold contaminants
source burning 100 percent waste oil. From a regulatory stand-
point, these estimates are of interest in the establisment of
limitations for emissions and-waste-oil feed stock contaminants." "
The ISC Model was used in its long-term mode. The meteoro-
logical input for this model is the STability ARray (STAR) data,
a joint-frequency distribution of windspeed, wind direction, and
atmospheric stability class. Meteorological data characteristic
of an urban location (John F. Kennedy International Airport) for
the years 1974 to 1978 were used. During this period, 3720
observations were tabulated for the month of January. Figure
4-1 presents a January wind rose combining all stability classes
for JFK Airport. As this figure indicates, winds from the west
and northwest were typical for this period.
January was selected for the analysis of pollutants with a
threshold response because of high oil consumption during this
coldest month. A monthly concentration was selected to give a
worst-case example for contaminants with threshold responses
when compared with lifetime or annual response limits. Short-
term (1- to 24-hour averaging periods) estimates of concentra-
tion were judged to be too high for comparison with long-term
environmental exposure. Such concentrations were not likely to
4-4
-------
WIND SPEED CLASSES <•/»)
03
c
o
(D
O
-s
01
ri-
^.
O
3
o>
•a
o
CD
PERCENT FREQUENCY OF OCCURRENCE
-------
be sustained long enough or to occur frequently enough to pro-
vide representative estimates of environmental exposure limits.
Other meteorological data used in this analysis included
monthly average mixing heights and temperatures. These were
4
estimated by using available mixing height climatologies and
Local Climatological Data summaries for New York. All meteoro-
logical inputs are presented in Appendix B.
Because the analysis is assumed to take place in an urban"' •
area, the option that considers the atmospheric effects of the
urban area was used. Essentially, the effect of this option on
the ISC calculations is to treat very stable atmospheric condi-
tions as neutral stability and thereby account for the effect of
heat losses and gains in an urban area on local atmospheric
stability.
After the modeling parameters had been specified, the ISC
Model was used to perform calculations for individual small
boilers and space heaters, medium-sized industrial boilers, and
multiple small boilers and space heaters.
4.2.2 Small Boilers and Space Heaters
Table 4-2 presents the 13 point sources analyzed in the ISC
Model, which represent typical small boilers and space heaters.
As the table shows, these sources include a range of boiler
sizes, feed rates, and stack heights. The invariability of
other stack parameters was justified by the fact that the plume
rise was small compared with physical stack height. Section 3
describes the sources reviewed in determining the representative
sources in Table 4-2.
4-6
-------
S-fr
WIND SPEED CLASSES (•/«)
C
-s
3
Q.
-5
O
fD
-s
O)
c-h
o'
Q)
3>
^<
-J
O
IO
00
PERCENT FREQUENCY OF OCCURRENCE
-------
be sustained long enough or to occur frequently enough to pro-
vide representative estimates of environmental exposure limits.
Other meteorological data used in this analysis included
monthly average mixing heights and temperatures. These were
estimated by using available mixing height climatologies4 and
Local Climatological Data summaries for New York. All meteoro-
logical inputs are presented in Appendix B.
Because the analysis is assumed to take place in an urban
area, the option that considers the atmospheric effects of the
urban area was used. Essentially, the effect of this option on
the ISC calculations is to treat very stable atmospheric condi-
tions as neutral stability and thereby account for the effect of
heat losses and gains in an urban area on local atmospheric
stability.
After the modeling parameters had been specified, the ISC
Model was used to perform calculations for individual small
boilers and space heaters, medium-sized industrial boilers, and
multiple small boilers and space heaters.
4.2.2 Small Boilers and Space Heaters
Table 4-2 presents the 13 point sources analyzed in the ISC
Model, which represent typical small boilers and space heaters.
As the table shows, these sources include a range of boiler
sizes, feed rates, and stack heights. The invariability of
other stack parameters was justified by the fact that the plume
rise was small compared with physical stack height. Section 3
describes the sources reviewed in determining the representative
sources in Table 4-2.
4-6
-------
TABLE 4-2. SOURCE CHARACTERISTICS FOR SINGLE AND MULTIPLE SOURCE ANALYSIS:
SMALL BOILERS AND SPACE HEATERS
Source
identification
number
1
2
3
4
5
6
7
8
9
10
11
12
13
Source type
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Space heater
Space heater
Capacity,
10° Btu/h
0.7
0.7
2.1
2.1
2.1
5.0
5.0
5.0
9.3
9.3
9.3
0.1
0.1
Feed rate,
liters/h
19
19
57
57
57
132
132
132
246
246
246
4
4
Feed
rate,
gal/h
5
5
15
15
15
35
35
35
65
65
65
1
1
Stack
height,
m
5
10
10
15
20
10
15
20
10
15
20
2
5
Stack
diameter,
m
0.5
0.5
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.2
0.2
Stack gas
temperature,
K
420
420
420
420
420
470
470
470
520
520
520
625
625
Stack gas
velocity,
m/s
0.5
0.5
1.2
1.2
1.2
2.0
2.0
2.0
4.5
4.5
4.5
1.8
1.8
•u
I
a For identification in the ISC Model.
-------
Emissions for the small boilers and space heaters were
based on several factors:
0 Only contaminants with a threshold reponse were con-
sidered in the single source analysis.
0 Contaminant concentrations in the waste oil were
assumed to be 90th percentile contaminant levels.
0 January boiler usage was assumed to be at 50 percent
of capacity for small boilers and 100 percent of
capacity for space heaters.
0 Feed rates were assumed to be 4, 19, 57, 132, and 246
liters/hour at 100 percent boiler capacity. In all
modeling, however, emission calculations are based on
50 percent boiler capacity, so actual feed rates are
half of these values.
0 Assumed emission rates of heavy metals were as follows;
Small boilers: 75 percent of input (50 percent
of input for barium and lead)
Space heaters, atomizing: 5 to 95 percent of
input.
0 Organic emission rates
Small boilers and space heaters: 3 percent of
input (i.e., 97% destruction removal efficiency)
Table 4-3 lists the emission factors used throughout the small
boiler and space heater modeling analyses.
The results of modeling waste oil contaminants from individ-
ual small boilers and space heaters (described in Table 4-2) are
presented in Table 4-4. The concentrations shown represent the
maximum estimated for each single source case. These estimated
concentrations are compared with the Environmental Exposure
Limit (EEL) or National Ambient Air Quality Standard (NAAQS) for
each contaminant. (The EEL's are described more fully in Sec-
tion 5.) The comparison with the EEL's indicates that barium,
4-8
-------
TABLE 4-3. CONTAMINANT EMISSION RATES USED IN ISC MODELING ANALYSIS:
SMALL BOILERS AND SPACE HEATERS
Contain nint
with threshold
level, EEL*
Barium
Cadmium
Chromium
Lead
Zinc
Toluene"
1,1.1-Trichloro-
ethane
Chlorinated
orgamcs
Naphthalene"
Concent ration
in waste oil,
ppm
485"
4b
28b
1000"
1150b
1200
1300
6150
580
Boiler
capicity,
106 Btu/h
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
0.1
0.7
2.1
5.0
9.3
Source
waste oil
bum rates,
I1t«rs/h
4
19
57
132
246
4
19
57
132
246
4
19
57
132
246
4
19
57
132
246
4
19
57
132
246
4
19
57
132
246
4
19
57
132
246
«
19
57
132
246
4
19
57
132
246
Contaminant
Input, ug/s
512
2.560
7.680
17,900
33.100
4.21
21.1
63.3
147.
273.
30.0
148.
443
1.031
1.910
1.055
5.280
15.800
36.600
68,400
1.210
6,070
18,200
42,200
78.600
1.270
6.340
19.000
43,900
82,000
1.370
6,860
20.590
47,570
88.530
6,470
32.335
97 ,000
226.320
420.250
610
3,050
9,150
21,350
39,650
Emissions at
1001 capacity.
ug/s
384.
1.920.
5.760.
13.400.
24,800.
3.16
15.8
47.5
110.
205.
22.5
111.
332.
773.
1.430.
791.
3,950.
11.900.
27.500.
51.300.
910.
4,550.
13,700.
31.700.
59.000.
38.
190.
570.
1,320.
2.460.
41.1
206.
618.
1.427.
2.856.
6,470.
32,335.
97,000.
226,320.
420.250.
18.3
92.
274.
640.
1,190.
Emissions in
January at
501 capacity,
u9/s
c
960.
2,880.
6.700.
12.400.
C
7.9
23.8
55.0
103.
c
55.5
166.
387.
715.
c
1,980.
5.930.
13,750.
25,600.
c
2.275.
6.850.
15,850.
29.500.
c
95.
285.
666.
1.235.
c
103.
309.
721.
1.339.
c
16,170.
48.500.
113,160.
210,125.
C
45.9
137.
321.
595.
See Environmental Exposure Level discussion.
Weight per volume; netal emissions are assumed to equal 75i of input emissions.
Space heaters only modeled at 1001 capacity level for the January analysis.
Destruction efficiency is assumed to be equal to 97S.
" Weight per volume; HC1 emissions ar« assumed equal to 1001 of Input chlorinated organic levels.
4-9
-------
TABLE 4-4. AMBIENT AIR IMPACT FROM COMBUSTION OF WASTE OIL CONTAINING CONTAMINANTS
WITH A THRESHOLD RESPONSE IN INDIVIDUAL BOILERS AND SPACE HEATERS SMALLER
THAN 15 x 106 Btu/h
Substance
Barium
Cadmium
Chromium
Lead
Zinc
Toluene
Trlchloroethane
Hydrogen chloride
Napthalene'
Man* mum ]0-day concentrations, uq/m'
Source number'
1
0.061
0.001
0.004
0.126
0.145
0.006
0.006
1.03
0.003
2
0.017
<0.001
0.001
0.034
0.040
0.001
0.001
0.28
< 0.001
3
0.041
<0.001
0.002
0.085
0.098
0.005
0.005
0.69
0.002
4
0.018
-------
lead, and chlorinated organics are the major contaminants of
concern; they contribute 23, 14, and 14 percent, respectively,
of the EEL's at the worst-case receptors. At a rate of 50
percent of input, emissions levels of barium and lead drop to 15
and 9 percent of their respective EEL's. The impact of the
other contaminants on the ambient air quality was no greater
than 1 percent of their respective EEL's. Maximum concentra-
tions were due to sources with the high emission rates (feed
rates of 246 liters/h) and sources with 10-meter stack heights
(Source No. 9). For each set of boilers or space heaters, the
maximum concentration occurred with the lowest stack height.
Figures 4-2 and 4-3 present the isopleths of lead concen-
trations for two scenarios to illustrate how the maximum concen-
tration and the overall concentration distribution depend on
stack height and final plume height. Figure 4-2 shows that the
maximum impact of a boiler with a feed rate of 19 liters/h and a
5-meter stack is within 200 meters east of the stack, primarily
downwind of the predominant wind directions. Figure 4-3 in-
dicates that with a taller stack (10 m) and slightly higher
plume rise the impact is farther downwind (200 to 400 meters
east of the source). Although the maximum concentration occurs
farther downwind as stack height increases, the concentration
does not necessarily decrease with distance because the larger
boilers with the taller stacks also generate more emissions. In
this and all subsequent tables, reported concentrations represent
the maximum 30-day averages at the highest receptor for each
case.
4-11
-------
30-day maximum lead ~
concentrations, ug/m
feed rate » 19 i/h
stack height = 5 m
0.06 0.03 0.015
0.004
0.004
0.004
0 100 200 300 400 500
i I I I I I
meters
Figure 4-2. Isopleths of lead concentration over a 30-day averaging
period for a 19 d/h (5 gal/h) small boiler.
4-12
-------
0.015
30-day maximum lead ~
concentrations, ug/m
feed rate « 246J./h
stack height = 10 m
0.015
0.038
0.038
0 100 200 300 400 500
I I II I I
meters
Figure 4-3. Isopleths of lead concentration over a 30-day averaging period
for a 246 £/h (65 gal/h) small boiler.
4-13
-------
For determination of the levels of waste oil contaminants
that would produce no significant impact at any individual
source (here defined as a waste oil concentration resulting in
ambient concentrations less than some specified percentage of
the EEL), concentrations were estimated at variable waste oil
levels of barium, lead, and chlorinated organics. Tables 4-5,
4-6, and 4-7 present these estimates for barium, lead, and
hydrogen chloride, respectively. Table 4-5 shows that waste oil
containing 50 ppm of barium produces ambient air concentrations
at or below 3 percent of the EEL; at 10 ppm, concentrations fall
below 1 percent. Table 4-6 shows that waste oil containing 100
ppm of lead produces ambient air concentrations at or below 1
percent of the NAAQS. Table 4-7 shows that waste oil with 3000
ppm of chlorinated organics results in ambient air quality con-
centrations of hydrogen chloride at or below 1 percent of the
EEL.
As described in Section 3, several types of space heaters
are commonly in current use. Because the firing mechanisms of
space heaters may differ from those on small boilers, the impact
of total metal emissions assumed in Tables 4-4 through 4-7 could
differ somewhat. In these tables, the emission rate for metal
contaminants was assumed to be 75 percent of the input contam-
inant levels (50 percent for selected barium and lead cases).
For some types of space heaters this estimate is low, but for
others it is high. In the subsequent analysis (represented by
Tables 4-8 through 4-10), emissions of metals due to the combus-
tion of waste oil in space heaters varied from 5 to 95 percent.
4-14
-------
TABLE 4-5. AMBIENT AIR IMPACT FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS INPUT
CONCENTRATIONS OF BARIUM IN INDIVIDUAL BOILERS AND SPACE HEATERS SMALLER THAN 15 x 106 Btu/h
I
M
Ul
Source
No.«.
1
2
3
4
5
6
7
B
9
10
11
12
13
Source
type
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Space
heater
Space
heater
Feed
rate.
t/h6
19
19
57
57
57
132
132
132
246
246
246
4
4
Stack
height.
M
5
10
10
15
20
10
15
20
10
15
20
2
5
Barium concentrations In waste oil feed stock, ppm
485C
X.d M9/»'
0.061
0.017
0.041
0.018
0.009
0.080
0.036
0.020
0.098
0.050
0.029
0.043
0.024
I EELe
14
4
10
4
2
19
8
5
23f
12
7
10
6
200
X. M9/W*
0.025
0.007
0.017
0.007
0.004
0.033
0.015
0.008
0.040
0.021
0.012
0.018
0.010
S EEL
6
2
4
2
<1
8
3
2
10
5
3
4
2
50
X. M9/»J
0.006
0.002
0.004
0.002
0.001
0.008
0.004
0.002
0.010
0.005
0.003
0.005
0.003
% EEL
2
<1
1
<1
<1
2
1
«1
3
1
<1
1
<1
25
KI wg/»J
0.002
0.001
0.002
0.001
<0.001
0.004
0.002
0.001
0.005
0.003
0.002
0.003
0.002
I EEL
1
<1
<1
<1
<1
1
<1
<1
2
<1
<1
<1
<1
10
X. wg/n*
0.001
<0.001
<0.001
<0.001
<0.001
0.002
<0.001
<0.001
0.002
0.001
<0.001
0.001
<0.001
I EEL
<1
<1
<1
<
<
<
<
<
<
<
<
<1
<1
' For Identification In the ISC Model.
b 19 llters/h • 5 gal/h; 57 llters/h • 15 gal/h; 132 llters/h - 35 gal/h; 246 llters/h « 65 gal/h;
4 11ters/h • 1 gal/h.
c 90th percentlle of barium concentrations in waste oil.
X = ambient concentration.
6 Environmental Exposure Limit = 0.43 ug/m^ for barium.
At 50 percent emissions, the % EEL is 15% at 485 ppm, 7% at 200 ppm, 21 at 50 ppm, 1% at 25 ppm.
-------
TABLE 4-6. AMBIENT AIR IMPACT FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS INPUT
CONCENTRATIONS OF LEAD IN INDIVIDUAL BOILERS AND SPACE HEATERS SMALLER THAN 15 x 106 Btu/h
Source
No. «
1
2
3
4
5
6
7
8
9
10
11
12
13
Source
type
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Space
heater
Space
heater
F»»H
rcco
rate,
l/hb
19
19
57
57
57
132
132
132
246
246
246
4
4
Stack
height.
m
5
10
10
15
20
10
15
20
10
15
20
2
5
Lead concentrations In waste oil, ppm
3000
X.d M9/m>
0.378
0.103
0.254
0.109
0.058
0.491
0.223
0.121
0.608
0.308
0.179
0.267
0.152
NAAQS"
25
7
17
7
4
33
15
8
41f
21
12
18
10
1000°
x. ug/«*
0.126
0.038
0.085
0.036
0.019
0.164
0.074
0.040
0.203
0.103
0.050
0.089
0.051
I
NAAQS
8
2
6
2
1
11
5
3
14
7
4
6
3
500
x. ijg/"J
0.063
0.019
0.042
0.018
0.010
0.082
0.037
0.020
0.101
0.051
0.029
0.044
0.026
t
NAAQS
4
1
3
1
<1
6
3
1
7
4
2
3
2
100
xi wg/ni*
0.013
0.004
0.009
0.004
0.002
0.016
0.007
0.004
0.020
0.010
0.006
0.009
0.005
t
NAAQS
1
<1
1
<1
<1
1
1
<1
1
1
<1
1
<1
25
x. wg/»J
0.002
<0.001
0.002
0.001
0.001
0.004
0.002
0.001
0.005
0.002
0.001
0.002
0.001
I
NAAQS
<1
<1
<1
<1
<1
<1
«1
<1
<1
<1
<1
<1
<1
" For Identification In the ISC Model.
b 19 I1ters/h * 5 gal/h; 57 llters/h = 15 gal/h; 132 Hters/h = 35 gal/h; 246 llters/h * 65 gal/h; 4 Hters/h = 1 gal/h.
c Most representative In terms of recent measurements and estimates of lead In waste oil; 90th percentlle.
x= ambient concentration.
e NAAQS =1.5 ug/m3 for lead.
At 50 percent emissions, the t of NAAQS Is 27T at 3000 ppm, 9* at 1000 ppm, 5X at 500 ppm, It at 100 ppm, <1J at 25 ppm.
-------
TABLE 4-7. AMBIENT AIR IMPACT OF HC1 FROM THE COMBUSTION OF WASTE OIL CONTAINING
VARIOUS INPUT CONCENTRATIONS OF CHLORINATED ORGANICS IN INDIVIDUAL BOILERS
AND SPACE HEATERS SMALLER THAN 15 x 106 Btu/h
Source
No. a
1
2
3
4
5
6
7
8
9
10
11
12
13
Source
type
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Boiler
Space
heater
Space
heater
Feed
rate,
l/hb
19
19
57
57
57
132
132
132
246
246
246
4
4
Stack
height,
m
5
10
10
15
20
10
15
20
10
15
20
2
2
Chlorinated organic concentrations (ppm) in waste oil
30 ,000
X,d M9/m3
5.03
1.39
3.38
1.45
0.76
6.58
3.00
1.60
8.10
4.10
2.40
3.57
2.01
I EELe
8
2
6
2
1
11
5
3
14
7
4
6
3
10.000
x. wg/m1
1.68
0.46
1.13
0.48
0.25
2.19
1.00
0.53
2.70
1.37
0.80
1.19
0.67
1 EEL
3
1
2
1
<1
4
2
1
4
2
1
2
1
b , 1 bOc
X. wg/m1
1.03
0.29
0.70
0.30
0.15
1.34
0.61
0.33
1.66
0.84
0.49
0.73
0.41
I EEL
2
<1
1
-------
TABLE 4-8. AMBIENT AIR IMPACTS FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS
INPUT RATES OF METAL CONTAMINANTS IN SPACE HEATERS
Maximum 30-day average concentrations, wg/m1
CO
Pollutant
Barium
Cadmium
Chromium
Lead
Zinc
Percentage of waste oil contaminants emitted
51
Stack ht., m*
2
0.003
<0.001
<0.003
0.006
0.007
5
0.002
<0.001
<0.001
0.003
0.004
1SX
Stack ht., ma
2
0.009
<0.001
<0.004
0.018
0.021
5
0.005
<0.001
<0.001
0.010
0.012
25t
Stack ht.. ma
2
0.014
<0.001
0.001
0.030
0.034
5
0.008
<0.001
<0.001
0.017
0.019
35%
Stack ht.. ma
2
0.020
<0.001
0.002
0.042
0.048
5
0.011
<0.001
<0.001
0.024
0.027
45X
Stack ht. . m'
2
0.026
<0.001
0.002
0.054
0.062
5
0.014
<0.001
<0.001
0.030
0.035
Maximum 30-day average concentrations, ug/m1
Pollutant
Barium
Cadmium
Chromium
Lead
Z1nc
Percentage of waste oil contaminants emitted
55%
Stack ht.. ma
2
0.031
<0.001
0.002
0.065
0.075
5
0.017
<0.001
<0.001
0.037
0.042
65%
Stack ht., ma
2
0.037
<0.001
0.003
0.077
0.089
5
0.021
<0.001
<0.001
0.044
0.050
751
Stack ht., m*
2
0.043
<0.001
0.003
0.089
0.103
5
0.024
<0.001
0.001
0.051
0.058
852
Stack ht., na
2
0.049
<0.001
0.003
0.101
0.117
5
0.027
<0.001
0.001
0.057
0.066
95%
Stack ht. , •*
2
0.054
<0.001
0.004
0.113
0.130
5
0.030
<0.001
0.001
0.064
0.073
a These sources have stack characteristics corresponding to Sources 12 and 13 In previous tables.
-------
TABLE 4-9. AMBIENT AIR IMPACT FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS
INPUT CONCENTRATIONS OF BARIUM IN INDIVIDUAL SPACE HEATERS
*>.
i
»-•
VD
Percent of
metals
emitted
5
5
IS
15
75
75
95
95
Feed
rate,
l/h
4
4
4
4
4
4
4
4
Stack
height.
m
2
5
2
5
2
5
2
5
Barium concentrations in waste oil feed stock, ppm
485*
X. M9/»'
0.003
0.002
0.009
0.005
0.043
0.024
0.054
0.030
I EEL
<1
<1
2
1
10
6
13C
7
200
X. P9/«J
0.001
<0.001
0.004
0.002
0.018
0.010
0.022
0.012
I EEL
«1
<1
<1
<1
4
2
5
3
100
x. pg/"1
<0.001
<0.001
0.002
0.001
0.009
0.005
0.011
0.006
I EEL
<1
<1
<1
<1
2
1
3
1
25
X. M9/«'
<0.001
<0.001
<0.001
<0.001
0.002
0.001
0.003
0.002
X EEL
<
<
<
<
<
<
<1
«1
10
K! Mg/«J
<0.001
<0.001
<0.001
<0.001
<0.001
<0.001
0.001
<0.001
X EEL
<1
<1
<1
<1
<1
<1
<1
<1
a Host representative in terms of recent measurements and estimates of barium in waste oil; 90th percentile.
EEL « Environmental Exposure Limit.
c At 50 percent emissions, the I EEL at each contaminant level is: 485 ppm - 9X; 200 pom - 31; 100 ppn - 2X 25 pp* -
-------
TABLE 4-10. AMBIENT AIR IMPACT FROM COMBUSTION OF WASTE OIL CONTAINING VARIOUS INPUT
CONCENTRATIONS OF LEAD IN INDIVIDUAL SPACE HEATERS
Percent of
metals
emitted
5
5
15
15
75
75
95
95
Feed
rate,
gal/h
4
4
4
4
4
4
4
4
Stack
height,
m
2
5
2
5
2
5
2
5
Lead concentrations (ppm) in waste oil feed stock
3000
U
x, M9/m3
0.018
0.010
0.054
0.030
0.267
0.152
0.339
0.192
t NAAQSC
1
1
•4
2
18
10
23<<
13
1000a
x, ug/wi'
0.006
0.003
0.018
0.010
0.089
0.051
0.113
0.064
I NAAQS
<1
<1
1
1
6
3
8
4
500
x, ug/m3
0.003
0.002
0.009
0.005
0.044
0.026
0.057
0.032
1 NAAQS
<1
<1
<1
<1
3
2
4
2
100
X, u9/mj
<0.001
0.001
0.002
0.001
0.009
0.005
0.011
0.006
t NAAQS
<1
<1
<1
<1
1
<1
1
<1
25
x, ug/w3
<0.001
<0.001
<0.001
<0.001
0.002
0.001
0.003
0.002
X NAAQS
<1
<1
<1
<1
<1
<1
<1
<1
•U
I
N>
O
a Most representative in terms of recent measurements and estimates of lead in waste oil; 90th percentile.
X = ambient concentration of barium.
c NAAQS = 1.5 gg/m3 for lead.
d At 50 percent emissions, the % of NAAQS is 15T at 3000 ppm, 5", at 1000 ppm. 3* at 500 ppm,
-------
Emissions of all threshold organic contaminants were assumed to
be at the same levels as in the previous analysis, and thus were
not remodeled.
Table 4-8 presents the dispersion modeling results for
space heaters at varying emission levels. Two stack heights
were modeled at each emission level. All other source charac-
teristics were the same as those presented in Table 4-2. Model-
ing indicated that the maximum concentrations are found in
emissions at the 95 percent level from the 2-meter stack height.
Barium and lead had the greatest impacts on ambient air.
As a test of the sensitivity of air quality concentrations,
alternative contaminant levels were estimated for barium and
lead at selected emission rates (5, 15, 75, and 95 percent of
input). Tables 4-9 and 4-10 present these results. As shown in
Table 4-9, at levels of 100 ppm in waste oil feedstock, levels
of barium in the ambient air concentrations fell to 3 percent or
less of the EEL. Table 4-10 shows that the ambient concentra-
tions of lead due to waste oil combustion in space heaters fell
below 1 percent of the NAAQS of 1.5 pg/m3 at a feedstock input
of about 100 ppm. The 3000-ppm level was included to give an
indication of the impact at high levels of lead in waste oil.
4.2.3 Multiple Small Sources
Except for barium and lead, ambient air impacts from the
burning of waste oil in individual small boilers and space heat-
ers were minimal compared with the EEL's. Because these sources
are proximate to one another and have an additive ambient air
4-21
-------
impact, an entire urban area was modeled to assess the impact of
multiple sources. The results of this modeling are discussed
later (Section 4.3).
A limited group of multiple point sources was also modeled
to ascertain the combined-source impact of the individual short-
range maximum impacts of multiple point sources. Three groups
of multiple point sources were selected from the 13 small boiler
and space heater sources listed in Table 4-2. Each group con-
sisted of four identical sources spaced at equal distances of 50
m apart, with one source located at each corner of a square. A
receptor grid (similar to that in the single source analysis)
was used, along with all other inputs specified in the single
source analysis. The three source groups modeled were as fol-
lows:
Group 1: Four boilers with feed rates of 19 liters/h and
5-m stacks
Group 2: Four boilers with feed rates of 246 liters/h and
10-m stacks
Group 3: Four space heaters with feed rates of 4 liters/h
and 2-m stacks
The metal contaminants were modeled at the 75 percent emission
level for Groups 1 and 2 and at the 95 percent level for
Group 3. The organics were modeled at the 3 percent emission
level (97 percent destruction removal efficiency), and HC1 was
modeled at 100 percent emissions.
The results of this limited analysis showed that the com-
bined-source impacts due to a group of small sources were great-
er than those from an individual source. This result was not
unexpected because of the overlap of emission plumes in the
4-22
-------
atmosphere and the subsequent combined downwind impacts. Table
4-11 shows the concentration estimates for the three groups of
sources modeled. In each case, the combined impact of the four
boilers with the feed rate of 246 liters/h was greatest. For
barium and lead, the combined impacts yielded concentration
estimates that represented a significant percentage of the
applicable EEL's. Maximum hydrogen chloride levels are 10
percent of the EEL. For all other contaminants with a threshold
response, concentrations were less than 2 percent of the EEL.
Tables 4-12, 4-13, and 4-14 present estimates of ambient
concentrations for each group of sources at various waste oil
contaminant levels. At a barium level of 25 ppm, the combined
source impact falls to 4 percent or less of the EEL. Similarly,
at a lead level of 100 ppm, combined-source ambient concentra-
tions are 5 percent or less of the NAAQS. At HC1 levels of 1000
ppm, the combined-source concentration is 1 percent or less of
the applicable EEL.
As would be expected, combined-source impacts can be mini-
mized for any contaminant by reducing the level of that contami-
nant in the waste oil feed stock. The limited multiple-source
modeling performed here, however, shows the impact of only four
sources. Additional modeling of lead emissions from waste oil
combustion was performed for 16 boilers located on a 50-m grid
spacing over a 150-m by 150-m square. Estimates were made for
16 small boilers with feed rates of 19 liters/h and 16 small
boilers with feed rates of 246 liters/h. This source configura-
tion and the resulting estimates of ambient lead concentration
4-23
-------
TABLE 4-11. AMBIENT AIR IMPACTS FROM MULTIPLE-SOURCE COMBUSTION OF
WASTE OIL CONTAINING CONTAMINANTS WITH A THRESHOLD RESPONSE
Pollutant
Barium
Cadmium
Chromium
Lead
Zinc
Toluene
Trichloroethane
Hydrogen chlo-
ride
Naphthalene
Maximum 30-day
concentrations, ug/m3
Source group number
Group la
0.1484
0.0012
0.0086
0.3059
0.3517
0.0148
0.0159
2.50
0.007
Group 2b
0.3424
0.0028
0.0197
0.7078
0.8145
0.0341
0.0370
5.80
0.017
Group 3C
0.0958
0.0008
0.0056
0.1975
0.2271
0.0095
0.0102
1.61
0.004
EEL
concentration,
ug/m3
0.43
0.34
4.32
1.5
43.2
3,240.
16,420.
59.7
432.
Maximum
percentage
of EEL
80d
<1
<1
47e
2
<1
<1
10
<1
Group 1 consists of four boilers with a feed rate of 5 gal/h and a 5-m
stack height.
Group 2 consists of four boilers with a feed rate of 65 gal/h and a 10-m
stack height.
Group 3 consists of four space heaters with a feed rate of 1 gal/h and a
2-m stack height.
At 50 percent emissions, Group 2 yields 0.2283 ug/m and 53% of the EEL
for barium.
e 3
At 50 percent emissions, Group 2 yields 0.4119 ug/m and 31% of the EEL
for lead.
4-24
-------
I
K)
Ul
TABLE 4-12. AMBIENT AIR IMPACTS FROM MULTIPLE SOURCE COMBUSTION OF WASTE OIL
CONTAINING BARIUM AT VARIOUS INPUT CONCENTRATIONS
Source
group
No.
la
2b
3C
Barium concentrations in waste oil feed stock, ppm
485
X,d ug/m3
0.148
0.342
0.076
% EEL6
35
80f
18
200
X, pg/m3
0.061
0.141
0.031
% EEL
14
33
7
50
x. yg/m3
0.015
0.035
0.008
% EEL
4
8
2
25
X, pg/m3
0.008
0.018
0.004
% EEL
2
4
1
10
x. pg/m3
0.003
0.007
0.002
% EEL
<1
2
<1
Four boilers with a feed rate of 5 gal/h and 5-m stack heights.
Four boilers with a feed rate of 65 gal/h and 10-m stack heights.
c Four space heaters with a feed rate of 1 gal/h and 2-m stack heights.
Ambient air concentration of barium.
e Environmental Exposure Limit = 0.43 yg/m^ for barium.
At 50 percent emissions, the % EEL is 53% at 485 ppm, 22% at 200 ppm, 5% at 50 ppm, 3% at 25 ppm,
1% at 10 ppm.
-------
TABLE 4-13. AMBIENT AIR IMPACTS FROM THE COMBUSTION OF WASTE OIL CONTAINING LEAD AT
VARIOUS INPUT CONCENTRATIONS IN FOUR CLUSTERED SOURCES
Source
group
No.
lc
2d,e
3f
Lead concentrations'in waste oil feed stock, ppm
3000
X,a ug/m3
0.918
2.123
0.468
% NAAQSb
61
142
31
1000
X, ug/m3
0.306
0.708
0.156
% NAAQS
20
47
10
500
X, ug/m3
0.153
0.359
0.078
% NAAQS
10
24
5
100
X, ug/m3
0.031
0.072
0.016
% NAAQS
2
5
1
25
X, ug/m3
0.008
0.018
0.004
% NAAQS
<1
1
<1
.u
to
Ambient air concentrations of lead.
bNAAQS =1.5 pg/m3 for lead
cFour boilers with feed rates of 5 gal/h and 5-m stack heights.
dFour boilers with feed rates of 65 gal/h and 10m stack heights.
Assuming 50 percent of the lead is emitted, waste oil with 3000 ppm lead results in ambient air concen-
trations that are 95% of the NAAQS; for 1000 ppm, the result is 31% of the standard; for 500 ppm, the
results 16% of the standard; for 100 ppm, the result is 3% of the standard; and for 25 ppm, the ambient
air concentration is less than 1% of the NAAQS.
^Four space heaters with feed rates of 1 gal/h and 2-m stack heights.
-------
TABLE 4-14. AMBIENT AIR IMPACTS OF HC1 FROM MULTIPLE-SOURCE COMBUSTION OF WASTE OIL CONTAINING
CHLORINATED ORGANICS AT VARIOUS INPUT CONCENTRATIONS
Source
group
No.
ld
2e
3f
30,000
h
x, yg/m3
12.19
28.30
7.87
% EELC
20
47
13
10,000
X, yg/m3
4.06
9.43
2.66
% EEL
7
16
4
6,150a
x, yg/m3
2.44
5.66
1.55
% EEL
4
10
3
3,000
x. yg/m3
1.22
2.83
0.78
* EEL
2
5
1
1,000
x, wg/m3
0.41
0.94
0.26
% EEL
1
1
-------
are representative of several boilers located near one another,
such as in a large apartment complex or a small industrial area.
Ambient air impacts of lead are presented in Table 4-15.
At the 25-ppm input level, lead concentrations from the 19-
liter /h sources are at or below 1 percent of the NAAQS. At the
5-ppm input level, lead concentrations in emissions from the
246-liter/h boilers are estimated to be below 1 percent of the
NAAQS.
4.2.4 Medium-Size Boilers
Some waste oil burning also occurs in boilers with capaci-
ties of 50 to 100 x 10 Btu/h (in larger commercial/institution-
al or residential applications). For determination of the ambi-
ent air quality impact of burning 100 percent waste oil in these
medium-size boilers, the meteorological data, model options, and
the modeling receptor grid were the same as those for the small
boiler analysis. Typical stack characteristics for these boil-
ers were obtained by reviewing boiler population studies and
selecting representative values.
Table 4-16 presents the source characteristics chosen for
analysis and the appropriate stack parameters. Concentrations
of the contaminants eliciting a threshold response were based on
the 90th percentile levels in waste oil. Emissions resulting
from the burning of waste oil in these medium-size boilers are
presented in Table 4-17. Because these boilers were assumed to
be used primarily for heating, a load factor of 50 percent for
January modeling applications was also assumed.
4-28
-------
TABLE 4-15. AMBIENT AIR IMPACTS FROM SIXTEEN CLUSTERED SOURCES
BURNING WASTE OIL CONTAINING LEAD AT VARIOUS INPUT CONCENTRATIONS
Concentration
of lead in
waste oil ,
ppm
1000
500
250
100
75
50
25
10
5
Source qroup number
la
x,c yg/m3
0.755d
0.378
0.189
0.076
0.057
0.057
0.019
0.008
0.004
Percent NAAQS
50
25
13
5
4
3
1
<1
<1
2b
X, yg/m3
2.392e
1.196
0.598
0.239
0.179
0.120
0.060
0.024
0.012
Percent NAAQS
159
80
40
15
12
8
4
2
<1
Sixteen 5-gal/h boilers with 5-m stack heights on a 50-m grid spacing.
Sixteen 65-gal/h boilers with 10-m stack heights on a 50-m grid spacing.
c Ambient concentrations of lead averaged over a 30-day period.
Emission rate from a 5-gal/h boiler is 1,980 ug/s at 1000 ppm lead in
waste oil.
Emission rate from a 65-gal/h boiler is 25,680 ug/s at 1000 ppm lead in
waste oil.
4-29
-------
TABLE 4-16. SOURCE CHARACTERISTICS FOR SINGLE SOURCE ANALYSIS:
MEDIUM-SIZE BOILERS
Source
identifi-
cation
number
14
15
16
17
18
19
Capacity,
106 Btu/h
50
50
50
100
100
100
Feed rate,3
liters/h
1330
1330
1330
2660
2660
2660
Stack
height,
m
10
15
20
20
25
30
Stack
diameter,
m
0.7
0.7
0.7
1.0
1.0
1.0
Stack gas
temperature,
K
450
450
450
450
450
450
Stack gas
velocity,
m/s
18.3
18.3
18.3
18.3
18.3
18.3
1330 liters/h = 350 gal/h; 2660 liters/h = 700 gal/h.
4-30
-------
TABLE 4-17. CONTAMINANT EMISSIONS USED IN SINGLE SOURCE ISC MODELING ANALYSIS:
MEDIUM-SIZE BOILERS
Threshold
contaminant
Barium
Cadmium
Chromium
Lead
Zinc
Toluene
1,1,1-Trichloroethane
Chlorinated organics d
Naphthalene
Concentration
in waste oil ,
ppm
485 c
4c
28 c
1000 c
1150 c
1200 d
1300 d
6150
580
Source
waste oil
burn rates,3
liters/h
1330
2660
1330
2660
1330
2660
1330
" 2660
1330
2660
1330
2660
1330
2660
1330
2660
1330
2660
Contaminant
input,
pg/s x 103
178.
356.
1.47
2.94
10.3
20.6
366.
732.
422.
844.
439.
878.
476.
951.
2272.
4544.
213.
426
Emissions at
100% capacity ,b
pg/s x 103
134.
268.
1.10
2.20
7.73
15.46
275.
550.
317.
634.
13.1
26.3
14.3
28.5
2272.
4544.
6.40
12.8
Emissions in
January at
50% capacity,
pg/s x 103
67.
134.
0.55
1.10
3.87
7.73
138.
275.
159.
317.
6.57
13.1
7.15
14.3
1136.
4544.
3.18
6.40
I
u>
Liters/hour waste oil input (gal/hour) = 350 (700).
Assume 75% of metals and 3% of organics (97% destruction rate efficiency) are emitted.
Assume ppm = mg/liter.
HC1 emissions are assumed equal to 100% of input chlorinated organic levels.
-------
The climatological mode of the ISC model was applied to
obtain 30-day average ambient air concentrations of all contami-
nants with a threshold response. Table 4-18 presents the re-
sults for six medium-size boilers. Source 14 generates the
maximum concentration in each case, primarily because of its low
stack height. Sources 17, 18, and 19 had much higher emissions,
but have a lesser ground-level impact because they have higher
stacks. Figures 4-4 and 4-5 compare the ambient concentration
patterns of lead from a boiler with a feed rate of 1330 liters/h
and a 10-m stack with those from one with a feed rate of 2660
liters/h and a 20-m stack. Maximum concentrations for the
1330-liter/h boiler were estimated at 400 to 500 meters east of
the source (Figure 4-4). Because of its higher stack, the
impact of the 2660-liter/h boiler was much farther downwind, at
receptors 800 to 1000 meters east and east-southeast of the
source. Even at twice the lead emissions, a lower maximum
ground-level concentration resulted from the larger boiler
because of transport and dilution that result from release from
the higher stack.
Most of the pollutants from the burning of waste oil in
medium-size boilers (Table 4-17) amounted to less than 1 percent
of the corresponding EEL's. The impacts of barium, lead, and
HC1 concentrations, however, amounted to 36, 21, and 4 percent
of their respective EEL's or AAQS. The sensitivity of lower
levels of lead, HC1, and barium in waste oil burned in medium-
size boilers was tested by performing additional dispersion
modeling. Table 4-19 (barium), Table 4-20 (lead) and Table 4-21
4-32
-------
TABLE 4-18. AMBIENT AIR IMPACTS FROM THE COMBUSTION OF WASTE OIL CONTAINING
ALL CONTAMINANTS WITH"A THRESHOLD RESPONSE IN MEDIUM-SIZE BOILERS
Substance
Barium
Cadmium
Chromium
Lead
Zinc
Toluene
Trichloroethane
HC1
Naphthalene
Maximum 30-day concentrations, ug/m3
Source number
14a
0.153
0.002
0.009
0.314
0.361
0.015
0.016
2.58
0.007
15a
0.102
0.001
0.006
0.210
0.242
0.010
0.0011
1.73
0.005
16a
0.071
<0.001
0.004
0.146
0.168
0.007
0.007
1.20
0.003
17b
0.087
<0.001
0.005
0.178
0.205
0.009
0.010
1.47
0.004
18b
0.067
<0.001
0.004
0.137
0.158
0.007
0.007
1.13
0.003
19b
0.053
<0.001
0.003
0.108
0.124
0.006
0.006
0.89
0.003
EEL
concen-
tration,
ug/m3
0.43
0.34
4.32
1.5
43.2
3240
16420
59.7
4320
Maximum
percent-
age of
EEL
36C
<1
<1
21d
1
<1
<1
4
<1
1330 liters/hour.
2660 liters/hour.
At 50 percent emissions, Source 14 yields barium concentrations of 0.102
ug/m3 and 24% of EEL.
At 50 percent emissions, Source 14 yields lead concentrations of 0.209
and 14% of NAAQS.
4-33
-------
0.05
0.05
0.05
30-day raxlmun lead ,
concentrations. w
-------
0.025
0.01
30-day maximum lead ,
concentrations, u9/m
feed rate • 2660 i/hr
stack height • 20 m
0.05
0.05
0.1
0 200 400 600 800 1000
i i i i
meters
Figure 4-5. Isopleths of lead concentration over a 30-day averaging
period for a 2660i /h (700 gal/h) medium boiler.
4-35
-------
TABLE 4-19. AMBIENT AIR IMPACT FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS INPUT
CONCENTRATIONS OF BARIUM IN MEDIUM-SIZE BOILERS
Source
No.
14
15
16
17
18
19
Feed .
rate,b
fc/h
1330
1330
1330
2660
2660
2660
Stack
height,
m
10
15
20
20
25
30
Barium concentrations in waste oil, ppm
485C
X,dug/m3
0.153
0.102
0.071
0.087
0.067
0.053
% EEL6
36 f
24
17
20
16
12
200
X, ug/m3
0.063
0.042
0.030
0.036
0.028
0.022
% EEL
15
10
7
9
7
5
50
X, wg/m3
0.016
0.011
0.008
0.009
0.007
0.006
% EEL
4
3
4
2
2
2
25
X, iig/m3
0.008
0.006
0.004
0.005
0.004
0.003
% EEL
2
2
1
1
1
<1
10
X, vig/m3
0.003
0.002
0.002
0.002
0.002
0.001
% EEL
<1
<1
<1
<1
<1
<1
*.
OJ
For identification in ISC Model.
1330 liters/h = 350 gal/h; 2b60 llters/h = 700 gal/h.
90th percent!le concentration.
Ambient air concentrations of barium.
Environmental Exposure Limit = 0.43 yg/m3 for barium.
At 50 percent emissions, the % EEL is 24% at 485 ppm, 10% at 200 ppm, 3% at 50 ppm, 2% at 25 ppm,
at 10 ppm -
-------
TABLE 4-20. AMBIENT AIR IMPACT FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS
INPUT CONCENTRATIONS OF LEAD IN MEDIUM-SIZE BOILERS
Source
No.3
14
15
16
17
18
19
Feed .
rate,b
fc/h
1330
1330
1330
2660
2660
2660
Stack
height,
m
10
15
20
20
25
30
Lead concentrations in waste oil, ppm
3000
x. pg/m3
0.94
0.63
0.437
0.590
0.411
0.323
%
NAAQSd
63
42
29
36
28
21
1000e
x, pg/m3
0.314
0.210
0.146
0.178
0.137
0.108
%
NAAQS
21
14
10
12
9
7
500
Xt pg/m3
0.157
0.105
0.073
0.089
0.069
0.054
%
NAAQS
11
7
5
6
5
4
100
x» pg/m3
0.032
0.021
0.015
0.018
0.014
0.011
%
NAAQS
2
2
1
1
1
<1
25
X, pg/m3
0.008
0.005
0.004
0.005
0.004
0.003
%
NAAQS
<1
<1
<1
<1
<1
<1
For identification in ISC Model.
1330 liters/h = 350 gal/h; 2660 liters/h = 700 gal/h.
Ambient air concentration of lead.
NAAQS = 1.5 pg/m3 for lead.
90th percentile concentration.
-------
TABLE 4-21. AMBIENT AIR IMPACT OF HC1 FROM THE COMBUSTION OF WASTE OIL CONTAINING VARIOUS
INPUT CONCENTRATIONS OF CHLORINATED ORGANICS IN MEDIUM-SIZE BOILERS
Source
No.
14
15
16
17
18
19
Feed .
rate,6
f/n
1330
1330
1330
2660
2660
2660
Stack
height,
m
10
15
20
20
25
30
Chlorinated organics concentrations in waste oil, ppm
30,000
x,d pg/m3
12.58
8.43
5.86
7.17
5.52
4.34
% EEL6
21
14
10
12
9
7
10,000
Xt ug/m3
4.16
2.81
1.95
2.39
1.84
1.45
% EEL
7
5
3
4
3
2
6,150C
Xi ug/m3
2.52
1.69
1.17
1.43
1.10
0.87
% EEL
4
3
2
2
2
1
3,000
X, ug/m3
1.26
0.84
0.59
0.72
0.55
0.43
% EEL
2
1
1
1
1
1
1,000
Xt ug/m3
0.42
0.28
0.20
0.24
0.18
0.15
% EEL
1
1
<1
<1
<1
<1
I
Ul
co
For identification in ISC Model.
1330 liters/h = 350 gal/h; 2660 liters/h = 700 gal/h.
90th percentile concentration.
Ambient air concentrations of HC1.
Environmental Exposure Limit = 59.7 pg/m3 for HC1.
-------
(HC1) indicate the change in ambient impacts at variable contami-
nant input levels. Barium input levels of 25 ppm decreased
medium-size boiler impacts to 2 percent or less of the EEL in
all six of the cases analyzed (Table 4-19) . Lead input levels
of 100 ppm produced ambient lead concentrations from medium-size
boilers of less than 2 percent of the NAAQS (Table 4-20). A
chlorinated organics input level of 1000 ppm results in medium-
size boiler impacts of less than a 1 percent of the EEL for HC1
(Table 4-21) .
Although not necessarily indicative of every possible
boiler in the 50 to 100 x 10 Btu/h range, this ambient air
analysis demonstrates the potential impact of boilers of this
size. Further detailed analyses of individual boilers is re-
quired to improve the results. In lieu of modeling every boiler
that could possibly burn waste oil, this analysis provides an
estimate of the impact of the important threshold contaminants
of concern.
4.2.5 Lead Concentrations at Elevated Receptors
Of particular concern in this analysis is the concentration
of lead at receptors at or near the stack height of small and
medium-size individual and multiple sources. To obtain estimates
of the maximum 30-day concentrations of lead from such sources,
several source configurations were modeled with the ISC Model in
its long-term mode. The sources modeled were as follows:
106 Btu/h: 1 source
4 sources in a 2 x 2 grid 50 m apart
16 sources in a 4 x 4 grid 50 m apart
4-39
-------
10 x 10 Btu/h: 1 source
4 sources in a 2 x 2 grid 50 m apart
16 sources in a 4 x 4 grid 50 m apart
50 x 10 Btu/h: 1 source
4 sources in a 2 x 2 grid 100 m apart
Stack parameters for the 10 Btu/h boilers are those shown
for Source 1 in Table 4-2; for the 10 x 10 Btu/h boilers, those
for Source 9 in Table 4-2; and for the 50 x 10 Btu/h boilers,
those for Source 14 in Table 4-16. Emissions of lead in each
case are similar to those in Tables 4-3 and 4-17 at the appro-
priate boiler size, except in this case the concentration of
lead in the feed waste oil is fixed at 250 ppm (rather than 1000
ppm), which results in lower lead emissions.
All meteorological and modeling inputs are as described in
previous sections. A 200-m interval grid of .receptors was used
for the 50 x 10 Btu/h boilers and a 100-m interval grid for the
10 and 10 x 10 Btu/h boilers. Because of the source configu-
ration, the closest source/receptor distance is generally 100 m.
All receptors are located at stack heights that give the worst-
case concentrations because they are very close to the plume
centerlines.
Table 4-22 shows the maximum 30-day lead concentrations for
the various individual and multiple-source configurations.
Concentrations of lead in emissions from the small boilers (10
Btu/h) are all less than the NAAQS for lead (1.5 yg/m ). In
emissions from the 10 x 10 Btu/h boilers, however, concentra-
tions exceed the NAAQS for the 16-source array. Lead concentra-
tions in emissions from the medium-size boilers were less than
4-40
-------
TABLE 4-22. AMBIENT AIR LEAD IMPACTS AT ELEVATED RECEPTORS RESULTING
FROM THE COMBUSTION OF WASTE OIL CONTAINING LEAD
Boiler size,
106 Btu/h
1
1
1
10
10
10
50
50
Boiler configuration
1
2x2
4x4
1
2x2
4x4
1
2x2
Maximum 30-day
concentration, yq/m3
0.067
0.092
0.244
0.443
0.773
1.993
0.263
0.674
4-41
-------
the NAAQS for the two cases modeled, even though emissions were
greater. The lower concentrations were due primarily to higher
plume rise. The 16-source array of medium-size boilers was not
analyzed because of the low probability that this many 50 x 106
Btu/h boilers would be clustered so closely together.
4.2.6 Lead Concentrations at Close Receptors
Calculations were made to account for concentrations at re-
ceptors closer than 100 m (the minimum allowable source-receptor
distance in the ISC Model). These calculations involved the use
of Gaussian plume equations. The plume heights and receptor
heights were equal, and the estimates were made at the plume
centerline. The following equation resulted:
Qf
X =
IT ay a u
where x = ambient concentration
Q = local emissions in vg/s at 250 ppm in the feed-
stock
f = frequency of occurrence of specific windspeed,
wind direction, and stability class categories
ay = horizontal standard deviation of the plume
az = vertical standard deviation of the plume
u = windspeed
Concentrations were estimated for a receptor 25 m downwind
of a source. At this distance, the ay and a parameters are 2.1
and 1.4 m, respectively, for a neutral stability class (D). A
January average windspeed (see Table B-6 in Appendix B) of 5.8
m/s was used in the calculations. Because previous sections
indicate the maximum concentrations to the east of the sources,
4-42
-------
the frequency of occurrence toward 67.5, 90.0, and 112.5 degrees
was compiled, which yielded a combined frequency under all
stability classes (D stability accounted for over 90 percent of
the total) of 0.043 or 4.3 percent.
The impact of multiple sources was determined by modeling
two sources separated by 50 m; i.e., the impact of the first at
25 m and the second at 75 m, to determine the combined source
impact at one downwind receptor. Only two sources were modeled
(in contrast to the modeling described in Section 4.2.6) because
distances between receptors indicate little contribution from
off-axis sources. The results of this analysis for the combus-
tion of waste oil with a lead content of 250 ppm were single-
source lead concentrations of 0.395, 5.10, and 27.5 yg/m in
emissions from the 1, 10, and 50 x 10 Btu/h boilers; and con-
centrations of 0.45, 5.76, and 31.0 ug/m for the two-source 1,
10, and 50 x 10 Btu/h cases, respectively. Concentrations of
lead at these close-in receptors are therefore estimated to be
higher than the NAAQS for sources larger than 10 Btu/h.
4.2.7 Increased Organic Levels
The dispersion modeling presented so far focuses on a 90th
percentile level of organic contaminants in the composite waste
oil. Mixing of individual or composite organic solvents into
waste oil is not uncommon, however. For an evaluation of the
potential increased ambient air quality impact of single sources
burning waste oil high in organic compounds, concentrations of
toluene and 1,1,1-trichloroethane were increased to the 10,000-
ppm level (1%) in the waste oil feed stock. These levels
4-43
-------
of contaminants were considered in further analysis of the 13
small boilers and space heaters described earlier in Table 4-2.
The results of previous analyses showed concentrations to
be much less than the applicable EEL's for toluene and trichlo-
roethane. Even at almost 10 times the original contaminant
levels of 1200 and 1300 ppm for toluene and trichloroethane,
respectively, the air quality impacts are insignificant. Table
4-23 presents the results of the dispersion modeling for the 13
small boilers and space heaters. In all cases, the 30-day
maximum concentrations of contaminants in ambient air are much
less than 1 percent of the applicable EEL. The primary reasons
for the insignificant air quality impacts are the destruction
efficiencies of organics in waste oil and the high threshold
response concentrations required to elicit a health-related
response.
4.3 URBAN-SCALE ANALYSIS WITH THE HANNA-GIFFORD MODEL
4.3.1 Overview
The urban-scale analysis examines the contaminant levels
that may occur over a broad urban scale area (50 to 100 km).
All modeling in this section was performed with a Hanna-Gifford
model.
Based on the following constraints, a simplified modeling
procedure was selected for this analysis:
0 Waste oil is burned primarily in a large number of
small boilers and space heaters.
0 These sources are distributed evenly across an urban
area.
4-44
-------
TABLE 4-23. MAXIMUM 30-DAY CONCENTRATIONS ASSUMING 1 PERCENT LEVELS OF TWO
ORGANICS IN WASTE OIL AND 97 PERCENT DESTRUCTION REMOVAL EFFICIENCIES
(ug/m3)
Source
No.
1
2
3
4
5
6
7
8
9b
10
11
12
13
EEL,C yg/m'
Maximum
percentage
of EEL
Concentration of waste oil contaminant
Toluene
0.051
0.014
0.035
0.015
0.008
0.066
0.030
0.016
0.082
0.042
0.024
0.036
0.023
3,240
<1
1 ,1 ,1-Trichloroethane
0.050
0.013
0.034
0.015
0.007
0.066
0.030
0.016
0.082
0.042
0.024
0.036
0.021
16,420
<1
Identification in ISC model.
Source giving maximum concentrations.
c EEL = Environmental Exposure Limit.
4-45
-------
0 State and national inventories and summaries of sourc-
es do not contain records of small sources.
The EPA has used the Hanna-Gifford Model, which was devel-
oped by the Air Resource Atmospheric Turbulence and Diffusion
Laboratory at Oak Ridge, Tennessee, for previous hazardous
g
pollutant analyses. This simple but realistic dispersion model
has proved to be adequate for estimating pollutant concentrations
resulting from area source emissions in urban areas. It has
performed best in large urbanwide applications, f9'10 especially
in the modeling of low-level sources that are distributed fairly
evenly across such an area.
The Hanna-Gifford Model entails the use of emissions aver-
aged over time and space rather than those from specific indi-
vidual sources. The emissions due to waste oil burning were
estimated over an urban study area and distributed evenly across
the study grids in a manner consistent with the model require-
ments. Additional inputs were:
0 The windspeed for the averaging time of interest.
0 A specific constant calculated on the basis of study
area grid size.
0 The number of grids and grid size.
0 Two atmospheric stability constants.
For this analysis, an urban area was selected where a large
number of residential/conunercial small boilers and space heaters
were believed to be in use. The study area was broken into five
subareas based on population density, urban-versus-rural land
4-46
-------
use, and distance from the urban core. The area selected for
analysis is situated within Air Quality Control Region 43 on the
eastern coastline of the United States. The study area and
subareas are shown in Figure 4-6. Concentric rings at 5, 10,
15, 25, and 50 km were used to differentiate subareas for study.
For this analysis, each area bounded by the concentric rings was
treated as a separate calculation in the model. This approach
was considered reasonable because a receptor located at some
midpoint in each subarea was affected primarily by emissions in
that area or those immediately adjacent.
The model was applied in an iterative fashion by first cal-
culating the value of the constant in the model, then estimating
the appropriate seasonal or annual windspeeds, and finally,
estimating the subarea emissions due to waste oil burning. The
concentration estimate made for each contaminant in each subarea
was assumed to apply over that whole subarea. All model inputs
are discussed in Appendix B.
4.3.2 Urban Emissions
Different concentric circle subareas were assigned different
emission rates, depending on population, total area of subarea,
and availability of waste oil for burning. The area of each
subarea shown in Figure 4-6 was calculated to determine the
extent over which subarea-specific emissions were distributed.
The area of each subarea (bounded by the distances specified in
Figure 4-6) is shown in Table 4-24.
4-47
-------
Figure 4-6. Urban area and associated study grid used
in the Hanna-Gifford Model analyses.
4-48
-------
TABLE 4-24. AREAS OF EACH STUDY SUBAREA
Distance from study
area centroid, km
0-5
5-10
10-15
15-25
25-50
Area of subarea,
107 m2
7.85
23.6
39.3
134.
589.
The emissions of contaminant per liter of waste oil burned
were calculated and are presented in Table 4-25. These emission
estimates were used together with the estimates of waste oil
burned in each subarea to calculate the emission rates for each
contaminant. Table 4-26 presents the estimates of waste oil
generated, the combustion factors, the study period use factors,
and the calculated amounts of waste oil burned in each subarea.
For this urban-scale analysis, all waste oil contaminants
eliciting a threshold level response were modeled over a 30-day
averaging time. All carcinogenic contaminants were modeled over
an annual averaging time. So that the ambient air concentra-
tions would be expressed in a format consistent with model
specifications, the emission factors were determined for the
appropriate averaging time and expressed in mass per area per
time period.
The urban-scale subarea emission rates for threshold and
nonthreshold contaminants due to waste oil burning are presented
in Tables 4-27 and 4-28. The subareas are expressed in terms of
distance of the boundaries from the centroid of the urban study
area. These emission rates per unit area provided the necessary
4-49
-------
TABLE 4-25. ESTIMATED CONTAMINANT EMISSION RATES PER LITER OF
WASTE OIL BURNED
Waste oil
contaminant
Arsenic
Barium
Cadmium
Chromium
Lead
Zinc
Trichloroetnane
Tetrachloroethylene
Toluene
Trichloroethylene
Carbon tetrachloride
PCB's
Chlorinated organics
Benzene
Naphthalene
Concentration
in waste oil ,
ppm
16C
485
4
28
1000
1150
1300
1200
1200
1050
1000
50
6150
160
580
Emissions per waste
oil burned,
g/liter
0.012C
0.364
0.003
0.021
0.750
0.863
0.039
0.036
0.036
0.031
0.030
0.001
6.15
0.005
0.018
a Weight/volume.
Metal emissions are assumed at 75 percent of input metal rates
and organics at 3 percent of organic input rates.
c At 5 ppm concentration, arsenic emissions equal 0.004 g/liter.
HC1 emissions are assumed equal to 100 percent of input rates
of chlorinated organics.
4-50
-------
TABLE 4-26. WASTE OIL BURNING CHARACTERISTICS DERIVED
FOR THE URBAN-SCALE DISPERSION MODELING ANALYSIS
Distance from
study area
centroid, km
0-5
5-10
10-15
15-25
25-50
Waste oil
generated,
107 liters/yr
1.94
2.35
3.32
4.29
1.94
Combustion
factor
0.7
0.7
0.7
0.7
0.7
January
Monthly
use factor
0.21
0.21
0.21
0.20
0.20
Waste oil
burned,
106 liters/mo
2.85
3.45
4.88
6.31
2.85
Annual
Annual
use factor
1.0
1.0
1.0
1.0
1.0
Waste oil
burned,
107 liters/yr
1.36
1.65
2.32
3.00
1.36
I
en
-------
TABLE 4-27. AREA SOURCE JANUARY EMISSIONS FROM THE COMBUSTION OF WASTE OIL CONTAINING
CONTAMINANTS WITH A THRESHOLD RESPONSE
DIs trance from
study area
centrold. 1*
0-5
5-10
10-15
15-25
25-50
January emissions of threshold contaminants. 109 g/s-m2
Barium
4.94
2.00
1.69
0.609
0.062
CadmiiM
0.041
0.017
0.014
0.0050
0.00052
Chromium
0.285
0.115
0.098
0.035
0.0036
Lead
10.2
4.11
3.49
1.26
0.129
Zinc
11.7
4.73
4.01
1.45
0.149
Toluene
0.48
0.20
0.16
0.060
0.0060
1,1.2-Trlchloroethane
0.523
0.211
0.179
0.064
0.0070
Chlorinated
organics
83.4
33.6
28.5
11.
1.1
Naphthalene
0.237
0.096
0.079
0.030
0.0030
Ol
to
-------
TABLE 4-28. AREA SOURCE ANNUAL EMISSIONS FROM THE COMBUSTION OF WASTE OIL
FOR NONTHRESHOLD CONTAMINANTS
lU
in
Distance fro*
study area
centrold, km
0-5
5-10
10-15
15-25
25-50
Annual emissions of nonthreshold contanlnants, 10 g/s->
Arsenic*
6.58
2.66
2.55
0.855
0.088
Chromium
11.6
4.66
3.95
1.49
0.154
Carbon
tetrachlorlde
16.4
6.59
5.60
2.12
0.218
PCB's
0.828
0.333
0.282
0.107
0.011
Tetrachloro-
ethylene
19.7
7.95
6.72
2.54
0.262
Trlchloro-
ethylene
17.1
6.91
5.87
2.22
0.228
1.1.2-TH-
chloroethane
21.2
8.52
7.25
2.74
0.282
Benzene
2.50
1.01
0.856
0.324
0.033
* At 5 pp« Input, arsenic emissions equal 2.06, 0.83, 0.70, 0.267, and 0.028 x 10 g/s-m2 per study grid.
-------
emissions input in all subsequent urban-scale modeling calcula-
tions. These rates were considered best estimates of possible
urbanwide emissions when all available waste oil was utilized.
The derivation of windspeeds, which was based on local data
summaries, is presented in Appendix B.
4.3.3 Urban-Scale Air Quality Impacts
For threshold contaminants, the results of applying the
Hanna-Gifford Model to the subject urban area are expressed in
terms of monthly averaged concentrations. Nonthreshold contami-
nants are calculated on an annual basis, based on the assumption
that year-to-year variations are not significant and that the
average subarea and urbanwide concentrations and exposures are
comparable to lifetime exposures.
Table 4-29 presents estimated monthly emission concentra-
tions of threshold contaminants. A review of the ratio of the
highest concentrations to the applicable EEL's or NAAQS shows
that barium, lead, and HC1 are of primary significance, at 77,
44, and 9 percent of their respective EEL's or NAAQS. All other
contaminants are 2 percent or less.
Because the levels of chlorinated organics in waste oil
were assumed, alternative emission levels of HC1 were analyzed.
Table 4-30 presents the ambient concentrations of HC1 at several
concentrations of chlorinated organics in the waste oil. At the
1000-ppm level, emission concentrations of HC1 fall to 1 percent
of the applicable EEL in the innermost study area (0 to 5 km).
4-54
-------
TABLE 4-29. AMBIENT AIR IMPACTS FROM THE COMBUSTION OF WASTE OIL CONTAINING
ALL CONTAMINANTS WITH A THRESHOLD RESPONSE IN THE URBAN STUDY AREA
Pollutant
Barium
Cadmium
Chromium
Lead
Zinc
Toluene
1,1,1-THchloroethane
HC1
Naphthalene
Study area monthly average concentrations, ng/m1
Distance from study area centroid, km
0-5
0.323*
0.003
0.019
0.665b
0.765
0.024
0.026
5.45
0.016
5-10
0.108
0.001
0.062
0.222
0.256
0.008
0.009
1.83
0.005
10-15
0.091
0.001
0.005
0.189
0.217
0.007
0.007
1.55
0.004
15-25
0.033
<0.001
0.002
0.068
0.078
0.002
0.003
0.56
0.002
25-50
0.003
<0.001
<0.001
0.007
0.008
<0.001
0.001
0.06
<0.001
EEL
concentration
0.43
0.34
4.32
1.5
43.2
3,240.0
16.420.0
59.7
432.
Maximum
percentage
of EEL
77
1
<1
44
2
<1
<1
9
"
I
I/I
U1
1 At 50 percent emissions (rather than the assumed 75 percent) the maximum concentration of barium Is 0.215 wg/m or 50 percent
nf tf\
of EEL.
or ttL.
At 50 percent emissions (rather than the assumed 75 percent) the maximum concentration of lead is 0.443 wg/m or 301 of MAAQS.
-------
TABLE 4-30. AMBIENT AIR IMPACTS FROM THE COMBUSTION OF WASTE OIL CONTAINING
CHLORINATED ORGANICS AT VARIOUS PLANT INPUT CONCENTRATIONS IN THE URBAN STUDY AREA
Distance from
study area
centroid, km
0-5
5-10
10-15
15-25
25-50
Chlorinated organics concentrations (ppm) in the waste oil feed stock
30,000
x,a ug/m3
26.6
8.91
7.55
2.71
0.28
k
% EELD
145
15
13
4
1
10,000
x, pg/m3
8.87
2.97
2.52
0.90
0.09
% EEL
15
5
4
1
<1
6,150C
X. pg/m3
5.45
1.82
1.55
0.55
0.06
% EEL
9
3
3
<1
<1
3,000
X, pg/m3
2.66
0.89
0.76
0.27
0.03
% EEL
4
1
1
<1
<1
1,000
x. pg/m3
0.89
0.30
0.25
0.09
0.01
% EEL
1
1
<1
<1
<1
en
Ambient air concentration of HC1.
b EEL = 59.7 ug/m3 for HC1.
C 90th percent!le concentration.
-------
Similar concentration estimates were made for each subarea
for the nonthreshold contaminants associated with waste oil
burning. Tables 4-31 and 4-32 present these estimates. Table
4-31 shows annual ambient air concentrations of nonthreshold
contaminants based on an assumed 97 percent destruction removal
efficiency for organics and the emission of 75 percent of the
metals in the oil. Table 4-32 shows annual ambient air concen-
trations for the same contaminants in the innermost ring of the
urban area, based on an assumed 99 percent destruction removal
efficiency for organics and the emission of 50 percent of the
metals in the oil. In each case, the highest concentrations
occur in the 0- to 5-km area. Cancer risks associated with the
modeled concentrations are presented in Section 5. In each
case, the highest concentrations occurred in the 0- to 5-km
area.
4-57
-------
TABLE 4-31. CONCENTRATIONS OF NONTHRESHOLD CONTAMINANTS FROM WASTE
OIL COMBUSTION IN THE URBAN AREA, BASED ON 97 PERCENT DESTRUCTION REMOVAL
EFFICIENCIES OF ORGANICS AND EMISSION OF 75 PERCENT OF THE METALS
IN THE OIL
Pollutant
Arsenic
Benzene
Cadmium
Chromium
Carbon tetrachloride
PCB's
Tetrachl oroethyl ene
Trichl oroethyl ene
1,1,2-Trichloroethane
Annual concentrations, ug/m3
Distance from study area centroid, km
0-5
0.0050
0.0018
0.0010
0.0086
0.0122
0.0006
0.0147
0.0128
0.0158
5-10
0.0016
0.0008
0.0003
0.0027
0.0038
0.0002
0.0046
0.0040
0.0050
10-15
0.0013
0.0007
0.0003
0.0023
0.0033
0.0001
0.0036
0.0034
0.0042
15-25
0.0005
0.0002
0.0001
0.0009
0.0013
0.0001
0.0015
0.0013
0.0016
25-50
<0.0001
<0.0001
<0.0001
<0.0001
0.0001
<0.0001
0.0001
0.0001
0.0001
a At the 5-ppm arsenic level (rather than the 16-ppm level assumed), the
concentration in the 0- to 5-km area equals 0.0016 yg/m3
4-58
-------
TABLE 4-32. CONCENTRATIONS OF NONTHRESHOLD CONTAMINANTS IN THE URBAN
AREA, BASED ON 99 PERCENT DESTRUCTION REMOVAL EFFICIENCIES AND
EMISSION OF 50 PERCENT OF THE METALS IN THE OIL
Contaminant
Arsenic
Benzene
Cadmium
Carbon tetrachloride
Chromium
PCB's
Tetrachloroethylene
Trichloroethylene
1,1,2-Trichloroethane
Maximum
concentration in
0- to 5-km area,
ug/m3
0.0033
0.0006
0.0007
0.0041
0.0057
0.0002
0.0049
0.0043
0.0053
4-59
-------
REFERENCES FOR SECTION 4
1. Bowers, J. F., J. R. Bjorkland, and C. S. Cheny. Industrial
Source Complex (ISC) Dispersion Model User's Guide. Vols. 1
and 2. EPA-450/4-79-030 and EPA-450/4-79-031, December
1979.
2. U.S. Environmental Protection Agency. Guideline on Air
Quality Models. EPA-450/2-78-027, April 1978.
3. Gifford, F. A., and S. R. Hanna. Modeling Urban Air Pollu-
tion. Atmos. Environment, 2:131~136/ 1973.
4. Holzworth, G. C. Mixing Heights, Wind Speeds, and Potential
for Urban Air Pollution Throughout the Contiguous United
States. AP-101, 1972.
5. U.S. Department of Commerce. Local Climatological Data.
J. F. Kennedy International Airport, National Oceanic and
Atmospheric Administration, New York, New York, National
Climatic Center. Asheville, North Carolina. 1979.
6. PEDCo Environmental, Inc. Program Office Support: Environ-
mental Impacts of Fossil Fuel Utilization. Prepared for
U.S. Environmental Protection Agency under Contract 68-02-
3173, Task Order 5. (No date.)
7. Miller, C. W. An Application of the ATDL Simple Dispersion
Model. JAPCA, 28:(8), August 1978.
8. Anderson, G. E., C. S. Liu, H. Y. Holman, and J. P. Killus.
Human Exposure to Atmospheric Concentrations of Selected
Chemicals. U.S. Environmental Protection Agency. March
1980.
9. Hanna, S. R. A Simple Method of Calculating Dispersion From
Urban Area Sources. JAPCA, 21:(8), December 1971.
10. Hanna, S. R. Urban Diffusion Problems. Presented at the
AMS Workshop of Meteorology and Environmental Assessment.
Boston, Massachusetts.
4-60
-------
SECTION 5
ENVIRONMENTAL IMPACT AND HEALTH RISKS
FROM WASTE OIL BURNING
5.1 ENVIRONMENTAL IMPACT AND HEALTH CONSEQUENCE OF THRESHOLD
CONTAMINANTS
This section presents an assessment of the impact on air
quality and the risk to human health posed by airborne concentra-
tions of threshold (noncarcinogenic) contaminants from the com-
bustion of waste oil through a comparison of these concentrations
with both EEL's and existing air quality.
Two atmospheric dispersion models were used to estimate the
contribution of waste oil combustion to air quality, the ISC
Model (point-source emissions) and the Hanna-Gifford model (area
emissions). Table 5-1 presents the results that reflect the
greatest ambient air impacts from all of the modeling runs. A
review of these modeling results indicates that three waste oil
constituents (barium, hydrogen chloride, and lead) may have a
substantial impact on air quality when viewed in relation to
their corresponding environmental exposure limits.
The magnitude of impact can be further assessed by comparing
airborne concentrations with existing air quality measurements,
as discussed in the following subsections.
5-1
-------
TABLE 5-1. A COMPARISON OF ESTIMATED MAXIMUM EXPOSURE CONCENTRATIONS
FROM WASTE OIL BURNING WITH ENVIRONMENTAL EXPOSURE LIMITS
Barium
Cadmium
Chromium (II and III)
HC1
Lead
Zinc
Naphthalene
Toluene
1 ,1 ,1-Trichloroethane
Ambient air
concentration,
ug/m3
0.098b
0.15^
0.34d
0.323e
<0.001b
0.002CH
0.0028°
0.0036
0.006*!
0.009C,
0.0197°
0.062*
1.66b
2.58^
5.80d
5.45e
0.203b
°-314H
0.708°
0.665e
0.233*!
0.361^
0.814°
0.765e
0.004*!
0.007^
0.017d
0.0166
0.009*!
0.015CH
0.0341a
0.0246
0.0105
0.016C.
0.0370°
0.0266
Waste oil
contribution,
percent of
the EELa
23
36
79
75
«1
<1
1
1
<1
<1
<1
<1
3
4
10
9
14
21
47
44
<1
1
2
2
<1
<1
<1
<1
<1
<1
-------
5.1.1 Barium Impact
Dispersion modeling of barium emissions from single point
sources indicates that local airborne concentrations of barium
compounds can reach 0.098 to 0.153 yg/m3 or 23 to 36 percent of
the EEL. The EEL is derived from a TLV designed to protect
against excitability. Barium increases the excitability of the
muscles, particularly cardiac muscle. Modeling results of barium
emissions from multiple point sources and high-density urban
areas indicate that airborne concentrations of barium from waste
oil burning may reach 75 to 79 percent of the EEL. Airborne
concentrations of barium in the northeastern United States are
reported to range from a high of 2.0 yg/m3 to a low of 0.005
ug/m3 (mean concentration of 0.25 ug/m3). Based on these ambient
concentrations, barium emissions from waste oil burning appear to
account for at least 16 percent of the airborne concentration in
high-density urban areas.
5.1.2 Hydrogen Chloride Impact
The dispersion modeling of hydrogen chloride emissions from
single point sources indicates that airborne hydrogen chloride
concentrations could occur at 1.66 to 2.58 yg/m3, or between 3
and 4 percent of the EEL. The main health effect from exposure
to hydrogen chloride is irritation. Hydrogen chloride emissions
from multiple point sources and high-density urban areas indicate
that 5.45 to 5.80 ug/m3 or 9 to 10 percent of the hydrogen chlo-
ride EEL may be reached as a result of waste oil burning. Air-
borne ambient concentrations of 2 to 8 ug/m3 (1.3 to 5 ppb)
5-3
-------
chlorine (as HC1) were measured at 10 m above the surface of the
12 3
North Atlantic. Cholak and Katz have indicated chloride
concentrations in American cities ranging from 0.016 to 0.095
ppm, or 24 to 140 ug/m3; however, these data are from the 1950's.
As indicated by the air dispersion modeling of multiple sources
and urban areas, concentrations of HC1 from waste oil burning in
some cities can still be as high as those measured. More recent
ambient HC1 data are not available.
5.1.3 Lead Impact
Lead emissions resulting from the burning of waste oil at
single point sources are believed to account for local airborne
concentrations from 0.203 to 0.314 yg/m3, or 14 to 21 percent of
the National Ambient Air Quality Standard. The lead standard is
designed to protect against encephalopathy (brain disease) and
renal damage. Emissions from multiple sources or high-density
urban areas may produce airborne lead concentrations of 0.665 to
0.708 yg/m3, or 44 to 47 percent of the NAAQS. Measured ambient
air concentrations of lead in the northeastern United States
range from a high of 3.6 ug/m3 to a low of 0.06 ug/nt3 (mean
concentration of 0.79 ug/m3). Based on these values, emissions
from waste oil burning could account for at least 10 percent of
the airborne lead concentrations in high-density urban areas.
5.1.4 Summary
In summation, the modeling results show that concentrations
of barium, hydrogen chloride, and lead from the burning of waste
5-4
-------
oil in small or medium-size boilers or space heaters has a mea-
surable impact on air quality and that waste oil burning may
account for a significant portion of the existing ambient air-
borne concentrations of each of these substances. The other
threshold pollutants (cadmium, chromium, zinc, naphthalene,
toluene, and 1,1,1-trichloroethane) do not have a serious impact
on air quality near single or multiple sources or in high-density
urban areas.
5.2 ENVIRONMENTAL IMPACT AND HEALTH CONSEQUENCE OF NONTHRESHOLD
CONTAMINANTS
Potential cancer risks associated with waste oil combustion
were calculated by comparing the estimated airborne concentra-
tions previously presented (in Tables 4-30 and 4-31) with the
reference concentrations presented in Appendix C, Table C-7.
Tables 5-2 and 5-3 present the results of this comparison.
Based on the estimated ambient concentrations and the refer-
ence concentrations, a cancer risk estimate was determined for
each waste oil constituent being emitted into the atmosphere.
Cancer risk is calculated as a ratio of the modeled airborne
concentration to the reference concentration. The resulting
value, which is expressed in scientific notation, represents the
frequency of cancers per a given population; e.g., the cancer
risk from exposure to airborne arsenic is 2.0 x 10 or two
incidences of cancer per 100,000 population.
Another means of expressing cancer risk is to present the
value in terms of risk to a single individual; e.g., the cancer
5-5
-------
TABLE 5-2. LIFETIME CANCER RISK ASSOCIATED WITH WASTE OIL BURNING IN A
HIGH-DENSITY URBAN STUDY AREA, BASED ON 97 PERCENT DESTRUCTION REMOVAL
EFFICIENCY FOR ORGANICS AND THE EMISSION OF 75 PERCENT OF THE METALS IN THE OIL
Substance
, . b
Arsenic
Benzene
Cadmium
Chromium
Carbon tetrachloride
Polychlorinated
biphenols (PCB's)
Tetrachloroethylene
Trichloroethylene
1,1,2-Trichloroethane
Airborne annual
concentration ,
yg/m3
0.0050
0.0018
0.001
0.0086
0.0122
0.0006
0.0147
0.0128
0.0158
Cancer risk
2.0 x 10"5
2.7 x 10"8
1.9 x 10"6
1.1 x 10"4
4.5 x 10"7
7.4 x 10"7
2.2 x 10"7
4.6 x 10"8
2.6 x 10"7
Approximate
risk to an
individual
1:50,000
1:37,000,000
1:530,000
1:9,100
1:2,200,000
1:1,300,000
1:4,500,000
1:22,000,000
1:3,800,000
See Table 4-30. This is based on concentrations in the 0- to 5-km innermost
ring; thus, risk represents the worst case for the most exposed individuals.
At the 5-ppm arsenic level (rather than the assumed 16-ppm level), the con-
centration in the 0- to 5-km area equals 0.0016 yg/m3, which reduces the
individual risk to 6.4 x 10"6.
5-6
-------
TABLE 5-3. LIFETIME CANCER RISK ASSOCIATED WITH WASTE OIL BURNING IN A
HIGH-DENSITY URBAN STUDY AREA, BASED ON 99 PERCENT DESTRUCTION REMOVAL
EFFICIENCY OF ORGANICS AND THE EMISSION OF 50 PERCENT OF THE METALS IN THE OIL
Substance
Arsenic
Benzene
Cadmium
Carbon tetrachloride
Chromium
PCB's
Tetrachl oroethyl ene
Trichloroethylene
1,1,2-Trichloroethane
Airborne annua]
concentration,
ug/m3
0.0033
0.0006
0.0007
0.0041
0.0057
0.0002
0.0049
0.0043
0.0053
Cancer risk
1.3 x 10"5
8.9 x 10"9
1.3 x 10"6
1.5 x 10"7
7.1 x 10"5
2.5 x 10"7
7.4 x 10'8
1.5 x 10"8
8.7 x 10"8
Approximate risk
to an individual
1:77,000
1:110,000,000
1:760,000
1:6,600,000
1:14,000
1:4,000,000
1:13,000,000
1:65,000,000
1:12,000,000
See Table 4-31.
5-7
-------
risk from arsenic (2.0 x 10~ ) can be expressed as the risk to a
single individual in terms of 1 chance in 50,000 (1:50,000).
The significance of the results in Tables 5-2 and 5-3 depend
on the level of risk considered to be acceptable. For either set
of conditions (97% or 99% DRE's of organics and the emission of
75% or 50% of the metals in the ore), if a risk level of 10 is
established as acceptable, the risk to health from chromium and
arsenic would then be considered significant. On the other hand,
should a lower risk level of 10~ be established as acceptable,
chromium, arsenic, and cadmium would then be considered to be
presenting a significant health problem. Table 5-4 identifies
the waste oil constituents of concern for three levels of risk
10~4, 10"5, and 10~6.
TABLE 5-4. WASTE OIL CONSTITUENTS PRESENTING A POTENTIALLY
UNACCEPTABLE CANCER RISK3
Acceptable
risk level
Assuming 97 percent ORE of
organics and emission of
75 percent of metals
Assuming 99 percent ORE of
organics and emissions of
50 percent of metals
10
10
-4
-5
10
-6
Chromium
Chromium, arsenic
Chromium, arsenic, cadmium
Chromium, arsenic
Chromium, arsenic, cadmium
See Tables 5-2 and 5-3.
5-8
-------
REFERENCES FOR SECTION 5
1. U.S. Environmental Protection Agency. Hydrochloric Acid and
Air Pollution: An Annotated Bibliography. Environmental
Protection Agency, Research Triangle Park, North Carolina.
AP-100, 1971.
2. J. Cholak (ed.) In: Proceedings of the National Air Pollu-
tion Symposium, 2nd, 1952, p. 6.
3. Katz, M. Air Repair, 4:176, 1955.
5-9
-------
SECTION 6
CHLORINATED DIBENZODIOXINS AND DIBENZOFURANS
Polychlorinated dibenzodioxins (PCDD's) are a subset of a
larger group of related organic compounds collectively referred
to in the literature as polyhalogenated aromatic hydrocarbons.
Other members of the group include polychlorinated dibenzofurans
(PCDF's), polychlorinated biphenyls (PCB's), polybrominated
biphenyls (PBB's), polychlorinated naphthalenes (PCN's), and
polychlorinated terphenyls (PCT's). The PCDD subset alone has 75
individual cogeners. One of these, 2,3,7,8-tetrachlorodibenzo-
dioxin (2,3,7,8-TCDD), is of special concern because of its
extreme toxicity and potentially widespread exposure. This and
other PCDD's occur as unwanted byproducts of the manufacture and
combustion of certain chemicals, most notably the chlorophenols
and their derivatives.
6.1 HEALTH EFFECTS
The most acutely toxic halogenated aromatic compound known
is 2,3,7,8-TCDD. Laboratory studies have shown that less than
1 ug/kg is lethal to at least half of a test population of guinea
pigs. No other synthetic chemical is known to be so toxic. This
lethal effect has been observed to vary in intensity among dif-
ferent animal species. Though all species are adversely affect-
ed, some are more tolerant of the chemical than others. Monkeys,
6-1
-------
mice, rabbits, frogs, and dogs for example, survive equivalent
exposures to 2,3,7,8-TCDD much better than guinea pigs. Also,
although humans have been both inadvertantly and accidentally
exposed to 2,3,7,8-TCDD, not one death has been directly at-
tributed to acute poisoning by this substance. The extent of
human exposure is not well known. Thus, it is unclear whether
2,3,7,8-TCDD, which is so lethally toxic to some laboratory
animals, is also highly toxic to humans.
The acute toxicity of other PCDD's varies according to the
number and placement of chlorine atoms on the basic dioxin triple-
ring molecule. For example, dioxins with a more-or-less symmet-
rical distribution of chlorines (i.e., chlorines laterally placed
on both benzene rings of the molecule) and in which at least
three of the four positions (2, 3, 7, and 8) are occupied tend to
be much more toxic than those with asymmetrical distribution.
For example, 2,3,7,8-TCDD is highly toxic (rat oral LD 40
yg/kg), whereas 1,2,3,4-TCDD is nearly nontoxic (rat oral LD
>1000 pg/kg).1
ci
Furthermore, the degree of chlorination also affects toxici-
ty. Dioxins with three to six chlorine substituents symmetrical-
ly distributed tend to be quite toxic, whereas dioxins with
either no substituents (i.e., no chlorine atoms) or dioxins that
6-2
-------
are totally substituted (i.e., with eight chlorines) exhibit
little or no toxicity.
The toxicity of PCDF's has not been studied to the extent of
PCDD's. From what has been learned so far, however, it appears
that structurally analogous PCDF's may exhibit toxic effects
similar to—though less potent than—the corresponding PCDD's.
For example, the acute oral toxicity of 2,3,7,8-tetrachlorodi-
benzofuran (2,3,7,8-TCDF) is within an order of magnitude of
2,3,7,8-TCDD, as shown below:
Cl Cl
Cl Cl
2,3,7,8-TCDD
2, 3, 7, 8-TCDF
TCDF
'en/ My/^y
0.5-1.5
30-70
150-300
uw_ n i My / ^y
5-10
1000
6000
LD TCDD
4
20
25
Species
Guinea pig
Monkey
Mouse
As with 2,3,7,8-TCDD, the toxicity of 2,3,7,8-TCDF varies in
intensity by animal species, and the acute effect level for
humans remains an unknown. Both molecules elicit chloracne,
thymus atrophy, weight loss, and liver effects, but the dose of
2,3,7,8-TCDF needed to produce a given effect is often many times
(30 to lOOx) higher than that for 2,3,7,8-TCDD. The observed
similarity between the toxic properties of 2,3,7,8-TCDD and those
of 2,3,7,8-TCDF appears to be closely linked to the similar geo-
metric dimensions and chlorine substitution patterns of the two
molecules.
6-3
-------
Apart from its well-documented acute toxic effects in animals,
there is growing evidence that 2,3,7,8-TCDD is carcinogenic in
animals and possibly in humans. An increased incidence of liver
and lung/upper airway cancers has been described in one study
where rats were fed 2,3,7,8-TCDD. In another more recent study
4
in rats, tumors of a variety of organs were produced. Experi-
mental data concerning other dioxins are limited. Recently (July
1983) , a panel of some 50 scientists from around the world, all
of whom are experts on the chemistry and toxicology of 2,3,7,8-
TCDD and other dioxins, met with EPA's Environmental Criteria and
Assessment Office staff in Cincinnati, Ohio, to review available
data on dioxin health effects. In particular, they were asked to
evaluate the compound's potential as a human carcinogen. The
panel discussed at length the results of the animal studies
mentioned above, and considered additional recent and admittedly
limited epidemiological data on the incidence of soft tissue
sarcoma in exposed populations in Europe and the United States
(i.e., Midland, Michigan). At the meeting's conclusion, the
panel endorsed a statement to the effect that "2,3,7,8-TCDD is
probably carcinogenic for humans on the basis of animal carcinogen-
icity studies which were positive in multiple species and organs."
In addition to its acutely toxic and cancer-producing pro-
perties, 2,3,7,8-TCDD has been observed to elicit the following
toxic and biologic responses in humans; '
Dermatological
Severe and persistent chemically induced acne known as
chloracne. Characterized by many blackheads, cysts,
pustules, papules, and abscesses.
6-4
-------
Blistering and scarring of the skin believed to be due
to increased light sensitivity, known as porphyria
cutanea tarda.
Darkened skin areas/blotches, known as hyperpigmenta-
tion.
Excessive growth of hair of normal or abnormal distri-
bution, known as hirsutism.
Internal
Liver damage
Raised serum hepatic enzyme levels
Disorders of fat metabolism
Disorders of carbohydrate metabolism
Cardiovascular disorders
Urinary tract disorders
Respiratory tract disorders
Pancreatic disorders
Neurological
Peripheral
Multiple sites of abnormal nerve degeneration
(polyneuropathies)
Sensory impairments (sight, hearing, smell, taste)
Central
Fatigue, weakness, impotence
Loss of sexual drive or libido
Psychiatric
Depression
The following additional responses have been observed in
monkeys and/or other nonhuman animals exposed to 2,3,7,8-TCDD:
1) A wasting syndrome manifested by a progressive weight
loss and decreased food consumption by treated animals.
2) Skin disorders, especially acneform eruptions or chlor-
acne, alopecia, edema (swelling), hyperkeratosis (thick-
ened and hardened, crusty skin), and hypertrophy of
Meibomian glands.
3) Progressive decline of the thymus gland (lymphoid
involution and atrophy).
6-5
-------
4) Porphyria resembling porphyria cutanea tarda (an abnor-
mal metabolism condition characterized by excess blood
metabolites—porphyrins—in the urine, and extreme
sensitivity to light).
5) Endocrine and reproductive disorders including abnormal
and toxic effects on the fetus (teratogenic and feto-
toxic effects).
6) Adverse effects on the immune system (immunotoxicity).
7) Abnormal induction of numerous enzymes.
Weight loss, chloracne, liver effects, and thymus atrophy
have also been observed in animals treated with 2,3,7,8-TCDF.
The two effects of exposure to TCDD's that occur at the lowest
doses and therefore represent the greatest concern to public
health are carcinogenic (10 ng/Kg/day) and reproductive effects.
6.2 POTENTIAL FOR DIOXIN AND DIBENZOFURAN FORMATION FROM THE
COMBUSTION OF WASTE OIL IN BOILERS
The PCDD's (and PCDF's) emitted to the air from combustion
processes are associated primarily with (i.e., sorbed to) air
particulate matter.
Several investigators have reported finding TCDD's and other
dioxins in both the particulate emissions and fly ash of munici-
pal incinerators. For example, fly ash from several such incin-
erators in Canada, Europe, and the United States has been found
Q
to contain the following range of dioxin concentrations:
Dioxin Concentration, ng/g
Total TCDD's 3.2 - 110
PeCDD's* 3.4 - 488
HxCDD's* 2.2 - 1200
* Pentachlorodibenzo-p-dioxin or hexachlorodibenzo-p-
dioxin.
6-6
-------
The formation of PCDD's during combustion of fuels (includ-
ing waste oil) is a controversial subject. One theory holds that
9 10
precursors (e.g., chlorinated phenols and chlorinated benzenes )
must be available in the fuel feed before PCDD's can be formed.
Dioxins can be formed from these precursors, as shown by the
following.
From chlorinated phenols:
Cl
From chlorinated benzenes:
02 . 620° C
Clr
2m
Another possible precursor is polyvinyl chloride, a compound that
produces chlorobenzenes during pyrolysis. '
Another theory, developed by Dow and others, proposes that
low concentrations of either inorganic or organic chlorides in
the fuel can be expected to produce traces of PCDD. This
hypothesis, which has been called "trace chemistries of fire," is
based on the following:
6-7
-------
(1) Combustion processes are seldom more than 99.9 percent
efficient in converting the carbon content of fuel to
carbon dioxide.
(2) The remaining 0.1 percent of the fuel is converted to a
large number of trace organic chemicals, including
chlorinated hydrocarbons, only a few of which have been
identified.
(3) Fossil fuels are extremely complex mixtures of many
chemicals and compounds, some of which are present at
low concentrations.
(4) The chlorine content of fuels can range from 1000 to
5000 ppm.
(5) Particles emitted from the combustion of oil contain
vanadium and nickel. These, together with silicon and
unburned carbon, can serve as catalysts in the combus-
tion process.
(6) Chemical reactions that occur in flames include pyrol-
ysis, oxidation, reduction, and acidolysis. Ions,
electrons, free radicals, and free atoms interact in a
continuously changing environment.
(7) In the Dow study, PCDD's were reportedly found in all
particulate matter taken from areas close to combustion
processes.
A significant criticism of the two theories discussed is
that often the fuel that is combusted is not well characterized
chemically. Thus, while dioxins may be detected in the combus-
tion products, it is usually not known whether dioxins are pre-
sent in the fuel itself; this makes it impossible to rule out
still another "theory": dioxins in, dioxins out. Although this
may not be true in all instances, it cannot be ignored.
Theories similar to those regarding PCDD's (i.e., precursor
and trace chemistries of fire hypotheses) can be developed for
the formation of PCDF's during combustion, with PCB as the major
14
precursor. This can be shown as follows:
6-8
-------
PCBs
CV-K
PCBPS
These theories can be used to predict that small quantities
of PCDD and PCDF may be formed during waste oil combustion. The
chemical reaction sequence involved in the combustion of organic
constituents present in waste oil is a complex process. This
sequence consists of a series of decomposition, polymerization/
and free radical reactions. The presence of chlorobenzene,
chlorinated phenols, and other chlorinated hydrocarbons (aromat-
ics and alphatics) appears to promote the formation of products
of incomplete combustion, such as PCDD and PCDF. These unwanted
byproducts of combustion have been detected in some fly ash and
flue gas samples from the combustion of waste oil, traditional
fossil fuels, and municipal refuse. The EPA has analyzed at
least seven coal-fired power plants, however, and has not detect-
ed any PCDD's or PCDF's, which suggests that the general contami-
nation of coal by PCDD's and PCDF's is not likely.
Analyses of waste oil samples have shown the presence of
inorganic and organic constituents that may serve as precursors
to PCDD and PCDF. Although data on chlorophenols are not
available, a recent composition survey of 49 waste oil samples
from a wide variety of sources showed a median phenol content of
18 ppm, and 90 percent of the samples had less than 110 ppm.
6-9
-------
It is not known, however, whether nonchlorinated phenols can
serve directly as dioxin precursers. Another set of 62 waste oil
samples had a median total chlorine value of 1400 ppm, and 90
percent of these samples had less than 6150 ppm. Other pre-
cursors for PCDD's (e.g., chlorinated benzenes) were not included
in the statistical analysis. Preliminary findings from a second
set of waste oil samples revealed detectable concentrations of
chlorobenzene in 5 out of 24 samples analyzed. The detectable
chlorobenzene concentrations ranged between 2 and 78 ppm.
Based on a set of 264 waste oil samples from a wide variety
of sources, the median PCB concentration was 9 ppm, and 90 per-
cent of the samples had less than 50 ppm PCB. The presence of
PCB, chlorine, and other chlorinated organics in waste oil,
creates a potential for formation of PCDF. Yields of PCDF were 1
14
to 5 percent during laboratory pyrolysis of known PCB isomers.
The presence of PCDF's was also detected in the ash from PCB
14
fires in Skovke, Sweden, and Stockholm, Sweden.
A PCDD concentration of 640 ppb and a TCDD concentration of
100 ppb were measured in fly ash from an industrial heating fa-
cility burning large quantities of waste oils. Unfortunately,
samples of the waste oil feed were not analyzed during these
tests.
More recent emission tests were performed at four boiler fa-
cilities burning waste oil spiked with trichlorophenol, trichlo-
robenzene, and other organic chlorides (possible PCDD precur-
18
sors). Species of PCDD's were detected in 6 of the 25 flue gas
6-10
-------
samples analyzed. The detectable concentrations of these com-
pounds ranged from 0.18 to 17 yg/m3. Bulk samples of fly ash
collected at one site contained four species of PCDD with detect-
able concentrations ranging between 33 and 230 yg/kg. Although
the results of these tests show that PCDD's were formed under the
test conditions, it is not known whether they would be formed
without the additions of the PCDD precursors.
The formation of PCDF's during waste oil combustion has not
been studied in detail. Data on the emission of PCDF's from
waste oil combustion are even more limited than the data for
PCDD's. GCA analyzed samples collected at four waste oil boiler
18
facilities for the presence of PCDF. The waste oil feed appar-
ently was not analyzed for PCB's. [The waste oil analysis did
include six organics (chloroform, trichloroethane, trichloro-
ethylene, tetrachloroethylene, trichlorobenzene, and dichloro-
naphthalene) and six inorganics (chlorine, arsenic, cadmium,
chromium, lead, and zinc).] Analyses of flue gas samples revealed
18
PCDF concentrations ranging between 0.07 and 62 yg/m3. Bulk
samples of fly ash collected at one site contained seven species
of PCDF with detectable concentrations in the range of 11 to 1000
., 18
yg/kg.
In summary, some limited evidence seems to suggest that
small amounts of PCDD's and PCDF's may be formed by the burning
of waste oils in some boilers. The temperatures in a typical
small, oil-fired, 15 million Btu/h boiler [i.e., a 4.4-MW boiler
6-11
-------
with a flue gas temperature of approximately 1315°C (2400°F)
19
exiting the combustion zone at 100 percent load ] are sufficient
to initiate a wide variety of reactions such as those suspected
of forming PCDD's and PCDF's. The destruction of these com-
pounds, however, depends on such factors as the location in the
combustion chamber where the compounds are formed, turbulence in
the combustion chamber, and the temperature profile of the cham-
ber. Further research and test data are required to attempt to
correlate the concentrations of PCDD's and PCDF's in the flue gas
with waste oil composition and boiler operating conditions.
6.3 AIR DISPERSION MODELING OF DIOXINS
Atmospheric transport of PCDD's and PCDF's from point-source
emissions can be predicted from dispersion modeling equations.
For example, when a chemical cloud containing small amounts of
TCDD's was accidentally emitted from a trichlorophenol manufac-
turing plant in Italy, it was experimentally demonstrated that
the TCDD deposition from air to soil followed an exponential
decay pattern along the downward wind direction. Therefore,
TCDD, other PCDD's, and PCDF's sorbed to particulate matter and
emitted from combustion sources can be expected to be dispersed
and deposited in the surrounding local area in a pattern reflect-
ing local wind patterns, stack height, and terrain.
Only limited test data are available on TCDD's emissions
from the combustion of waste oils. GCA reported measuring flue
18
gas concentrations of 1.4 ug/m3 in 1 of 25 samples. (This is
for all TCDD isomers, not just 2,3,7,8-TCDD.) The remaining 24
6-12
-------
flue gas samples fell below the detection limit of O.Io yg/m3.
The fly ash samples that GCA analyzed for TCDD's were also below
the detection limit of 0.5 yg/kg (ppb). In addition to the GCA
test data, TCDD concentrations of 100 ppb have been measured in
fly ash from an industrial facility burning waste oils.
These data can be used to estimate reasonable flue gas
concentrations for ambient air modeling of TCDD emissions. The
GCA flue gas data indicate TCDD concentrations generally fall
below 0.10 yg/m3, with high concentrations in the magnitude of
1.0 yg/m3.
This range can be substantiated by converting the reported
fly ash concentrations of 100 ppb into approximate flue gas
concentrations. Because particulate concentrations for the
industrial heating facility were not reported, the particulate
concentration was assumed to be 180 mg/dsm3, which is the allow-
able particulate emission rate from municipal and hazardous waste
incinerators.
Conversion of the 180 mg/dsm3 (standard temperature, 70°F,
and standard pressure, 1 atm) into an actual emission concentra-
tion (reported stack gas temperature of 392°F and pressure equal
to 1 atm) yields 113 mg/m3, assuming constant moisture content.
Based on a TCDD fly ash concentration of 100 ppb and assuming a
particulate concentration of 113 mg/m3, the estimated flue gas
TCDD concentration would be 0.011 yg/m3.
The test data and calculations just discussed indicate that
TCDD flue gas concentrations fall within an order of magnitude of
6-13
-------
0.01 and 1.0 yg/m3 and the majority of TCDD concentrations fall
below the detection limit of 0.10 ug/m3. Therefore, a TCDD
concentration range of 0.001 and 10 ug/m3 has been selected for
air quality modeling purposes. On a volume-per-volume basis,
this range corresponds to the following:
TCDD flue gas TCDD flue gas
concentrations, pg/m3 concentrations, ppt
10 1000
1 100
0.1 10
0.01 1
0.001 0.1
The selected range reflects the probable levels of TCDD concen-
trations in waste oil boiler flue gases.
Dispersion modeling was performed for dioxin emissions
resulting from the combustion of waste oil in space heaters and
small and medium-size boilers. A complete discussion and the
rationale of the modeling methodology and models selected for
this analysis are presented in Section 4.1 and Appendix B. This
analysis is based on the same assumptions concerning individual
boilers and source characteristics as those presented in Sections
3 and 4. The specific models used in this analysis were the
Industrial Source Complex (ISC) Model for short-range air impacts
and the Hanna-Gifford Model for urban-wide impacts. The specific
applications of these models were:
0 The short-range modeling concentrated on worst-case
estimates resulting from individual or multiple point
sources, assuming each source burned 100 percent waste
oil; concentrations of contaminants eliciting a
threshold response were of particular concern.
6-14
-------
0 The urban-wide modeling focused the area-wide disper-
sion of emissions from the burning of all waste oil
generated in that area; urban-wide impacts of contami-
nants eliciting a threshold response were of consider-
able concern.
6.3.1 Point Source Analysis
Individual or small group source modeling was performed to
estimate the possible air quality impact from a particular source
of dioxin emissions during the burning of 100 percent waste oil.
From a potential regulatory standpoint these estimates are of
interest in establishing emission limitations and waste oil feed
stock contaminant limitations. The ISC Model was used in con-
junction with meteorological data to estimate concentrations of
dioxins in the ambient air resulting from various size boilers
and space heaters.
Small Boilers and Space Heaters—
Table 4-2 (Section 4) presented the 13 point sources analyzed
in the ISC Model that represent typical small boilers and space
heaters (as discussed in Section 3). As the table shows, these
sources represented a range of boiler sizes, feed rates, and
stack heights. The invariability of other stack parameters was
justified by the fact that plume rise was small compared with
physical stack height.
Emissions of dioxins from small boilers were based on mea-
sured levels in fly ash from an industrial heating facility burn-
ing large quantities of waste oil. Because quantifiable levels
of dioxin emissions are not available over many types of small
boilers at variable feed rates, several levels of TCDD emissions
6-15
-------
in flue gas were analyzed, including 1000, 100, 10, 1, and 0.1
ppt. Table 6-1 presents the emission factors used throughout
this analysis of small boilers and space heaters.
Table 6—2 presents the results of modeling dioxins from the
burning of waste oil in individual small boilers and space heat-
ers for each source scenario described in Table 4-2. These
concentrations represent the maximum estimated for each single
source case. Obviously, as stack emissions decrease from the
1000 ppb level to the 0.1 ppt level, concentrations also de-
crease.
Medium-Size Boilers—
Some waste oil also may be burned in commercial/institutional
boilers with capacities of 50 to 100 x 10 Btu/h. An analysis
(similar to that described in Section 4.3.4) was performed with
the ISC Model to examine the air quality impacts of dioxin emis-
sions from waste oil burning in medium-size boilers. Typical
source characteristics chosen for this analysis are presented in
Table 4-6 (Section 4). Dioxin emissions resulting from the
burning of waste oil in these medium-size boilers are presented
in Table 6-3, and the results of the dispersion analysis are
presented in Table 6-4. As shown in Table 6-4, the maximum
concentrations for a given source occur at the 1000 ppt dioxin
level. Source 14 always has the maximum ground-level (ambient
air) concentration because it has the lowest stack height.
6-16
-------
TABLE 6-1. CONTAMINANT EMISSIONS USED IN SINGLE-SOURCE ISC
MODELING ANALYSIS: SMALL BOILERS AND SPACE HEATERS
TCDD flue gas
concentration,
ppt
1000
100
10
1
0.1
Source
waste oil
burn rates,
liters/h
4
19
57
132
346
4
19
57
132
346
4
19
57
132
346
4
19
57
132
346
4
19
57
132
346
Emissions at
100% Capacity,
•10 * yg/s
10.6
52.8
158.0
366.0
684.0
1.06
5.28
15.80
36.60
68.40
0.11
0.53
1.58
3.66
6.84
0.011
0.053
0.158
0.366
0.684
0.001
0.005
0.016
0.037
0.069
Emissions in
January at
50% capacity,
10-* ug/s
5.3
26.4
78.9
183.0
342.0
0.53
2.64
7.89
18.30
34.20
0.05
0.26
0.79
1.83
3.42
0.005
0.026
0.079
0.183
0.342
0.001
0.003
0.008
0.018
0.034
6-17
-------
TABLE 6-2. AMBIENT AIR IMPACT FROM WASTE OIL COMBUSTION IN INDIVIDUAL BOILERS
AND SPACE HEATERS (<15 x 106 Btu/h) FOR VARIABLE TCDD CONCENTRATIONS
Dioxin (TCDD)
flue gas
concentrations,
ppt
1000
100
10
1
0.1
o p
Maximum 30-day concentrations, v»g/m x 10
Source number
1
12.600
1.260
0.126
0.013
0.001
2
3.400
0.340
0.034
0.003
0.000
3
8.500
0.850
0.085
0.009
0.001
4
3.600
0.360
0.036
0.004
0.000
5
1.900
0.190
0.019
0.002
0.000
6
16.400
1.640
0.164
0.016
0.002
7
7.400
0.740
0.074
0.007
0.001
8
4.000
0.400
0.040
0.004
0.000
9
20.303
2.030
0.203
0.020
0.002
10
10.300
1.030
0.103
0.010
0.001
11
6.000
0.600
0.060
0.006
0.001
12
8.900
0.890
0.089
0.009
0.001
13
5.100
0.510
0.051
0.005
0.001
CTl
I
00
Identification in ISC Model.
-------
TABLE 6-3. DIOXIN EMISSIONS USED IN SINGLE-SOURCE ISC
MODELING ANALYSIS: MEDIUM-SIZE BOILERS
TCDD
concentration,
ppt
1000
100
10
1
0.1
Source
waste oil
burn rates,
liters/h
1330
2660
1330
2660
1330
2660
1330
2660
1330
2660
Emissions at
100% capacity,
10"* ug/s
3661
7320
366
732
36.6
73.2
3.7
7.3
0.4
0.7
Emissions in
January at
50% capacity,
10'* yg/s
1830
3661
183
366
18.3
36.6
1.8
3.7
0.2
0.4
TABLE 6-4. MAXIMUM 30-DAY DIOXIN CONCENTRATIONS IN AMBIENT AIR
AS A RESULT OF WASTE OIL COMBUSTION IN MEDIUM-SIZE BOILERS
(10~8 yg/m3)
TCDD
flue gas
concentration, ppt
1000
100
10
1
0.1
Source number
14
31.392
3.139
0.314
0.031
0.003
15
21.000
2.100
0.210
0.021
0.002
16
14.600
1.460
0.146
0.015
0.002
17
17.800
1.780
0.178
0.018
0.002
18
13.700
1.370
0.137
0.014
0.001
19
10.800
1.080
0.108
0.011
0.001
6-19
-------
6.3.2 Urban Scale Analysis
A dispersion modeling analysis based on the same assumptions
and methodology as presented in Section 4.4 was performed to
estimate the urban-wide impact of dioxins generated by waste oil
burning. Estimated dioxin concentrations in waste oil combustion
emissions were used to estimate dioxin emissions per liter of
fuel burned (Table 6-5). These dioxin emissions were used in
conjunction with estimates of waste oil burned in each subarea of
the urban study area to calculate the rate of emissions at each
dioxin concentration level in waste oil.
Emissions were estimated for a 30-day averaging period [con-
sistent with the short-term (30-day) maximum waste oil impacts]
and for an annual averaging period (consistent with long-term
impacts from nonthreshold substances. Tables 6-6 and 6-7 present
estimates of dioxin emissions over each subarea at each dioxin
level in the waste oil combustion effluents.
The results of applying the Hanna-Gifford Model over the
study area (Tables 6-6, 6-7, and 6-8) show that 30-day average
concentrations are higher than those averaged over a whole year.
These higher concentrations result from the higher emission rates
averaged over a winter month rather than over a whole year (be-
cause emissions are low in the summer).
6.4 RISK ASSESSMENT
The risk associated with human exposure to dioxins in par-
ticulate matter resulting from combustion is not yet known. The
a
National Research Council of Canada has estimated that limiting
6-20
-------
TABLE 6-5. DIOXIN EMISSION RATES PER WASTE OIL BURNED
USED IN URBAN SCALE MODELING
TCDD
concentration
in flue gas,
ppt
1000
100
10
1
0.1
TCDD
concentration
in flue gas,
yg/m3
10
1.0
0.1
0.01
0.001
Emissions per waste
oil burned,
ng/liter
1.0
0.10
0.01
0.001
0.0001
TABLE 6-6. AREA SOURCE 30-DAY-AVERAGED EMISSIONS FOR TCDD
IN WASTE OIL (JANUARY)
(10"17 g/s-m2)
Distance from
TCDD flue gas concentration, ppt
centroid, km
0-5
5-10
10-15
15-25
25-50
1000
1360.0
548.0
465.4
168.0
17.2
100
136.00
54.80
46.54
16.80
1.72
10
13.600
5.480
4.654
1.680
0.170
1
1.360
0.548
0.465
0.168
0.017
0.1
0.13
0.055
0.047
0.017
0.002
TABLE 6-7. AREA SOURCE ANNUAL-AVERAGED EMISSIONS FOR TCDD IN WASTE OIL
(10~17 g/s-m2)
Distance from
TCDD flue gas concentration, ppt
centroid, km
0-5
5-10
10-15
15-25
25-50
1000
552.3
222.0
188.0
70.9
7.3
100
55.23
22.20
18.80
7.09
0.73
10
5.523
2.220
1.880
0.709
0.073
1
0.552
0.222
0.188
0.071
0.007
0.1
0.055
0.022
0.019
0.007
0.001
6-21
-------
TABLE 6-8. AVERAGE ANNUAL TCDD CONCENTRATIONS FOR WASTE OIL
BURNING IN THE URBAN AREA
(10"B yg/m3)
TCDD
flue gas
roncentration .
ppt
1000
100
10
1
0.1
Distance from study area centroid, km
0-5
30.714
3.071
0.307
0.031
0.003
5-10
9.643
0.964
0.096
0.010
0.001
10-15
8.214
0.821
0.082
0.008
0.001
15-25
3.214
0.321
0.032
0.003
0.0003
25-50
0.357
0.036
0.004
0.0004
<0.0001
6-22
-------
ingestion of TCDD's to no more than 2.1 to 6.3 pg/day should be
sufficient to reduce the cancer risk to the 10 risk level
(i.e., 1 cancer in 1,000,000 people). The virtually safe dose
for HxCDD's is estimated at about twice that calculated for
TCDD's. The EPA's Carcinogen Assessment Group has estimated
that lifetime oral exposure to 1.6 pg/day TCDD's may pose a human
cancer risk of 10 (1 in 100,000 people). This is the best
estimated value currently available for estimating a reference
concentration. Assuming that the body uptake by inhalation
corresponds to uptake by oral exposure, and assuming a daily
breathing volume of 20 m3 (Appendix E, Equation 2), the reference
concentration for TCDD's is:
(1'6 day*(2Q1m3) = °'08 pg/m3 = 8 x 10~8 y9/m3
Tables 6-1 through 6-8 presented the results of air disper-
sion modeling. The point-source modeling of small boilers
(Sources 1 through 11), space heaters (Sources 12 and 13), and
medium-size boilers (Sources 14 through 19) was based on TCDD
concentrations in the particulate emissions of 1000, 100, 10, 1,
and 0.1 parts per trillion from waste oil burning. The available
data suggest that actual values are in the middle of this range.
This section presents an estimate of cancer risk from exposure to
the modeled point-source emissions.
It should be noted that for all the other waste oil contami-
nants modeled (metals and organics), cancer risk was calculated
only for urban-scale (area) modeling and not for point-source
modeling. Health effects (and therefore risk) from nonthreshold
6-23
-------
substances should be calculated for long-term exposure" rather
than for short-term (30-day average) maximums around a point
source. The point-source modeling was added to this section on
dioxins, however, because of the widespread attention now being
focused on dioxins.
Table 6-9 presents the maximum ground-level (ambient air)
concentrations of dioxin (TCDD) obtained in the point-source and
urban area modeling. Of the 19 point sources modeled, the two
resulting in the greatest ambient air impacts (Sources 9 and 14,
a small and a medium-size boiler) are included in the risk assess-
ment in Table 6-9. Also included are the modeled annual concen-
trations from the innermost ring of the urban area (0 to 5 km),
the portion with the greatest ambient air impact.
As the modeling results show, ground-level concentrations
from the small boiler, from the medium-size boiler, and within
the urban area core are all comparable in magnitude. Variation
resulting from the dioxin concentration in the particulate emis-
sions is far greater than that resulting from the source configu-
ration. The uncertainty of selecting numbers for the dioxin
concentration results in comparable uncertainty in estimating
risk.
The last two columns in Table 6-9 show the cancer risk ex-
pressed both as a risk level to a population as a whole (i.e.,
2.54 x 10~ ) and as a risk to the individual (i.e., 1 in 39,400
or 1:39,400). The cancer risk number is obtained by dividing the
ground-level concentration by the reference concentration, and
6-24
-------
TABLE 6-9. ESTIMATED RISK FROM DIOXIN EMISSIONS
Modeled source
Single-source small boiler
(Source 9)
Single-source medium-size
boiler (Source 14)
Urban-scale area source
(annual concentrations
in 0-5 km area)
Dioxin (TCDD)
flue gas
concentrations,
PPt
1000
100
10
1
0.1
1000
100
10
1
0.1
1000
100
10
1
0.1
Maximum
ground level
concentration,
10"8 yg/m3
20.303
2.030
0.203
0.020
0.002
31.393
3.139
0.314
0.031
0.003
30.714
3.071
0.307
0.031
0.003
Cancer risk
2.54 x 10';?
2.54 x 10"?
2.54 x 10";
2.54 x 10"°
2.54 x 10"y
3.93 x 10"!?
3.93 x 10"H
3.93 x 10";
3.93 x 10"°
3.93 x 10"y
3.84 x 10~;>
3.84 x 10"°
3.84 x 10";
3.84 x 10"°
3.84 x 10~9
Lifetime cancer
risk to an
individual
1:39,400
1:394,000
1:3,940,000
1:39,400,000
1:394,000,000
1:25,400
1:254,000
1:2,540,000
1:25,400,000
1:254,000,000
1:26,000
1:260,000
1:2,600,000
1:26,000,000
1:260,000,000
I
N)
Ul
-------
multiplying that number by 10 (because the reference concen-
tration is based on a risk level of 10~ ). Its inverse is the
lifetime cancer risk to an individual.
Cancer risks from the modeled dioxin ground-level concentra-
-9 -5
tions range from 10 to 10 or one cancer in 254 to 394 million
to one in 25,400 to 39,400. The dioxin concentrations at the
lower end of the modeled range result in risks that are at ac-
ceptable levels according to criteria most commonly applied to
risk assessment (i.e., risks less than 1 cancer in a million).
Concentrations in the upper end of the modeled range, however, do
result in risks that are potentially significant (i.e., in the
area of 10~ and 10~ ). The results cannot be accurately quanti-
fied to any greater precision at this time because very limited
data are available.
These results should be interpreted with the modeling limi-
tations in mind. Several of the assumptions could contribute to
erring on the safe side, i.e., overestimating the risk. The
urban-area modeling assumes that dioxins are formed and emitted
in 100 percent of the sources. Actual formation varies with pre-
cursors in the waste oil, with combustion time, and with tempera-
ture. It is likely, however, that dioxins are not always formed.
If they were formed 75 percent of the time rather than 100 per-
cent of the time, concentrations would be 25 percent less than
those shown in Table 6-8. The corresponding risks would decrease
to 2.9 x 10~5, 2.9 x 10"6, 2.9 x 10~7, 2.9 x 10~8, and 2.9 x 10~9
for the five concentrations modeled. In terms of risk, the
6-26
-------
reduced concentrations do not change the order of magnitude of
risk. If TCDD's were formed 10 percent of the time, the concen-
trations would be only 10 percent of those shown in Table 6-8.
The corresponding risks would decrease for the four urban scale
modeling concentrations by one order of magnitude, and range from
3.84 x 10~6 to 3.84 x 10~10.
Finally, it should be kept in mind that the numbers included
in Table 6-9 are for the greatest concentrations calculated in
the models. Much lower concentrations result in slightly lower
risks. The methodology used to generate the figures involves
"upper limit" numbers; i.e., the risks are not likely to be any
higher than these figures, and they may be much lower. Suffi-
cient data are not available at this time to improve the range of
risk estimates.
In summary, the modeling indicates that if the underlying
assumptions are correct, cancer risk from dioxin formation during
combustion of waste oil may be potentially significant in some
cases. The data base for the assumptions is extremely limited,
however.
6-27
-------
REFERENCES FOR SECTION 6
1. Esposito, M. P., T. 0. Tiernan, and F. E. Dryden. Dioxins.
EPA 600/2-80-197, 1980.
2. Reggiani, G. Toxicology of TCDD and Related Compounds. In:
Workshop on the Impact of Chlorinated Dioxins and Related
Compounds in the Environment, Rome, Italy, October 22-24,
1980.
3. Van Miller, J. J. Lalich, and J. R. Allen. Increased Inci-
dence of Neoplasms in Rats Exposed to Low Levels of 2,3,7,8-
Tetrachlorodibenzo-p-dioxin. Chemosphere, 6:537-544, 1977.
4. Kociba, R. J., et al. Results of a Two-Year Chronic Toxic-
ity and Oncogenicity Study of 2 ,3,7 ,8-Tetrachlorodibenzo-p-
dioxin (TCDD) in Rats. Toxicology and Applied Pharmacology,
46:279-303, 1978.
5. Homberger, E., et al. The Seveso Accident: Its Nature,
Extent and Consequences. Annals of Occupational Hygiene,
22:3272-80, 1979.
6. U.S. Environmental Protection Agency. PCB Disposal by
Thermal Destruction. Appendix H: Consideration of Risks
and Benefits/Alternatives Concerning PCB Incineration in
Region VI. EPA-906/9-82-003, PB82-241860, 1982.
7. National Research Council of Canada (NRCC). Polychlorinated
Dibenzo-p-dioxins: Limitations to Current Analytical Tech-
niques. Publication No. 18576. NRCC/NRC, Ottawa, Canada.
1981.
8. National Research Council of Canada. Polychlorinated
Dibenzo-p-dioxins: Criteria for Their Effects on Man and
His Environment. Publication No. 18574, 1981.
9. Rappe, C., et al. Formation of Polychlorinated Dibenzo-p-
dioxins (PCDD's) and Dibenzofurans (PCDF's) by Burning or
Heating Chlorophenates. Chemosphere, 7:269-281, 1978.
10. Buser, H. R. Formation of Polychlorinated Dibenzofurans
(PCDF's) and Dibenzo-p-dioxins (PCDD's) From the Pyrolysis
of Chlorobenzenes. Chemosphere, 8:415-424, 1979.
6-28
-------
11. lida, I., M. Nakoshini, and R. Goto. Evolution of Aromatics
on Pyrolysis of Poly(vinylchloride) and Its Mechanism. J.
Polym. Sci. Chem. Ed., 12:737, 1964.
12. O'Mara, M. M. J. Polym. Sci., Al(8):1887, 1970.
13. Cruiranett, W. G. Environmental Chlorinated Dioxins From
Combustion—The Trace Chemistries of Fire Hypothesis.
Michigan Division Analytical Laboratories, Dow Chemical USA,
Midland, Michigan. Pergamon Press. 1982.
14. Rappe, C., et al. Polychloronated Dioxin (PCDDs), Dibenzo-
furans (PCDFs) and Other Polynuclear Aromatics (PCPNA's)
Formed During PCB Fires. Chemica Scripta, 20:56-61, 1982.
15. Franklin Associates Limited and PEDCo Environmental, Inc.
Survey of the Waste Oil Industry and Waste Oil Composition.
(Draft report) April 1983.
16. Personal communication from B. Bider, Franklin Associates
Limited, October 3, 1983.
17. Buser, H. R., H-P. Bosshardt, and C. Rappe. Identification
of Polychlorinated Dibenzo-p-dioxin Isomers Found in Fly
Ash. Chemosphere, 7:165-172, 1978.
18. GCA Corporation. Draft Data Summary Waste Oil Incineration
Study - Sites A, C, E, F. GCA 1-619-068, June 1983.
19. Castaldini, C., et al. A Technical Overview of the Concept
of Disposing of Hazardous Wastes in Industrial Boilers.
Prepared by Acurex Corporation, Mountain View, California,
for U,.S. Environmental Protection Agency, Cincinnati, Ohio.
January 1981. pp. 5-36, 5-37.
20. Arthur D. Little, Inc. Study on State of the Art of Dioxin
From Combustion Sources. Prepared for American Society of
Mechanical Engineers. New York, New York. 1981.
21. U.S. Environmental Protection Agency. Health Assessment
Document for Dioxins. Research and Development Peer Review
Workshop Draft document prepared by EPA Environmental Cri-
teria and Assessment Office, Cincinnati, Ohio. ECAO-CIN-
302A, July 1983.
22. U.S. Environmental Protection Agency. Health and Environ-
mental Effects Profile for Tetra-, Penta- and Hexachlorodi-
benzo-p-dioxins. Research and Development Peer Review Draft
Document prepared by Environmental Criteria and Assessment
Office, Cincinnati, Ohio. ECAO-CIN-P004, July 1983.
6-29
-------
APPENDIX A
OIL-FIRED BOILER CHARACTERIZATION
A-l
-------
APPENDIX A
OIL-FIRED BOILER CHARACTERIZATION
One of the approaches to the regulation of waste oil is to
limit the number, types, and locations of boilers allowed to
burn waste oil as fuel. For example, one possibility would be
to allow the use of waste oil as fuel only in oil-fired utility
boilers equipped with air pollution control devices. Since this
approach is under consideration, it is necessary to gather some
information on numbers, types, and locations of boilers (partic-
ularly utility, industrial, and commercial/institutional boil-
ers) . The figures and tables in this appendix summarize the
data gathered on boilers in these three classes.
Utility Residual-Oil-Fired Boilers
Table A-l summarizes the utility boiler population for
which residual oil is the primary fuel in 10 states (California,
Florida, Illinois, Louisiana, New Jersey, New York, Ohio,
Pennsylvania, Texas, and Virginia). The source of data was
"Power Directory, 1981, An Environmental Directory of U.S. Steam
Electric Power Plants." This directory was prepared for the
Edison Electric Institute by the Utility Data Institute, Inc.
The data presented in the "Power Directory" are based on the
Edison Electric Institute's POWER Data Base as of January 1981.
A-2
-------
TABLE A-l. POPULATION OF UTILITY BOILERS WITH RESIDUAL OIL
AS THE PRIMARY FUEL IN URBAN AND RURAL AREAS OF TEN STATES
Urban I
Rural I
Urban II
Rural II
Totals
Number
of boilers
257 (66.4)
130 (33.6)
357 (92.2)
30 (7.8)
387 (100.0)
Capacity, MWea
44,460.1 (64.8)
24,197.9 (35.2)
63,842.9 (93.0)
4,815.1 (7.0)
68,658.0 (100.0)
Number of
boilers with
particulars
controls
91C (35.4)
57d (43.8)
141e (39.5)
7 (23.3)
148 (38.2)a
Capacity of
boilers with
particulate ^
controls, MWe
14,717.2C (33.1)
13,524.4d (55.9)
26,134.6e (40.9)
2,107.0 (43.8)
28,241.6 (41.1)a
Numbers in parentheses are percentages of ten-state totals.
Numbers in parentheses are percentages of group totals.
c Forty-one boilers with a capacity of 8,483.3 MWe are believed to have
electrostatic precipitators.
Fifteen boilers with a capacity of 2,197.8 MWe are believed to have electro-
static precipitators.
e Fifty-six boilers with a capacity of 10,681.1 MWe are believed to have
electrostatic precipitators.
A-3
-------
The 1980-1981 "Electrical World Directory of Electric Utilities"
provided supplemental information on the location of some power
2
plants.
The data on the boiler populations of these 10 states were
reviewed. For each unit in which residual or No. 6 fuel oil was
indicated as the primary fuel, the following data were compiled:
company, plant name, unit number, county, town, capacity (MWe),
and pollution control device information.
Each plant was categorized into one of four groups. Any
city listed in the "1977 Census of Manufacturers, Volume III,
Geographic Area Statistics, Parts 1 and 2, General Summary,"
which provides manufacturing data for cities with 450 manufac-
turing employees or more, was considered appropriate for our
classification. State maps contained in the 1977 Census of
Manufacturers were used to determine whether a plant was located
in a Standard Metropolitan Statistical Area (SMSA). The four
groups defined for this effort are:
Group A - In an SMSA and in a city
Group B - In an SMSA but not in a city
Group C - In a city outside of an SMSA
Group D - Not in a city or an SMSA
Further categories were established to indicate the number
of plants in urban or rural areas. In the first of two defini-
tions, it was decided that the Urban I classification would
include all plants located within a city; thus, Urban I includes
the plants in Group A plus Group C, and Rural I includes the
A-4
-------
plants in Groups B and D. In the second broader definition, it
was decided that Urban II would include all plants in cities or
SMSA's; thus, Urban II includes the plants in Groups A, B, and
C, and Rural II includes only Group D.
Table A-l summarizes the total population of utility boil-
ers using residual oil as primary fuel in the 10 states exam-
ined. The 10-state population of residual oil utility boilers
is 387, representing a capacity of 68,658.0 MWe. Of this total,
148 boilers (41 percent of the total capacity) are equipped with
particulate control devices. Two hundred and fifty-seven boil-
ers (nearly 65 percent of the total capacity) fall into the Ur-
ban I classification, and particulate emissions from 33 percent
of this capacity are controlled by some type of device. In the
broader Urban II classification, there are 357 boilers (92% of
the total), representing a capacity of 63,842.9 MWe (93% of the
total); particulate emissions from 40.9 percent of this capacity
are controlled. Thus, nearly all (92%) of the oil-fired utility
boilers are located in a city or SMSA. Less than half of these
boilers are equipped with pollution control devices. Of the
total population, only 56 boilers (14% of the number), represent-
ing a capacity of 10,681.1 MWe, appear to have electrostatic
precipitators. Many of these boilers are probably former coal-
fired units that have been converted to residual-oil-fired
units.
Industrial Residual-Oil-Fired Boilers
The population of industrial boilers is presented graphi-
cally in Figures A-l and A-2. Nearly 90 percent of the total
A-5
-------
100,000
10.000
1,000
cc
—1
o
CO
LL.
o
at
UJ
* 100
10
40.714
-
-
45,633
23.649
—
5.774
4.216
1,654
1,039
261
56
8 "
BOILER HEAT INPUT CAPACITY RANGES, 10° Btu/h
Figure A-l. Number of industrial, residual-oil-fired boilers.
A-6
-------
1OU
150
140
130
^120
3
£ no
n
0
"* 100
*
5 90
a.
3 80
5
z 70
| 60
|
40
30
20
10
.
-
_
-
-
-
-
1
1
1
4
0,257
2,432
17,846
S
8,653
54,09
? 1
1
?1,650
52,88
0
-
_
-
8
6,830
-
-
42,660
;
19,200 .
-
0-0.4 0.4-1.5 1.5-10 10-25 25-50 50-100 100-250 250-500 500-1,500 >1,500
BOILER HEAT INPUT CAPACITY RANGES, 106 Btu/h
Figure A-2. Heat input capacity of industrial, residual-oil-fired boilers.
-------
number of industrial boilers have a heat input capacity of less
than 10 million Btu/h. As shown in Figure A-2, however, this
group represents only 20 percent of the total industrial boiler
capacity. Fifty percent of the capacity is generated by boilers
in the 25 to 250 million Btu/h range, and 82 percent of the
capacity is generated by boilers with heat input capacities of
less than 250 million Btu/h. Tables A-2 through A-7 provide
detailed data on the numbers and capacities of the various sized
industrial boilers by type (i.e., water-tube, fire-tube, etc.).
Because of their large number, characterizing the distribu-
tion of industrial boilers according to urban vs. nonurban
locations would be a formidable and time-consuming task. The
number of hours required to complete such an inventory was
beyond the scope of this task; therefore, we looked for another
parameter that could be used to indicate fuel oil use (and
ultimately boiler distribution) by capacity. The parameters
considered for estimating boiler distribution were 1) number of
manufacturing employees, 2) number of manufacturing establish-
ments, and 3) number of production employees.
The first parameter includes sales and management employees
as well as workers, and the second parameter includes brokers or
sales establishments as well as actual manufacturers. There-
fore, it was decided that the third parameter, number of produc-
tion employees, would be the least biased of the three because
it includes only those employees actually involved with produc-
tion. The number of production employees was believed to be
A-8
-------
TABLE A-2. THE 1977 POPULATION OF INDUSTRIAL WATER-TUBE
BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
11,872
0
744
2,443
2,021
'3,616
1,654
1,039
261
56
8
Capacity, 106 Btu/h
601,160
920
12,540
32,860
131,620
121,650
152,880
86,830
42,660
19,200
A-9
-------
TABLE A-3. THE 1977 POPULATION OF INDUSTRIAL SCOTCH
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
18,268
0
9,445
6,955
1,580
288
0
0
0
0
0
Capacity, 10 Btu/h
85,805
9,455
37,867
27,704
10,789
A-10
-------
TABLE A-4. THE 1977 POPULATION OF INDUSTRIAL FIREBOX
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
24,171
0
15,833
6,682
1,368
288
0
0
0
0
0
Capacity, 10 Btu/h
87,283
15,831
36,671
23,992
10,789
A-ll
-------
TABLE A-5. THE 1977 POPULATION OF INDUSTRIAL HRT
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
5,371
0
2,953
1,731
687
0
0
0
0
0
0 '
Capacity, 106 Btu/h
24,873
2,953
9,907
12,013
A-12
-------
TABLE A-fi. THE 1977 POPULATION OF OTHER INDUSTRIAL
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
3,428
0
2,305
981
118
24
0
0
0
0
0
Capacity, 106 Btu/h
10,678
2,304
5,391
2,084
899
A-13
-------
TABLE A-7. THE 1977 POPULATION OF INDUSTRIAL CAST IRON
BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
59,894
40,714
14,323
4,857
0
0
0
0
0
0
0
Capacity, 106 Btu/h
36,706
10,257
10,979
15,470
A-14
-------
roughly indicative of the percentage of fuel burned in urban and
nonurban areas. These numbers were taken from the Census of
Manufacturers, which lists people according to where they work,
rather than from the Census of Population, which lists people
according to where they live.
Another note of explanation is necessary. The Census of
Manufacturers defines a city as a political division with 450 or
more manufacturing employees. This very inclusive definition
results in many relatively small towns being defined as cities.
For this reason, we have used a combination of that definition
and the Bureau of Census definition of a city (i.e., population
of 25,000 or more). Thus, a city is defined as having a popula-
tion of 25,000 or more and 450 or more manufacturing employees.
Table A-8 summarizes the number of production workers in
3 4
urban and nonurban areas for selected states. ' As requested
by EPA, we have included two definitions of an urban area.
Urban I includes all cities having a population greater than or
equal to 25,000 and at least 450 manufacturing employees; urban
II includes all SMSA's and those cities (as defined above) not
located in SMSA counties.
As shown in Table A-8, the number of production workers in
urbanized areas varies significantly from state to state.
California, New York, and Pennsylvania are predominantly urban
in character; whereas Maine, Vermont, Georgia, and South Caro-
lina are predominantly nonurban.
In the 22 states investigated, approximately 43.7 percent
A-15
-------
TABLE A-8. SUMMARY OF PRODUCTION WORKERS IN URBAN AND NONURBAN AREAS.
STATE
ALABAflA
CALIFORNIA
CONNECTICUT
DELAWARE
FLORIDA
6EOR6IA
ILLINOIS
LOUISIANA
KAINE
MARYLAND
MASSACHUSETTS
NEK HAMPSHIRE
NEK JERSEY
NEK YORK
N. CAROLINA
OHIO
PENNSYLVANIA
RHODE ISLAND
S. CAROLINA
TEIAS
VERMONT
VIR6INIA
N. VIRGINIA
TOTAL
PRODUCTION
WORKERS
IN STATE
273.00
1142.60
253.30
32.30
249.20
376.20
857.80
145.20
84.20
162.90
407.90
72.00
489.60
958.10
611.30
924.40
934.10
94.70
299.80
600.70
28.70
302.00
89.20
9391.20
URBAN
i OF PRODUC-
TION WORKERS
86.10
783.30
106.80
3.90
109.00
66.90
456.30
49.90
11.40
58.40
224.20
23.40
178.80
.555.80
123.70
398.20
282.80
60.20
37.00
360.30
0.00
107.40
21.30
4105.10
I3
PERCENT
31.54
68.55
41.83
12.07
43.74
17.78
53.19
34.37
13.54
35.85
54.96
32.50
36.52
58.01
20.24
43.08
30.28
63.57
12.34
59.98
0.00
35.56
23.88
43.71
RURAI
t OF PRODUC-
TION WORKERS
186.90
359.30
148.50
28.40
140.20
309.30
401.50
95.30
72.80
104.50
183.70
48.60
310.80
402.30
487.60
526.20
651.30
34.50
262.80
240.40
28.70
194.60
67.90
5286.10
PERCENT
66.46
31.45
58.17
87.93
56.26
82.22
46.81
65.63
86.46
64.15
45.04
67.50
63.48
41.99
79.76
56.92
69.72
36.43
87.66
40.02
100.00
64.44
76.12
56.29
URBAN ]
I QF PRODUC-
TION WORKERS
150.40
1091.90
236.00
20.70
224.00
154.60
736.00
94.10
22.90
128.80
362.50
31.10
477.70
857.40
260.00
753.80
759.30
92.90
139.40
509.10
0.00
153.40
56.50
7312.50
:ib
PERCENT
55.09
95.56
92.44
64.09
89.89
41.10
85.80
64.81
27.20
79.07
88.87
43.19
97.57
89.49
42.53
81.54
81.29
98.10
46.50
84.75
0.00
50.79
63.34
77.87
RURAL
t OF PRODUC-
TION WORKERS
122.60
50.70
19.30
11.60
25.20
221.60
121.80
51.10
61.30
34.10
45.40
40.90
11.90
100.70
351.30
170.60
174.80
1.80
160.40
91.60
28.70
148.60
32.70
2078.70
II
PERCENT
44.91
4.44
7.56
35.91
10.11
58.90
14.20
35.19
72.80
20.93
11.13
56.81
2.43
10.51
57.47
18.46
18.71
1.90
53.50
15.25
100.00
49.21
36.66
22.. 13
PRODUCTION WORKERS IN CITIES WITH A POPULATION OF 25,000 OR MORE AND 450 NANUFACTURIN6 EMPLOYEES
b
PRODUCTION WORKERS IN SRSA'S AND CITIES (AS DEFINED ABOVE) NOT LOCATED IN SHSA COUNTIES
A-16
-------
of all production workers are located in cities having a popula-
tion of 25,000 or more and at least 450 manufacturing employees,
and 56.3 percent are in rural areas. In the second definition
of urban areas (SMSA's and cities not in SMSA counties), about
77.9 percent of all production workers are in urban areas and
22.1 percent in rural areas. Thus, assuming that boiler capaci-
ty distribution is directly proportional to the number of pro-
duction workers, the boilers generating a little more than
three-quarters of the industrial boiler capacity are located in
cities or SMSA's; the remainder are in rural locations.
Commercial/Institutional Residual-Oil-Fired Boilers
Commercial/institutional oil-fired boilers are those used
in hospitals, greenhouses, shopping malls, and similar applica-
tions. The very nature of this type of boiler makes their dis-
tribution correlate well with population. No attempt was made
to locate actual commercial/institutional boilers and categorize
them as being in an urban or rural location.
Figures A-3 and A-4 present data on the distribution of
commercial/institutional boilers by number and by heat input
capacity. Most commercial boilers are small: 88 percent (by
number) have heat input capacities of less than 1.5 million
Btu/h, and 98 percent have heat input capacities of less than 10
million Btu/h. Much of the heafinput capacity of commercial/
institutional boilers (i.e., about 25 percent) is found in the
1.5 to 10 million Btu/h range. About half of the total generat-
ed capacity is represented by boilers with heat input capacities
of less than 10 million Btu/h. Tables A-9 through A-14 provide
A-17
-------
100.000
10,000
1,000
100
10
162,855
-
_
73,421
25,362
-
2.029
1,639
-
551
198
-
39
6
0-0.4 0.4-1.5 1.5-10 10-25 25-50 50-100 100-250 250-500500-1,500
•OILER HEAT INPUT CAPACITY RANGE. 106 Stu/h
Figure A-3. Number of commercial/institutional, residual-oil-fired boilers.
A-18
-------
6T-V
TOTAL HEAT INPUT CAPACITY, 109 Btu/h
n>
fO
01
o
O»
•o
Ql
O
_*•
4?
o
fO
o
CU
^ rvj
O V1
tn
m
to
O
3
O)
o.
c
o>
I
o
O
a>
cr
o
re
*~rN}u»^m0s-*jCD
^3 ^y ^3 1^1 c^ Q ^3 ^3
S 8
-------
TABLE A-9. THE 1977 POPULATION OF COMMERCIAL WATER-TUBE
BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
4,081
0
399
772
710
1,406
551
198
39
6
0
Capacity, 10 Btu/h
154,540
480
3,960
11,540
51,180
40,550
29,120
12,970
4,740
A-20
-------
TABLE A-1.0. THE 1977 POPULATION OF COMMERCIAL/INSTITUTIONAL SCOTCH
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
7,729
0
4,866
2,196
555
112
0
0
0
0
0
Capacity, 10 Btu/h
30,753
4,865
11,958
9,734
4,196
A-21
-------
TABLE A-11. THE 1977 POPULATION OF COMMERCIAL/INSTITUTIONAL FIREBOX
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
10,589
0
8,156
2,110
481
112
0
0
0
0
0
Capacity, 106 Btu/h
32,362
8,156
11,580
8,430
4,196
A-22
-------
TABLE A-12. THE 1977 POPULATION OF COMMERCIAL/INSTITUTIONAL HRT
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
2,309
0
1,522
546
241
0
0
0
0
0
0
Capacity, 106 Btu/h
8,871
1,522
3,128
4,221
A-23
-------
TABLE A-13. THE 1977 POPULATION OF OTHER COMMERCIAL/INSTITUTIONAL
FIRE-TUBE BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
1,548
0
1,187
310
42
9
0
0
0
0
0
Capacity, 106 Btu/h
3,971
1,187
1,702
732
350
A-24
-------
TABLE A-14. THE 1977 POPULATION OF COMMERCIAL/INSTITUTIONAL
CAST IRON BOILERS FIRING RESIDUAL OIL
Class population
Distribution by capacity ranges, 10 Btu/h
0 to 0.4
Over 0.4 to 1.5
Over 1.5 to 10
Over 10 to 25
Over 25 to 50
Over 50 to 100
Over 100 to 250
Over 250 to 500
Over 500 to 1,500
Over 1,500
Number
239,574
162,855
57,291
19,428
0
0
0
0
0
0
0
Capacity, 10 Btu/h
146,821
41,029
43,914
61,878
A-25
-------
more detailed data on the numbers and capacities of various
sizes of commercial/institutional boilers by type of boiler
(i.e., water-tube, fire-tube, etc.).
Geographical Distribution of Residual-Oil-Fired Boilers
Although it was not an objective of this study to survey
small boilers, it is useful to provide some indication of the
relative potential use of waste oil across the Nation. Two
approaches were considered for obtaining an estimate of the
small boiler distribution. The first approach was to base the
distribution on residual fuel oil deliveries to the commercial/
institutional sector by state. These data are available from
the U.S. Department of Energy. This approach indicated that
three states (New York, New Jersey, and Louisiana) accounted for
54 percent of the fuel oil deliveries. This large percentage
results from the DOE figures reflecting fuel deliveries rather
than fuel use. States with large ports suitable for oil barges
would naturally show the largest amounts of fuel delivered.
This obvious bias eliminated the use of fuel oil delivery data
from further consideration.
A second approach was based on data obtained from a survey
of states and waste oil collectors/processors. Franklin Associ-
ates Ltd. conducted the survey and used the information obtained
to estimate the amount of waste oil generated by each state and
the approximate portion of the generated waste oil that is
burned as fuel. These numbers are shown in Table A-15. Based
on the assumption that about 50,000 small boilers in the country
A-26
-------
TABLE A-15. AN ESTIMATE OF THE DISTRIBUTION OF SMALL BOILERS
(<15 million Btu/h) WITH THE POTENTIAL FOR BURNING WASTE OIL
State
Alabama
Alaska
Arkansas
Arizona
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
•Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
Nevada
Mew Hampshire
New Jersey
New Mexico
New York
North Carolina
North Dakota
Ohio
Oklahoma
Oregon
Pennsylania
Rhode Island
South Carolina
South Dakota
Tennessee
Texas
Utah
Vermont
Virginia
Washington
Washington, D.C.
West Virginia
Wisconsin
Wyomi ng
Total
Waste oil
generated,
1000 gallons
14,500
1,400
10,600
9,100
74.500
9,600
6,900
2,000
18,800
18,600
1,600
2,800
51,800
25,500
12,600
15,600
15,400
29,100
3,000
10,600
13,000
48,200
17,300
9.600
24,300
3,800
9,100
1,700
1,400
34.500
4,900
31,400
18,100
2,900
54,500
20,500
11,500
58,700
2,000
7,800
2,800
18,500
78,200
4,300
1.100
14,100
10,800
1,100
9,400
15,300
3,100
867.900
Proportion
burned
per state. I
60
30
80
70
60
70
30
90
80
60
60
30
70
30
40
80
70
80
80
90
90
80
70
70
80
40
80
70
90
80
70
80
60 •
60
40
70
90
80
90
70
70
80
70
50
80
85
90
90
60
50
50
69 (average)
Amount burned
per state,
1000 gallons
8,700
400
8,500
6,700
44,700
6,700
2,100
1.800
15.000
11.200
1.000
800
36,300
7,600
5,000
12,500
10,800
23,300
2,400
9,500
11,700
38,600
12,100
6,700
19,400
1,500
7,300
1,200
1,300
27,600
3.400
25,100
10,900
1.700
21.800
14,400
10,400
47,000
1,800
5.500
2,000
14,800
54.700
2.100
900
12.000
9,700
1,000
5.600
7,600
1.600
596,400
Proportion
burned,
% of total
1.46
0.07
1.43
1.12
7.49
1.12
0.35
0.30
2.52
1.88
0.17
0.13
6.09
1.27
0.84
2.10
1.81
3.91
0.40
1.59
1.96
6.47
2.03
1.12
3.25
0.25
1.22
0.20
0.22
4.63
0.57
4.21
1.83
0.29
3.66
2.41
1.74
7.88
0.22
0.92
0.34
2.48
9.17
0.35
0.15
2.01
1.63
0.17
0.94
1.27
0.27
Number of
small boilers
730
35
715
560
3,745
560
175
150
1,260
940
85
65
3,045
635
420
1,050
905
1,955
200
795
980
3,235
1.015
560
1,625
125
610
100
110
2,315
285
2,105
915
145
1,830
1,205
870
3,940
150
460
170
1,240
4,585
175
75
1,005
815
85
470
635
135
-50,000
A-27
-------
could be burning waste oil (as estimated in Reference 5), these
50,000 were distributed according to the amount of waste oil
estimated to be burned in each state. The resulting distribu-
tion is shown in Table A-15. This method indicated that New
Jersey, New York, Pennsylvania, Illinois, Michigan, Texas, and
California have the largest number of small boilers with the
potential for burning waste oil. These seven states account for
almost 46 percent of the 50,000 boilers. Illinois, Michigan,
Texas, and California alone account for 29 percent of the total.
Other states with a significant number of small boilers with the
potential for burning waste oil are Ohio, Virginia, Tennessee,
Florida, Louisiana, Oklahoma, Kansas, Nebraska, and Minnesota.
The accuracy of this second method is questionable for two
reasons: 1) the number of small boilers with the potential for
burning waste oil ("50,000) is only an estimate; and 2) the
distribution shown in Table A-15 is based on the premise that
the amount of waste oil burned correlates directly with the
number of small boilers. The actual relationship is probably
less direct.
A-28
-------
REFERENCES FOR APPENDIX A
Utility Data Institute, Inc. Power Directory, 1981 - An
Environmental Directory of U.S. Steam Electric Power Plants.
Prepared for Edison Electric Institute. 1981.
Electrical World Directory of Electric Utilities (1980-1981).
89th ed. McGraw-Hill, Inc., New York. 1980.
U.S. Department of Commerce. 1977 Census of Manufacturers.
Volume III, Geographic Series. Bureau of Census, Washington,
D.C. 1981.
U.S. Department of Commerce. County and City Data Book
1977. Bureau of Census, Washington, D.C. 1978.
Franklin Associates Limited and PEDCo Environmental, Inc.
Survey of the Waste Oil Industry and Waste Oil Composition
Draft Report. April 1983.
A-29
-------
-------
APPENDIX B
AIR QUALITY MODELING TECHNIQUES
AND
ASSUMPTIONS
B-l
-------
APPENDIX B
AIR QUALITY MODELING TECHNIQUES AND ASSUMPTIONS
B.I INTRODUCTION
Two air dispersion models were selected for the calculation
of air quality impact due to emissions from waste oil burning.
Modeling was performed at two different spatial scales of analy-
sis to investigate air quality impacts from estimates of concen-
trations from single sources and areawide sources over a year.
The three models selected and used in the Section 4 analysis
were:
1) The Industrial Source Complex Model for monthly
impacts within 10 km.
The Hanna-Gifford Modi
impacts within 50 km.
2) The Hanna-Gifford Model for monthly and annual
3) The Brookhaven/Pacific Northwest Model for monthly
and annual impacts beyond 50 km.
Section 4 focused on the use of these models and the resulting
estimates. This appendix describes each model in more detail
and presents the modeling assumptions, techniques, and input
variables and how they were specified.
B.2 INDUSTRIAL SOURCE COMPLEX (ISC) MODEL
The Industrial Source Complex (ISC) Model is a steady-
state Gaussian plume model that can be used to assess pollutant
concentrations from a wide variety of sources associated with an
B-2
-------
industrial source complex. This model can account for settling
and dry deposition of particulates , downwash, area, line, and
volume sources; plume rise as a function of downwind distance;
physical separation of point sources; and limited terrain
adjustment. It operates in both long-term (monthly to annual)
and short-term (1- to 24-hour) averaging time modes. The ISC
concentration model for point sources uses the steady-state
Gaussian plume equation for a continuous ground-level or
elevated source. For each stack and each hour, the hourly
ground-level concentration at downwind distance x and crosswind
distance y is given by:
X = — - exp [- h (--)] [V] [D]
IT u ay az y
where Q = pollutant emission rate
K = a scaling coefficient for units
a , a = standard deviation of the lateral and vertical
^ concentration
u = mean windspeed at stack height
V = accounts for vertical plume spread, reflection
from the ground and upper mixing height limits,
and gravitational settling
D = accounts for simple pollutant removal by physical
or chemical processes
Specifications of variables, model options, and output are
summarized in Table B-l.
The long-term version of the ISC Model was applied to
calculate monthly average concentrations of contaminants with a
B-3
-------
TABLE B-l. ISC MODEL SPECIFICATIONS
Parameter
Description
Input requirements
Output
Model options
Model limitations
Pollutant types
Source-receptor
relationships
Plume behavior
Horizontal wind
field
Vertical wind field
Dispersion
Chemistry/reaction
Mechanism
Physical removal
Emissions data: Location, emission rate, pollutant
decay coefficient, elevation of source (M5L), stack
height, stack exit velocity, stack Inside diameter,
stack exit temperature, particle size distribution
with corresponding settling velocities, surface
reflection coefficient, dimensions of adjacent
buildings.
Meteorological data: Short-term—hourly surface
weather data, including cloud ceiling, wind direc-
tion, windspeed, temperature, opaque cloud cover.
Daily mixing height is also reguired.
Long-term—stability wind rose (STAR deck), average
afternoon mixing height, average morning mixing height.
and average air temperature.
Concentration or deposition for any averaging time for
any combination of sources. Table of high, second-
high values and highest 50.
Site-specific wind profile exponents, site-specific
vertical temperature gradients, dry deposition,
terrain effects (limited), variable emission rates,
stack and building downwash.
Flat or gently rolling terrain.
Nonreactive. Particulates with or without signif-
icant settling velocities. Reactive pollutants, 1f
they can be accounted for by the exponential decay
term.
Arbitrary location for point, line, area, and volume
sources. Arbitrary receptor locations or receptor
rings. Receptors at ground level at elevation not
exceeding stack height.
Briggs plume-rise formulas. Building downwash and
stack tip downwash. If plume height exceeds mixing
height, ground-level concentration set to zero.
Does not treat fumigations.
Uses user-supplied hourly windspeeds. Uses user-
supplied hourly wind direction (nearest 10 degrees).
Internally modified by addition of a random integer
value between -4 degrees and +5 degrees. Windspeeds
corrected for release height based on power law varia-
tion, different exponents for different stability
classes, reference height « 10 meters. Constant,
uniform (steady-state) wind assumed within each hour.
Assumed equal to zero.
Semi-empirical/Gaussian plume. Six hourly stability
classes. Dispersion coefficient for urban or rural.
Exponential decay, user Input time constant. Surface
deposition.
Settling and dry deposition of participates.
B-4
-------
threshold response resulting from the burning of waste oil in
single or multiple point sources. The long-term version of ISC
calculates concentrations for specific meteorological categories
and then determines the weighted concentration average at each
receptor, based on climatological frequency distributions.
All point-source characteristics for medium-size and small
boilers and space heaters were specified in Sections 3 and 4.
Three options in the ISC Model provide methods for including
variations of the Gaussian plume approximation: 1) plume down-
wash, 2) particle deposition, and 3) physical and chemical
removal.
Downwash effects due to building wake were not included
because the low plume heights lead to conservative concentration
estimates even if downwash effects are not included. A second
reason for not including these effects was the ill-defined
nature of typical stack-building configurations.
The importance of considering the particulate size distri-
bution of emissions due to waste oil burning was assessed by
applying the ISC Model for sources both with and without deposi-
tion. The input parameters used in this sensitivity analysis
are presented in Table B-2. The methods suggested in the ISC
Model were used to calculate the mass mean diameter, the set-
tling velocity, and the reflection coefficient (how much is
reflected from the surface versus how much is retained) for each
particle size category. The particle size distribution was
4
derived from previous emissions testing and represents a
B-5
-------
particle size distribution of lead particulate emissions due to
waste oil burning. Results of this analysis for the whole range
of small boiler sizes (capacities of 19 to 246 liters/h) indi-
cate a maximum difference of less than about 2 percent in a
comparison of the highest downwind concentrations with and
without deposition.
TABLE B-2. PARTICLE DISTRIBUTION AND CALCULATED CHARACTERISTICS
Particle size
category, ym
<1
1-10
10-50 '
Percentage
by weight
in each
category, %
76
16
8
Mass mean
diameter, ym
0.63
6.52
33.9
Settling
velocity, cm/s
0.002437
0.25295
6.838
Reflection
coefficient
(unitless)
1.0
0.88-
0.57
The sensitivity of the ISC Model calculations to physical
and chemical reactivity was also tested by using the first-order
decay coefficient algorithm included in the ISC Model. The
equation, which adjusts for pollutant reactivity, decreases the
emissions term as a function of downwind distance and pollutant
half-life and takes the form:
D = exp [-\l> x/u]
where D = decay term in Gaussian equation
x = downwind distance, m
u = the average wind speed, m/s
i|i = decay coefficient, liters/s - 0.693/T,
*5
T, = pollutant half-life, s
B-6
-------
A contaminant half-life of 5 hours yielded concentration esti-
mates less than 1 percent different from those concentrations
without reactivity. Because most of the organic contaminants of
interest had half-lives greater than 5 hours (indicating even
less model sensitivity), all subsequent ISC calculations ne-
glected contaminant reactivity.
Meteorological data used in the ISC Model analysis of waste
oil burning included windspeed, wind direction, atmospheric
stability, monthly averaged mixing heights, and monthly averaged
ambient temperatures. The primary meteorological input for the
long-term ISC Model is the STability ARray (STAR), a joint
frequency distribution of windspeed, wind direction, and atmo-
spheric stability class. For this analysis, meteorological data
characteristic of an urban location were selected, i.e. from the
John F. Kennedy International Airport (National Weather Service,
Station No. 94789) for the years 1974-1978. For this 5-year
period, 3720 observations were tabulated for the month of Jan-
uary. Figure B-l presents a wind rose combining all stability
classes for January at the JFK Airport. Of considerable promi-
nence are winds from the west and northwest at all atmospheric
stabilities.
The mixing heights and temperatures for the study area were
estimated by use of available mixing height climatologies and
Local Climatological Data summaries. All meteorological inputs
were determined as specified by the ISC Model. Values used in
the analysis are presented in Table B-3.
B-'
-------
M
I
CO
Figure B-l. Wind rose for JFK International Airport, 1974-1978.
-------
TABLE B-3. AVERAGE JANUARY TEMPERATURES AND MIXING HEIGHTS
FOR NEW YORK USED IN ISC ANALYSIS
Average January temperature, K
Average January mixing heights, m
Stability class
A
276
1350
B
276
900
C
276
900
D
273
925
E
269
950
F
269
950
Preliminary analysis of the characteristics of stacks of
small and medium-size boilers and space heaters indicated maxi-
mum concentrations within the first 1000 meters of the lower-
stack sources and within 1500 for the taller-stack sources. An
array of 320 receptor locations was selected to ensure that the
maximum monthly concentrations were located for each individual
source analysis. These receptors were arranged in concentric
circles at 20 specified distances every 22.5 degrees (16 direc-
tions in all) around the source. The receptors were located at
distances of 100, 150, 200, 250, 300, 350, 400, 450, 500, 600,
700, 800, 900, 1000, 1500, 2000, 2500, 3000, 4000, and 5000
meters distant from the source. The receptor grid used in each
analysis is shown in Figure B-2.
B.3 HANNA-GIFFORD MODEL
The Hanna-Gifford Model was developed by the Air Resources
Atmospheric Turbulence and Diffusion Laboratory at Oak Ridge,
Tennessee, and has been used by the EPA for previous hazardous
pollutant analyses. The model is a simple but realistic
dispersion model that has proved to be adequate for estimating
B-9
-------
HNU
HSU
ssw
SSE
NOT SHOWN: 800, 900, 1000, 1500, 2000, 2500, 3000,
4000, 5000m.
RECEPTORS ARE LOCATED AT INTERSECTION OF RAYS AND
CONCENTRIC CIRCLES.
Figure B-2. Receptor grid used in ISC Model analysis
of single and multiple sources.
B-10
-------
pollutant concentrations due to widely distributed low-level and
8 9 10
area source emissions in urban areas. ' ' The surface concen-
tration of each pollutant was directly proportional to the local
area (or grid) source strength and inversely proportional to
average windspeed as indicated by the following equation:
X = CQi/u
where x = the average surface concentrations over the analy-
sis grid
Q. = the total distributed emissions of a given
1 pollutant over a specified area (i)
u = the average windspeed (averaged over the period
of interest)
C = a constant that relates the ambient concentra-
tions to source contributions within the grid of
the analysis 'and those upwind
The constant C may be defined in terms of the grid size, the
average number of upwind grids, and two stability-dependent
empirical coefficients. The constant C is expressed by:
-> % 1 Av 1~b N IK IK
C = (T}
-------
five subareas that were used in this analysis and the 5 km x 5
km grids in each subarea. The major divisions between subareas
are the concentric circles at intervals of 5, 10, 15, 25, and 50
km, as indicated.
For calculation of the factor C, the constants a and b as
specified, the average number of upwind grids (N = 9), and the
grid size (5 km) were used. The resulting constant C was 314.
As a test of the sensitivity of the Hanna-Gifford Model to
particle deposition and contaminant reactivity, the factor C was
modified and tested for each effect. Testing the factor C was
most appropriate because of its dependence on upwind emission
grids, which can be adjusted for deposition and reactivity
effects.
For the case of particle depositon, the factor C was modi-
fied by incorporating a plume depletion technique (as summarized
in Slade). The factor C (including the effects of deposition)
took the form:
2 * ±_ AX 1~b M
j = l
N M - . . .
I Z [((2i+l)i"D-(2i-l)1"D) (f . (F.)iAx)]]
J J
where ^FI^A 74 ~ the depletion factor for each particle
3 AX/ 4 category j taken in the grid wherein the
receptor lies at half the distance from
the center of the grid to its edge (Ax/ 4)
B-12
-------
0 5 10 15 20
i i i i i
Only grids within 50-km ring are considered in analysis.
Figure B-3. Urban study area with subareas and 5 km x 5 km grids.*
B-13
-------
(F.). = the plume depletion factor for each particle
-* size category j in each grid at a distance
iAx from the receptor
f . = percent of particles in particle size
-1 category j
Based on a particle distribution similar to that in Section A. 2
(90 percent are less than or equal to 10 ym and 10 percent are
greater than 10 ym) , a factor of C equal to 310 was calculated.
This, and the effects of deposition made only a 1 percent
difference over the entire urban study area and was considered
insignificant in terms of calculating subarea contaminant
concentration estimates.
Factors to adjust upwind concentrations for reactive
physical and chemical processes also may be estimated by .using
the first-order rate decay term suggested in the ISC Model. The
suggested expression (with symbols modified for consistency with
the previous equation) was:
where D.. = the decay factor for grid i at distance iAx
IAX
Ax = grid spacing in m
u = monthly, seasonal, or annual average windspeed
in m/s
T, = half-life of the given pollutant in seconds
The resulting expression for C was:
C - ^ (¥Tl^bT)(^)1"b [DAx/4 +
where D /4 = depletion factor for grid where receptor is
AX/4 located
B-14
-------
A half-life of 8 hours was used with a windspeed of about 6 m/s
(an average value from climatological records) to calculate a C
factor of 302. Using the value of C (302) rather than the
original 314 results in a concentration decrease of about 4
percent at 50 km. Because most of the organic contaminants had
characteristic half-lives in ambient air that were longer than 8
hours, and because concentration differences were less than 4
percent, effects of reactivity were considered insignificant and
therefore received no further consideration.
The Hanna-Gifford Model was applied in an iterative fashion
by using the calculated value of the parameter C, estimating the
appropriate monthly or annual windspeeds, and estimating the
subarea emissions due to waste oil burning. Because upwind
grids of emissions are considered to be the same as the grid
under consideration (with the uniform emissions assumption),
wind direction was of little consequence. The contaminant
concentration estimate for air that was made in each ring was
assumed to be applicable over the whole subarea.
The area (in square meters) of each subarea in Figure B-3
was calculated to determine the extent over which subarea-
specific emissions were distributed. The areas of each subarea
bounded by the distances specified in Figure B-3 are as shown in
Table B-4.
B-15
-------
TABLE B-4. AREAS OF EACH STUDY SUBAREA
Distance from study
area centroid, km
0-5
5-10
10-15
15-25
25-50
Area of subarea,
10 7 m2
7.85
23.6
39.3
134.
589.
Based on the estimates of emissions of waste oil contami-
nants, emissions per liter of waste oil burned were calculated
for the contaminant levels assumed at the 90th percentile values.
These emissions estimates were used with the estimates of waste
oil burned in each subarea to calculate the rates of emissions.
Expressed in equation form:
WOGi' * EF* CF * HDDF.^
Qi = AREA~7* ATF
where Q. = the emission rate in which each subarea i, g/s-ma
WOG. = the waste oil generated in each subarea i, liters/yr
EF = the contaminant emission factor, g/liter
CF - the combustion factor, i.e., the portion of waste
oil generated that is burned
HDDF. = the monthly or annual use factor based on heating
degree-days in each subarea i
ATF = the averaging time factor to convert time units to
seconds
AREA. = the area of each subarea, m2
The emission factors for each contaminant are specified in
Table 4-24. Distribution of waste oil generated within the
urban area is described in Appendix C. The state's available
B-16
-------
waste oil per year was apportioned to the individual urban study
subareas by distribution of multiple-dwelling units. The use of
multiple dwellings is thought to provide the best distribution
of waste oil consumption because these are the most likely
locations for small boilers and space heaters.
The combustion factors (CF) were based on state and local
estimates of waste oil burning as described in Appendix A (see
Table A-15) and set equal to 70 percent for AQCR 43. The use
factors (HDDF.) were calculated on the basis of heating degree-
days. This is a reasonable basis for calculation because the
waste oil burned was proportional to the demand for space heat-
ing. On an annual basis, HDDF. equaled 1.0 because all the
waste oil that was available for burning was burned sometime
during the year. For the monthly analysis, HDDF. was estimated
as a function of the ratio of heating degree-days for January
and the annual heating degree-days. This ratio was assumed to
reflect the use of waste oil for burning in January as a percent
of the total available in a year. Table B-5 describes the
12 f\
National Weather Service (NWS) ' locations and the associated
heating degree-day totals for each subarea of the urban analysis,
At least 10 or more years of observations were used in the
climatological summaries of heating degree-days. All emissions
for January totals of threshold contaminants and annual totals
of nonthreshold contaminants were presented in Section 4.4.2.
B-17
-------
TABLE B-5. HEATING DEGREE-DAY TOTALS IN EACH URBAN SUBAREA
AND ASSOCIATED USE FACTOR (HDDF^)
Distance from
study area
centroid, km
0-5
5-10
10-15
15-25
25-50
NWS sites
Central Park
La Guardia
Newark
JFK International
Bridgeport
Heating degree-day totals
January
1017
1020
1042
1042
1079
Annual
4848
4909
5034
5184
5461
HDDF.a
0.21
0.21
0.21
0.20
0.20
a HDDF., the ratio of January to annual heating degree-days for subarea (i).
One other critical element in the application of the Hanna-
Gifford Model to the urban study area was the specification of
windspeed averaged over the two averaging times of interest
(i.e., January and annual). The five NWS stations used to
derive the heating degree-day totals were examined to determine
monthly and annual average windspeeds. Outside of the 0-5 km
subarea, windspeeds were nearly identical with the reduced
windspeeds evident within the 0-5 km subarea. Table B-6 de-
scribes the derived windspeeds used in all subsequent calcu-
lations with the Hanna-Gifford Model for urban-scale modeling.
TABLE B-6. JANUARY AND ANNUAL AVERAGE WINDSPEEDS FOR
THE URBAN-SCALE MODELING
Distance from
study area
centroid, km
0-5
5-10
10-15
15-25
25-50
Average windspeeds, m/s
January
4.8
5.8
5.8
5.8
5.8
Annual
4.2
5.4
5.4
5.4
5.4
B-18
-------
Because all grids within a given subarea had uniform emis-
sions, only five receptor points were needed to describe the
concentrations in the urban study area, i.e., one in each sub-
area. In the Hanna-Gifford Model, a receptor is usually placed
in each grid; in this case, however, the uniformity of emissions
(i.e., adjacent grids had the same concentrations) was unneces-
sary. The assumption of uniform emissions in each subarea was
reasonable because of the ill-defined nature of the sources and
emissions.
B-19
-------
REFERENCES FOR APPENDIX B
1. Bowers, J. F., J. R. Bjorkland, and C. S. Cheney. Indus-
trial Source Complex (ISC) Dispersion Model User's Guide,
Volumes 1 and 2. EPA-450/4-79-030 and EPA-450/4-79-031,
December 1979.
2. Gifford, F. A., and S. R. Hanna. Modeling Urban Air Pollu-
tion. Atmos. Environment, 7:131-136, 1973.
3. Rowe, M. D. Human Exposure to Particulate Emissions From
Power Plants. Brookhaven National Laboratory. Upton, Long
Island, New York. BNL 51305, May 1981.
4. U.S. Environmental Protection Agency. Report to Congress;
Waste Oil Study. Accession No. PB-257693, April 1974.
5. Holzworth, G. C. Mixing Heights, Wind Speeds, and Poten-
tial for Urban Air Pollution Throughout the Contiguous
United States. Environmental Protection Agency, Research
Triangle Park, North Carolina. AP-101, January 1972.
6. National Oceanic and Atmospheric Administration.. Local
Climatological Data for J. F. Kennedy International Air-
port. National Climatic Center, Asheville, North Carolina.
1979.
7. Anderson, G. E., et al. Human Exposure to Atmospheric
Concentrations of Selected Chemicals. U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina.
March 1980.
8. Hanna, S. R. A Simple Method of Calculating Dispersion
from Urban Area Sources. JAPCA, 21(8), December 1971.
9. Miller, C. W. An Application of the ATDL Simple Dispersion
Model. JAPCA, 28(8), August 1978.
10. Hanna, S. R. Urban Diffusion Problems. Presented at the
AMS Workshop of Meteorology and Environmental Assessment,
Boston, Massachusetts, October 1975.
11. Atomic Energy Commission. Meteorology and Atomic Energy
1968. Slade, D. H., ed. Oak Ridge, Tennessee. July 1971.
12. National Oceanic and Atmospheric Administration. Local
Climatological Data. Central Park, New York, New York,
1979; La Guardia Airport, New York, New York, 1979; Newark,
New Jersey, 1979; Bridgeport, Connecticut, 1975. National
Climatic Center, Asheville, North Carolina.
B-20
-------
APPENDIX C
ESTIMATES OF WASTE OIL BURNED
IN URBAN STUDY AREA
C-l
-------
APPENDIX C
ESTIMATES OF WASTE OIL BURNED IN URBAN STUDY AREA
The Hanna-Gifford model was chosen to estimate airborne
emissions and their dispersion over a medium-range area (0 to 50
km in diameter). The exposures from the ground-level concentra-
tions of waste oil contaminants (as calculated by the modeling)
were subsequently used to estimate risk to human populations.
The amount of waste oil available for use as a fuel source
was taken from a recent report that gave a breakdown of total
2
waste oil generation by state. These numbers were adjusted by
deducting the amount of oil that is not recycled back into the
collection system (i.e., waste oil that is unavailable to col-
lectors and processors), and the remaining volume of waste oil
was then distributed among all Air Quality Control Regions
(AQCR's) in the contiguous United States according to the popu-
4
lation figures given in the 1980 Census. This approach was
used for both the Hanna-Gifford and the Brookhaven Long Range
Transport models.
Because the intent of the study was to determine whether
uncontrolled burning of waste oil presents an environmental
problem requiring regulation, an exposure scenario was developed
that is believed to represent maximum concentrations of waste
oil contaminants to which high-density population areas typical
C-2
-------
of populated northeastern cities are exposed. A hypothetical
study area was constructed by using a pattern of five concentric
rings with increasing radial distances (5, 10, 15, 25 and 50
kilometers) from the study area center. The amount of waste oil
generated, the portion of that waste oil burned, and the popula-
tion densities were then calculated for each ring, based on
statistics from the northeastern United States.
The distribution of waste oil among the five rings was
based on the estimated demand for fuel oil from multiple-family
dwelling units.
The estimation of demand was calculated by modifying the
"heating degree-day formula" outlined in the EPA's Guide for
Compiling a Comprehensive Emission Inventory. The modified
formula is as follows:
Demand for heating oil (gallons) = 0.18 a(^)c
where 0.18 = gallons burned per dwelling unit per heating
degree-day
a = heating degree-days per year
b = rooms per dwelling unit
c = number of dwelling units in study area
The demand for waste oil as a heating fuel was estimated by
considering only multiple-family dwelling units equipped to burn
oil because waste oil is generally not burned in the type of
small heating furnaces used in single-family homes. The rooms
per dwelling unit (b) and the number of dwelling units in the
C-3
-------
study area (c) were adjusted to reflect only multiple-family
dwellings. Table C-l presents the results. It is assumed that
the percent demand for heating fuel oil by multiple-family
dwelling in each ring reflects the same level of demand for
waste oil resources. An estimated total amount of waste oil
available for heating and the percent demand for oil from
multiple-family housing were used to calculate annual estimates
of waste oil burned in the study area (Table C-2). This table
also presents monthly and seasonal estimates of waste oil.
These estimates are based on portional reductions in the heating
degree-days for the highest-demand month (January) and the
highest-demand season (winter: December, January, and Febru-
ary) .
C-4
-------
TABLE C-l. CALCULATION OF DEMAND FOR HEATING OIL BY MULTI-FAMILY DWELLING
UNITS IN THE STUDY AREA
Study
area
ring
1
2
3
4
5
Distance
from center,
km
0-5
5-10
10-15
15-25
25-50
Average
heating
degree
days/year
4848
4909
5075
5241
5696
Average
rooms/multi-
family dwel-
ling unit
3.2
3.5
3.6
3.7
3.8
Total
Number
of multi-
family dwelr
ling units
535,126
679,251
885,990
1,116,475
449,528
3,666,370
Percent
of dwel-
ling units
equipped
to burn
heating oil
63
58
56
55
54
Estimated annual
need for heating oil
for multi-family buildings
103 gal
188,300
243,700
326,300
428,700
189,200
1,376,200
percentage
14
17
24
31
14
100
o
I
U1
a Annual heating degree-days (these are averages of heating degree-day data from weather stations
located in each ring).6
Reference 4.
-------
TABLE C-2. CALCULATION OF WASTE OIL USAGE IN THE STUDY AREA
o
Study
area
ring
1
2
3
4
5
Total
Distance
from center,
km
0-5
5-10
10-15
15-25
25-50
Waste oil
generated,
1000
gal/yr
36.572
Waste oil
available for.
burning (70%),
1000 gal/yr
25,600
Percentage
demand for
heating oil from
multl -family
dwelling units
14
17
24
31
14
100
Annua 1
estimate
of waste
oil burned,
1000 gal/yr
3,584
4,352
6,144
7,936
3,584
25,600
Monthly and seasonal
estimates of waste oil
January
Id
0.21
0.21
0.21
0.20
0.20
1000 gal /mo.
753
914
1,290
1,587
717
5.261
Winter
td
0.58
0.58
0.58
0.56
0.55
1000 gal/
Dec. -Feb.
2.079
2.524
3.564
4.444
1.971
14,582
' Reference 2.
b Estimate obtained from Franklin Associates Limited. Personal communication, C. Jarvls, PEOCo Environmental with ft. Hunt.
February 1983.
c See Table A-I.
January percentage values represent the number of heating degree-days In January divided by the annual number of heating degree-days.
Seasonal percentage values represent the number of heating degree-days for the winter months of December, January, and February divided by
the annual number of heating-degree-days.6
-------
REFERENCES FOR APPENDIX C
1. Hanna, S. R. A Simple Method of Calculating Dispersion
from Urban Area Sources. JAPCA, 21(12), December 1971.
2. Franklin Associates Limited. Evaluation of the Use of
Waste Oil as a Dust Suppressant. Draft Report. January
1983.
3. Franklin Associates Limited. Waste Oil Management and
Composition. Draft Report. February 1983.
4. U.S. Department of Commerce. 1980 Census Summary Tape File
(STF3). Bureau of Census, Washington, D.C. January 1983.
5. Eadie, W. J., and W. E. Davis. The Development of a Na-
tional Inter-Regional Transport Matrix for-Respirable
Particulates. Pacific Northwest Laboratories, Richland,
Washington. PNL-RAP-37, 1979.
6. U.S. Environmental Protection Agency. Guide for Compiling
a Comprehensive Emission Inventory. Washington, D.C.
APTD-1135, 1976.
C-7
-------
APPENDIX D
COMPARISON OF CONTAMINANT
EMISSION RATES FROM THE COMBUSTION
OF WASTE OIL WITH THOSE FROM OTHER
COMBUSTION SOURCES
D-l
-------
APPENDIX D
COMPARISON OF CONTAMINANT EMISSION RATES FROM THE COMBUSTION OF
WASTE OIL WITH THOSE FROM OTHER COMBUSTION SOURCES
Atmospheric emissions of contaminants from the burning of
waste oil may have some adverse environmental impact. The
potential contaminant emissions discussed here are:
0 Lead
0 Arsenic
0 Polynuclear aromatic hydrocarbons
0 Benzene
0 Naphthalene
Emission rates of these contaminants from waste oil combustion
are compared with their emission rates from other combustion
sources.
LEAD EMISSIONS
Lead emissions from the following sources are compared: 1)
a space heater burning waste oil, 2) an idling car burning leaded
fuel, 3) an idling car burning unleaded fuel, 4) a small boiler
(9 x 10 Btu/h) burning waste oil, and 5) a small boiler (9 x 10
Btu/h) burning residual fuel oil.
Lead contamination in waste oil is estimated to be 220
yg/g, and in automotive waste oil, it can be as high as 1000
yg/g. Lead concentrations can be as high as 2.65 g/gal in
3 2
leaded gasoline and 0.5 g/gal in unleaded gasoline. Since July
1983, the lead in leaded gasoline has been limited to 1.1 g/gal.
D-2
-------
Both this value and the higher estimated value (2.65 g/gal) were
used in this comparison because currently available waste oil
could have been contaminated prior to the effective date of this
limitation. Assuming a density of 3.3 kg/gal for gasoline, the
lead concentration converts to 800 ug/g and 150 wg/g for leaded
and unleaded gasoline, respectively. The concentration of lead
in residual or No. 6 fuel oil ranges from 0.4 to 2.0 ug/g. For
all combustion processes except space heaters, it is assumed that
75 percent of the lead is emitted through the exhaust to the
atmosphere. An atomizing space heater is estimated to emit 50
percent of the lead in the fuel. A vaporizing space heater
retains more of the lead in the pot residue, and 4 to 30 percent
of the lead is estimated to be emitted. In this comparison, a
conservative number of 25 percent will be used for this source.
Another critical assumption for this comparison is that an
idling car will burn 2 gallons of gasoline per hour. This esti-
mate may be high; however, because of the wide range of fuel
economy (1 to 3 gallons per hour) in automobile engines, 2 gal-
lons was arbitrarily selected and is probably a conservative
value. This theoretical approach was taken to estimate lead
emissions from automobiles because no specific data are available
on the amount of lead emitted by an "average" automobile in the
course of an hour's operation.
Based on these assumptions, lead emission rates were calcu-
lated and are presented in Table D-l. The emission rate at the
source does not directly relate to risk because dispersion model-
ing parameters were not considered. These data only indicate the
D-3
-------
TABLE D-l. ESTIMATE OF LEAD EMISSION RATES
Atomizing
space heater,
100% waste oil
Vaporizing
space heater,
100% waste oil
Id! ing car,
leaded fuel
Idling car,
unleaded fuel
Small boiler,
100% waste oil
Small boiler,
residual fuel oil
Fuel feed
rate,
g/h
3,580
3,500
6,700
6,700
220,000
220,000
Lead
concentration,
ug/g
220-1000
220-1000
3
330-800
- »
150
.
220-1,000
0.4-2
Percent
lead
emitted
50
25
75
75
75
75
Lead
emission
rate,
ug/s
110-500
55-250
460-1,100
210
10,000-46,000
18-92
As of July 1983, refineries will be limited to 1.1 g/gal of lead.'
D-4
-------
potential rate of lead emissions from one particular isolated
source.
ARSENIC EMISSIONS
Arsenic contamination in waste oil is 11 yg/g (median) and
16 ug/g (90th percentile). The concentration in residual fuel
1 4
oil ranges from 0.2 to 0.8 ug/g. ' It is estimated that 75
percent of the arsenic in the waste oil or fuel is emitted to the
atmosphere from combustion in boilers. In space heaters, 50
percent of the arsenic is emitted from atomizing burners and 25
percent from vaporizing burners. Based on these assumptions, the
arsenic emission rates were calculated and reported in Table D-2.
The arsenic emission rate from space heaters is 3 to 8 ug/s. The
arsenic emission rate from small boilers burning 100 percent
waste oil is 504 vg/s (median) and 733 ug/s (90th percentile),
and the emission rate from the same source while burning residual
oil is 9 to 37 ug/s.
POLYNUCLEAR AROMATIC HYDROCARBONS
Polynuclear aromatic hydrocarbons (PNA's) have been reported
to be contaminants in waste oil. This group of compounds is
also a product of incomplete combustion. Typical combustion
process temperatures up to 800°C (1500°F) provide the environment
for the formation of PNA's. Thus, the burning of waste oil may
cause the emission of PNA's to the atmosphere. These emissions
are of concern because several members of this class of organics
compounds are considered to be nonthreshold contaminants.
D-5
-------
TABLE D-2. ESTIMATE OF ARSENIC EMISSION RATES
Atomizing
space heater,
100% waste oil
Vaporizing
space heater,
100% waste oil
Small boiler,
100% waste oil
Small boiler,
residual fuel oil
Fuel feed
rate,
g/h
3,580
3,580
220,000
220,000
Arsenic
concentration,
pg/g
11-16
11-16
11-16
0.2-0.8
Percent
arsenic
emitted
50
25
75
75
Arsenic
emission
rate,
yg/s
5-8
3-4
504-733
9-37
D-6
-------
Sources of PNA emissions under consideration for this dis-
cussion are waste oil space heaters (atomizing and vaporizing
pot), small boilers burning waste oil, and small boilers burning
residual fuel oil.
Based on the literature estimates, emissions from vaporizing
pot space heaters are higher in PNA's than those from atomizing
Q
space heaters. Table D-3 presents emission test data for several
PNA compounds for these two combustion sources. The average
total PNA emission rate from vaporizing pot space heaters is 75
yg/m3, and the average from atomizing space heaters is 15 yg/m3.
Benzo (a)pyrene (BaP) emissions average 7 yg/m3 from vaporizing
pot space heaters and 1 yg/m3 from atomizing space heaters.
Assuming a volumetric flow rate* of 0.12 m3/s for the vaporizing
pot heater and 0.044 m3/s for the atomizing heater, the emission
rates are as follows:
Space heater type
Atomizing
Vaporizing pot
Emission rates
(yg/s)
PNA
0.7
9
BaP
0.04
0.8
Emission factors reported in the literature for residual-oil-
fired boilers with capacities of less than 250 x 10 Btu/h are 46
a in
ng/Btu for PNA's and 0.4 to 1.0 ng/Btu for BaP. ' Benzo(a)pyrene
is also included because specific data are available for this
particular PNA compound. Based on these emission factors, it is
The assumed volumetric flow rates are based on velocities
measured during tests reported in Reference 8; a stack diameter
of 20 cm is assumed.
D-7
-------
TABLE D-3. TEST RESULTS OF PNA
EMISSIONS FROM SPACE HEATERS BURNING WASTE OIL3
Compound
Naphthalene
Acenaphthene
Acenaphthylene
Fluorene
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(
-------
estimated that PNA emissions from a boiler burning residual oil
at 9 x 10 Btu/h would be approximately 110 yg/s and BaP emissions
would be 1 to 2.5 yg/s.
Data are insufficient to indicate that PNA emissions from
the combustion of waste oil differ from those from combustion of
virgin oil. Concentrations of BaP in waste oil are similar to
those in regular fuel oils. Table D-4 presents the BaP concen-
trations in various oils. Concentrations of BaP have been found
in residual fuel oil at levels from 10 to 320 ppm and at levels
up to 400 ppm in waste oil. Typical values in residual and waste
4
oil are 10 to 15 ppm. Other PNA compounds that the literature
indicates are found in waste oil are presented in Table D-5.
The uncontrolled emission factor for PNA from the combustion
of waste oil is reported as 0.0075 lb/103 gal (0.9 mg/liter).
Assuming a waste oil feed rate of 220,000 g/h (246 liters/h), the
PNA emission rate would be 60 yg/s based on this emission factor.
The two estimated emission rates (60 yg/s and 110 yg/s) for small
boilers burning waste oil and residual oil, respectively, are
generally equivalent. Table D-6 summarizes the estimated emission
rates for the PNA compounds and BaP just discussed.
BENZENE AND NAPHTHALENE
Estimates of benzene and naphthalene emissions are based on
theoretical calculations and available test data. The three
sources evaluated are 1) space heaters burning waste oil, 2)
small boilers burning waste oil, and 3) small boilers burning
residual fuel oil.
D-9
-------
TABLE D-4. BENZO(a)PYRENE CONCENTRATIONS IN VARIOUS OILS'
(yg/g)
011 type
Concentrations
Virgin No. 2 oils
Virgin No. 4 oil
Virgin No. 5 oils
Virgin No. 6 oils (residual oils)
Unused motor oil basestocks
Used motor oils and waste oils
Used diesel motor oil
Used synthetic motor oil
Used oil (new car dealer)
Unused re-refined motor oil basestock
Used industrial oil
Reprocessed used oil
Used oil/fuel oil blends
0.03-0.6
2.1
2.8-3.3
2.9-44
0.03-0.28
3.2-28
<0.15
16
0.7
2.1
5.9
10.5
1.6-3.0
a Reference 11.
D-10
-------
TABLE D-5. TEST RESULTS OF ANALYSIS OF WASTE OIL SAMPLES FOR PNA COMPOUNDS0
(pg/g)
Compound
Naphthalene
Phenanthrene/ Anthracene
Pyrene
Benz(a)anthracene/Benz(a
Benzo(a)pyrene
)phenanthrene
Total
Average concentration
390
190
32
18
12
640
Reference 12.
Average of analysis results on 14 waste oil samples.
D-ll
-------
TABLE D-6. ESTIMATE OF PNA EMISSION RATES
(ug/s)
PNA compounds
BaP
Atomizing
space heater,
100% waste oil
Vaporizing pot
space heater,
100% waste oil
Small boiler,
100% waste oil
Small boiler,
residual fuel oil
0.7
9
60
110
0.04
0.8
1-2.5
TABLE D-7. CONCENTRATIONS OF BENZENE AND NAPHTHALENE IN
RESIDUAL AND WASTE OIL SAMPLES
(ppm)
Residual oil
Waste oil
Benzene
Naphthalene
<2-37d
20 (avg)
18-890U
70.(avg)
110-1400b
460 (avg)
Reference 11.
Reference 5.
D-12
-------
An analysis of residual oil and waste oil samples is pre-
sented in Table D-7 for the two compounds of interest. These
data are from the available literature sources. Based on these
data, a direct correlation is made to compare benzene and naph-
thalene emissions from small boilers of the same type and feed
rate, with one burning 100 percent waste oil and the other burn-
ing residual fuel oil. Assuming the two combustion processes
have the same destruction efficiency, emissions would be directly
proportional to the pollutant concentration in the oil.
The assumed destruction removal efficiency for vaporizing
pot space heaters was 40 percent, and that for the atomizing
space heaters was 53 percent. Destruction removal efficiency
was assumed to be 97 percent'for small boilers. Based on these
assumptions and the concentrations presented in Table D-7, the
estimated emission rates for benzene and naphthalene were calcu-
lated and are presented in Table D-8.
A summary of all emissions discussed in this report is
presented in Table D-9.
D-13
-------
TABLE D-8. CALCULATED BENZENE AND NAPHTHALENE EMISSION RATES
Source
Vaporizing pot
space heater, .
100% waste oil
Atomizing
space heater,
100% waste oil
Small boiler,
100% waste oil
Small boiler,
residual oil
Fuel feed
rate, g/h
3,580
3,580
220,000
220,000
Destruction
efficiency,
%
40
53
97
97
Emission rate, ug/s
Benzene
42
33
290
37
Naphthalene
270
215
1,060
D-14
-------
TABLE D-9. SUMMARY OF ESTIMATED EMISSION RATES0
(pg/s)
Source
Atomizing
space heater,
100% waste oil
Vaporizing pot
space heater,
100% waste oil
Idling car,
leaded fuel
Idling car,
unleaded fuel
Small boiler,
100% waste oil
Small boiler,
residual oil
Pollutant
Lead
110
55
460
210
10,000
55
Arsenic
5
3
-
-
504
23
PNA
0.7
9
-
-
60
110
BaP
0.04
0.8
-
-
1.8
Benzene
33
42
-
-
290
37
Naphthalene
215
270
-
.
-
1,060
Only average or median values reported on this table. Maximum or 90th
percentile values were reported in text and on previous tables.
D-15
-------
REFERENCES FOR APPENDIX D
1. PEDCo Environmental, Inc. A Risk Assessment of Waste Oil
Burning in Boilers and Space Heaters. (Preliminary draft)
U.S. Environmental Protection Agency Office of Solid Waste.
May 1983.
2. Letter to Docket Clerk from Timothy A. Vanderver, Jr.,
Counsel for Walter Kroll GmbH and Heating Alternatives, Inc.
August 14, 1980.
3. Robertson, W. Implications of the Report "Chemical Analysis
of Waste Crankcase Oil Combustion Samples" and Other Studies
on the Burning of Used Oil. Robertson Company. May 12,
1983.
4. U.S. Environmental Protection Agency. Memo to Stephen A.
Lingle from Michael Petruska. Subject: Comparison of Fuel
Oils and Waste Oils. June 8, 1983.
5. GCA Corporation. The Fate of Hazardous and Nonhazardous
Wastes in Used Oil Recycling. (Draft final report) Pre-
pared for U.S. Department of Energy. April 15, 1983.
6. Davies, I. W., et al. Municipal Incinerator as Source of
Polynuclear Aromatic Hydrocarbons in Environment. Environ-
mental Science and Technology, 10(5):May 1976.
7. National Academy of Sciences. Particulate Polycyclic Organic
Matter. 1972.
8. Cooke, M., et al. Emissions From Waste-Oil-Fired Heaters.
Prepared by Battelle-Columbus Laboratories for U.S. Environ-
mental Protection Agency. May 1983.
9. Krishnan, E. R., and G. V. Hellwig. Trace Emissions from
Coal and Oil Combustion. Environmental Progress, l(4):Novem-
ber 1982.
10. Surprenant, N. F., et al. Emissions Assessment of Conven-
tional Stationary Combustion Systems: Vol. IV. Commercial/-
Institutional Combustion Sources. EPA-600/7-81-003b, 1981.
D-16
-------
REFERENCES (continued)
11. Recon Systems, Inc., and ETA Engineering, Inc. Used Oil
Burned as a Fuel. 1980.
12. Brinkman, D. W., P. Fennelly, and N. Surprenant. The Fate
of Hazardous Wastes in Used Oil Recycling (Abstract) U.S.
Department of Energy and GCA Corporation. April 1983.
13. Hall, R. R. Comparative Analysis of Contaminated Heating
Oils. (Draft Final Report) Prepared for U.S. Environmental
Protection Agency by the GCA Corporation. May 1983.
D-17
-------
-------
APPENDIX E
HEALTH EFFECTS ASSESSMENT METHOD
E-l
-------
APPENDIX E
HEALTH EFFECTS ASSESSMENT METHOD
INTRODUCTION
Assessing the effects of waste oil burning on human health
requires an individual analysis of the impact of each waste oil
contaminant. Analysis of waste oil emissions as a single waste
stream is not practical because of the wide range of health
effects produced by the various individual emission components.
When the impact on human health is examined, two general classes
of effects can be distinguished: threshold and nonthreshold.
The traditional approach to the establishment of acceptable
exposure levels to chemical substances is to identify concentra-
tions that will have no adverse effects in target populations.
This approach assumes the existence of a threshold dose below
which no deleterious effects will occur. Indeed, many chemical
substances have been characterized as eliciting a threshold-type
4
response, e.g., irritants and simple poisons. In this report,
concentration limits believed to protect public health from acute
adverse reactions resulting from chronic exposure to toxic emis-
sions eliciting a threshold effect are referred to as Environ-
mental Exposure Limits (EEL's).
E-2
-------
Safe exposure levels are not easily identified for some
chemical substances. Any exposure to these chemicals elicits a
response, no matter how small the concentration. Such substances
are said to produce nonthreshold responses in their target popu-
lations. Several substances that appear to elicit a nonthreshold
response have been identified. A pathological end-point of great
public concern that results from a nonthreshold response is
cancer. In the case of carcinogens, evidence indicates that
these substances have the potential to produce deleterious ef-
fects regardless of the quantity of the chemical present in the
body; i.e., one molecule can initiate the process of carcinogene-
sis (one-hit theory). Although a debate still goes on within the
regulatory community on how best to regulate cancer-producing
chemicals, it is generally accepted that the weight of scientific
data clearly supports the existence of the nonthreshold phenom-
enon.
Because threshold doses have not been established for car-
cinogens, the practice of "risk estimation" has gained wide
acceptance. ' Estimates of cancer risk involve the use of
animal toxicological data, human epidemiological data, and math-
ematical models to estimate the cancer incidence rates associated
with exposures to suspected carcinogens. This risk estimation
method entails the use of a carcinogen's exposure-response rela-
tionship to estimate the health impact of the substances. These
estimates are generally expressed as the number of excess cancers
E-3
-------
per unit of population (e.g., one excess cancer per 100,000
people, referred to as a risk of 10~ ) or the lifetime risk to
the highest exposed individual. For the purposes of this report
it was convenient to express the exposure-response relationship
as specific reference concentrations (i.e., at what concentration
could a risk of cancer of 10 ,10 , 10~ , etc., be expected).
This appendix describes the data and assumptions used to deter-
mine both the EEL's and the reference concentrations.
MODEL FOR ESTIMATING ENVIRONMENTAL EXPOSURE LEVELS
The structure of the model chosen to estimate EEL's for use
in the waste oil risk assessment study is similar to that of
several models currently used in the health risk assessment
community. The major premise behind all of these models (a
premise that is not universally accepted) is that workplace
threshold limit-values (TLV's) published by the American Confer-
ence of Governmental Industrial Hygienists (ACGIH) can be ad-
justed mathematically for use in assessing nontraditional work-
4
place or environmental exposures. These mathematical adjust-
9
ments have ranged from simple time adjustments to a few sophis-
ticated models that incorporate uptake and excretion coeffi-
7 8
cients. ' The success of each attempt depends on how well the
authors have accounted for the limitations inherently associated
with the use of TLV's. The preface to the ACGIH publication
clearly states the limitations associated with the TLV's as
identified by the committee:
E-4
-------
"Threshold limit values refer to airborne concentrations of
substances and represent conditions under which it is be-
lieved that nearly all workers may be repeatedly exposed day
after day without adverse effect. Because of wide variation
in individual susceptibility, however, a small percentage of
workers may experience discomfort from some substances at
concentrations at or below the threshold limit; a smaller
percentage may be affected more seriously by aggravation of
a pre-existing condition or by development of an occupa-
tional illness.
"Threshold limits are based on the best available informa-
tion from industrial experience, from experimental human and
animal studies, and when possible, from a combination of the
three. The basis on which the values are established may
differ from substance to substance; protection against im-
pairment of health may be a guiding factor for some, whereas
reasonable freedom from irritation, narcosis, nuisance, or
other forms of stress may form the basis for others.
"The amount and nature of the information available for
establishing a TLV varies from substance to substance;
consequently, the precision of the estimated TLV is also
subject to variation, and the latest documentation should be
consulted in order to assess the extent of the data avail-
able for a given substance."
This preface identifies five important caveats that should
be addressed when TLVs are adjusted to account for environmental
exposures: 1) the exposure duration, 2) the population at great-
est risk (susceptibility), 3) pre-existing conditions or ill-
nesses in the exposed population, 4) the basis for determining
the original TLV, and 5) the type of protection intended.
All models developed to date (including the model presented
herein) are only partially successful in addressing each of these
caveats or limitations. Mathematical models are usually devel-
oped for specific purposes (e.g., to establish exposure limits
for 10- or 12-hour workdays, overtime, the additive effects of a
second job, avocational exposures to toxic agents, or chronic
E-5
-------
environmental exposures). These needs have limited the past
application to the most applicable or important caveats (e.g.,
accounting for the duration of exposure when the 8-hour TLV is
used to derive a 12-hour workplace exposure limit). Also, models
usually were designed to address only those limitations for which
corrective information was readily available. Despite these
deficiencies, outputs from these modified models are of greater
utility than the original TLV's simply because the adjusted value
accounts for one or more of the limitations. The greater the
number of limitations addressed, the more confidence one can
place in the model.
The model used to calculate TLV-derived EEL's for use during
this waste oil risk assessment is presented in Equation 1.
TLV (D f) (M -)
EEL = f= — x 103 (Eq. 1)
bf
where EEL = environmental exposure limit, yg/m3
TLV = 8-hour time-weighted average threshold limit value,
mg/m3
D - = duration of exposure adjustment factor (0.12),
a nondimensional
M f = magnitude of exposure adjustment factor (0.72),
a nondimensional
Sf = safety factor (10-1000), nondimensional
This model adjusts for differences in duration and magnitude
of exposure. Also, through the selection of a safety factor, it
accounts for differences in the documentation used to develop
each TLV and the type of protection the TLV is intended to pro-
vide.
E-6
-------
Duration of Exposure Adjustment Factor (D -]_
The ACGIH TLV's were developed to provide protection "...for
a normal 8-hour workday and a 40-hour workweek, to which nearly
all workers may be repeatedly exposed, day. after day, without
adverse effect." This excerpt defines the length of a "normal"
weekly work schedule and also implies a normal working lifetime.
Because environmental exposures are not limited to an 8-hour
day, a 40-hour workweek, or a working lifetime, an adjustment in
exposure was made by estimating the ratio of a "normal" work
exposure duration to a public lifetime exposure. The normal
working lifetime of an adult male worker* was calculated to be
8.0 x 10 hours.** This value represents the likely duration of
an occupational exposure.
Environmental exposures have the potential of occurring over
an entire lifetime. The value used to define the duration of a
biological lifetime must account for variations in longevity
within the general population. Consideration of this variation
is important because a person with a long life span will experi-
ence a greater total exposure and ultimately more stress than a
person with a shorter life span. Within the United States aver-
age life expectancy varies greatly with gender. An American
*
The term "adult male workers" is sometimes used when referring
to TLV's. This distinction is made because the vast majority
of industrial experience and human exposure data cited in the
ACGIH documentation is based on adult male subjects.
**
Value is based on 8 hours/day, 5 days/week, 50 weeks/year,
and a working lifetime of 40 years. The selection of 40
years assumes a starting age of 25 years and a retirement age
of 65. This work period provides some allowance for job
changes, college, and early retirement, which are not con-
sidered in a 47-year working lifetime (18 to 65 years old).
E-7
-------
female born in 1979 has a longer average life expectancy than an
American male born at the same time (77.8 years for females
12
versus 69.9 years for males). This difference results in a
longer lifetime exposure duration for females (6.8 x 10 hours
5 *
versus 6.1 x 10 hours). If the gender difference is taken into
account, the resulting adjustment factor for the change in the
4 5
duration of exposure is 8.0 x 10 hours/6.8 x 10 hours, or
approximately 0.12. This adjustment factor addresses, in part,
the first caveat and accounts for the cohort in the general
population with the longest life expectancy.
Magnitude of Exposure Adjustment Factor (M f)
—,—^ ^—
Identifying population groups at greatest risk is difficult.
The ACGIH noted this difficulty in describing the limitation of
the TLV's: "...Because of wide variation in individual suscepti-
bility. .. a small percentage of workers may experience discomfort
from some substances at concentrations at or below the threshold
limit..." The reasons for this discomfort may be differences in
morphology, physiology, behavior, or genetics among certain
members of an exposed population. It is not possible to lower
threshold limits to levels that presumably would protect all
workers at all times; nor is it possible to reduce EEL's to
levels that presumably would protect every portion of the popu-
lation, regardless of size. The data needed to make such de-
cisions are not available.
*
Female value is based on 24 hours/day, 7 days/week, and 50
weeks/year over a lifetime of 77.8 years. Male value is
based on the same hours/day, days/week, and weeks per year
but over a lifetime of 69.9 years.
E-8
-------
Nevertheless, because environmental exposures, unlike work-
place exposures, affect a larger and more heterogenous popula-
tion, EEL's derived from workplace TLVs must strive to account
for and protect those portions of the population that are at
risk.
Based on a comparison of daily volumes of air breathed with
the body weights of four cohorts of the general population (i.e.,
adult males, adult females, children, and infants), airborne
contaminants present the greatest risk to a 10-year-old child
(Table E-l). A magnitude-of-exposure adjustment factor was
developed to account for this increased risk to a 10-year-old
child. Workplace TLVs are determined from data on adult males
4 "
(70-kg reference/man) with a daily air volume of 2.3 x 10
liters, which results in a ratio of 3.28 x 10 liters of air/kg
of body weight. The daily volume of air breathed by a child
4
(33-kg reference/10-year-old) is 1.5 x 10 liters, which results
in a ratio of 4.54 x 10 liters of air/kg body weight.
An adjustment factor of 0.72 (3.28 x 102/4.54 x 102) ac-
counts for the greater ventilation rate per unit body weight of a
10-year-old child compared with that of an adult male.
Safety Factor (Sf)
Despite attempts to adjust for differences in exposure
duration and to account for large population cohorts known to be
at the greatest risk, much uncertainty is still associated with
the estimated EEL's. The uncertainty associated with each EEL is
directly related to the paucity and quality of information used
E-9
-------
TABLE E-l. DAILY AIR VOLUMES, REFERENCE BODY WEIGHTS,
AND ESTIMATED ADJUSTMENT FACTORS FOR VARIATIONS IN THE
LEVEL OF EXPOSURE EXPERIENCED BY VARIOUS POPULATION COHORTS
Reference
individual
(cohort)
Adult male
Adult female
Child (10 years)
Infant (1 year)
Daily air
volume
breathed,
liters
2.3 x 104
2.1 x 104
1.5 x 104
0.38 x 104
Reference
body .
weight,
kg
70
58
-33C
-ioc
Exposure per
unit body
weight,
liters/kg
328
362
454
380
Ratio of adult male
value to value for
reference individual
(nondimensional )
1.00
0.91
0.72
0.86
Reference 13, p. 346. Daily air volumes breathed by adult men and women and
the 10-year-old child are based on 8 hours of working ("light activity"),
8 hours of nonoccupational activity, and 8 hours of resting. The value for
an infant is based on 8 hours of "light activity" and 16 hours of resting.
Reference 13, p. 13.
Reference 13, p. 11. Reference body weights for a 10-year-old child and a
one-year-old infant were taken as an average of both sexes for each age
group. In both age groups the actual sex-specific mean body weights vary
less than 0.5 kg from the values given above.
E-10
-------
in the definition of the original TLV's. An attempt has been
made to account for this source of uncertainty by including
safety factors in the estimation procedure. Table E-2 presents
the safety factors used to define the uncertainty associated with
specific conditions of information or experimental data. These
factors, which were developed for the determination of water
14
quality criteria, are also applicable to the estimating of
environmental exposure limits.
In the use of these safety factors it is recognized that the
amount and nature of the information available for establishing a
TLV varies from substance to substance. Within these limita-
tions, the information that forms the basis of the ACGIH documen-
tation and the nature of the illness or disease the TLV is de-
signed to protect against are presented in Table E-3. This table
also presents the safety factors that best describe the uncer-
tainty associated with each TLV.
Results
Table E-4 presents the estimated airborne EEL's (based on
Equation 1) for ten substances found in waste oil. Specific
adjustments were made for expected duration differences between
workplace and environmental exposures and exposures of a popula-
tion cohort at great risk. A safety factor was used to account
for the condition and quality of information used to develop the
workplace TLV's and for the type of protection they are meant to
provide.
E-ll
-------
TABLE E-2. UNCERTAINTY FACTORS ASSOCIATED WITH SPECIFIC
CONDITIONS OF THE EXPERIMENTAL DATA
Nature and conditions of experimental data'
Uncertainty ,
(safety) factor
Valid experimental results of chronic exposure
studies on man
Valid results of chronic exposure studies on experi-
mental animals; human exposure data limited to acute
studies
Acute exposure studies on experimental animals; no human
data available
10
100
1000
Data that present no indication of carcinogenicity.
Reference 14. Uncertainty factors developed by the National Academy of
Sciences during a study of Drinking Water and Health.
E-12
-------
TABLE E-3. SUMMARY OF ACGIH DOCUMENTATION FOR SPECIFIC TLV's AND SELECTED UNCERTAINTY (SAFETY) FACTORS14'15
Substance
Barium
Cadmium
compounds
Chromium
(II and III)
Hydrogen chloride
Lead
Naphthalene
Xylene
Zinc
(as zinc oxide)
ACGIH TLV
8-h TWA,
mg/m3
0.5
0.05
0.5
-7.0
(5 ppm)
0.15
50.0
~435
(100 ppm)
5.0
Type of information forming
basis of ACGIH documentation
Industrial experience related to
barium nitrate exposures
Epidemiological and occupational
exposure studies
Clinical studies of exposed
workers
Occupational exposure studies,
and animal studies
Occupational exposure studies,
clinical studies of exposed
workers, and animal studies
Industrial experience, occupa-
tional exposure studies, and
animal studies
Industrial experience, occupa-
tional exposure studies, clinical
studies of exposed workers, and
animal studies
Occupational exposure studies
and animal studies
Targeted
prevention
Excitability
Proteinuria, pulmonary
edema, and emphysema
Pulmonary edema and
irritation
Irritation
Encephalopathy and renal
damage
Irritation
Narcosis, chronic
Reduced incidence of
metal fume fever
Selected
uncertainty
(safety) factor
100
10
10
10
10
10a
10
10
w
I
(-J
Co
(Continued)
-------
TABLE E-3 (Continued)
Substance
Toluene
Trichloroethane
(1.1,1-)
ACGIH TLV
8-h TWA,
mg/m3
-375
(10 ppm)
-1900
(350 ppm)
Type of information forming
basis of ACGIH documentation
Occupational exposure studies,
clincial studies of exposed
workers, and animal studies
Occupational exposure studies,
clincial studies of exposed
workers, and animal studies
Targeted
prevention
Loss of muscle coordina-
tion and cardiomuscular
changes
Anesthetic effects and
objectionable odor
Selected
uncertainty
(safety) factor
10
10
There is some support within the scientific health effects community for applying a safety factor of
1 to those substances identified as irritants. This practice appears to be reasonable for those sub-
stances for which no other health effects have been observed. The ACGIH TLV for naphthalene was estab-
lished to protect against ocular irritation.** Although this end-opint is still a major concern, acute
exposures to airborne naphthalene are recognized to produce direct hemolytic effects in vivo, and oral
exposure may result in the development of cataracts.6 Because naphthalene exposures may result in toxic
end-points other than irritation, an uncertainty factor of 10 has been selected for use in determining
a TLV-derived EEL.
-------
TABLE E-4. ESTIMATED AIRBORNE ENVIRONMENTAL EXPOSURE LIMITS
Substance
Barium
Cadmium
Chromium (II and III)
Hydrogen chloride
Lead
Zinc
Naphthalene
Toluene
1,1,1-Trichloroethane
Xylene
Environmental
exposure level,
vg/m3
0.43
0.34
4.32
59.7
1.30e
43.2
432
3,240
16,416
3,758
The ambient air quality standard of 1.5 yg/m3
was used instead of the estimated environmental
exposure limit of 1.3 yg/m3.
E-15
-------
APPROACH USED TO DETERMINE REFERENCE CONCENTRATIONS FROM CARCINO-
GENIC POTENCY FACTORS
The reference concentrations provide reference points against
which to assess the relative impact of air quality on health and
to calculate the cancer risks attributable to that exposure; they
are neither estimates of safety nor statements of acceptable
levels of risk. The EPA procedures used to evaluate the toxico-
logical data were consistent with the Agency's objective of
estimating a risk so as to err on the side of safety. The
carcinogenic risk factors were developed from data sets that gave
the highest estimate of a lifetime cancer risk. This estimated
risk probably errs on the side of safety. The reference concen-
trations were determined from the carcinogenic potency factors
developed for the EPA Water Quality Criteria Documents and up-
dated in the Health Effects Assessment Summary for 300 Hazardous
Organic Constituents.
Chemicals eliciting a carcinogenic response are assessed by
use of a linear nonthreshold dose-response model. Use of this
model is based on the following assumptions: 1) a nonthreshold
relationship exists for carcinogens, 2) the dose-response rela-
tionship developed from animal and human studies at relatively
high exposure levels can be extrapolated to low exposure levels
likely to be experienced by the general public over an entire
lifetime, and 3) the dose-response relationship is linear. These
linear nonthreshold models are used by the Interagency Regulating
Liaison Group (IRLG) and the EPA Carcinogen Assessment Group
(CAG) ' to evaluate risks posed by potentially carcinogenic
substances.
E-16
-------
In this study, reference concentrations have been developed
by the use of the carcinogenic potency factor q,* and equivalent
dosage estimates.
The EPA developed the q * factors from lifetime animal
experiments or human epidemiological studies. ' Because of
the variety of studies accessed for data, EPA had to correct for
differences in metabolism between species and for variable ab-
sorption rates via different routes of administration. The
resulting q * factors are therefore based on exposures likely to
produce a given cancer incidence rate.
Table E-5 presents potency factors for carcinogenic sub-
stances found in waste oil. Equation 2 presents the method used
to derive airborne reference concentrations from the established
carcinogenic potency factors.
C = K (7° k X 103 (Eq. 2)
a qx* (20 mJ)
where C = reference concentration in air for a lifetime
a -53
risk to cancer of 10 , ug/m
K = risk level, 10~5
q,* = carcinogenic potency factor, risk per mg/kg per day
1 (Table B-6)
Again, the value of 70 kg represents the weight of a refer-
ence adult male. The value of 20 m is an estimate of the
total daily volume of air ventilated by an adult male. The
derived values for the airborne reference concentrations are
presented in Table E-6.
E-17
-------
TABLE E-5. CARCINOGENIC POTENCY FACTORS
Substance
Arsenic
Benzene
Cadmium
Chromium (VI)
Carbon tetrachloride
Polychlorinated biphenols (PCB's)
>
Tetrachloroethylene
1,1,2-Trichloroethane
Trichloroethylene
Risk, mg/kg per day"
14. Oa
0.52b
6.65b
41. Ob
0.13b
4.34a
0.0531b
0,0573a
0.0126a
Reference 10, p. 3.
Reference 17.
TABLE E-6. REFERENCE CONCENTRATIONS FOR A 10"5 RISK LEVEL
Substance
Arsenic
Benzene
Cadmium
Chromium (VI)
Carbon tetrachloride
Polychlorinated biphenols (PCBs)
Tetrachloroethylene
1,1,2-Trichloroethane
Trichloroethylene
Air,*
vg/m3
0.0025°
0.6731
0.0053
0.0008
0.2692
0.0081
0.6591
0.61
2.78
Airborne reference concentration was determined by use of published carcino-
genic potency factors (Table E-5) and a conversion methodology (Equation 2),
E-18
-------
REFERENCES FOR APPENDIX E
1. Stokinger, H. E., and R. L. Woodward. Toxicologic Methods
for Establishing Drinking Water Standards. J. American
Water Works Association, 50:515-529, 1958.
2. U.S. Environmental Protection Agency. Water Quality Criteria
Documents: Availability. Federal Register, 45 (231):79318-
79379.
3. American Conference of Governmental Industrial Hygienists.
TLV's Threshold Limit Values for Chemical Substances and
Physical Agents in the Work Environment With Intended Changes
for 1982. Cincinnati, Ohio. 1982.
4. American Conference of Governmental Industrial Hygienists.
Documentation of the Threshold Limit Values. 4th Ed.
Cincinnati, Ohio. 1982.
5. Report of the Interagency Regulatory Liaison Group, Work
Group for Risk Assessment. Federal Register, 44(131), July
6, 1979.
6. Casarett and Doull's Toxicology. The Basic Science of
Poisons, 2d Ed. MacMillian Publishing Co., Inc., New York,
New York. 1980. pp. 26-27.
7. Hickey, J. L. S., and P. C. Reist. Application of Occupa-
tional Exposure Limits to Unusual Work Schedules. American
Industrial Hygiene Association Journal 38:613-621, November
1977.
8. Hickey, J. L. S., and P. C. Reist. Adjusting Occupational
Exposure Limits for Moonlighting, Overtime, and Environmental
Exposures. American Industrial Hygiene Association Journal,
40:727-733, August 1979.
9. luliucci, R. L. 12-Hour TLV's. Pollution Engineering,
November 1982. pp. 25-27.
10. U.S. Environmental Protection Agency. Health Effects Assess-
ment Summary for 300 Hazardous Organic Constituents. Envi-
ronmental Criteria and Assessment Office, Cincinnati, Ohio.
August 18, 1982.
E-19
-------
11. U.S. Environmental Protection Agency. Multimedia Environ-
mental Goals for Environmental Assessment. Vol. 1.
EPA-600/7-77-136a, November 1977.
12. U.S. Department of Commerce, Statistical Abstract of the
United States: 1981. 102d Ed. Bureau of the Census,
Washington, D.C. 1981. p. 69.
13. International Commission on Radiological Protection No. 23.
Report of the Task Group on Reference Man. Pergamon Press,
New York. 1975. pp. 11, 13, 346.
14. U.S. Environmental Protection Agency. Water Quality Criteria
Documents: Availability. Federal Register, 45 (231) : 79353-
79355.
15. American Conference of Governmental Industrial Hygienists.
Documentation of the Threshold Limit Values. 4th Ed.
Cincinnati, Ohio. 1982.
16. U.S. Environmental Protection Agency. Interim Procedures
and Guidelines for Health Risk and Economic Impact Assess-
ments of Suspected Carcinogens. Federal Register, Vol. 41.
May 25, 1976.
17. Personal communication with Marie Pfaff, Carcinogen Assess-
ment Group, Office of Health and Environmental Assessment.
Assistant Administrator for Research and Development, U.S.
Environmental Protection Agency, Washington, D.C., July 7,
1983.
U.S. Environmental Protection Agency
Region V, Library
230 South Dehorn Street
Chicago, Illinois 60604
E-20
------- |