December 2012
Emission Estimation Protocol
for Iron and Steel Foundries
Version 1
Final
Submitted to:
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Submitted by:
RTI International
3040 Cornwallis Road
Research Triangle Park, NC 27709-2194
HRTI
INTERNATIONAL
-------
[This page intentionally left blank.]
-------
Version 1
Draft
Contents
CONTENTS
Section Page
List of Acronyms and Abbreviations vi
1. Introduction 1-1
1.1 Need for Emission Inventories 1-1
1.2 History of Foundry Emission Factors 1-2
1.3 Ranking of Emission Estimation Methods 1-2
1.4 Completeness 1-2
1.4.1 General Foundry Emission Sources and Pollutants 1-3
1.4.2 Special Considerations for PM Emission Inventories 1-11
1.4.3 Consideration of Capture Efficiencies 1-12
1.5 Data Quality 1-15
1.6 Calculations and Significant Digits 1-15
1.7 Sections of Protocol Document 1-16
2. Fugitive Dust Sources 2-1
2.1 Methodology for Material Handling 2-2
2.2 Methodology for Paved and Unpaved Roads 2-4
2.2.1 Methodology for Paved Roads 2-4
2.2.2 Methodology for Unpaved Roads 2-6
2.3 Methodology for Estimating HAP Metals Emissions from Fugitive Dust 2-7
3. Melting Operations 3-1
3.1 Melting Furnaces 3-1
3.1.1 Methodology Rank 1 for Melting Furnaces 3-2
3.1.2 Methodology Rank 2 for Melting Furnaces 3-5
3.1.3 Methodology Rank 3 for Melting Furnaces 3-8
3.1.4 Methodology Rank 4 for Melting Furnaces 3-9
3.2 Scrap Handling, Preparation and Preheating 3-19
3.2.1 Methodology Rank 1 and 2 for Scrap Handling, Preparation, and Preheating 3-20
3.2.2 Methodology Rank 3 for Scrap Handling, Preparation, and Preheating 3-20
3.2.3 Methodology Rank 4 for Scrap Handling, Preparation, and Preheating 3-20
3.3 Metallurgical Treatment of Molten Metal 3-23
3.3.1 Methodology Rank 1 and 2 for Metallurgical Treatment of Molten Metal 3-23
3.3.2 Methodology Rank 3 for Metallurgical Treatment of Molten Metal 3-23
3.3.3 Methodology Rank 4 for Metallurgical Treatment of Molten Metal 3-24
3.4 Holding Furnaces 3-24
3.4.1 Methodology Rank 1 and 2 for Holding Furnaces 3-25
3.4.2 Methodology Rank 3 for Holding Furnaces 3-25
3.4.3 Methodology Rank 4 for Holding Furnaces 3-25
4. Mold and Core Making 4-1
4.1 Methodology Ranks 1 and 2 for Mold and Core Making 4-2
4.2 Methodology Rank 3 for Mold and Core Making 4-5
4.3 Methodology Rank 4 for Mold and Core Making 4-11
4.3.1 Pollutant Emissions from Chemical Binder Systems 4-11
4.3.2 PM Emissions from Sand Handling Operations Associated with Mold and
Core Making 4-13
in
-------
Version 1
Draft Contents
5. Pouring, Cooling, and Shakeout 5-1
5.1 Methodology Ranks 1 and 2 for PCS Operations 5-2
5.2 Methodology Rank 3 for PCS Operations 5-4
5.3 Methodology Rank 4 for PCS Operations 5-6
5.3.1 Organic Pollutant Emission from PCS Operations 5-6
5.3.2 PM Emissions from PCS Operations 5-11
5.3.3 HAP Metal Emissions from PCS Operations 5-12
5.3.4 Other Criteria Pollutant Emissions from PCS Operations 5-12
6. Finishing Operations 6-1
6.1 Methodology Ranks 1 and 2 for Finishing Operations 6-2
6.2 Methodology Rank 3 for Finishing Operations 6-2
6.3 Methodology Rank 4 for Finishing Operations 6-2
6.3.1 Organic Pollutant Emission from Finishing Operations 6-3
6.3.2 PM Emissions from Finishing Operations 6-3
6.3.3 Metal HAP Emissions from Finishing Operations 6-5
7. References 7-1
Appendix A: Foundry Glossary Terms
Appendix B: Development of Emission Factors for Section 3 Melting Operations
Appendix C: Development of Emission Factors for Section 4 Mold and Core Making
Appendix D: Development of Emission Factors for Pouring, Cooling, and Shakeout
Appendix E: Development of Emission Factors for Finishing Operations
Appendix F: Control Efficiency and Particulate Matter Size Distribution
Appendix G: List of Suggested SCC for Iron and Steel Foundry Operations
LIST OF FIGURES
Number Page
1-1 Simplified schematic of a typical iron and steel foundry operation 1-3
LIST OF TABLES
Number Page
1-1 Summary of Pollutants and Emission Sources Inclusion in a Foundry's Emission
Inventory 1-4
IV
-------
Version 1
2-1 Default Values for Moisture Content for Fugitive Dust Emission Estimates from
Materials Handling 2-3
2-2 Particle Size Multipliers for Paved Road Equation 2-5
2-3 Default Values for Fugitive Dust Emission Estimates from Paved Roads 2-5
2-4 Constants for Equation 2-3 2-7
2-5 Default Values for Fugitive Dust Emission Estimates from Unpaved Roads 2-7
3-1 Summary of Typical Hierarchy of Melting Operations Emission Estimates 3-1
3-2 F-factor Exhaust Volumes and Heat Content of Common Fuels 3-6
3-3 Summary of PM Emission Factors for Melting Furnace Operations 3-11
3-4 Typical Collection Efficiencies of Various Particulate Control Devicesa (%) 3-12
3-5 Summary of Non-PM Criteria Pollutant Emission Factors for Melting Furnace Operations 3-15
3-6 Default Metal Composition for PM from Melting Furnace Operationsa 3-16
3-7 Congener-specific Profile for Ferrous Foundries 3-19
3-8 Particulate Emission Factors for Scrap and Charge Handling, Heating at Iron and Steel
Foundries 3-21
3-9 Particulate Emission Factors for Metallurgical Treatment at Iron and Steel Foundries 3-24
3-10 Particulate Emission Factors for Holding Furnaces at Iron and Steel Foundries 3-25
4-1 Summary of Typical Hierarchy of Mold and Core Making Emission Estimates 4-1
4-2 HAP Emitted from Chemical Binder Systems Used for Sand Cores and Molds (AFS and
CISA, 2007) 4-5
4-3 Default Content of Sand Binder System Components 1, 2 4-7
4-4 Default Pollutant Emission Factors for Sand Binders 4-11
4-5 Default PM Emission Factors for Sand Handling Operations Associated with Mold and
Core Making 4-14
5-1 Summary of Typical Hierarchy of Pouring, Cooling, and Shakeout Emission Estimates 5-2
5-2 VOC Emission Factors by Casting Type for PCS Operations 5-7
5-3 Default HAP Composition Profiles for PCS Operations 5-8
5-4 Summary of PM Emission Factors for PCS Lines 5-11
5-5 Summary of HAP Content of PM from PCS Components 5-12
5-6 Summary of Non-PM Criteria Pollutant Emission Factors for PCS Operations 5-13
6-1 Summary of Typical Hierarchy of Finishing Operations Emission Estimates 6-1
6-2 Default PM Emission Factors for Finishing Operations 6-4
-------
Version 1
Draft
List of Acronyms and Abbreviations
List of Acronyms and Abbreviations
(im
micrometers
Acfm
actual cubic feet per minute
AFS
American Foundry Society
Atm
atmosphere
BID
Background Information Document
Btu/mol
British thermal units per mole
Btu/scf
British thermal units per standard cubic feet
CAA
Clean Air Act
CEMS
continuous emission monitoring system
CERP
Casting Emission Reduction Program
CFR
Code of Federal Regulations
CO
carbon monoxide
CS
Core system
Cr+6
hexavalent chromium
dscf/MMBtu
dry standard cubic feet per million British thermal unit
dscfm
dry standard cubic feet per minute
EAF
Electric arc furnace
EIF
Electric induction furnace
EPA
U.S. Environmental Protection Agency
ESP
electrostatic precipitator
FID
flame ionization detector
GHG
greenhouse gases
HAP
hazardous air pollutants
HHV
Higher heating values
Hg
Mercury
hr/yr
hours per year
ICR
information collection request
in.
Inches
kg/kg-mol
kilogram per kilogram mole
kg/yr
kilograms per year
lb/dscf
pounds per dry standard cubic foot
lb/hr
pounds per hour
VI
-------
Version 1
Draft
List of Acronyms and Abbreviations
lb/kg
lb/ton
LOI
MACT
MDI
mg/kg
min/hr
Mph
MMBtu/scf
MS
MSDS
MW
Ni
NATA
NESHAP
Ni
NO
no2
NOx
02
OCDD
OCDF
OCMA
PAH
PCB
PCS
PM
PMio
PM10-FIL
PMio-PRI
pm25
pm25-fil
pounds per kilogram
pounds per ton
Loss on ignition
Maximum Achievable Control Technology
Methylene diphenyl diisocyanate
milligrams per kilogram
minutes per hour
Miles per hour
million British thermal units per standard cubic foot
Mold system
Material safety data sheets
molecular weight
Nickel
National-scale Air Toxics Assessment
National Emission Standards for Hazardous Air Pollutants
Nickel
nitric oxide
nitrogen dioxide
nitrogen oxides
Oxygen
octachloro-dibenzo-p-dioxin
octachloro-dibenzo-furan
Ohio Cast Metals Association
polycyclic aromatic hydrocarbons
polychlorinated biphenyls
Pouring, cooling, and shakeout
particulate matter
PM emissions that are 10 (.un in diameter or less
filterable (or front-half catch) portion of the PM emissions that are 10 |_im in
diameter or less
"primary" PM emissions that are 10 |_im in diameter or less
PM emissions that are 2.5 |_im in diameter or less
filterable (or front-half catch) portion of the PM emissions that are 2.5 |_im in
diameter or less
Vll
-------
Version 1
Draft List of Acronyms and Abbreviations
PM2 5-PRI "primary" PM emissions that are 2.5 (.un in diameter or less
PM-CON condensable PM (or back-half catch)
POM polycyclic organic matter
Ppmv parts per million by volume
ppmw parts per million by weight
Psi pounds per square inch
Psia pounds per square inch absolute
Psig pounds per square inch gauge
PUCB Phenolic urethane cold box
QA quality assurance
QC quality control
RMT Residuals Management Technology
SCC Standard Classification Code
SIP State implementation plan
S02 sulfur dioxide
TCDD tetrachlorodibenzo-p-dioxin
TDS total dissolved solids
TEA Triethyl amine
TEF Toxic Equivalent factors
TEQ toxic equivalents
THC total hydrocarbons
TOC total organic compounds
tons/hr tons per hour
tons/kg tons per kilogram
tons/yr tons per year
VMT Vehicle miles traveled
VOC volatile organic compounds
vol% volume percent
wt% weight percent
WHO World Health Organization
°C degrees Celsius
°F degrees Fahrenheit
°R degrees Rankine
Vlll
-------
Version 1
Draft
Section 1—Introduction
1. Introduction
This Foundry Emissions Protocol document is intended to provide guidance and instructions to iron and
steel foundry owners and operators for the purpose of improving emission inventories, and emission
reporting in general, by source owners and operators. This document presents a hierarchy of emission
measurement or estimation methods for various foundry emission sources and provides a listing of
pollutants that may be emitted by each source type. This document may also be used by other entities,
such as federal, state, and local agencies, for consistency in estimating emissions from iron and steel
foundries.
This Foundry Emissions Protocol document provides methods for estimating criteria pollutant and
hazardous air pollutant (HAP) emissions from foundry operations. Methodologies for estimating
emissions of greenhouse gases (GHG) are not included in this Foundry Emissions Protocol document.
Methodologies for estimating emissions of GHG are provided in the Mandatory Greenhouse Gas
Emissions Reporting Rule (Final Rule, 74 Federal Register [FR] 56260) for selected foundry processes.
1.1 Need for Emission Inventories
Air emission inventories are a fundamental first step in developing air quality and emission control
strategies. Section 172, Part C, of the Clean Air Act (CAA) as amended in 1990, which addresses state
implementation plan (SIP) requirements, states that ". . .plan provisions shall include a comprehensive,
accurate, current inventory of actual emissions from all sources or the relevant pollutants or pollutants in
such area . . ." Regulatory agencies and industrial facilities rely on emission inventories on an ongoing
basis as indicators of air quality changes and for setting permit requirements.
Section 112 of the CAA establishes a two-stage regulatory process to address emissions of hazardous air
pollutants (HAP) from stationary sources. In the first stage, Section 112 of the CAA directs the
Environmental Protection Agency (EPA) to develop technology-based "maximum achievable control
technology" (MACT) standards to control all major and some area sources emitting HAP. Emission
inventories were used to identify source categories with facilities that are major sources of HAP emission
(i.e., emit at least 10 tons per year (tons/yr) of a single HAP or 25 tons/yr of any combination of HAP).
Iron foundries and steel foundries were two of the listed source categories and MACT standards for major
source iron and steel foundries were established on April 22, 2004 (69 FR 21906). Section 112(k)(3)(B)
of the CAA calls for the EPA to identify at least 30 air toxics that pose the greatest potential health threat
in urban areas, and section 112(c)(3) requires the EPA to regulate the area source categories that represent
90 percent of the emissions of the 30 "listed" air toxics. Again, emission inventories were used to
identify these pollutants and source categories and iron and steel foundries were two of the listed source
categories. The EPA published national emission standards for area source iron and steel foundries (i.e.,
facilities that emit less than 10 tons per year (tons/yr) of a single HAP and less than 25 tons/yr of any
combination of HAP) on January 2, 2008 (73 FR 226).
The second stage in standard-setting focuses on reducing any remaining "residual" risk according to CAA
section 112(f). Section 112(f)(2) requires the EPA to determine for source categories subject to certain
section 112(d) standards, such as the foundry rules identified above, whether the emissions limitations
protect public health with an ample margin of safety. Once again, the emission inventory will be used as
a key input to assess the risk from iron and steel foundries and to determine if the MACT standards for
HAP provide an ample margin of safety to protect public health and prevent an adverse environmental
effect.
The importance placed on emission inventories requires that they be of the highest quality obtainable
considering their end use. Since they are the foundation of many air quality decisions, inventory quality is
critical to defining realistic regulations and attainment strategies. Deficiencies and inconsistencies in
1-1
-------
Version 1
Draft
Section 1—Introduction
existing compilation processes accentuate the need for developing and implementing more uniform and
systematic approaches to collecting and reporting data. One of the primary goals of this Foundry
Emissions Protocol document is to improve the quality of inventory data so that it is a reliable source of
information for sound decision making regarding the iron and steel foundry source categories.
1.2 History of Foundty Emission Factors
The EPA has developed air emission factors to support inventory development and publishes these factors
in the Compilation of air pollutant emission factors (commonly referred to as "AP-42") (U.S. EPA, 1995
and 2003). The emission factors available in AP-42 are primarily limited to criteria pollutants (although
lead is both a criteria pollutant and an HAP). In 1998, the EPA conducted a detailed survey of iron and
steel foundries to support HAP regulatory development. This data, along with other information collected
during this study, are summarized in the Foundry Background Information Document (BID) (U.S. EPA,
2002). The 2002 BID provides additional data regarding particulate matter (PM) emissions and some
default HAP emission factors for selected processes. For example, it provides typical chemical binder
compositions and usage rates by which HAP emissions were estimated for mold and core making
processes.
The methods and factors presented in this Foundry Emissions Protocol document rely heavily on factors
and methods established in AP-42 and the 2002 foundry BID. When emission factors were not available
for certain pollutants or sources in these documents, available source test data were reviewed to develop
default emission factors for these sources/pollutants. These default emission factors are expected to
represent emissions during normal operations. As many processes in the foundry operate in a batch or
cyclic nature, the default emission factors were developed, to the extent practical, to include the available
source test data occurring during typical cyclic operation. The emission factors developed specifically for
this Foundry Emissions Protocol document used source test data located in US EPA publications and peer
reviewed journals. Additional source test data was identified but not used due to the difficulty to assess
the representativeness of the process operations and to quality assure the test procedures. The test data
used was evaluated and compiled following the general guidelines presented in the 1997 Procedures for
Preparing Emission Factor Documents, EPA-454/R-95-015 (U.S. EPA, 1997). To accommodate future
emissions factors development efforts, the individual test data used in this Foundry Emissions Protocol
document will be submitted electronically using the alternative procedure described on the Electronic
Reporting Tool web page (http://www.epa.gov/ttn/chief/ert/index.html), as resources allow.
1.3 Ranking of Emission Estimation Methods
For each emission source, the various emission measurement or estimation methods specific to that
source are ranked in order of preference, with "Methodology Rank 1" being the preferred (most accurate)
method, followed by "Methodology Rank 2," and so on. The highest ranked method (with Methodology
Rank 1 being the highest) for which data are available should be used. Methodology Ranks 1 or 2
generally rely on continuous emission measurement data for pollutant concentrations and either
continuous measurement data for flow rates (Methodology Rank 1) or engineering calculations for flow
rates (Methodology Rank 2). When continuous measurement data are not available, but site-specific
emission source test data are available, emission estimates based on site-specific emission factors
(Methodology Rank 3) are specified. When site-specific measurement or test data are not available,
default emission factors (Methodology Rank 4) are provided.
1.4 Completeness
While this Foundry Emissions Protocol document attempts to identify and provide methodologies for
each emission source at a typical foundry, there may be certain sources located at the foundry facility
(i.e., that are owned or under the common control of the foundry owners or operators) that are not
specifically addressed within the Foundry Emissions Protocol document. Additionally, there are sources
1-2
-------
Version 1
Draft
Section 1—Introduction
included in this Foundry Emissions Protocol document for which no emission data are available to
provide default emission factors (Methodology Rank 4) methods. The absence of a default emission
factor is not necessarily sufficient grounds to exclude these sources from an emissions inventory. The
emission inventory should be as complete as possible. If there are known emission release points within
the facility for which no default emission factors are available, the emissions from these sources should
be estimated using available process data, product knowledge, and engineering judgment, and these
emission estimates should be included in the facility's emissions inventory.
1.4.1 General Foundry Emission Sources and Pollutants
There are five or six primary operations conducted at a typical iron and steel foundry, as illustrated in
Figure 1-1. Foundry operations start with two parallel paths: one path includes the scrap and material
handling and metal melting, and one path includes mold and core making. The two paths merge and
continue the production with pouring of the molten metal into the mold, followed by cooling of the
casting, and separation of the solid casting from the mold (commonly referred to as ""shakeout" or
"knockout")- For sand mold foundries, sand is typically recovered from the shakeout operations,
reconditioned, and reused in the mold and core making operations. The cast metal recovered from the
shakeout operations goes to the cutting, grinding, finishing, and cleaning processes or can also be reused
in the melting operations. Additional description of foundry processes and their emission points is
provided in the foundry BID (U.S. EPA, 2002).
Figure 1-1. Simplified schematic of a typical iron and steel foundry operation.
The types of pollutants emitted vary by the type of foundry operation. Table 1-1 provides a listing of the
pollutants expected to be emitted by various sources at iron and steel foundries described in this protocol
1-3
-------
Version 1
Draft
Section 1—Introduction
document. While Table 1-1 is intended to provide a comprehensive list of pollutants for each foundry
emission source, the inventory developer should take into consideration facility-specific operations or
conditions. In particular, emissions from mold and core making can vary widely based on the chemical
binder system used at the facility. There may be numerous pollutants identified in Table 1-1 as being
potentially emitted from mold- and core-making operations that are not released by a given facility based
on the binder system employed. Likewise, it is possible that there may be pollutants that are not indicated
in Table 1-1, but that are released from a given binder system. Thus, Table 1-1 should be used as a guide
to identify pollutants to include in the emissions inventory, but site-specific factors should also be
evaluated in developing a final pollutant list for the emissions inventory. To the extent possible, pollutant
groupings such as volatile organic compounds (VOCs) and particulate matter (PM) should be speciated
for a comprehensive and accurate emission inventory.
Table 1-1. Summary of Pollutants and Emission Sources Inclusion
in a Foundry's Emission Inventory
Melting
Pouring/ Cooling
CAS
Number or
Pollutant
Code
Substance
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
Criteria Pollutants
630-08-0
Carbon monoxide
•
•
O
7439-92-1
Lead
•
•
•
PM-PRI
Particulate matter (PM)
of any particle size
PM-FIL
Filterable PM
PM10-PRI
PM < 10 micrometers
(Mm)
PM10-FIL
Filterable PM < 10 |jm
PM25-PRI
PM < 2.5 |jm
PM25-FIL
Filterable PM < 2.5 |jm
PM-CON
Condensable PM
10102-44-0
Nitrogen dioxide
O
•
•
O
NOx
Nitrogen oxides
O
•
•
o
7446-09-5
Sulfur dioxide
o
•
•
o
VOC
Volatile organic
compounds
o
•
•
o
Specific VOC Constituents (Compounds listed below plus those listed under "Volatile Organic HAPs")
74-85-1
Ethylene
o
O
O
o
O
O
O
O
O
O
O
O
O
74-86-2
Acetylene
o
o
O
o
O
O
O
O
O
O
O
O
O
74-98-6
Propane
o
o
o
o
o
O
o
o
o
o
o
o
O
1-4
-------
Version 1
Draft
Section 1—Introduction
Melting
Pouring/ Cooling
CAS
Number or
Pollutant
Code
Substance
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
115-07-1
Propylene
O
o
O
O
O
O
O
O
O
O
O
O
O
463-49-0
Propadiene
O
o
o
O
O
O
o
o
o
O
o
O
O
106-97-8
n-Butane
o
o
o
o
o
O
o
o
o
o
o
o
O
75-28-5
Isobutane
o
o
o
o
o
O
o
o
o
o
o
o
O
106-98-9
1-Butene
o
o
o
o
o
O
o
o
o
o
o
o
O
107-01-7
2-Butene
o
o
o
o
o
O
o
o
o
o
o
o
O
115-11-7
Isobutene
o
o
o
o
o
O
o
o
o
o
o
o
O
590-19-2
1,2-Butadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
109-66-0
n-pentane
o
o
o
o
o
O
o
o
o
o
o
o
O
78-78-4
2-Methylbutane
o
o
o
o
o
O
o
o
o
o
o
o
O
287-92-3
Cyclopentane
o
o
o
o
o
O
o
o
o
o
o
o
O
591-95-7
1,2-Pentadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
1574-41-0
1-cis-3-Pentadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
2004-70-8
1 -trans-3-Pentadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
591-93-5
1,4-Pentadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
591-96-8
2,3-Pentadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
598-25-4
3-Methyl-1,2-butadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
78-79-5
2-Methyl-1,3-butadiene
o
o
o
o
o
O
o
o
o
o
o
o
O
542-92-7
Cyclopentadiene
o
o
o
o
o
O
o
o
o
o
o
o
o
110-82-7
Cyclohexane
o
o
o
o
o
O
o
o
o
o
o
o
o
108-87-2
Methylcyclohexane
o
o
o
o
o
O
o
o
o
o
o
o
o
142-82-5
Heptane (and isomers)
o
o
o
o
o
O
o
o
o
o
o
o
o
111-65-9
Octane (and isomers)
o
o
o
o
o
O
o
o
o
o
o
o
o
67-64-1
Acetone
o
o
o
o
o
o
o
o
o
o
o
o
o
78-93-3
Methyl ethyl ketone
o
o
o
o
o
o
o
o
o
o
o
o
o
25551-13-7
T rimethylbenzene(s)
o
o
o
o
o
o
o
o
o
o
o
o
o
Hazardous Air Pollutants (HAPs)
Volatile Organic HAPs
75-07-0
Acetaldehyde
o
o
o
o
•
•
•
o
•
•
•
o
•
107-02-8
Acrolein
o
o
o
o
o
o
o
o
o
o
o
o
o
1-5
-------
Version 1
Draft
Section 1—Introduction
CAS
Number or
Pollutant
Code
Substance
Melting
Pouring/ Cooling
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
62-53-3
Aniline
O
o
O
O
•
O
O
O
O
O
O
O
O
71-43-2
Benzene
O
o
o
O
•
•
•
O
•
•
•
O
O
67-66-3
Chloroform
o
o
o
o
O
O
o
o
o
o
o
o
O
74-87-3
Chloromethane
o
O
98-82-8
Cumene
o
o
o
o
O
O
o
o
o
o
o
o
O
111-42-2
Diethanolamine
O
100-41-4
Ethylbenzene
o
o
o
o
•
•
•
o
•
•
•
o
O
50-00-0
Formaldehyde
o
o
o
o
•
•
•
o
•
•
•
o
•
110-54-3
n-Hexane
o
o
o
o
•
•
•
o
•
•
•
o
•
67-56-1
Methanol
o
o
o
o
o
•
•
o
•
•
•
o
•
108-10-1
Methyl isobutyl ketone
o
o
o
o
o
o
o
o
o
o
o
o
o
100-42-5
Styrene
o
o
o
o
•
o
o
o
o
•
o
o
o
79-34-5
1,1,2,2-
Tetrachloroethane
o
127-18-4
Tetrachloroethylene
o
108-88-3
Toluene
o
o
o
o
•
•
•
o
•
•
•
o
o
79-01-6
Trichloroethylene
o
121-44-8
Triethylamine
o
95-47-6
o-Xylene
o
o
o
o
•
•
•
o
•
•
•
o
o
108-38-3
m-Xylene
o
o
o
o
•
•
•
o
•
•
•
o
o
106-42-3
p-Xylene
o
o
o
o
•
•
•
o
•
•
•
o
o
1330-20-7
Xylenes (total)
o
o
o
o
•
•
•
o
•
•
•
o
o
Semi-volatile and Non-volatile Organic HAPs (except POMS, dioxins, furans, andpolychlorinated biphenyls [PCBs])
108-39-4
m-Cresol
o
•
o
o
o
o
o
o
o
95-48-7
o-Cresol
o
•
o
o
o
o
o
o
o
106-44-5
p-Cresol
o
•
o
o
o
o
o
o
o
1319-77-3
Cresols (total)
o
•
o
o
o
o
o
o
o
Glycol ethersf
o
101-68-8
4,4'-Methylene
diphenyl diisocyanate
(MDI)
o
o
o
o
o
o
o
o
1-6
-------
Version 1
Draft
Section 1—Introduction
Melting
Pouring/ Cooling
CAS
Number or
Pollutant
Code
Substance
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
26447-40-
5*
4,4'-
Methylenebis(phenyl
isocyanate) [aka, MDI]
O
O
O
O
O
O
O
O
108-95-2
Phenol
O
O
O
O
•
•
•
o
•
•
•
O
O
POMs: Compounds that meet the HAP definition of polycyclic organic matter (POM). These compounds are often referred
to as polycyclic aromatic hydrocarbons or "PAH."
83-32-9
Acenaphthene 16"PAH
o
o
O
o
o
o
o
o
208-96-8
Acenaphthylene 16"PAH
o
o
o
o
o
o
o
o
120-12-7
Anthracene 1&PAH
o
o
o
o
o
o
o
o
56-55-3
Benzo(a)anthracene 7"
PAH, 16-PAH
o
o
o
o
o
o
o
o
50-32-8
Benzo(a)pyrene 7"PAH'16"
PAH
o
o
o
o
o
o
o
o
205-99-2
Benzo(b)fluoranthene7"
PAH, 16-PAH
o
o
o
o
o
o
o
o
192-97-2
Benzo(e)pyrene
o
o
o
o
o
o
o
o
191-24-2
Benzo(g,h,i)perylene16"
PAH
o
o
o
o
o
o
o
o
207-08-9
Benzo(k)fluoranthene7"
PAH, 16-PAH
o
o
o
o
o
o
o
o
92-52-4
Biphenyl
o
o
o
o
o
o
o
o
91-58-7
2-Chloronaphthalene
o
o
o
o
o
o
o
o
218-01-9
Chrysene7"PAH'16"PAH
o
o
o
o
o
o
o
o
53-70-3
Dibenz(a,h)
anthracene7"™'1&PAH
o
o
o
o
o
o
o
o
57-97-6
7,12-Dimethylbenz(a)
anthracene
o
o
o
o
o
o
o
o
206-44-0
Fluoranthene 16"PAH
o
o
o
o
o
o
o
o
86-73-7
Fluorene 16"PAH
o
o
o
o
o
o
o
o
193-39-5
lndeno(1,2,3-cd)
pyrene PAH' 16
o
o
o
o
o
o
o
o
56-49-5
3-Methylcholranthrene
o
o
o
o
o
o
o
o
91-57-6
2-Methyl naphthalene
O
o
o
O
•
•
•
o
•
•
•
O
O
218-01-9
Chrysene
o
o
o
o
o
o
o
o
91-20-3
Naphthalene 16"PAH
o
o
o
o
•
•
•
o
•
•
•
o
O
1-7
-------
Version 1
Draft
Section 1—Introduction
CAS
Number or
Pollutant
Code
Substance
Melting
Pouring/ Cooling
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
198-55-0
Perylene
o
O
O
O
O
O
O
O
85-01-8
Phenanthrene 16"PAH
o
O
O
O
O
O
O
O
129-00-0
Pyrene 16"PAH
o
o
O
o
o
o
o
o
Dioxins/Furans/PCBs
1746-01-6
2,3,7,8-
Tetrachlorodibenzo-p-
dioxinJ
O
o
O
O
1746-01-6
2,3,7,8-
Tetrachlorodibenzo-p-
dioxin
O
o
o
O
40321-76-4
1,2,3,7,8-
Pentachlorodibenzo-p-
dioxin
o
o
o
o
39227-28-6
1,2,3,4,7,8-
Hexachlorodibenzo-p-
dioxin
o
o
o
o
57653-85-7
1,2,3,6,7,8-
Hexachlorodibenzo-p-
dioxin
o
o
o
o
19408-74-3
1,2,3,7,8,9-
Hexachlorodibenzo-p-
dioxin
o
o
o
o
35822-46-9
1,2,3,4,6,7,8-
Heptachlorodibenzo-p-
dioxin
o
o
o
o
3268-87-9
Octachlorodibenzo-p-
dioxin
o
o
o
o
132-64-9
Dibenzofuransk
o
o
o
o
51207-31-9
2,3,7,8-
Tetrachlorodibenzofura
n
o
o
o
o
57117-41-6
1,2,3,7,8-
Pentachlorodibenzofur
an
o
o
o
o
57117-31-4
2,3,4,7,8-
Pentachlorodibenzofur
an
o
o
o
o
70648-26-9
1,2,3,4,7,8-
Hexachlorodibenzofura
n
o
o
o
o
1-8
-------
Version 1
Draft
Section 1—Introduction
CAS
Number or
Pollutant
Code
Substance
Melting
Pouring/ Cooling
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
57117-44-9
1,2,3,6,7,8-
Hexachlorodibenzofura
n
O
o
O
O
72918-21-9
1,2,3,7,8,9-
Hexachlorodibenzofura
n
o
o
o
O
60851-34-5
2,3,4,6,7,8-
Hexachlorodibenzofura
n
o
o
o
o
67562-39-4
1,2,3,4,6,7,8-
Heptachlorodibenzofur
an
o
o
o
o
55673-89-7
1,2,3,4,7,8,9-
Heptachlorodibenzofur
an
o
o
o
o
39001-02-0
Octachlorodibenzofura
n
o
o
o
o
1336-36-3
Polychlorinated
biphenyls (total)
o
o
o
o
Metal HAPs
7440-36-0
Antimony
O
o
o
o
O
o
O
O
O
O
O
O
O
O
O
O
O
7440-38-2
Arsenic
O
o
o
o
o
o
o
O
O
O
O
O
O
O
o
O
o
7440-41-7
Beryllium
o
o
O
O
o
7440-43-9
Cadmium
o
o
O
O
o
18540-29-9
Chromium (hexavalent)
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
O
o
o
7440-47-3
Chromium (total)
o
o
O
o
o
7440-48-4
Cobalt
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
O
o
o
7439-92-1
Lead
o
o
o
o
o
7439-96-5
Manganese
o
o
o
o
o
7439-97-6
Mercury
O
o
o
o
o
o
o
o
o
o
7440-02-0
Nickel
o
o
o
o
o
7782-49-2
Selenium
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
o
o
o
Other Inorganic HAPs
75-15-0
Carbon disulfide
o
463-58-1
Carbonyl sulfide
o
1-9
-------
Version 1
Draft
Section 1—Introduction
CAS
Number or
Pollutant
Code
Substance
Melting
Pouring/ Cooling
Shakeout/ Knockout
Finishing/ Grinding
Cleaning/ Coating
Mold & Core Making
Sand Handling
Materials Handling
Scrap Management
Slag Piles
Electric Induction
Electric Arc
Cupola
Reverberatory
Crucible
Scrap Preheating
Inoculation
Green Sand Molds
Chemically-bonded
Sand Molds
Centrifugal Casting
Permanent Casting
Investment Casting
Expendable
Pattern Casting
7647-01-0
Hydrogen chloride
O
o
O
O
O
O
O
O
O
74-90-8
Hydrogen cyanide (&
cyanide compounds)
o
o
O
O
O
O
o
O
o
7723-14-0
Phosphorus
O
o
o
O
O
O
O
o
O
o
o
o
o
o
O
Other Compounds of Interest
74-84-0
Ethane
o
o
o
O
o
O
o
o
o
o
o
O
O
74-82-8
Methane
o
o
o
o
o
O
o
o
o
o
o
O
O
7440-39-3
Barium
O
o
o
o
o
o
o
o
O
o
o
o
o
o
O
O
O
O
7440-50-8
Copper
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
O
O
o
7439-95-4
Magnesium
•
O
7439-98-7
Molybdenum
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
O
O
o
7440-62-2
Vanadium
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
O
O
o
7440-66-6
Zinc
o
o
o
o
o
o
o
o
O
o
o
o
o
o
o
o
o
o
598-56-1
N,N-
dimethylethylamine
(DMEA)
o
O
996-35-0
N,N-
dimethylisopropylamine
(DMIPA)
o
107-21-1
Ethylene glycol
O
78-59-1
Isophorone
o
o
O
o
o
o
o
o
16984-48-8
Fluorides
•
•
•
•
•
• Designates pollutants expected to be emitted from foundry unit,
o Designates pollutants potentially emitted from foundry unit.
7 PAH Designates compounds part of the list of seven polycyclic aromatic hydrocarbons
16"PAH Designates compounds part of the list of sixteen polycyclic aromatic hydrocarbons
* While the HAP list only includes CAS No. 101-68-8 for "Methylene diphenyl diisocyanate (MDI)," some chemical
manufacturers may label this product under CAS No. 26447-40-5, but these are the same compounds.
' The listed HAP is 2,3,7,8-Tetrachlorodibenzo-p-dioxin (2378-TCDD); other dioxin isomers are listed because they can be used
to calculate a 2378-TCDD toxicity equivalence.
k The listed HAP is dibenzofurans; other furan isomers are listed because they can be used to calculate furan toxicity
equivalence.
f Includes mono- and di- ethers of ethylene glycol, diethylene glycol, and triethylene glycol R-(OCH2CH2)n -OR' where
n = 1, 2, or 3
R = alkyl or aryl groups
R' = R, H, or groups which, when removed, yield glycol ethers with the structure: R-(OCH2CH)n-OH. Polymers are excluded from
the glycol category.
1-10
-------
Version 1
Draft
Section 1—Introduction
Some criteria pollutants, such as PM10 or PM2 5 have special reporting nomenclatures to indicate the size
range of the PM and the fraction of the PM emissions that are filterable or condensable. Other criteria
pollutants, such as nitrogen dioxide (N02), are often determined or regulated as a combination of
chemicals. For example, some foundries may have emission limits in terms of nitrogen oxides (NOx),
which is the sum of N02 and nitric oxide (NO) emissions. The inclusion of these additional
nomenclatures or groupings in Table 1-1, above, is not intended to suggest that these compounds are
criteria pollutants, but that these "pollutants" are emitted by foundries and should be included for
completeness in the emissions inventory.
1.4.2 Special Considerations for PM Emission Inventories
This section provides specific information to help inventory developers better understand PM
measurements and PM emission inventory nomenclature. In particular, there are a variety of PM test
methods that may be used to assess PM emissions and it is important to understand what is measured by
the different test methods so that the test results can be used properly in developing emission inventory
estimates. As noted previously, there is a special reporting nomenclature for a PM inventory; the PM
Emission Inventory Nomenclature text box below provides an overview of this nomenclature. A
complete PM emissions inventory would include emission estimates for each of the PM2 5 and PM10
fractions listed in this text box. Following is a discussion of typical PM test methods and how their
measurements relate to this PM nomenclature.
EPA Method 5 (including its variations in Methods 5A through 51) is the most commonly used test
method for measuring PM emission from stationary sources. A typical Method 5 sampling train consists
of a sampling probe, a heated line and filter, and a series of impingers that are kept in an ice bath.
Method 5 sampling measures PM that is contained in the sampling probe and filter, which is often
referred to as the "front-half or "filterable" PM catch. PM that condenses in the impinger section of the
sampling train is often referred to as the "back half' catch or the "condensable" PM. The front-half catch
of Method 5 actually measures total filterable PM (PM-FIL) as there is no prefilter or cyclone to remove
particulates with a mean aerodynamic diameter of greater than 10 |_im. Nonetheless, Method 5 results are
often reported as PMi0. For controlled sources, this may be a reasonable estimate, but it is likely to
overestimate the actual PM10-FIL emission for some sources that might emit larger particles. If EPA
Method 5 test data are used, it is good practice to consider this PM-FIL and use the size distribution
guidance provided in this Foundry Emissions Protocol document.
PM Emission Inventory Nomenclature
PM emissions inventories have their own nomenclature and structure. A complete PM emissions inventory
includes the following components:
¦ PM-io-PRI: "Primary" PM emissions that are 10 |jm in diameter or less. PM-io-PRI = PM-io-FIL + PM-CON.
¦ PM o-FIL: Filterable (or front-half catch) portion of the PM emissions that are 10 |jm in diameter or less.
¦ PM-CON: Condensable PM (or back-half catch). All condensable PM is assumed to be less than 2.5 microns
(|jm) in diameter (PM2.s)-
¦ PM2.5-PRI: "Primary" PM emissions that are 2.5 |jm in diameter or less. PM2.5-PRI = PM25-FIL + PM-CON.
¦ PM2.5-FIL: Filterable (or front-half catch) portion of the PM emissions that are 2.5 |jm in diameter or less.
Although a complete PM emissions inventory includes PM emissions that are 10 |jm in diameter or less, some
measurement methods also collect PM particles that are greater than 10 |jm in diameter. The following
nomenclature is used to designate PM emissions that include PM greater than 10 |jm in diameter:
¦ PM-PRI: "Primary" PM emissions of any particle size. PM-PRI = PM-FIL + PM-CON.
¦ PM-FIL: Filterable (or front-half catch) portion of the PM emissions of any particle size.
1-11
-------
Version 1
Draft
Section 1—Introduction
EPA Method 17 is similar to Method 5, except that the filter in the Method 17 sampling probe is within
the stack so that the "filterable" PM content is measured at the stack temperature. EPA Methods 201 and
201A are similar to Method 17, except that there is also a cyclone or other sizing device to remove
particles greater than 10 |_im in diameter prior to the filter so that Methods 201 and 201A determine PM10-
FIL directly. Any of the Method 5, 17, or 201 variant methods describe only the procedures to determine
the front-half or filterable PM catch. EPA Method 202 specifies the procedures to determine the mass of
PM that condenses in the impingers (i.e., PM-CON).
Although Method 202 generally references the use of Method 17 (or 201 or 201 A) sampling trains, it may
also be used in conjunction with EPA Methods 5, 5B, or 5F. As the filter temperature in Method 5 is
different from Method 5B and also likely different from the filter temperature when using Method 17, the
fraction of PM that is filterable versus condensable, which is a function of the sampling temperature, will
also vary depending on the PM sampling method used. To further complicate PM-CON measurements,
test specifying the use of Method 202 that were conducted prior to 2010 would measure the PM that
condensed in the wet impingers. In 2010, the EPA revised Method 202 to include a dry impinger for the
measurement of PM-CON. All Method 202 data used to develop PM-CON emission factors presented in
this Foundry Emissions Protocol document are based on the wet impinger method,
1.4.3 Special Considerations for VOC Emission Inventories
This section provides specific information to help inventory developers better understand VOC
measurements. As defined in 40 CFR 51.100, VOC means "any compound, excluding carbon monoxide,
carbon dioxide, carbonic acid, metallic carbides or carbonates, and ammonium carbonate, which
participates in atmospheric photochemical reactions." Certain compounds, which have been determined to
have negligible photochemical reactivity, are specifically excluded from the definition of VOC, including
methane, ethane, methylene chloride, 1,1,1-trichloroethane, and a number of chlorofluorocarbons. There
are a variety of test methods that may be used to assess VOC emissions, and it is important to understand
what is measured by the different test methods so that the test results can be used properly in developing
emission inventory estimates.
EPA Method 25 measures total gaseous nonmethane organics. This method uses a gas chromatography
to separate carbon monoxide, carbon dioxide, and methane from the gas sample, oxidizing the remaining
compounds to carbon dioxide, reducing this carbon dioxide to methane, and measuring the methane using
a flame ionization detector (FID). EPA Method 25 results are typically reported "as carbon" in units of
either mass carbon per unit dry gas volume or parts per million by volume (ppmv). Since EPA Method
25 converts the organics to methane prior to analysis, it provides an accurate measure of the mass of
carbon without needing to consider the variable response of the FID to different types of organics. EPA
Method 25 will include response from ethane, so EPA Method 25 could overestimate VOC
concentrations if ethane is present. However, EPA Method 25 is more likely to underestimate the mass
concentration of VOC compounds because it considers only the mass of carbon and does not include the
mass of hydrogen, oxygen, or other elements that may be in the gaseous compounds.
EPA Method 25A determines the total gaseous organic concentration of a sample by direct FID analysis;
EPA Method 25B is similar to EPA Method 25A, but uses a nondispersive infrared analyzer. The results
from either method are reported on a volume concentration equivalent to the calibration gas, e.g., "ppmv
as propane" or "ppmv as methane." While different calibration gases have different FID responses, it is
common practice to convert concentrations based on the number of carbon atoms, so a concentration
reported as 5 ppmv as propane would be equivalent to 15 ppmv as methane (or sometimes reported "as
carbon" or "as carbon equivalents"). It is best to use the concentration as measured for the calibration gas
and the molecular weight of the calibration gas to convert the concentration to a mass per volume
estimate. EPA Method 25 A will typically have a high bias as a measure of VOC if methane and ethane
1-12
-------
Version 1
Draft
Section 1—Introduction
are present in the gas stream, since the FID analyzer response will include responses for these non-VOC
compounds.
EPA Method 18 uses a gas chromatograph to separate the gaseous compounds and uses an FID or another
type of analyzer to determine the concentration of each separate compound. The concentrations are
generally reported in volumetric concentrations (i.e., ppmv), as determined for each specific compound.
Therefore, the volumetric response of the analyzer to a specific compound is specifically determined
when using EPA Method 18. It should be noted that Method 18, as implemented by many testers, may not
be sufficient to measure all of the VOC present in the gas stream. The reason is that not all reagents (e.g.,
charcoals, silica gells, liquid media) retain and then release for analysis all organic compounds present in
the gas stream. In addition, not all analytical finishes (analyzers) are sensitive to all of the organic
compounds. As a result, a pre-test survey usually is required to select the correct reagent(s) and analytical
finish(es) such that an estimated 90+% of the compounds are measured. In addition, Method 18 is
typically not used for more than about five individual compounds, and as a result, may yield lower VOC
estimates than EPA Method 25 or 25A.
When calculating VOC emissions, it is important to note the test method used and the units for which
these VOC concentration data are reported. Volume concentration data will need to be converted to a
mass basis to determine the mass VOC emissions. If EPA Method 18 is used, the molecular weight of the
specific compound is needed to convert the volume concentration to a mass concentration. If EPA
Method 25A or 25B is used and the concentrations are reported "as propane," then the molecular weight
of propane should be used to convert the concentration to a mass basis, regardless of the molecular weight
of the compounds actually present in the gas stream. If separate methane and ethane concentrations are
available, it is possible to correct the total gaseous organic carbon results from EPA Methods 25A or 25B
to VOC concentrations by subtracting out these concentrations, but care must be taken to ensure the units
of measure are consistent (i.e., ensure all concentrations are reported "as propane," for example, or
convert the concentrations to "as carbon equivalents" before adjusting the Method 25A/25B results). If a
VOC concentration from EPA Method 25 is reported in a mass concentration "as carbon," this
concentration can be used directly determine the mass emission of VOC "as carbon." Alternatively, if the
typical hydrocarbon composition is known, then the concentration "as carbon" can be converted to a mass
concentration "as VOC" by multiplying by the molecular weight of the representative compound divided
by the mass of carbon per mole of representative compound. For example, if a gas stream contained
mostly aromatic compounds reasonably represented by toluene (C7H8), then the mass concentration "as
carbon" can be converted to a mass concentration "as VOC" by multiplying by 1.095 [(7x 12+8)/(7x 12)].
These examples illustrate the need to understand the differences in VOC measurement methods and how
these differences may impact the VOC emissions calculations.
1.4.4 Consideration of Capture Efficiencies
Many of the emission estimation methodologies provided in this Foundry Emissions Protocol document
consider controlled and uncontrolled emission estimates. Control efficiencies are provided by which
uncontrolled emission factors can be used to determine the controlled emissions based on the type of
control device used. As a general rule, the examples provided in this document apply primarily to either
uncontrolled sources or systems that have excellent capture systems venting the emissions to a control
device. For systems with less than 99 percent capture efficiency, the emission inventory should
specifically account for the uncaptured emissions. This section describes the basic methods for
estimating these emissions.
Methodology Ranks 1, 2, or 3 will directly account for the emissions vented to the atmosphere from the
capture and control system. If the capture efficiency is low or moderate, a significant portion of the
source's emissions may not be measured. If the capture system does not contain a control device or if a
control device is present but is not expected to reduce the emissions of a measured pollutant (e.g., if VOC
1-13
-------
Version 1
Draft
Section 1—Introduction
is measured at the outlet of a baghouse), then the measured emissions can be adjusted for the capture
efficiency directly as shown in Equation 1-1.
-¦Total-
where:
E'[Oit;i
E\Icms
CapEff
CapEff
Eqn. 1-1
total emissions for the source, pounds per hour (lbs/hr)
the emissions measured within the stack for the source, lbs/hr
estimated capture efficiency, mass fraction
Thus, if the emissions rate from the stack of an uncontrolled source was measured to be 10 lbs/hr and the
capture efficiency of the hooding is estimated to be 70 percent, the total emissions would be 10/0.7 or
14 lbs/hr.
If the capture system does have a control device that would influence the emission rate of the measured
pollutant, then the total emissions would be calculated using Equation 1-2.
^Total- Ejvieas ^Uncap X (1 — CapEff)
Eqn. 1-2
where:
ETotal = total emissions for the source, pounds per hour (lbs/hr)
EMeas = the emissions measured within the stack for the source, lbs/hr
Euncap = the emissions expected from the source if it were entirely uncaptured, lbs/hr
CapEff = estimated capture efficiency, mass fraction
The uncontrolled emissions from the source can be estimated using the default uncaptured emission factor
for the source. Alternatively, the measured emissions can be divided by the expected control device
efficiency and the capture efficiency to "back-calculate" the uncontrolled emissions from the source.
Note for the PM sources, where not all of the uncaptured emissions are expected to be emitted to the
atmosphere (i.e., settling of PM within the foundry building), back-calculating the emissions from the
measured stack emissions may overstate the total emissions to the atmosphere.
When emissions are estimated based on emission factors, a general approach for estimating the total
emissions is presented by Equation 1-3.
^Total ^Uncap x (1 CapEff) + Ecap x CapEff
Eqn. 1-3
where:
ETotal = total emissions for the source, pounds per hour (lbs/hr)
Euncap = the emissions expected form the source if it were entirely uncaptured, lbs/hr
ECap = the emissions expected from the source if it were entirely captured and emitted through
the control device, if applicable, lbs/hr
CapEff = estimated capture efficiency, mass fraction
Equation 1-3 will account for the difference in PM emissions when settling occurs within the foundry
building area. For organic emissions, the uncaptured emissions are the same as the captured but
uncontrolled emissions. In this special case, the total emissions can be calculated using Equation 1-4.
1-14
-------
Version 1
Draft
Section 1—Introduction
where:
E'[Oit;i
Euncntrld
CapEff
CE,
^Total- ^uncntrld * (1 CapEff X CEj)
Eqn. 1-4
total emissions for the source, pounds per hour (lbs/hr)
the emissions expected from the uncontrolled source (for pollutants that have the same
uncontrolled emissions regardless of presence of a capture system), lbs/hr
estimated capture efficiency, mass fraction
estimated control efficiency of the control device, if present, for pollutant "i", mass
fraction
The methods to calculate the various emission terms in Equations 1-1 through 1-4 are provided in this
Foundry Emissions Protocol document
1.5 Data Quality
The consistent use of standardized methods and procedures is essential in the compilation of reliable
emission inventories. Quality assurance (QA) and quality control (QC) of an emission inventory are
accomplished through a set of procedures that ensure the quality and reliability of data collection and
analysis. These procedures include the use of appropriate emission estimation techniques, applicable and
reasonable assumptions, accuracy/logic checks of computer models, and checks of calculations and data
reliability. Depending upon the technical approach used to estimate emissions, a checklist with all of the
particular data needs should be prepared to verify that each piece of information is used accurately and
appropriately.
Appropriate metadata (data about the data) should be maintained to assist data users with assessing the
accuracy of the reported emissions. QA/QC and other metadata records should also be maintained to
allow verification of the reported emissions, although this information does not need to be reported unless
specifically requested. For continuous emission measurement systems, these metadata include
manufacturer's design specifications for accuracy, initial calibrations, periodic calibration checks, and
other QA/QC procedures used to ensure the accuracy of the measurement device(s). For source tests used
to develop site-specific emission factors, the metadata include the specific sampling and analysis
procedures used, the results of field and laboratory blanks, duplicate analyses, method detection limits,
isokinetic and cyclonic flow checks (if applicable), and key process operating data (e.g., throughput,
temperature, material processed). For some pollutants, there may be different methods by which the
emissions can be determined. For example, VOC emissions may be determined using a "total organics"
method (e.g., using EPA Method 25, 25A through 25E, or 305) and subtracting any non-VOCs present or
by speciating individual VOCs and summing the emissions of these compounds to determine the overall
VOC emissions. When reporting VOC emissions, therefore, it should be clearly indicated how the
emissions were determined. If the emissions are determined as TOC or from a TOC concentration
measurement, it is important to indicate how the emissions are being reported, i.e., "as methane" (or as
whatever compound was used to calibrate the total organic analyzer). There are also several PM test
methods and the method used can greatly influence the emissions measured due to differences in probe
and filter temperatures for the different methods and whether or not condensable PM is measured (see PM
Test Method Consideration text box). These metadata assist users of the inventory data and help to
ensure that the inventory data are correctly used when performing subsequent analyses.
1.6 Calculations and Significant Digits
The methodology ranking presented in this Foundry Emissions Protocol document is designed to
highlight and promote those methods that are expected to yield the most accurate emission data. We
recognize that the Methodology Rank 5 methodologies may only provide emission estimates that are
1-15
-------
Version 1
Draft
Section 1—Introduction
within a factor of 2 or 3 from the actual emission rate. Nonetheless, the emission factors presented in this
document are generally presented with two significant digits. The two significant digits should not be
construed as an expectation that these emission factors are more accurate. The emission factors are
provided with two significant digits because it is recommended that all calculations be performed carrying
at least one additional significant digit to minimize round-off errors. The emissions calculated using
default emission factors may be rounded to one significant digit when reporting the emissions, but at least
two significant digits should be carried in the calculations. For methodologies that may have uncertainties
in the range of ±10 to 20 percent, at least three significant digits should be carried when performing the
calculations, even though the final emission estimate may only warrant two significant digits.
1.7 Sections of Protocol Document
The iron and steel foundry protocol document is organized in the following sections:
Section 1. Introduction
Section 2. Fugitive Dust Sources
Section 3. Melting Operations
Section 4. Mold and Core Making
Section 5. Pouring, Cooling, and Shakeout
Section 6. Finishing Operations
Section 7. References
Appendix A Glossary of Foundry Definitions Applicable to this Protocol Document
Appendix B Development of Emission Factors for Melting Operations
Appendix C Development of Emission Factors for Mold and Core Making
Appendix D Development of Emission Factors for Pouring, Cooling, and Shakeout
Appendix E Development of Emission Factors for Finishing Operations
Appendix F Control Efficiency and Particulate Matter Size Distribution
Appendix G List of Suggested SCC for Iron and Steel Foundry Operations
1-16
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
2. Fugitive Dust Sources
There are numerous sources of fugitive dust (or PM) emissions at iron and steel foundries. For the
purposes of this Foundry Emissions Protocol document, fugitive dust sources relate to outdoor sources of
dust emissions and do not include "uncaptured" emissions from indoor sources of PM emissions.
Therefore, this section covers fugitive dust emissions from outdoor materials handling (e.g., material
loading and transfer points, scrap piles, slag piles, sand handling systems) and roadways (e.g., vehicle
movement on paved and unpaved roads).
Material handling operations include receiving, unloading, storing, and conveying materials for the iron
and steel foundry, including metallic raw materials, fluxes, fuels, and sand. Metallic raw materials used
by iron and steel foundries include pig iron, iron and steel scrap, foundry returns, metal turnings, and
alloys. Fluxes used by iron and steel foundries include limestone, dolomite, soda ash, fluoride (fluorspar),
and calcium carbide. Fuels used by iron and steel foundries include coal and coke (iron and steel
foundries may also use oil and/or natural gas as fuel, but these are not expected to generate fugitive dust
emissions). Other materials used or managed at foundries include sand, sand additives (binders), slag, and
scrap returns (gates, risers, and defective castings). Fugitive particulate emissions from materials handling
operations are generated from the receiving (unloading), storing, and conveying of these materials. The
quantity of dust emissions from materials handling operations varies with the volume of materials
processed, the number of transfer points, and the characteristics of the material (moisture content and silt
content).
There are no direct emission measurement methodologies commonly employed by foundries for routine
measurement of emissions from outdoor fugitive dust sources. Fugitive dust sources are described in
Sections 13.2.1, 13.2.2, 13.2.3, and 13.2.4 of Chapter 13 of AP-42 (U.S. EPA, 2006b, 2006c, 2006d,
201 lb), which presents general correlation equations for estimating fugitive dust emissions. As such,
there is only essentially one methodology (equivalent to a Methodology Rank 4) available for these
sources.
As noted in AP-42 (U.S. EPA, 2006c), total dust emissions from aggregate storage piles result from
several distinct source activities within the storage cycle:
1. Loading of aggregate onto storage piles (batch or continuous drop operations).
2. Equipment traffic in storage area.
3. Wind erosion of pile surfaces and ground areas around piles.
4. Loadout of aggregate for shipment or for return to the process stream (batch or continuous drop
operations).
Section 2.1, Methodology for Material Handling, provides emission methodologies for estimating PM
emissions from aggregate loading or mixing (items 1 and 4, above). Section 2.2, Methodology for Paved
and Unpaved Roads, provides methodologies for estimating PM emissions from equipment traffic (item
2, above). Section 2.3, Methodology for Estimating HAP Metals Emissions from Fugitive Dust, describes
methods for estimating HAP emissions related to emitted PM from material handling.
For most foundries, wind erosion (item 3) is not expected to be a significant fugitive dust emission source
because materials that are likely to be stored outdoors (like scrap metal or slag) have very limited
amounts of dust-sized particles (particles less than 30 nm in diameter). As such, the threshold friction
velocity needed to make the dust become airborne is typically quite large and is not generally exceeded
under typical meteorological conditions. If a foundry stores sand, coke breeze, or other materials that may
have a significant dust content, then the methodologies outlined in Section 13.2.5, Industrial Wind
2-1
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
Erosion, of Chapter 13 of AP-42 (U.S. EPA, 2006d) should be used to estimate the PM emissions from
wind erosion (item 3 above), and the wind erosion PM emission estimates would be added to the loading
and loadout emissions to determine the total PM emissions for that materials handling source. For scrap
and slag piles, dust emissions from wind erosion may be assumed to be zero. For emissions from
equipment traffic (e.g., trucks, front-end loaders, dozers) traveling between or on piles (item 4, above), it
is recommended that the equations for vehicle traffic on unpaved surfaces be used (see Section 2.2 of this
Protocol document).
2.1 Methodology for Material Handling
Either adding aggregate material to a storage pile or removing it usually involves dropping the material
onto a receiving surface. Truck dumping on the pile or loading out from the pile to a truck with a front-
end loader are examples of batch-drop operations. Adding material to the pile by a conveyor stacker is an
example of a continuous-drop operation.
Emissions from a material handling "drop" operation (either batch or continuous) are estimated using
Equation 2-1. If the moisture content is known, the actual moisture content of the material should be
used (equivalent to a Methodology Rank 4). If the moisture content is not known, then the default
moisture contents found in Table 2-1 should be used (Methodology Rank 5). The default moisture
contents presented in Table 2-1 are highly uncertain as these data are from the late 1970's for integrated
iron and steel plants. As such, measurements of moisture content are recommended for more accurate
results. While it is reasonable to expect that silt content and emission factors are interrelated, no
significant correlation between the two was found during the derivation of the equation, probably because
most tests with high silt contents were conducted under lower winds, and vice versa (U.S. EPA, 2006c).
E=QxkX» X®!!
^ V 2000 J (—J
Eqn. 2-1
where:
E =
emissions for a given materials drop, tons per year (tons/yr)
Q =
quantity of material transferred, tons/yr
k =
particle size multiplier, dimensionless (see values provided below for k)
u =
mean wind speed, miles per hour (mph)
M =
material moisture content (percent)
2,000 =
Conversion factor, lbs/ton.
The particle size multiplier in the equation, k, varies with aerodynamic particle size range, as follows:
Aerodynamic Particle Size Multiplier (k) For Equation 2-1
< 30 urn
< 15 nm
< 10 nm
< 5 nm
< 2.5 urn
0.74
0.48
0.35
0.20
0.053
The mean wind speed can be obtained from on-site meteorological station (preferred) or the nearest
meteorological station or airport data. Appendix C-2 of AP-42 provides procedures for laboratory
analysis of dust samples, including moisture analysis (U.S. EPA, 1993).
2-2
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
Table 2-1. Default Values for Moisture Content for Fugitive Dust Emission Estimates from
Materials Handling
Source/Variable Description
Moisture Content
Reported Range
Value
Moisture Content
Recommended
Default Value
Scrap metal/pig iron
0.2
Slag piles
0.25-2.0a
1.0
Sand
0.3
Coal
2.8-11 a
5.0
Coke breeze
CO
CM
CT>
I
CD
8.0
a Based on iron and steel production facility factors in Table 13.2.4-1 in AP-42 (U.S. EPA, 2006c).
Example 2-1: Estimation of PM Emissions from Material Handling
In this example, the required PM emission inventory data are calculated for emissions from
sand being transferred from an outdoor sand pile to an open silo. A facility transfers 80,000
tons per year of sand, with a moisture content of 0.3 percent, from an outdoor sand pile to an
open silo, where the average wind speed is 7 miles per hour.
Start with Equation 2-1:
flA13
ton
E = Q x k x
0.0032^
2000 )
111
&
year
where:
E = emissions for a given materials drop, tons/yr
Q = quantity of material transferred, tons/yr
k = particle size multiplier, dimensionless
Aerodynamic Particle Size Multiplier (k) For Equation 2-1
< 30 urn
0.74
< 10 |im
0.35
< 2.5 urn
0.053
U = mean wind speed, miles per hour (mph)
M = material moisture content (percent)
For fugitive dust sources, PM-CON = 0. Therefore, PMX-PRI = PMX-FIL. Using the k-values
for PMio and PM2 5, the following PM emission inventory data are calculated:
PM10-PRI = PM10-FIL = 80,000 X 0.35 X (-
\ c
0032\ x_©
2000 /
PM25-PRI = PM25-FIL = 80,000 X 0.053 X (-
v t
0032\
2000 )
m
= 0.96
ton
year
x-2L=0.14-!2S-
year
2-3
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
2.2 Methodology for Pawed and Urspaved Roads
Emissions from paved and unpaved roads are estimated using equations from Section 13.2.1.3 of AP-42
(U.S. EPA, 201 lb) for paved roads and Section 13.2.2.2 of AP-42 for unpaved roads (U.S. EPA, 2006b)
and recommended silt contents found in Table 2-3 and Table 2-5.
2.2.1 Methodology for Paved Roads
The quantity of particulate emissions from resuspension of loose material on the road surface due to
vehicle travel on a dry paved road may be estimated using the following empirical expression:
E = VMT x k x (si)091 x (W)102 x —
v y v y 2000
Eqn. 2-2
where:
E = particulate emissions, tons/yr
VMT = vehicle miles traveled per year, miles/yr
k = particle size multiplier from Table 2-2, pounds per vehicle mile traveled (lb/VMT)
sL = road surface silt loading, grams/square meter (g/m2)
W = average weight of the vehicles traveling the road, tons.
2,000 = Conversion factor, lbs/ton.
It is important to note that Equation 2-2 calls for the average weight of all vehicles traveling the road. For
example, if 99 percent of traffic on the road are 2 ton cars/trucks, while the remaining 1 percent consists
of 20 ton trucks, then the mean weight "W" is 2.2 tons. More specifically, Equation 2-2 is not intended to
be used to calculate separate emissions for each vehicle weight class. Instead, emissions should be
calculated to represent the "fleet" average weight of all vehicles traveling the road.
The particle size multiplier (k) above varies with aerodynamic size range, as shown in Table 2-2. To
determine particulate emissions for a specific particle size range, use the appropriate value of k shown in
Table 2-2.
Table 2-2. Particle Size Multipliers for Paved Road Equation
Size range''
Particle Size Multiplier kb
lb/VMT
PM-2.5
0.00054
PM-10
0.0022
a Refers to airborne particulate matter (PM-x) with an aerodynamic
diameter equal to or less than x micrometers.
b Units shown are pounds per vehicle mile traveled (lb/VMT).
If the road surface silt loading on the paved road is known (determined), the determined silt loading
should be used in Equation 2-2 (equivalent to a Methodology Rank 4). If the road surface silt loading is
not known, then a default silt loading from Table 2-3 should be used (Methodology Rank 5). Factors for
asphalt batching and concrete batching are provided for area near sand piles. The default silt loading
values presented in Table 2-3 are highly uncertain because these data are from the late 1970's and are not
specific to foundry facilities. Consequently, measurements of silt content are recommended for more
accurate results. Appendix C-2 of AP-42 provides procedures for laboratory analysis of dust samples,
including silt analysis (U.S. EPA, 1993).
2-4
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
Table 2-3. Default Values for Fugitive Dust Emission Estimates from Paved Roads
Source Description
Potential Range for Silt
Loading (g/m2)
Recommended Value for
Silt Loading (g/m2)
Iron and Steel production
0.09-79
9.7
Asphalt batching
76 -193
120
Concrete batching
11-12
12
a Based on factors in Table 13.2.1-3 in AP-42 (U.S. EPA, 2011b).
Equation 2-2 was developed to predict only the emissions from dust resuspension from paved roads.
Additional PM emissions occur from vehicle exhaust, brake wear, and tire wear. The additional emissions
from these sources can be obtained from either EPA's MOBILE6.2 (http://www.epa.gov/oms/m6.htm) or
MOVES2010 (http://www.epa.gov/otaq/models/moves/index.htm) models and added to the emissions
from Equation 2-2 to determine total PM emissions from paved roads.
Example 2-2: Estimation of PM Emissions from Paved Roads
In this example, the PM-PRI, PM10-PRI, and PM2 5-PRI emissions from paved roads at a
facility are calculated. A facility has an average weight of vehicles of 20 tons traveling a paved
road with surface silt loading of 15 g/m2. The vehicle miles traveled per year on this road is
approximately 5,000 miles/yr.
Start with Equation 2-2:
E = VMT x k x (si)091 x (W)102 x -—
y J y J 2000
where:
E = particulate emissions, tons/yr
VMT = vehicle miles traveled per year, miles/yr
k = particle size multiplier from Table 2-2, pounds per vehicle mile traveled (lb/mile)
Size range
Particle Size Multiplier k
Ib/VMT
PM-2.5
0.00054
PM-10
0.0022
sL = road surface silt loading, grams/square meter (g/m2)
W = average weight of the vehicles traveling the road, tons.
For fugitive dust sources, PM-CON = 0.
Since PM-CON = 0, PM10-PRI = PM10-FIL. Using the k-value for PM-10:
PM1(I-PRI = 5,000 x 0.0022 x (15)091 x (20)102 x — = 1.4 —
v J v J 2000 year
Since PM-CON = 0, PM2.5-PRI = PM2 5-FIL. Using the k-value for PM-10:
PM25-PRI = 5,000 x 0.00054 x (15)091 x (20)102 x — = 0.34 —
2-5
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
2.2.2 Methodology for Unpaved Roads
The following empirical expressions may be used to estimate the quantity in tons of size-specific
particulate emissions from an unpaved road, per vehicle mile traveled (VMT):
For vehicles traveling on unpaved surfaces at industrial sites, emissions are estimated from the
following Equation 2-3:
E = VMT x k x (—)3 x f-)b x —
V12/ V 3 / 2000
Eqn. 2-3
where:
E =
size-specific emission (tons/yr),
VMT =
vehicle miles traveled per year,
k =
particle size multiplier from Table 2-4, lb/VMT)
s =
surface material silt content (percent) [use area-specific factors],
W =
mean vehicle weight (tons).
2,000 =
Conversion factor, lbs/ton.
The source characteristics s and W are referred to as correction parameters for adjusting the emission
estimates to local conditions. As noted previously, W refers to the average weight of all vehicles traveling
the road. Equation 2-3 is not intended to be used to calculate separate emissions for each vehicle weight
class. Instead, only one value for W should be calculated to represent the "fleet" average weight of all
vehicles traveling the road.
The constants for Equation 2-3 based on the stated aerodynamic particle sizes are shown in Table 2-4.
Table 2-4. Constants for Equation 2-3.
Constant
Industrial Roads
PM-2.5
PM-10
k (lb/VMT)
0.15
1.5
a
0.9
0.9
b
0.45
0.45
For vehicles traveling primarily on the storage pile, the silt value of the stored material should be used;
for vehicles traveling between or around storage piles, the road silt content in these areas should be
determined and used because they are expected to differ from the silt values for the stored materials
(equivalent to a Methodology Rank 4). If the silt content is not known, then the default silt content for
iron and steel production facilities found in Table 2-5 should be used (Methodology Rank 5). Defaults
for sand and gravel processing plants and stone quarrying and processing plants are also provided for
foundries that may have similar raw materials to those types of facilities.
2-6
-------
Version 1
Draft
Section 2—Fugitive Dust Sources
Table 2-5. Default Values for Fugitive Dust Emission Estimates from Unpaved Roads
Source/Variable Description
Unpaved road: silt content %
Potential Range
Value3
Recommended
Value3
Iron and steel production
0.2-19
6.0
Sand and gravel processing plant road
o
CD
I
4.8
Stone quarrying and processing plant road
2.4-16
10
a Based on factors in Table 13.2.2-1 in AP-42 (U.S. EPA, 2006b).
Equation 2-3 was developed to exclude emissions from vehicle exhaust, brake wear, and tire wear.
However, PM emissions from dust resuspension on unpaved road are expected to be significantly higher
than any additional PM emissions from vehicle exhaust, brake wear, and tire wear, so it is acceptable to
ignore these sources when determining PM emissions from unpaved roads. Alternatively, PM emissions
from vehicle exhaust, brake wear, and tire wear can be estimated from either EPA's MOBILE6.2 or
MOVES2010 models and added to the emissions from Equation 2-3 to determine total PM emissions
from unpaved roads.
2.3 Methodology for Estimating HAP Metals Emissions from Fugitive Dust
The metal HAP composition of the fugitive dust from materials handling operations can be assumed to be
equivalent to the composition of the material stored if the material stored is generally composed of small
particles (e.g., sand piles). It is expected that only scrap metal or slag piles will contain significant
quantities of HAP, but these materials are generally not composed primarily of fine particles, so the dust
emissions from these piles may have significantly different metal HAP composition than the stored
material. Sand or coke piles, which are materials containing fine particles, are not expected to contain
significant quantities of HAP, so no speciation is necessary for these fugitive dust sources.
In the absence of site-specific dust analysis, the average metal HAP concentration of PM emitted from
scrap piles can be determined based on the average composition of the melted metal produced from the
furnace, as a worst-case assumption. It is reasonable to adjust that composition based on alloying
materials added to the furnace but not stored in the outdoor scrap pile. That is, if you alloy with chromium
and add 1 lb of chromium per 99 lb of metal melted and your final melted steel contains 2 weight percent
chromium, then you can assume the scrap pile contains approximately 1 weight percent of chromium and
your chromium emissions would be lpercent of the PM-FIL emissions determined from the scrap pile
(rather than 2 percent).
In the absence of site-specific dust analysis, the chemical composition of the produced slag should be
used to speciate the metal HAP emissions from slag piles, as a worst-case assumption. If chemical
composition data are not available for the produced slag, then the average chemical composition of the
melted metal can be used as a proxy for estimating metal HAP emissions from the slag pile.
2-7
-------
Version 1
Draft
Section 3—Melting Operations
3. Melting Operations
Iron and steel foundries include several melting operations, the most common of which are scrap cleaning
and preheating, melting furnaces, inoculation, and holding furnaces. These melting operations include
point and fugitive sources. Some of these operations are uncontrolled, and some may be controlled using
a wet scrubber, fabric filter, or other air pollution control techniques.
The emission estimation methods for melting operations are presented in Table 3-1. These methods are
ranked according to anticipated accuracy.
Table 3-1. Summary of Typical Hierarchy of Melting Operations Emission Estimates
Rank
Measurement Method
Additional Data Needed
1
Direct measurement (continuous emission monitoring
systems [CEMS]) for both flow rate and gas
composition
¦ Pressure, temperature, and moisture
content (depending on the monitoring
system)
2
Direct measurement (CEMS) for gas composition
Use engineering calculations to calculate flow rate
¦ Blast air rates, coke addition rates, fan
power consumption, etc. as needed for the
engineering flow estimates
3a
Source-specific stack testing to calculate source-
specific emission factors
¦ Metal melting rate
3b
Source-specific PM emission factors from baghouse
catch data
¦ Metal melting rate; baghouse catch data;
control device collection efficiency estimates
4a
Default PM emission factors with site-specific metal
chemistry
¦ Metal melting rate
4b
Default PM emission factors with default metal
chemistry
¦ Metal melting rate
As all foundries have melting furnaces, the methodologies for melting furnaces will be presented first in
Section 3.1, Melting Furnaces. Methodologies for scrap pretreatment and preheating are presented in
Section 3.2, Scrap Handling, Preparation and Preheating. Methodologies for inoculation are presented
in Section 3.3, Metallurgical. Treatment of Molten Metal. Methodologies for holding furnaces are
presented in Section 3.4, Holding Furnaces.
3.1 Melting Furnaces
There are several types of melting furnaces used in the iron and steel foundry industry, including, but not
limited to electric arc furnace (EAF), electric induction furnace (EIF), cupola furnace, reverberatory
furnace, and crucible furnace. The most commonly used melting furnaces are the EAF, EIF, and cupola.
EAFs are large, refractory-lined cylindrical vessels made of heavy welded steel plates. They are equipped
with a removable roof through which carbon electrodes mounted on a superstructure above the furnace
can be raised and lowered through holes in the furnace roof. For alternating current furnaces, the
electrodes are lowered through the roof of the furnace and are energized by three-phase alternating
current, creating arcs that melt the metallic charge with their heat. Additional heat is produced by the
resistance of the metal between the arc paths. A direct-current furnace uses only one electrode and
provides stable electrical current to the metal scrap with less electrode consumption. Once the melting
cycle is complete, the carbon electrodes are raised and the roof is removed. The vessel can then be tilted
to pour the molten iron (U.S. EPA 2002, 2003).
EIFs are cylindrical or cup-shaped refractory-lined vessels that are surrounded by electrical coils either
around or below the main body of the furnaces. Furnaces with the coil around the furnace body are called
3-1
-------
Version 1
Draft
Section 3—Melting Operations
coreless induction furnaces, and those with the coil below the body are called channel induction furnaces.
When these coils are energized with high-frequency alternating current, they produce a fluctuating
electromagnetic field that heats the metal charge. The metal is melted by resistance heating produced by
the current. Consequently, EIFs may also be referred to as electric resistance furnaces. The coils carrying
the electric current are typically cooled with water. Most induction furnace lids are closed when not
charging, skimming, or tapping to reduce heat loss. The molten metal is tapped by tilting and pouring
through a hole in the side of the vessels (U.S. EPA 2002, 2003).
The cupola is a cylindrical steel shell with a refractory-lined or water-cooled inner wall. Hinged doors at
the bottom allow the furnace to be emptied when not in use. When charging the furnace, the doors are
closed and a bed of sand is placed at the bottom of the furnace, covering the doors. A charge consisting of
coke for fuel, scrap metal, alloying materials, and flux is loaded into the furnace. Flux, often chloride or
fluoride salts, is added to the furnace to remove impurities. The flux reacts with impurities to form dross
or slag, which rises to the surface of the molten metal and helps to prevent oxidation of the metal. The
presence of coke in the melting process raises the carbon content of the metal to the casting specifications.
Heat from the burning coke melts the scrap metal and flux, which both then drop to the bottom of the
cupola to be tapped. As the melt proceeds, new charges are added at the top. A hole that is level with the
top of the sand bed allows molten metal to be drawn off, or tapped. A higher hole allows slag to be
tapped. Additional charge is added as needed (U.S. EPA 2002, 2003).
3.1.1 Methodology Rank 1 for Melting Furnaces
Though uncommon, some melt furnaces may have a CEMS for NOx, CO, and/or SO2. A CEMS is a
comprehensive unit that continually determines gaseous or PM concentrations or emission rates using
pollutant analyzer measurements and a conversion equation, graph, or computer program to produce
results in the desired units.
There are two main types of CEMS: in-situ and extractive. An in-situ CEMS measures and analyzes the
emissions directly in a stack. There is less sample loss associated with the in-situ CEMS compared to the
extractive CEMS because the sample lines of an extractive system can leak, freeze, or clog, or pollutants
can be lost because of adsorption, scrubbing effects, or condensation. In an extractive CEMS, the sample
gas is extracted from the gas stream and transported to a gas analyzer for the measurement of the
contaminant concentration. Because an extractive CEMS is located outside the stack, the sampling
instruments are not affected by stack conditions, maintenance and replacement are generally simpler, and
the cost is lower than with an in-situ CEMS, although extra costs are incurred for the sampling and
conditioning system for an extractive CEMS (Levelton Consultants, 2005).
The pollutant concentration recorded by a CEMS is generally on a concentration basis, such as parts per
million. The CEMS may also include a diluent monitor (e.g., O2) for correcting the concentrations to a
fixed excess air concentration. For in-situ CEMS, these measurements are made at stack conditions so
that the concentrations would be determined on a "wet basis." That is, the concentrations are based on the
total amount of gas, including water vapor. For extractive CEMS, the gas is often conditioned to remove
water vapor before analysis, so the concentrations are commonly determined on a "dry basis." A flow rate
monitor is also needed in order to determine mass emission rates directly using a CEMS (Methodology
Rank 1). Gas flow measurements are made at stack conditions, so the flow rate will be in terms of actual
gas volume on a wet basis. If the gas composition is determined on a dry basis, then a moisture content
measurement is needed to convert the flow rate to a dry basis (or convert the composition to a wet basis)
so that both measurements are on the same basis, and many gas flow monitors contain temperature and
pressure monitors to allow conversion of the flow to standard conditions for this purpose. It is important
to note that care must be taken to ensure that the gas and flow measurements are made on the same basis
and in the same terms as the permitted limits, if applicable, or that appropriate ancillary measurements are
made to perform the necessary unit conversions.
3-2
-------
Version 1
Draft
Section 3—Melting Operations
The following general equation (Equation 3-1) is used for determining a mass emission rate from a
CEMS:
E, = ES=1 ((V)„ X [1 - (f„20)„] X X x @ x (£) x K)
100%
Eqn. 3-1
where:
E, = Emission rate of pollutant "i" (tons/yr).
N = Number of measurement periods per year (e.g., for hourly measurements, N = 8,760 as
there are 8,760 hours in a non-leap year).
n = Index for measurement period.
(V)n = Cumulative volumetric flow for measurement period "n" (actual cubic feet per
measurement period [acf]). If the flow rate meter automatically corrects for
temperature and pressure, then replace "T0 + Tn x Pn + P0" with "1." If the pollutant
concentration is determined on a dry basis and the flow rate meter automatically
corrects for moisture content, replace the term [l-(fH2o)n] with 1.
(fmcOn = Moisture content of exhaust gas during measurement period "n." volumetric basis
(cubic feet water per cubic feet exhaust gas).
(C,)n = Concentration of pollutant "i" in the exhaust gas for measurement period "n" (volume
percent, dry basis). If the pollutant concentration is determined on a wet basis, then
replace the term [l-(fH2o)n] with 1.
MW; = Molecular weight of pollutant "i" (kilogram per kilogram mole [kg/kg-mol]).
MVC = Molar volume conversion factor = 849.5 standard cubic feet per kilogram mole
(scf/kg-mol) at 68°F (528°R) and 1 atmosphere (atm).
T0 = Temperature at "standard conditions" = 528 °R.
Tn = Temperature at which flow is measured during measurement period "n" (°R).
Pn = Average pressure at which flow is measured during measurement period "n" (atm).
P0 = Average pressure at "standard conditions" = 1 atm.
K = Conversion factor = 2.2046/2,000 (tons per kilogram [tons/kg] = 0.0011023 tons/kg.
A CEMS records multiple measurements per hour; the frequency depends on the pollutant being
measured and the type of CEMS. For example, a CEMS monitoring benzene concentration using gas
chromatography may only sample and record a measurement every 15 minutes, while a CEMS
monitoring S02 concentration may measure concentration multiple times per second and combine these
concentrations into a recorded output on a minute basis. These individual measurements can be used to
calculate annual emissions in two ways. The most common method is for the CEMS to average the
measurements within each hour and develop 8,760 hourly average concentrations and flow rates that can
be summed. Example 3-1 demonstrates the calculation of CO emissions for 1 hour for a cupola melting
furnace based on an hourly average concentration and flow rate. This method is best suited for
measurements that are fairly consistent and stable over the course of an hour. The other method is to
determine the emission rate for each recorded measurement based on the concentration and flow rate for
that measurement. In other words, if the CEMS records measurements every minute, then the emission
rate is determined per minute and hourly emissions are determined by summing the 60 applicable
emission rates; if the CEMS records measurements every 5 minutes, then the emission rate is determined
for each 5-minute interval and hourly emissions are determined by summing the 12 applicable emission
rates. This method is expected to be more accurate than using hourly averages of the individual
measurement parameters (i.e. concentration and flow) if the source's flow rate and concentration vary
independently within an hourly time frame.
3-3
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-1: Calculation of CO Emissions Using a OEMS
The following example shows the calculation for 1 hour (60 minutes); the total emissions
during any period (i.e., day, month, quarter, or year) may be calculated as the sum of the
hourly emissions determined by the CEMS. In terms of Equation 3-1, "i" is CO and the index
for the period in this example is 1, so the result "E" is in tons per hour.
Calculate hourly CO emissions for a cupola melting furnace given that the following data have
been collected:
¦ The hourly average CO concentration calculated by the CEMS for this hour is 600 parts per
million by volume (ppmv), dry basis
¦ The hourly average flow rate calculated by the CEMS for this hour is 200,000 acfm, wet basis
¦ The unit continuously operated for the full hour
¦ The dry bulb temperature and pressure at the monitoring location was 400°F and 0.5 psig, and
the moisture content was 3.9 volume percent.
The cumulative volume for the 1 hour time period (60 minutes) is:
(V)n = 50,000 acfm x 60 min = 3,000,000 acf
Since the CO concentration was measured on a dry basis and the flow was determined on an
actual (wet basis), the moisture content (fmo) correction must be used. The moisture content
term in Equation 3-1 is in units of volume of water per volume of wet gas (i.e., volume of
water plus volume of wet gas). From the Ideal Gas Law, a molar ratio is equivalent to a
volume ratio. The molecular weight of dry air is 29 kg/kg-mol, and the molecular weight of
water is 18 kg/kg-mol. Thus, fH2o can be determined from the mass-based moisture content as
follows:
e _ Moles Water _ (0.082 kg water)/18 _ n n->n
IutA — u.u^y
Moles Water+Moles Dry Gas (0.082 kg water)/18+(l kg air)/29
[1 - (fH2o)n] = 1 - 0-039 = 0.961
600
(Cj)n = X 100% = 0-06%
V Un 1,000,000
The temperatures must be converted to an absolute scale (degrees Rankine) as follows:
Tn = Tdrybulb = (400 + 460) = 860°R
The gauge pressure must be converted to an absolute pressure as follows:
Pn- [o.5 psi x 1 atm + 1 atml = 1.034 atm
L ^ 14.7 psi J
The molecular weight of carbon monoxide (CO) is 28. Therefore,
N (
E(IO = ^ ( 00n X [1 — (fH2o)n]
n=l ^
100% MVC VV VPQ
0.06% 28 /528\ /1.034\
Eco = 3,000,000 x 0.961 x x x x x 0.0011023
co 100% 849.5 V860/ VI/
ECo = 0.040 tons per hour (tons/hr).
If the cupola melting furnace operated steadily and continuously for an entire year and the
emission rate remained perfectly constant over that year, annual emissions would be
0.040 tons/hr x 8,760 hours per year (hr/yr) = 350 tons/yr.
-------
Version 1
Draft
Section 3—Melting Operations
3.1.2 Methodology Rank 2 for Melting Furnaces
Even though some foundries may have CEMS installed to measure a pollutant concentration, many of
these foundries may not have a continuous flow monitor. When pollutant concentration data are available
from a CEMS, but not flow rates, exhaust flow rates can often be calculated or estimated based on
engineering calculations. For example, exhaust gas flow rates can be calculated from fuel combustion
measurements. This method is commonly referred to as the "F-factor" method, and procedures for
conducting a fuel analysis and calculating estimated emissions are described in EPA Method 19 (40 CFR
Part 60, Appendix A-7) and in Appendix F of 40 CFR Part 75. This method is applicable to blast type
furnaces, such as a reverberatory furnace, whose flow is primarily combustion air. This method may also
be applied for cupola melting furnaces, but with greater uncertainties because not all of the coke used in
the process is combusted (some of the carbon remains in the iron). For furnaces where the flow is
generated primarily by the use of a capture/ventilation system exhaust fan, such as for EAF and EIF, fan
curves may be used to estimate exhaust gas flow rates depending on the design of the ventilation system
and presence of dampers (use of variable speed fans and dampers limit the ability to use fan curves for
estimating flow rates).
For most common foundry melting furnaces, the engineering methods described in this section are
expected to have greater uncertainties than when applied to other sources at the foundry. These methods
are described here as Methodology Rank 2, but for cases where these methods have greater uncertainty,
the assigned rank level is degraded to a Methodology Rank 3. If a facility has source test data
(Methodology Rank 3a for Melting Furnaces), then stack flow rate measurement data will be available for
the test runs. These measured flow rates should be compared to the engineering estimates to assess the
accuracy of the engineering estimates. If the engineering estimates of the flow rates provide a reasonably
accurate estimate of the flow rate as measured during the source tests (within approximately 20 or 25
percent), then the pollutant CEMS concentration and engineering estimates should be used preferentially
to the site-specific emission factors. However, if the engineering estimates do not provide a reasonable
estimate of the sources exhaust rate, then the site-specific emission factors (Methodology 3) should be
used. Even with the uncertainties of these engineering methods as applied to melting furnaces, these
engineering estimates are generally expected to provide a more accurate site-specific emission estimate
than is obtained by using the default emission factors (Methodology Rank 4).
For blast-type melting furnaces such as reverberatory furnaces and (with greater uncertainties) cupolas,
the F-factor method can be used to estimate the stack gas flow rate. When available, ultimate analysis of
the fuel used can be used to determine a site-specific F-factor using Equation 3-2:
[3.64 (%H) + 1.53 (%C) + 0.57 (%S) + 0.14 (%N) - 0.46 (%0)] X 106
Fd =
where:
Fd =
%H =
%C =
%S =
%N =
%0 =
HHV =
HHV
Eqn. 3-2
Volume of combustion components per unit of heat content (dry standard cubic feet
per million British thermal units; dscf/MMBtu)
hydrogen content of fuel (weight percent)
carbon content of fuel (weight percent)
sulfur content of fuel (weight percent)
nitrogen content of fuel (weight percent)
oxygen content of fuel (weight percent)
Higher heating value of fuel (Btu/lb)
3-5
-------
Version 1
Draft
Section 3—Melting Operations
Table 3-2 provides default F-factors and higher heating values (HHV) for typical fuels used at foundries.
It is important to note that these values are for estimating flow on a dry basis (at 0 percent excess air) and
are based on the higher heating value of the fuel (referred to as gross calorific value in EPA Method 19).
Table 3-2. F-factor Exhaust Volumes and Heat Content of Common Fuels
Constituent
F-factora (dscf/mmBtu)
HHVb
Coal - Bituminous
9,780
24.9 MMBtu/ton
Byproduct Coke (from coke ovens)
10,400 c
26.4 MMBtu/ton c
Petroleum Coke
9,830
30.0 MMBtu/ton
Natural Gas
8,710
1.028x10"3 MMBtu/scf
From Table 1 to Appendix F of 40 CFR Part 75, unless otherwise noted.
From Table C-1 to Subpart C of 40 CFR Part 98, unless otherwise noted .
Determined from ultimate analysis and HHV values reported by Avallone and Baumeister (1978).
The volumetric flow rate of the exhaust gas, (Q)n, can be estimated using the F-factor, the amount of fuel
combusted, and the heat content of the fuel using Equation 3-3.
20.9
Qn = Fd x Qf x HHV x
(20.9 -%02d)
Eqn. 3-3
where:
Qn = Volumetric flow rate for measurement period ""n" (dry standard cubic feet per hour
[dscfh])
Fd = Volume of combustion components per unit of heat content (dscf/MMBtu)
Qf = Fuel input rate (ton/hour or dscfh)
HHV = Higher heating value of fuel (million British thermal units [MMBtu] per ton or
MMBtu/scf])
%02d = Concentration of 02 on a dry basis (percent)
If multiple fuels are consumed, apply Equation 3-3 for each fuel type and add the results together to
determine the overall volumetric flow rate. Once the volumetric flow rate is known, the volume of gas
exhausted over a time period can be determined and the emissions can be calculated using Equation 3-1
as in Methodology Rank 1 for Melting Furnaces. When using the F-factor method as indicated here, the
exhaust gas flow rate will be in units of dry standard cubic feet per minute, so the temperature and
pressure correction terms are not needed. A moisture correction term is not needed when the
concentration measurement is also made on a dry basis. If the concentration measurements are made on a
wet basis, then they must be corrected to a dry basis by dividing by the [l-(fH2o)n] term (rather than
multiplying by this term as shown in Equation 3-1). Example 3-2 demonstrates how to calculate the
exhaust flow rate from an F-factor.
As discussed previously, exhaust flow rates for EAF and EIF may be estimated using fan curves.
Depending on the design of the ventilation systems, the use of fan curves to estimate flow rates may be
difficult or inaccurate. Side draft hoods or canopy hood systems generally operate with constant flow
(constant pressure drop), allowing the use of fan curves. Close fitting hoods that are raised and lowered
and/or dampers that are frequently open and closed will significantly alter the pressure drop of the
ventilation system, making it more difficult and inaccurate to use fan curves. An example of the use of
fan curves is provided in Section 5.1 Methodology Ranks 1 and 2 for PCS Operations.
3-6
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-2: Calculation of Exhaust Flow Rate from t~ t-actor
A cupola is equipped with an S02 CEMS and 02 diluent monitor, and outputs S02
concentrations in units of ppmv, dry basis, corrected to 0% oxygen, but the cupola does not
have a flow monitor. The average S02 concentration over a 1-hour period was 22 ppmv (dry
basis; corrected to 0% 02). During this 1-hour period, the cupola consumed 8,000 lbs of coke
and the afterburner consumed 2,600 scf of natural gas. What are the mass emissions of S02
during this 1 hour period?
Given only the above information (with no site-specific analysis of the coke used), the default
factors in Table 3-2 are used with Equation 3-2. Since the concentration is already corrected to
0 percent 02, the 02 correction term in Equation 3-3 must be applied assuming 0 percent 02
(the flow rate and concentration measures must be determined on the same basis).
The exhaust flow attributable to the coke combustion is as follows:
Fd = 10,400 (default from Table 3-2)
_ 8,000 lbs of coke 1 ton 8,000 tons
Qf = ^ x
hour 2,000 lbs 2,000 hour
HHV = 26.4^^ (default from Table 3-2)
Qn = Fd x Qf x HHV x 20 9
(20.9—%02d)
20 9
Using %02d = 0, the 02 correction term goes to unity, i.e., g = 1
Qn = 10,400 X — X 26.4 X 1 = 1,098,240 dscfh
2,000
The exhaust flow attributable to the natural gas combustion is as follows:
Fd = 8,710 (default from Table 3-2)
2600 scf
Qf =
hour
o MMRtu
HHV = 1.028 x 10~3 (default from Table 3-2)
The 02 correction term is 1 (since %02a = 0).
Qn = 8,710 X 2,600 X 1.028 X 10~3 X 1 = 23,280 dscfh
Total Qn = 1,098,240 + 23,280 = 1,121,520 dscfh
In 1 hour, (V)n = Qn x lhr = 1,121,520 dry standard cubic feet (dscf) of gas would be
exhausted (corrected to 0% excess 02). The emissions for the hour are determined using
Equation 3-1 (without pressure, temperature, or moisture content corrections):
(Ci)n = —-— X 100% = 0.0022%
v 1Jn 1,000,000
MW; = 32.066+ (2 X 15.9994) = 64.06
K |j^J = 2.2046 (to output the mass in lbs rather than tons)
ES0 = In-lf(V)n X^X^XkPI)
2 Jn 100o/o MVC [kgjy
„ 0.0022% 64.06 „ „ „ . , . . lb
Eso = 1,121,520 x x x 2.2046 = 4.1 —
bu2 100% 849.5 hr
3-7
-------
Version 1
Draft
Section 3—Melting Operations
3.1.3 Methodology Rank 3 for Melting Furnaces
Source testing can provide useful data for developing site-specific emission correlations or emission
factors. Source testing provides a measurement of the emissions at a particular point in time, and most
tests are performed at conditions representative of normal operation, in which case the emission
measurement can provide an estimate of emissions at similar operating conditions. It is important to note
that this methodology will be less reliable when the unit is operating at conditions other than those tested
Emission factors developed directly from source test data are considered to be Methodology Rank 3a.
For melting furnaces that are controlled with a baghouse, site-specific emission factors can be developed
based on the baghouse catch data (the measurement of the mass of PM collected over a period of time).
While baghouse catch data can be determined over longer periods of time, and thereby can represent the
average PM (uncontrolled) emissions over a range of operating conditions, this method requires
additional assumptions in order to be used to estimate emissions. Consequently, this methodology is
considered Methodology Rank 3b.
3.1.3.1 Site-specfic Emission Factors from Source Test Data
Emission factors are developed by dividing the emission rate by a process parameter such as fuel usage or
metal melting rates. Generally, one source test consisting of three runs is performed at a specific set of
conditions, and the results for each run can be averaged to determine an emission factor that is assumed to
apply at all heat input rates using Equation 3-4. The annual emissions can be calculated using Equation
3-5. Example 3-3 demonstrates a sample emission factor calculation based on one test with three test
runs.
N
1 V" Eir
EmFi = NXX q7
r=l
Eqn. 3-4
where:
Em F, = emission factor of pollutant "i", lbs/ton
N = number of test runs
Ei r = emissions of pollutant "i" during run "r", lbs/hr
Qr = quantity of metal melted (or other relevant processing rate), tons/hr
where:
„ _ QAnnual * EmF;
1 ~ 2,000
Eqn. 3-5
E; = emissions of pollutant "i", tons/yr
QAnnuai = quantity of metal melted (or other relevant processing rate) during the inventory year,
tons/yr
Em F, = emission factor of pollutant "i", lbs/ton
2,000 = Conversion factor, lbs/ton.
3-8
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-3: Site-Specific Emissions Factor from Source Test Data
During a source test, three test runs were conducted to determine the PM emission rate for a
controlled EAF melting furnace. The PM emissions rate measured during the source tests were
4.16, 5.29, and 5.33 lbs/hr for tests 1, 2, and 3, respectively. During the three runs, the EAF
melted 20.6, 22.0, and 23.5 tons/hr of steel.
First, calculate the appropriate emissions factor for each individual run, and then average the
emissions factors with the following calculations:
¦ Run 1: Emissions/throughput = 4.16 [lbs/hr] 20.6 [tph] = 0.202 lbs PM/ton
¦ Run 2: Emissions/throughput = 5.29 [lbs/hr] + 22.0 [tph] = 0.240 lbs PM/ton
¦ Run 3: Emissions/throughput = 5.33 [lbs/hr] 23.5 [tph] = 0.227 lbs PM/ton
The average PM emission factor, using Equation 3-4, is
EmFpM = (0.202 + 0.240 + 0.227) 3 = 0.223 lbs PM/ton metal melted
The annual operating rate for the EAF was 46,750 tons metal melted for the year. Using
Equation 3-5, the annual PM emissions for the EAF is calculated as:
g _ QAnnual * EmF;
1 ~ 2,000
46,750X0.223 _ „„ tons PM
h: = = 5.2.1
2,000 year
3.1.3.2 Site-specfic Emission Factors from Baghouse Catch Data
This method is only useful for estimating PM emissions and can only be used for sources controlled using
a baghouse. PM collection quantities for wet collection systems are generally too inaccurate since the dry
quantity of PM is difficult to determine in wet collection systems and because the PM collection
efficiency of wet systems is generally less than for baghouses. The dry mass of PM collected over a set
time period is measured along with the total quantity of metal melted over the same time period.
Generally, a longer time period will provide a more accurate estimate of the average emission factor than
a shorter time period.
This baghouse catch method provides a site-specific estimate of the captured but uncontrolled PM-FIL
emissions; it does not provide a means to determine PM-CON emissions. The uncontrolled PMio-FIL and
PM2 5-FIL emission as 90 percent and 70 percent of the PM-FIL emissions (based on the default size
distribution for melting furnaces as presented in the following section for Methodology Rank 4). The
actual emissions must still be calculated based on the control device PM collection efficiency; default
values for these are provided in the following section for Methodology Rank 4. Example 3-4
demonstrates the calculation of site-specific emission factors using baghouse catch data. These site-
specific factors would then be used in-place of the default uncontrolled emission factors presented in
Methodology Rank 4, but would use the general Rank 4 calculation methodologies to determine the actual
emissions from the source.
3-9
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-4: Site-Specific Emissions Factor from Baghouse Catch Data
A cupola melting furnace is controlled by a baghouse. Over a 1,000-hour period, the mass of
PM collected in the baghouse was 16,000 lbs. Over the same 1,000-hour period, 2,000 tons of
iron were melted in the cupola. Determine the site-specific "uncontrolled" PM emission
factors for this cupola melting furnace.
The uncontrolled PM-FIL emission factor is simply the mass of PM collected divided by the
mass of metal melted over the time period.
16,000 [lbs]
PM-FIL uncontrolled emissions factor = = 8.0 lbs/ton
2,000 [tons] '
The PM10-FIL and PM2 5-FIL uncontrolled emission factor is estimated based on the default size
distribution for melting furnaces, which is 90% and 70%, respectively (regardless of furnace type).
PMio-FIL uncontrolled emissions factor = 8.0 [lbs/ton] x 90% = 7.2 lbs/ton
PM2 5-FIL uncontrolled emissions factor = 8.0 [lbs/ton] x 70% = 5.6 lbs/ton
3.1.4 Methodology Rank 4 for Melting Furnaces
When direct emission monitoring or site-specific emission factors are not available, then default emission
factors may be the only way to estimate emissions. This section presents the default emission factors
applicable to melting furnaces and is divided into four subsections: 1) PM emission factors; 2) "other
criteria" (non-PM) emission factors; 3) metal HAP emission factors (or, more accurately, default metal
HAP composition of emitted PM); and 4) organic HAP emission factors.
3.1.4.1 PM Emissions Inventory Default Factors
The EPA has developed PM emission factors for various types of melting furnaces, which are provided in
AP-42 (U.S. EPA, 1995; 2003), and provides size distribution analysis for iron foundry sources (U.S.
EPA, 2003). The EPA also compiled PM emission factors as part of the Background Information
Document (BID) to support standards development for iron and steel foundries. Comparing these
emission factors, there are several discrepancies between the default factors. The emission factors and
methodologies presented in this section reconcile, to the extent possible, these disparate emission factors.
Appendix B provides additional details about the PM emission factors developed for melting furnaces.
Due to the importance of providing emissions for the appropriate size of PM emitted, general
uncontrolled emission factors and size distribution data are provided in this Protocol document, along
with default control efficiencies based on the type and design of the control device. This approach is
recommended due to the variability in design and performance of different control devices. Table 3-3
provides the default filterable (by size) and condensable PM emission factors for different types of
melting furnaces. For EAF and EIF, two separate emission factors are provided: one for melting and one
for charging/tapping. These are provided in the event charging and tapping are conducted without control
or with limited control (e.g., the control system may be operating, but the capture system has significantly
lower capture efficiencies).
3-10
-------
Version 1
Draft Section 3—Melting Operations
Table 3-3. Summary of PM Emission Factors for Melting Furnace Operations
Suggested
SCC for
Iron
Foundries
Suggested
SCC for
Steel
Foundries
Emission category
PM-FIL3
emissions
factor,
lb/ton
PMio-FILa
emissions
factor,
lb/ton
PM2.5-FILa
emissions
factor,
lb/ton
PM-CONb
emissions
factor,
lb/ton
30400301
Cupolas uncontrolled (or
prior to controls)
13.8
12.4
9.7
0.05
30400304
30400701
EAF melting uncontrolled c
11.0
9.9
7.7
0.05
30400316
EAF charging & tapping
uncontrolledc
1.8
1.6
1.3
0.01
30400303
30400705
Induction melting
uncontrolled c
1.5
1.4
1.1
0.05
30400316
Induction charging &
tapping uncontrolledc
0.5
0.5
0.4
0.01
30400302
Reverberatory uncontrolled
2.1
1.9
1.5
0.05
a Emission factors, in lb of pollutant per ton of grey iron melted, based primarily on data reported in AP-42 (U.S.
EPA, 2003), except used an average particle size distribution for all furnace types (e.g., 70 percent of PM-FIL is
less than 2.5 urn in diameter).
b Derived from test data received from 1998 Foundry ICR.
c Separated EAF and EIF emissions between melting and charging/tapping based on date received from 1998
Foundry ICR.
Table 3-4 provides default control efficiencies for different types of emission control systems. When
applying the control efficiencies from Table 3-4 to the emission factors of Table 3-3, use the control
efficiency reported for the particle size range of 0 to 2.5 nm for the fraction of PM2.5-FIL and use the
control efficiency reported for the particle size range of 2.5 to 10 pim for the fraction of PM10-FIL.
Generally, it can be assumed that the PM collection efficiencies for PM greater than 10 nm in diameter
are 100 percent (provided the PM collection efficiency for the 2.5 to 10 pim is 90 percent or greater).
Together, the default emission factors in Table 3-3 and control efficiencies from Table 3-4 yield
controlled emission factors that agree well with the controlled melting furnace emissions reported in the
Iron and Steel Foundry BID (U.S. EPA, 2002). To calculate emissions from the default emission factors
and control efficiencies, use Equation 3-6 for each size range of PM.
Note that the PM emission factors in Table 3-3 for PM10-FIL include emissions of PM2 5-FIL. PM-coarse—
FIL is the PM between 2.5 and 10 nm in diameter and the emissions factor specifically for PM between
2.5 and 10 pim in diameter is the emissions factor for PM10-FIL minus the emissions factor for PM2 5-FIL.
Therefore, to calculate PMm-FIL emissions, the emissions for PM coarse -FIL is added to the emissions for
pm25-fil.
EmFi
Ei = TTTTTTr x Q x (1 - CEi)
1 2,000 v v lJ
where:
Eqn. 3-6
E; = Emissions of pollutant "i" (tons/yr)
EmF, = Emission factor for pollutant "i" (lb/ton metal melted)
Q = Metal melt rate (tons/yr)
CE, = Control efficiency (fraction).
2,000 = Conversion factor, lbs/ton.
3-11
-------
Version 1
Draft
Section 3—Melting Operations
Condensable PM (PM-CON) is always assumed to be less than 2.5 nm in diameter and it is often not
controlled efficiently in PM emission control devices. The key to controlling condensable PM is to
sufficiently cool the gases prior to the control device so that the condensable PM is in particulate form as
it enters the control device rather than gaseous form. As exhaust gases from melting furnaces with closed
capture systems are hot, dry PM control systems may not effectively remove condensable PM. Because
wet systems generally have a water quench to cool the gases prior to scrubbing/PM removal, more of the
condensable PM is expected to be in particulate phase and amendable to control. The default control
efficiencies reported for PM-CON in Table 3-4 generally assume that 30 percent of the PM-CON is in
particulate form for the control systems that are designed to operate at approximately 200°F. In addition,
it also assumes that 50 percent PM-CON of the PM-CON is in particulate form for controls that are
designed to operate at approximately 165°F. So the condensable PM control efficiency is 30 to 50
percent of the control efficiency of PM in the 0 to 2.5 pim range. Example 3-5 demonstrates how to
estimate PM emissions using default emission factors.
Table 3-4. Typical Collection Efficiencies of Various Particulate Control Devices3 (%)
AIRS
Codeb
Type of Collector
Collection Efficiency
Condensable
PM
Filterable Particle
Size (pm)
0-2.5
2.5-10
001
Wet scrubber - hi-efficiency
30
90
97
002
Wet scrubber - med-efficiency
10
25
90
003
Wet scrubber - low-efficiency
7
20
85
004
Gravity collector - hi-efficiency
0
3.6
5
005
Gravity collector - med-efficiency
0
2.9
4
006
Gravity collector - low-efficiency
0
1.5
3.4
007
Centrifugal collector - hi-efficiency
0
80
95
008
Centrifugal collector - med-efficiency
0
50
80
009
Centrifugal collector - low-efficiency
0
10
42
010
Electrostatic precipitator - hi-efficiency
0
95
99
011
Electrostatic precipitator - med-efficiency
0
80
93
012
Electrostatic precipitator - low-efficiency
0
70
85
014
Mist eliminator - high velocity >250 FPM
0
10
92
015
Mist eliminator - low velocity <250 FPM
0
5
57
016
Fabric filter - high temperature (>250 °F)
0
99
99.5
017
Fabric filter - med temperature (180 °F 30 inches of water)
33
95
98
053
Venturi scrubber (AP < 30 inches of water)
30
88
96
054
Process enclosed
0
1.5
3.4
055
Impingement plate scrubber
0
25
97
056
Dynamic separator (dry)
0
90
97
3-12
-------
Version 1
Draft
Section 3—Melting Operations
AIRS
Codeb
Type of Collector
Collection Efficiency
Condensable
PM
Filterable Particle
Size (pm)
0-2.5
2.5-10
057
Dynamic separator (wet)
15
50
80
058
Mat or panel filter - mist collector
0
92
95
059
Metal fabric filter screen
0
10
17
061
Dust suppression by water sprays
0
40
77
062
Dust suppression by chemical stabilizer or wetting agents
0
40
77
063
Gravel bed filter
0
0
42
064
Annular ring filter
0
80
93
071
Fluid bed dry scrubber
0
10
55
075
Single cyclone
0
10
42
076
Multiple cyclone w/o fly ash reinjection
0
80
95
077
Multiple cyclone w/fly ash reinjection
0
50
80
085
Wet cyclonic separator
15
50
80
086
Water curtain
0
10
67
a Data represent an average of actual efficiencies. Efficiencies are representative of well-designed and well operated
control equipment. Site-specific factors (e. g., type of particulate being collected, varying pressure drops across
scrubbers, maintenance of equipment) will affect collection efficiencies. Efficiencies shown are intended to provide
guidance for estimating control equipment performance when source-specific data are not available. Table derived
from Table B.2-3 Typical Collection efficiencies of various particulate control devices of Appendix B.2 of AP-42,
Volume I, Fifth Edition (U.S. EPA, 1996)
b Control codes in Aerometric Information Retrieval System (AIRS), formerly National Emissions Data Systems.
3-13
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-5: Estimating PM Emissions from Melting Furnace
A facility melts 40,000 tons of gray iron in EIFs. Each EIF has a lid that is ducted to a
baghouse. Emissions from the EIF are uncontrolled when the lid is off, i.e., during charging
and tapping. Calculate the PM emissions from the melting furnace.
The relevant factors from Table 3-3 are as follows:
Pollutant
Uncontrolled Melting
Emission Factor
(lb/ton metal melted)
Uncontrolled
Charging/Tapping
Emission Factor
(lb/ton metal melted)
PM-FIL
1.5
0.5
PMio-FIL
1.4
0.5
PM2.5-FIL
1.1
0.4
PM-CON
0.05
0.01
First, the emissions from the control device must be determined. Without any other
information, it is assumed the baghouse operates at moderate temperatures. The default control
efficiency for a moderate temperature baghouse (fabric filter) is 33% for PM-CON, 99% for
PM < 2.5 pirn, and 99.5% for PM between 2.5 and 10 pirn. It is generally assumed that the
control efficiencies for particles greater than 10 nm in diameter is 100%.
Next, use Equation 3-6 with the melting emission factors:
EmF:
Ei = TTTTTTr x Q x (1 - CEi)
1 2,000 v v lJ
EpM-coN,meit = (0.05/2000) x 40,000 tons/yr x (1-0.33) = 0.67 tons/yr
EpM2.5-FiL.meit= (1.1/2000) x 40,000 tons/yr x (1-0.99) = 0.22 tons/yr
Note to calculate EpM"coarse"-FiL,meit subtract the PM25-FIL emission factor from the PM10-FIL
emission factor and use the control efficiency for particles from 2.5 - 10 pirn.
EPM"coarSe"-FiL,meit= ((1.4-1. l)/2000) x 40,000 tons/yr x (1-0.995) = 0.03 tons/yr
Recall that to calculate PMio-FIL emissions, calculate PM2 5-FIL emissions and PM coarSe -FIL
emissions and add together.
EpM10-FIL,melt= EpM2.5-FIL,melt + EpM"coarse"-FIL,melt =0.22 + 0.03 = 0.25 tOns/yr
Next, emissions from uncontrolled charging and tapping (C/T) are calculated:
Epm-con ,c/t = (0.01/2000) x 40,000 tons/yr x (1) = 0.2 tons/yr
EpM2.5-fil,c/t = (0.4/2000) x 40,000 tons/yr x (1) = 8 tons/yr
Epm-fil,c/t= Epmio-fil,c/t = (0.5/2000) x 40,000 tons/yr x (1) = 10 tons/yr
Adding the melting and charging/tapping emissions together, a complete PM inventory for the
EIF would be as follows:
PM-CON = EPM_coN,meit + Epm-con,c/t = 0.67+0.2 = 0.87 tons/yr
PMio-FIL = EpMio-FiL,meit+ EPMio-fil,c/t = 0.25 + 10 = 10.25 tons/yr
PMio-PRI = PMio-FIL + PM-CON = 10.25 + 0.87 = 11.12 tons/yr
PM2 5-FIL = EPM2.5_FiL,meit+ EPM2.5-fil,c/t = 0-22 + 8 = 8.22 tons/yr
PM2 5-PRI = PM2 5-FIL + PM-CON = 8.22 + 0.87 = 9.09 tons/yr
-------
Version 1
Draft Section 3—Melting Operations
3.1.4.2 Other Criteria Pollutant Emissions Inventory Default Factors
For other criteria pollutants (except for lead, which is also a HAP), default emission factors are presented
in Table 3-5.
Table 3-5. Summary of Non-PM Criteria Pollutant Emission Factors
for Melting Furnace Operations3
Suggested
SCC for Iron
Foundries
Suggested
SCC for Steel
Foundries
Emission category
CO,
lb/ton
so2,
lb/ton
NOx,
lb/ton
voc,
lb/ton
30400301
Cupolas with afterburner and
either fabric filter or
uncontrolled for PM
3
0.155b
ND
ND
30400301
Cupolas with afterburner and
wet scrubber control
3
0.019b
ND
ND
30400301
Cupola with no afterburner and
either fabric filter or
uncontrolled for PM
145
0.155b
30400301
Cupolas with no afterburner
and wet scrubber control
145
0.019b
30400304
30400701
EAF
18
Negligible
0.3
0.15
30400316
Electric induction furnace (EIF)
Negligible
Negligible
ND
ND
30400302
Reverberatory
ND
ND
ND
ND
a Emission factors, in lb of pollutant per ton of grey iron melted, based on data reported in AP-42 (U.S. EPA, 2003),
unless otherwise specified; used mid-range value when a range was reported. ND = no data.
b Derived from data received during the development of the MACT standards for iron and steel foundries; see EPA
Docket No. EPA-HQ-OAR-2002-0034 and Appendix B of this Protocol document.
3.1.4.3 Metal HAP Emissions Inventory Default Factors
Most metal HAP emissions will be associated with the filterable PM emissions. To estimate melting
furnaces emissions of these metallic HAP, site-specific metal chemistry should be used to estimate the
metal content of the emitted (filterable) PM (Methodology Rank 4A). In the absence of site-specific metal
chemistry data, the default metal chemistries provided in Table 3-6 can be used to estimate HAP metal
emissions (Methodology Rank 4B). As noted in Table 3-6, default metal compositions are provided for
both filterable and condensable PM. For certain metal HAP, such as mercury (Hg), the metal HAP
emissions are expected to be in vapor form and will not be controlled or well correlated with the PM
emissions. These pollutants will act like condensable PM and will have limited levels of control in most
particulate control devices used with melting furnaces. It is inappropriate to use the chemistry of the
melted metal as the composition for the condensable PM. Use Equation 3-7 to determine the emissions
of specific HAP metals from the melting furnace PM emission estimates.
%PM-FILi %PM-C0Ni
Ei = x PM-FIL + „„„n/ X PM-CON
1 100% 100%
Eqn. 3-7
where:
E; = Emissions of pollutant "i" (tons/yr)
%PM-FIL, = Percent of filterable PM mass contributed by pollutant "i" (weight percent)
PM-FIL = Total filterable PM emissions rate (tons/yr)
% PM-CON = Percent of condensable PM mass contributed by pollutant "i" (weight percent)
PM-CON = Condensable PM emissions rate (tons/yr).
3-15
-------
Version 1
Draft
Section 3—Melting Operations
Table 3-6. Default Metal Composition for PM from Melting Furnace Operations3
Suggested SCC
for Iron
Foundries
Suggested SCC
for Steel
Foundries
CAS No.
Metal Constituent
% of PM-
FIL
%of P A/I-
CON
30400301
30400302
30400303
30400304
30400701
30400705
7440-36-0
Antimony
0.01
0.02
30400301
30400302
30400303
30400304
30400701
30400705
7440-38-2
Arsenic
0.003
0.01
30400301
30400302
30400303
30400304
30400701
30400705
7440-39-3
Barium
0.03
0.1
30400301
30400302
30400303
30400304
30400701
30400705
7440-41-7
Beryllium
0.0001
0.007
30400301
30400302
30400303
30400304
30400701
30400705
7440-43-9
Cadmium
0.02
0.01
30400301
30400302
30400303
30400304
18540-29-9
Chromium (hexavalent)
Iron Foundry Melting
0.0024 b
0.0015 b
30400701
30400705
18540-29-9
Chromium (hexavalent)
Steel Foundry Melting
0.0096 b
0.006 b
30400301
30400302
30400303
30400304
30400701
30400705
7440-47-3
Chromium (total)
0.08
0.05
30400301
30400302
30400303
30400304
30400701
30400705
7440-48-4
Cobalt
0.001
0.003
30400301
30400302
30400303
30400304
30400701
30400705
7439-92-1
Lead
1
0.3
30400301
30400302
30400303
30400304
30400701
30400705
7439-96-5
Manganese
3
2.9
3-16
-------
Version 1
Draft
Section 3—Melting Operations
Suggested SCC
for Iron
Foundries
Suggested SCC
for Steel
Foundries
CAS No.
Metal Constituent
% of PM-
FIL
%of P A/I-
CON
30400301
30400302
30400303
30400304
30400701
30400705
7439-97-6
Mercury
0.02
0.5-2.0°
30400301
30400302
30400303
30400304
30400701
30400705
7440-02-0
Nickel
0.2
0.04
30400301
30400302
30400303
30400304
30400701
30400705
7723-14-0
Phosphorus
0.2
0.2
30400301
30400302
30400303
30400304
30400701
30400705
7782-49-2
Selenium
0.0015
0.015
30400301
30400302
30400303
30400304
30400701
30400705
7440-66-6
Zinc
9
2
a Derived from test data received from 1998 Foundry ICR. See Appendix B for more details on the development of
the default metal composition of PM.
b Assume hexavalent chromium is 3% of total chromium emissions for Iron foundry melting and 12% oftotal
chromium emissions for steel foundry melting based on Chromium hexavalent percentages reported for iron and
steel foundry Standard Classification Codes (SCC) in Appendix D of National-scale Air Toxics Assessment
(NATA). For other non-melting operations at Iron and Steel foundries using this table for default values, use 3% of
the total chromium emissions (U.S. EPA, 2011a).
c Use the lower value if no automobile scrap is used and the higher value if automobile scrap is used as part of the
charge material to the furnace.
3-17
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-6: Estimating Metal HAP Emissions with Site-specific Metal Chemistries
For the EIF in Example 3-5, what are the metal HAP emissions given the following site-
specific metal chemistry details for the melted metal within the furnace?
% Pb
% Mn
% Cd
% Cr
% Ni
% Hg
% Selenium
1.10
7.28
0.0013
0.078
0.013
0.037
0.0094
Use Equation 3-7 with the site-specific composition data above and the default composition
data from Table 3-6. From Example 3-5, PMio-FIL = 10.25 tons/yr and PM-CON = 0.87
tons/yr. Since the melting portion is controlled and the charging emission factores for PM and
PM10 are identical, PM-FIL = PM10-FIL = 10.25 tons/yr
%PM-FIL: %PM-CONj
E; = X PM-FIL + — X PM-CON
Elead —
100%
1.1%
100%
100%
0.3%
x 10.25 + x 0.87 = 0.115 tons/yr
-manganese - 10qo/o
0.0013%
100%
7.28% 2.0%
x 10.25+ x 0.87 = 0.764 tons/yr
-¦cadmium
100%
0.078%
-chromium
100%
0.02%
X 10.25 H X 0.87 = 0.000307 tons/yr
100% n
0.05%
x 10.25 + -tttttt- x 0.87 = 0.00843 tons/yr
100% 100%
ECr+6 = 3% (for iron melting) x Echromium = 0.000253 tons/yr
0.013% 0.04%
Enickei = x 10.25 H X 0.87 = 0.00168 tons/yr
nickel 1.00% 100% n
If automobile scrap is melted in the EIF:
E
-"mercury
0.037% 2.0%
x 10.25 + x 0.87 = 0.0212 tons/yr
100% 100%
If automobile scrap is NOT melted in the EIF:
E
-"mercury
0.037% 0.5%
x 10.25 + x 0.87 = 0.00814 tons/yr
-selenium
100%
0.0094%
100%
100%
0.01%
x 10.25 H x 0.87 = 0.00105 tons/yr
100% n
3.1,4,4 Organic HAP Emissions Inventory Default Factors
Organic HAP emission data for cupolas that use afterburners indicate negligible organic HAP emissions,
with most HAP present below analytical detection limits (U.S. EPA, 2002) although small, detectable
quantities of dioxin have been measured at iron and steel foundries. Recommended dioxin/furan emission
factors are provided in Table 3-7. Note that the emission factors in Table 3-7 are in units of nanogram
(10~9 grams) of dioxin toxicity equivalence (TEQ) per kg of metal melted. Other organic compounds
generally emitted during the combustion of coke (e.g., benzene, toluene, ethylbenzene, xylenes,
naphthalene, and 2-methylnaphalene) are expected to be emitted from a cupola if there is no or inefficient
afterburning; however, there are inadequate test data to develop organic HAP emission factors for cupolas
with no or poor afterburning.
3-18
-------
Version 1
Draft
Section 3—Melting Operations
Table 3-7. Congener-specific Profile for Ferrous Foundries3
Suggested SCC
for Iron
Suggested SCC
for Steel
Mean emission factor (2 facilities)
(ng TEQ-WHOo5/kg of metal melted)
Nondetect set to %
Foundries
Foundries
Congener
Nondetect set to zero
detection limit
30400301
30400701
2,3,7,8-TCDD
0.11
0.11
30400302
30400705
1,2,3,7,8-PeCD
0.30
0.30
30400303
1,2,3,4,7,8-HxCDD
0.012
0.012
30400304
1,2,3,6,7,8-HxCDD
0.023
0.023
1,2,3,7,8,9-HxCDD
0.028
0.028
1,2,3,4,6,7,8-
0.0033
0.0033
HpCDD
0.048
0.048
OCDD
30400301
30400701
2,3,7,8-TCDF
0.084
0.084
30400302
30400705
1,2,3,7,8-PeCDF
0.65
0.65
30400303
2,3,4,7,8-PeCDF
0.13
0.13
30400304
1,2,3,4,7,8-HxCDF
0.1
0.1
1,2,3,6,7,8-HxCDF
0.0079
0.0079
1,2,3,7,8,9-HxCDF
0.075
0.075
2,3,4,6,7,8-HxCDF
0.0082
0.0082
1,2,3,4,6,7,8-
0.0014
0.0014
HpCDF
0.00009
0.00009
1,2,3,4,7,8,9-
HpCDF
0.00007
0.00007
OCDF
30400301
30400701
Total TEQ-WHOos
1.57
1.57
30400302
30400705
30400303
30400304
a Data from U.S. EPA, 1999a and 1999b, as reported in U.S. EPA, 2006a. Toxic Equivalent (TEQ) values reported
using 2005 World Health Organization (WHO) Toxic Equivalent Factors (TEF).
3.2 Scrap Handling, Preparation, and Preheating
Iron and steel foundries use recycled scrap metals as their primary source of metal and resort to metal
ingots as a secondary source when scrap is not available. Scrap metals may require some type of
preparation prior to melting, such as cutting or sizing, shot or sand blasting to remove coatings, cleaning
with organic solvents to remove oils and grease, and drying. (U.S. EPA, 2002)
Another form of scrap preparation that is commonly used is specification of quality. Iron and steel
foundries may use some type of scrap selection, cleaning, or inspection program to ensure the quality of
scrap metal used by the foundry. The scrap selection, cleaning, or inspection programs included
specifications on the types or grades of scrap used; limits or bans on oil, grease, and/or paint in the scrap;
and restrictions on lead, galvanized metals (a source of cadmium), and certain alloys (a source of
chromium, nickel, or high manganese) (U.S. EPA, 2002).
Foundries can use preheaters to increase the temperature of the scrap prior to being melted in the furnace
(most common for electric induction furnaces). Mechanical processes associated with scrap preheaters
(e.g., loading of scrap) generate PM emissions that are of concern only in the work area. Scrap preheating
itself can produce both PM and organic emissions. Over 90 percent of preheaters are direct-fired with
natural gas. Metal HAP content of the PM is expected to be a function of the composition of the scrap.
Organic HAP emissions, which arise from oil and grease contaminants, are assumed to include products
3-19
-------
Version 1
Draft
Section 3—Melting Operations
of incomplete combustion. Scrap preheating entails loading, heating, and discharging of the scrap, and
each individual step can be uncontrolled or controlled with a fabric filter, cyclone, afterburner, or
scrubber. (U.S. EPA, 2002)
3.2.1 Methodology Rank 1 and 2 for Scrap Handling, Preparation, and Preheating
Emissions from Scrap Handling, Preparation and Preheating can be directly measured at the stack or
outlet of the control device using a CEMS. If CEMS are available for both a pollutant concentration and
flow rate, the CEMS data should be used to determine the pollutant emissions as Methodology Rank 1 for
Scrap Handling, Preparation, and Preheating. If a CEMS is used to determine a pollutant's concentration,
but direct flow measurement is not available, flow rates can often be determined using engineering
estimates, such as fan amperage-to-flow correlations, to determine the pollutant emissions as
Methodology Rank 2 for Scrap Handling, Preparation and Preheating. Follow the calculation methods
provided in Section 3.1, Melting Furnaces, of this Protocol document as an example.
3.2.2 Methodology Rank 3 for Scrap Handling, Preparation, and Preheating
Source testing can provide useful data for developing site-specific emission correlations or emission
factors. Source testing provides a measurement of the emissions at a particular point in time, and most
tests are performed at conditions representative of normal operation, in which case, the emission
measurement can provide an estimate of emissions at similar operating conditions. Site-specific emission
factors are developed by dividing the emission rate by a process parameter, such as metal charge rate or
metal processed rate using Equation 3-4. The annual emissions are then estimated using Equation 3-5
following the same steps as in Example 3-3, above. It is important to note that this methodology will be
less reliable when the unit is operating at conditions other than those tested.
3.2.3 Methodology Rank 4 for Scrap Handling, Preparation, and Preheating
When direct emission monitoring or site-specific emission factors are not available, then default emission
factors may be the only way to estimate emissions. The EPA has developed a PM emission factor for
scrap and charge handling; heating for iron and steel foundries in AP-42 Section 12.10; and from scrap
handling for steel foundries in AP-42 Section 12.13 (U.S. EPA 1995, 2003). The appropriate default
emission factors are provided in Table 3-8. For scrap handling (scrap piles indoors) and/or scrap
preheating, the emission factors in Table 3-8 should be used. For scrap handling, preparation, and
preheating, it is assumed that there are no condensable PM emissions. While scrap preheating may
generate some condensable PM, the default condensable PM emission factor for melting furnaces is
expected to cover the condensable emissions from the combined preheating/melting operations.
To develop HAP-specific emission estimates from scrap handling, preparation, and preheating, the PM
emission factors in Table 3-8 should be used in conjunction with site-specific metal chemistries to
estimate HAP metal emissions when site-specific metal chemistry data are available (Methodology Rank
4A). When site-specific metal chemistry data are not available, use the default metal compositions for PM
from melting furnace operations provided previously in Table 3-6 (Methodology Rank 4B).
3-20
-------
Version 1
Draft
Section 3—Melting Operations
Table 3-8. Particulate Emission Factors for Scrap and Charge Handling,
Heating at Iron and Steel Foundries
Suggested
SCC for Iron
Foundries
Suggested
SCC for Steel
Foundries
Emission Source
Pollutant
Emission
Factor
(lb/ton metal
processed)
30400315
30400712
30400768
Scrap and charge handling, captured and
ducted uncontrolled to atmosphere
PM-FIL
PM10-FIL
PM2.5-FIL
0.6 a
0.5 b
0.4 b
30400314
30400741
Scrap and charge preheating, captured and
ducted uncontrolled to atmosphere
PM-FIL
PM10-FIL
PM2.5-FIL
0.6 a
0.5 b
0.4 b
30400315
30400712
30400768
Scrap and charge handling, uncaptured
fugitive dust to atmosphere
PM-FIL
PM10-FIL
PM2.5-FIL
0.2 a
0.18 c
0.17 c
30400314
30400741
Scrap and charge preheating, uncaptured
fugitive dust to atmosphere
PM-FIL
PM10-FIL
PM2.5-FIL
0.2 a
0.18 c
0.17 c
a Used the default factor for iron foundries (U.S. EPA, 2003).
b Assumes approximately 90% of PM is PM-io and 70% of PM is PM2.5, similar to melting furnace size distribution.
c Assumes 90% of PM is PM10, so that the PM10-FIL factor matches that reported for steel foundries (U.S. EPA,
1995) and 85% of PM is PM2.5, as smaller particles are more likely to escape to the atmosphere.
In estimating organic emissions from organic solvents used to clean/prepare scrap, use the VOC and HAP
content found in the MSDS for the solvent and multiply by the amount of the solvent purchased for the
year. The procedure is similar to the procedures for estimating organic emissions from mold and core
making (as presented in Section 4 of this Protocol document); it is assumed that 100 percent of the
organics are emitted during scrap cleaning and preparation.
3-21
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-7: Default Emissions Factor with Site-specific Metal Chemistries
A facility lias an indoor scrap handling operation and preheater. The annual throughput for the scrap
handling and preheater was 29,865 tons/yr last year. The scrap handling area is uncontrolled; the scrap
preheater is controlled via a cyclone. What are the PM emissions for the scrap handling, preparation
and preheating operations?
The PM emissions can be calculated using the default uncontrolled PM emission factor and the default
PM control efficiencies. As noted above, for scrap handling and preheating PM-CON = 0 so PMm-PRI
= PMio-FIL. It is generally assumed that the control efficiencies for particles greater than 10 nm in
diameter is 100%,. The relevant factors from Table 3-8 are as shown below:
Emission Factor
Emission Source
Pollutant
(lb/ton metal processed)
Scrap and charge handling , uncaptured
PM-FIL
0.2
fugitive dust to atmosphere
_l
LL
o
CL
0.18
pm2.5-fil
0.17
Scrap and charge preheating, captured
PM-FIL
0.6
and ducted uncontrolled to atmosphere
_l
LL
0
CL
0.5
PM2.5-FIL
0.4
Next, use Equation 3-6 with the uncontrolled scrap handling (SH) emission factors:
EmF:
Ei = TTTTTTr x Q x (1 - CEi)
1 2,000 v v lJ
Note that EPM-con,sh = 0 tons/yr (so PM-PRI = PM-FIL) and CESH = 0 (uncontrolled)
Epm2.5-fil,sh= (0.4/2000) x 29,865 tons/yr x (1-0) = 5.97 tons/yr
Epm-fil,sh = Epmio-fil,sh= (0.6/2000) x 29,865 tons/yr x (1) = 8.96 tons/yr
Next, use Equation 3-6 with the uncontrolled scrap preheating (SP) emission factors and control
efficiency for medium efficiency cyclone. It is assumed the cyclone operates similarly to a centrifugal
collector - medium efficiency. The default control efficiency for a centrifugal collector - med-
efficiency is 50% for PM < 2.5 nm and 80% for PM between 2.5 and 10 nm:
Epm-con,sp = 0 tons/yr
Epm2.5-fil.sp = (0.4/2000) x 29,865 tons/yr x (1-0.50) = 2.99 tons/yr
Note to calculate EPM. coarse-.FIL SP subtract the PM2 5-FIL emission factor from the PM10-FIL emission
factor and use the control efficiency for particles from 2.5 - 10 nm
EpM"coarse"-FiL,sp= ((0-5 - 0.4)/2000) x 29,865 tons/yr x (1-0.8) = 0.30 tons/yr
Recall that to calculate PM10-FIL emissions, calculate PM2 5-FIL emissions and PM -coarse-FIL
emissions and add together.
Epm-fil,sp = Epmio-fil,sp= EpM2.5-fil,sp + EpM~coarse"-FiL,sp =2.99 + 0.30 = 3.29 tons/yr
Adding the scrap handling and scrap preheating emissions together, a complete PM inventory
for scrap handling and scrap preheating would be as follows:
PM-CON = Epm_COn,sh + Epm-con,sp = 0 + 0 = 0 tons/yr
PMiq-FIL = Epm10_fil,sh+ EPmio-fil,sp= 8.96 + 3.29 = 12.25 tons/yr
PMio-PRI = PM10-FIL + PM-CON = 12.25 + 0 = 12.25 tons/yr
PM2.5-FIL = EPm2.5-fil,sh+ EPM2.5-fil,sp = 5.97 + 2.99 = 8.96 tons/yr
PM2 5-PRI = PM2 5-FIL + PM-CON = 8.96 + 0 = 8.96 tons/yr
-------
Version 1
Draft
Section 3—Melting Operations
Example 3-8: Estimating Metal HAP Emissions with Site-specific Metal Chemistries
For the scrap handling and scrap preheating operations in Example 3-7, what are the metal
HAP emissions given the following site-specific HAP metal chemistries: Mn, % = 1.03; Pb, %
= 0.56; Ni, % = 0.05; and other HAP metals much less than 0.01 %.
Use Equation 3-7 with the site-specific composition data above. From Example 3-7, PM-FIL
= 12.25 tons/yr and PM-CON = 0 tons/yr.
%PM-FIL: %PM-CONj
Ei = „^n/ X PM-FIL + „„„n/ X PM-CON
1 100% 100%
0.56%
Eiead = x 12.25 = 0.069 tons/yr
lead 100% /y
1.03%
100%
^manganese 1 nnr,/ * 12.25 0.13 tons/yr
0.05%
^nickel = x 12.25 = 0.0061 tons/yr
nickel 10Q%
3.3 Metallurgical Treatment of Molten Metal
Before the molten metal is cast, the chemical composition is adjusted to meet product specifications by
inoculation, or refining. Inoculation is the process in which magnesium and other elements are added to
molten iron to produce ductile iron. Ductile iron is formed as a steel matrix containing spheroidal
particles (or nodules) of graphite. Ordinary cast iron contains flakes of graphite. Each flake acts as a
crack, which makes cast iron brittle. Ductile irons have high tensile strength and are silvery in appearance
(U.S. EPA, 2003). Metallurgical treatment also includes adding specific alloying materials to the molten
metal in a holding furnace or ladle.
3.3.1 Methodology Rank 1 and 2 for Metallurgical Treatment of Molten Metal
Emissions from metallurgical treatment of molten metal can be directly measured at the stack or outlet of
the control device using a CEMS. If CEMS are available for both a pollutant concentration and flow rate,
the CEMS data should be used to determine the pollutant emissions as Methodology Rank 1 for
Metallurgical Treatment of Molten Metal. If a CEMS is used to determine a pollutant's concentration, but
direct-flow measurement is not available, flow rates can often be determined using engineering estimates,
such as fan amperage-to-flow correlations, to determine the pollutant emissions as Methodology Rank 2
for Metallurgical Treatment of Molten Metal. Follow the procedures for estimating emissions using a
CEMS provided in Section 3.1, Melting Furnaces, of this Protocol document as an example.
3.3.2 Methodology Rank 3 for Metallurgical Treatment of Molten Metal
Source testing can provide useful data for developing site-specific emission correlations or emission
factors. Source testing provides a measurement of the emissions at a particular point in time, and most
tests are performed at conditions representative of normal operation, in which case the emission
measurement can provide an estimate of emissions at similar operating conditions. Emission factors are
developed by dividing the emission rate by a process parameter such as metal charge rate or fuel usage
using Equation 3-4. The annual emissions are then estimated using Equation 3-5 following the same steps
as in Example 3-3, above. It is important to note that this methodology will be less reliable when the unit
is operating at conditions other than those tested.
3-23
-------
Version 1
Draft
Section 3—Melting Operations
3.3.3 Methodology Rank 4 for Metallurgical Treatment of Molten Metal
When direct emission monitoring or site-specific emission factors are not available, then default emission
factors may be the only way to estimate emissions. The EPA has developed a PM emission factor for
metallurgical treatment of molten metal for iron foundries in AP-42 Section 12.10. Table 3-9 presents the
recommended emission factors for PM for metallurgical treatment of molten metal at both iron and steel
foundries. The "ducted" emission factor of 1.8 lb/ton from AP-42 (U.S. EPA, 2003) agrees reasonably
well with the baghouse catch data reported in the Iron and Steel Foundries BID (U.S. EPA, 2002), which
averaged 2.6 lb PM/ton metal processed, so the uncontrolled emission factors reported in AP-42 were
selected. Note that for "in pour" magnesium inoculation, the inoculation emissions are expected to be
negligible. Thus, the emission factors presented in Table 3-9 should not be used for "in pour" inoculation.
Table 3-9. Particulate Emission Factors for Metallurgical Treatment at Iron and Steel Foundries
Emission
Suggested
Suggested
Factor
SCC for Iron
SCC for Steel
(lb/ton metal
Foundries
Foundries
Emission Source
Pollutant
produced)
30400310
Inoculation/metal treatment,
PM-FIL
1.8 a
30400321
captured and ducted uncontrolled to
PM10-FIL
1.6 b
30400322
atmosphere
PM2.5-FIL
1.3 b
30400310
Inoculation/metal treatment,
PM-FIL
0.4 a
30400321
uncaptured fugitive dust to
PM10-FIL
0.38 c
30400322
atmosphere
PM2.5-FIL
0.34 c
a Used the default factor for iron foundries (U.S. EPA, 2003).
b Assumes approximately 90% of PM is PM-io and 70% of PM is PM2.5, similar to melting furnace size distribution.
c Assumes 95% of PM is PM10 and 85% of PM is PM2.5, as smaller particles are more likely to escape to the
atmosphere.
The metal HAP contents of the PM emitted from metallurgical treatment are expected to be weighted
toward the composition of the inoculants or alloying material with some emissions of the original melted
metal. To estimate the HAP emissions from metallurgical treatment, it is recommended that the PM
emissions be estimated as comprising 90 percent of the inoculants and 10 percent of the molten metal to
which the inoculants (or alloying materials) are added. The default metal composition for PM from
melting furnace operations provided previously in Table 3-6 can be used if site-specific metal chemistry
data are not available.
3.4 Holding Furnaces
After melting, the molten metal is tapped from the furnace into either a holding furnace or a transfer ladle.
A holding furnace is generally an EIF used to maintain the molten metal in the proper condition until the
foundry is ready to pour. Generally, emissions from holding furnaces are expected to be small because the
holding furnaces are generally covered to retain heat. At times, the melting furnace acts as a holding
furnace, in which case, the melting emission factors are expected to include emissions from this
temporary holding process. Transfer ladles are generally open, but the molten metal does not remain in
the ladles for long, and emissions from the ladle are expected to be small compared to tapping,
metallurgical treatment, and pouring emissions, which could generally be considered, at least in part,
emissions from the transfer ladle. This section should only be used to estimate emissions from dedicated
holding furnaces that are not covered or that are covered and vented to the atmosphere (with or without
PM emissions control).
3-24
-------
Version 1
Draft
Section 3—Melting Operations
3.4.1 Methodology Rank 1 and 2 for Holding Furnaces
Emissions from holding furnaces can be directly measured at the stack or outlet of the control device
using a CEMS. If CEMS are available for both a pollutant concentration and flow rate, the CEMS data
should be used to determine the pollutant emissions as Methodology Rank 1 for Holding Furnaces. If a
CEMS is used to determine a pollutant's concentration, but direct flow measurement is not available, flow
rates can often be determined using engineering estimates, such as fan amperage-to-flow correlations, to
determine the pollutant emissions as Methodology Rank 2 for Holding Furnaces. Follow the procedures
for estimating emissions using a CEMS provided in Section 3.1, Melting Furnaces, of this Protocol
document as an example. An example of the use of fan curves to estimate exhaust flow rates is provided
in Section 5.1, Methodology^ Ranks 1 and 2 for PCS Operations.
3.4.2 Methodology Rank 3 for Holding Furnaces
When source testing data are available, a site-specific emission factor can be developed based on either
the metal processing rate of the furnace or a direct emission rate per hours of operation of the furnace.
Emission factors are developed by dividing the emission rate by a process parameter such as metal charge
rate or fuel usage using Equation 3-4. The annual emissions are then estimated using Equation 3-5
following the same steps as in Example 3-3, above. It is important to note that this methodology will be
less reliable when the unit is operating at conditions other than those tested.
3.4.3 Methodology Rank 4 for Holding Furnaces
When direct emission monitoring or site-specific emission factors are not available, then default emission
factors may be the only way to estimate emissions. Table 3-10 presents the recommended emission
factors for PM for holding furnaces at both iron and steel foundries. If the holding furnace is ducted to a
control device, use the control efficiencies in Table 3-4 along with the captured/uncontrolled emission
factors in Table 3-10 to estimate the emissions to the atmosphere (exiting the control device).
Table 3-10. Particulate Emission Factors for Holding Furnaces at Iron and Steel Foundries
Emission
Suggested
Suggested SCC
Factor
SCC for Iron
for Steel
(lb/ton metal
Foundries
Foundries
Emission Source
Pollutant
produced)
30400303
30400701
Holding furnace, captured and ducted
PM-FIL
0.5 a
30400304
uncontrolled to atmosphere
_l
LL
0
CL
0.45 a
PM2.5-FIL
0.35 a
30400303
30400701
Holding furnace, uncovered fugitive
PM-FIL
0.3 b
30400304
dust to atmosphere
_l
LL
0
CL
0.29 c
PM2.5-FIL
0.26 c
a Used one-third the default factor for melting in induction furnaces (U.S. EPA, 2003).
b Assumes approximately 60% of PM released from furnace is emitted to atmosphere.
c Assumes 95% of PM is PM-io and 85% of PM is PM2.5, as smaller particles are more likely to escape to the
atmosphere.
The metal HAP contents of the PM emitted from holding furnaces are expected to be similar to the metal
HAP contents of the PM emitted from the melting furnace. To estimate the HAP emissions from holding
furnaces, the default metal composition for PM from melting furnace operations provided previously in
Table 3-6 can be used if site-specific metal chemistry data are not available.
3-25
-------
Version 1
Draft
Section 4—Mold and Core Making
4. Mold and Core Making
The predominant casting operations at iron and steel foundries include sand mold, centrifugal, permanent
mold, investment, and expendable pattern casting. Sand molds are bonded using resin-like chemicals or
clay plus other materials. Permanent and centrifugal casting operations use metal molds, and investment
casting operations use molds made from refractory material. Expendable pattern casting uses molds of
unconsolidated sand. A variety of cores can be used with each type of mold. Most cores are made from
chemically bonded sand. Others are made from plaster, collapsible metal, or soluble salts.
The emission estimation methods for mold and core making are presented in Table 4-1. Most equipment
for mold and core making does not include well-defined stacks. Exceptions are baking ovens and cold-
box machines that use catalyst gases. For emissions that are captured and emitted through well-defined
stacks, direct measurement methods may be used. Direct measurements may be continuous or routinely
frequent (e.g., daily or weekly) in nature so that the measurement data can be used directly to determine
emissions (Methodology Rank 1). For many pollutants, a one-time or annual source test may be
performed, but the emissions are not routinely measured. For these pollutants, the measured data can be
used to determine an emission factor based on process throughput (quantity of sand used or quantity of
binder/catalyst used), and the annual emissions can be determined using the annual material usage
multiplied by the site-specific emission factor (Methodology Rank 2). However, for most mold and core
making operations, no direct measurement data will be available. In such cases, emissions will be
estimated based on binder composition, usage rates, and default release factors (Methodology Rank 3) or
generic emission factors (Methodology Rank 4).
Table 4-1. Summary of Typical Hierarchy of Mold and Core Making Emission Estimates
Rank
Methodology
Description
Application
Data Requirements
1
Direct routine
measurement
Baking ovens, gas-cured
binder systems, and other
systems that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate
2
Site-specific emission
factor from one-time or
periodic emissions
source test data
Baking ovens, gas-cured
binder systems, and other
systems that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate; and material usage rate (or throughput)
during source test; annual material usage
rate (or annual throughput)
3A
Chemical-specific
release factors
Chemical binder systems
and coating materials
Composition of chemical binder or coating
material (from Material Safety Data Sheets)
and binder/coating material annual usage
quantities
3B
Chemical-specific
release factors
Chemical binder systems
and coating materials
Composition of chemical binder or coating
material (defaults) and binder/coating
material annual usage quantities
4
Default emission factors
All sources
Material usage rates
Emissions from mold and core making operations include PM emissions from sand handling and mixing
as well as organic emissions from the release or volatilization of constituents in organic binders or coating
materials used to make (or coat) the molds and/or cores. The bulk of this section focuses on emission
estimating procedures for organic chemical releases during the mixing, setting, and storage of the molds
and cores. These procedures are also applicable to mold and core coating materials. PM emissions from
the sand handling/mixing system are difficult to assess unless they are well-captured and vented through a
defined stack. Otherwise, particles that may become airborne (suspended) within the foundry during sand
4-1
-------
Version 1
Draft
Section 4—Mold and Core Making
handling/mixing are likely to redeposit within the foundry, and only a fraction of the PM that becomes
suspended will be released "to the atmosphere" (i.e., escape the foundry building). Default PM emission
factors (Methodology Rank 4) for sand handling operations are provided in Section 4.3 of this Protocol
document.
Different foundries can have significantly different emissions from mold and core making depending on
the type of casting system employed, binder system used to make cores, and whether the sand molds are
chemically bonded. For example, permanent and centrifugal molding operations may not use sand at all
(if no cores are needed), so they would have no PM emissions from sand handling or mixing. Even if
cores are used, permanent and centrifugal molding operations will handle very limited quantities of sand
compared to sand casting systems, so their PM emissions will be much smaller. Many sand molds are
made from clay-bonded sand, commonly called green sand, which uses clay and water as the binder.
While foundries that use green sand molds with little to no cores will have PM emissions from sand
handling/mixing, they may have very limited or no organic HAP releases from their mold making
operations. Molds that use chemically bonded sand or that require significant quantities of core materials
can have significant organic HAP emissions; however, the specific HAP released may vary significantly
based on the chemical binder system used. Chemical bonding systems work by polymerization reactions
that occur at ambient temperature or are induced by heat or catalysis. The major types of binder systems
used for core making are the oil-bake, shell, hot-box, warm-box, no-bake, and cold-box systems. The
major system used for mold making is the shell system (U.S. EPA, 2002).
4.1 Methodology Ranks 1 and 2 for Mold and Core Making
Emissions from vented mold and core making processes can be directly measured at the stack or outlet of
the control device using the direct measurement methods described in Section 3, Melting Operations, of
this Foundries Emissions Protocol document. If CEMS are available for a pollutant, the CEMS data
should be used to determine the pollutant emissions as Methodology Rank 1 for Mold and Core Making.
It is more likely, however, that only data from a single source test or a limited number of source tests are
available. For this situation, a site-specific emission factor can be developed and used to assess the
emissions from the mold and core making process, similar to the methods described in Section 3, Melting
Operations. For some mold and core making systems, the "activity" data are likely to be specific
chemical usage rates or sand usage rates (rather than metal charge or melting rates). Example 4-1
provides an example of developing a site-specific emission factor for estimating triethyl amine (TEA)
emissions from a wet scrubber used to control a cold-box catalyst gas sweep system.
4-2
-------
Version 1
Draft
Section 4—Mold and Core Making
Example 4-1: Development of Site-Specific Emissio tor
A source test was performed on an acidic wet scrubber used to control TEA gas catalyst
emissions from a cold box core making line. The average TEA emissions rate for the three test
runs was 0.03 pounds per hour (lb/hr). The cumulative TEA usage over the 5-hour period
encompassing the source test runs was 88 lbs. What is the site-specific emission factor for
TEA? If the annual TEA usage rate is 35,500 lbs, what are the annual emissions of TEA from
the acidic wet scrubber? What is the annual emissions ir the capture system only captures 90%
of the TEA from the cores?
The average TEA usage rate during the source test was: 88 (lbs)/5 (hrs) = 17.6 lbs/hr.
The site-specific emission factor is simply the emissions rate divided by the TEA usage rate:
EmFTEA = 0.03 (lbs/hr) - 17.6 (lbs/hr) = 0.00171 lb/lb or 3.42 lb/ton.
Given the annual TEA usage of 35,000 lbs = 17.5 tons (per year), the annual average emissions
rate for TEA is estimated as: 3.42 lb/ton x 17.5 tons/yr = 59.85 lbs/yr or 0.0299 tons/yr.
If the TEA capture system is only 90 percent, Equation 1-4 can be used to estimate the annual
emissions.
Elotal= Euncntrld X (1 — CapEff X CEj)
Eunctrid = emissions if the source was uncontrolled = TEA usage rate = 35,000 lbs/yr
The capture efficiency, CapEff, is given as 90% or 0.90.
To determine the collection efficiency, CEtea, the inlet mass loading rate to the scrubber must
be calculated given the assumed capture efficiency.
Mass TEA Inlet = TEA Usage Rate x CapEff = 17.6 [bls/hr] x 0.9 = 15.8 lbs/hr.
The control efficiency = 1 - Mass Out/Mass In = 1 - 0.03/15.8 = 0.998
ETotal = 35,000 [lbs/yr] x (1 - 0.9x0.998) = 3,560 lbs/yr or 1.78 tons/yr.
In this case, accounting for the capture efficiency of the TEA scrubber is critical for
determining the true emissions from thus core-making unit.
The Ohio Cast Metal Association (OCMA) funded a study to assess the emission losses during mold and
core-making, including the curing process, by measuring the mass reduction of chemically bonded molds
over time (RMT, 1998). While this testing method, commonly referred to as the "OCMA method,"
provides useful information for the specific binder formulation used by a facility, the data from these
studies are often misapplied and the emission factors derived from this misapplication of the data
understates the true emissions from the mold and core-making process. Example 4-2 provides an example
of developing site-specific emission factors for OCMA method mass-loss data. When properly applied,
emission factors derived from site-specific OCMA method mass-loss data are considered Methodology
Rank 2 for Mold and Core Making.
4-3
-------
Version 1
Draft
Section 4—Mold and Core Making
Example 4-2: Development of Site-Specific Emissions Factor from OCMA Test Data
A facility tested the mass loss from a phenolic urethane no-bake cores during the core-making
and setting process using the OCMA method (RMT, 1998). The amount of binder used for
each core was 180 grams. The mass loss from per core averaged 12 grams for the cores tested.
The aggregate composition of the binder, considering the resin and the coreactant, is as
follows:
Compound
Concentration (wt%)
Formaldehyde
0.3
Phenol
4.0
Xylene
0.2
Cumene
0.5
Naphthalene
1.0
1,2,4-Trimethylbenzene
1.0
Methylene phenylene isocyanate (MDI)
44.0
Biphenyl
0.1
The overall VOC emission factor is easily determined from the total mass-loss data from the
OCMA test.
VOC emission factor = 12 [grams]/180 [grams] = 0.0667 grams VOC/gram binder
To determine the mass loss of specific compounds, it is critical to first account for the
compounds that are not available for volatilization. Binders generally consist of reactive
compounds and "solvent" compounds. The Form R reporting guidance (AFS and CISA, 2007)
provides guidance regarding which compounds are reacted and therefore not available for
volatile loss. For phenolic urethane no-bake, the reactive components are formaldehyde,
phenol, and MDI (see Table 4-2). These compounds make up 48.3% of the aggregate binder.
Since these compounds are assumed to react, only 51.7% of the mass of the binder added is
projected to be available for volatile loss. The fraction of the "solvent" portion of the binder
that is emitted is calculated as follows:
Fraction "Solvent" Emitted = Mass Loss/"Solvcnt" Mass = r12 = 0.129
180 [grams] x 0.517
This emission factor (12.9%) would be used along with the composition data provided above to
determine the mass of each "solvent" compound (i.e., xylene, cumene, naphthalene, 1,2,4-
trimethylbenzene, and biphenyl) that is emitted per mass of binder used. Taking naphthalene
as an example, the site-specific naphthalene emission factor would be:
Naphthalene emission factor = 1 wt% *0.129 = 0.00129 lb/lb binder.
While some compounds in the "solvent" fraction may be more volatile than others, there is
generally no direct method to further differentiate the chemical-specific losses from OMCA
test data. In any case, the key to consideration when applying OMCA test data to specific
compounds in the binder system is to first determine the portion of the formulation that does
not react and is, therefore, available to be emitted.
4-4
-------
Version 1
Draft
Section 4—Mold and Core Making
4.2 Methodology Rank 3 for Mold and Core Making
Except for mold and core making equipment whose emissions are collected and released through a well-
defined stack, the EPA recommends that the emission estimation procedures described in this section be
used to calculate air pollutant emissions from
organic binders (or coating materials) used
when making the molds and cores. The
quantity of each chemical binder
material/component used should be available
from purchase records or direct material
usage meters. Binder component composition
should also be available from the chemical
supplier via material safety data sheets
(MSDS) or other information (see inset).
Together, this information provides a direct
means of determining constituent-specific
usage rates (Methodology Rank 3A).
The emission factors (reported as percent
emitted) in Table 4-2 can then be used to
determine the annual emissions for each
constituent from the mold and core making process using Equation 4-1:
E,=^(QxX^x%em'tteM
1 iLiX-l \^X.X 100% 100% J
Eqn. 4-1
where:
MSDS Data: Composition data are provided typically in
either Section 2 or 3 of an MSDS, depending on the
manufacturer. MSDS compositional data are commonly
reported in the form of a concentration range, which may
be fairly wide. Generally, the midpoint of the range is used
for concentration estimate for each pollutant. Check to see
if the midpoint concentrations add to 100 percent; and
adjust the compounds with widest reported concentration
ranges so the total concentrations sum to 100%. For
example, if a product is made up of two compounds and
the MSDS reports the concentration of Compound A as 60
- 100% and Compound B as 10-20% , use 15% for the
concentration of Compound B and 85% (rather than 80%)
for Compound A. Occasionally, Section 15 of the MSDS
sheet contains more specific compositional data (direct
values rather than ranges). The use of the direct
composition data is preferable to estimating
composition from ranges reported in MSDS sheets.
E; = emissions for compound i, tons/yr
Qx = quantity of binder material ""x" used at the foundry, tons/yr
Cx4 = concentration of compound "i" in binder material ""x". weight percent
%emittedx l = percent of compound "i" in binder material ""x" emitted (see Table 4-2)
Table 4-2. HAP Emitted from Chemical Binder Systems Used for Sand Cores and Molds
(AFS and CISA, 2007)
Binder system
HAP and component in which it is
used
Percent
reacted
Percent emitted
during core and
mold making1
Percent
remaining
in mold or
core
Alkyd oil
Methylene phenylene isocyanate,
coreactant
99.99
0.0012
0.0092
Cobalt, resin
0
0
100
Lead, resin3
0
0
100
Acrylic/Epoxy/SC>2
Cumene hydroperoxide
97
0.32
2.72
Cumene
0
1.5
98.5
Furan hotbox
Formaldehyde, resin
95
5
0
Furan nobake
Phenol, resin
98
0.22
1.82
Formaldehyde, resin
98
2
0
Methanol, resin
0
50
50
Methanol, catalyst
0
50
50
Sulfuric acid, catalyst
100
0
0
Furan/SC>2
Formaldehyde, resin
98
2
0
4-5
-------
Version 1
Draft
Section 4—Mold and Core Making
Percent
Binder system
HAP and component in which it is
used
Percent
reacted
Percent emitted
during core and
mold making1
remaining
in mold or
core
Methanol, resin
0
50
50
Dimethyl phthalate, oxidizer
0
50
50
Methyl ethyl ketone, oxidizer
0
50
50
Furan warmbox
Formaldehyde, resin
95
5
0
Methanol, catalyst
0
100
0
Phenolic baking
Phenol, Part 1
95
0.52
4.52
Formaldehyde, Part 1
95
5
0
Phenolic ester
Formaldehyde, resin
98
2
0
nobake
Phenol, resin
98
0.22
1.82
Phenolic ester
Formaldehyde, resin
98
2
0
coldbox
Phenol, resin
98
0.22
1.82
Glycol ethers, resin
0
50
50
Methanol, co-reactant
0
50
50
Phenolic CO2 cure
Diethylene glycol butyl ether (112-34-5),
resin
0
0.5
99.5
Ethylene glycol monophenyl ether (122-
99-6), resin
0
0.5
99.5
Phenolic hotbox
Formaldehyde, resin
95
5
0
Phenol, resin
95
0.52
4.52
Phenolic nobake
Phenol, resin
98
0.22
1.82
(acid catalyzed)
Formaldehyde, resin
98
2
0
Methanol, resin
0
50
50
Methanol, acid
0
50
50
Sulfuric acid, acid
100
0
0
Phenolic Novolac
Phenol, resin
95
0.52
4.52
flake (hot coating
operations)
Phenolic Novolac
Phenol, Part I
0
20
80
liquid (warm-coating
operations)
Formaldehyde, Part I
95
5
0
Methanol, Part I
0
100
0
Phenolic Novolac
Phenol, resin
99
0.12
0.92
flake (resin-coated
Ammonia, catalyst
50
50
0
sand)
(Assume ammonia = 40% of
hexamethylenetetramine)
Phenolic urethane
Formaldehyde, Part I
98
2
0
nobake
Phenol, Part I
98
0.22
1.82
Xylene, Part I
0
164
OO
Cumene, Part I
0
164
OO
Naphthalene, Part I
0
164
OO
1,2,4-Trimethylbenzene, Part I
0
164
OO
Methylene phenylene isocyanate, Part II
99.99
0.0012
0.0092
Xylene, Part II
0
164
00
Cumene, Part II
0
164
00
Naphthalene, Part II
0
164
00
1,2,4-Trimethylbenzene, Part II
0
164
00
Phenolic urethane
Formaldehyde, Part I
98
2
0
coldbox
Phenol, Part I
98
0.22
1.82
4-6
-------
Version 1
Draft
Section 4—Mold and Core Making
Binder system
HAP and component in which it is
used
Percent
reacted
Percent emitted
during core and
mold making1
Percent
remaining
in mold or
core
Xylene, Part I
0
94
914
Naphthalene , Part I
0
94
914
Cumene, Part I
0
94
914
1,2,4-Trimethylbenzene, Part I
0
94
914
Methylene phenylene isocyanate, Part II
99.99
0.0012
0.0092
Xylene, Part II
0
94
914
Naphthalene, Part II
0
94
914
Cumene, Part II
0
94
914
Biphenyl, Part II
0
94
914
Triethyl amine or diethyl amine, catalyst
gas
0
1005
0
Urea formaldehyde
Formaldehyde, Part I
98
2
0
1 Percent emitted up to the time that metal is poured.
2 Revised from Form R Reporting Guidance provided by AFS and CISA (2007); assumes 10% of unreacted
chemical is released during mold/core making and storage. Note that of the Phenolic Novolac liquid system, the
Form R Reporting Guidance suggests 20% of the unreacted phenol is emitted.
3 Compound not listed in 4th Edition (AFS and CISA, 2007) but listed in previous versions (AFS and CISA, 1998)
4 Revised from Form R Reporting Guidance provided by AFS and CISA (2007); the emission factors in the Form R
Reporting Guidance of 5.85% and 3.25% are based on weight loss measurements of the molds upon storage
(RMT, 1998). However, considering the components reacted that were not available for emissions, 16% and 9% of
the unreacted components for the nobake and coldbox systems, respectively, would have to be released to provide
the 5.85% and 3.25% weight losses observed.
5 Assumed 100% of catalyst used is emitted. If a sulfuric acid wet scrubber with pH less than 4 is used, a control
efficiency of 99% can be assumed.
If direct chemical composition data are available for the specific binder formulation used, that
information should be used in Equation 4-1 (Methodology Rank 3A). If specific chemical usage rates are
available, but chemical-specific composition data are not available, the default composition data in
Table 4-3 can be used in Equation 4-1 to estimate the emissions from the system (Methodology
Rank 3B). Tables 4-2 and 4-3 provide data for specific HAP, which will typically be only a small portion
of the total VOC emitted. As illustrated in Example 4-2, OMCA test data can be used to develop an
overall VOC emission factor. If OMCA test data are not available, the "solvent" faction of the binder
formulation (i.e., the portion of the binder not chemically reacted) can be determined and the HAP portion
of the "solvent" fraction calculated. Using the data from Example 4-2, the "solvent" fraction in this
example was 51.7 wt% and the HAP content of the "solvent" fraction was 5.4 wt% (2.8%/0.517). In this
example, the HAP emissions are expected to be 5.4% of the total VOC emissions. The VOC emissions
can be projected from the HAP emissions by dividing the "solvent" fraction's cumulative HAP emissions
associated by 0.054 (i.e, 5.4%).
Table 4-3. Default Content of Sand Binder System Components1'2
Binder system
Component
HAP present
Amount of pollutant in
component, percent
Range
Typical
Alkyd oil
coreactant
MDI
No data
00
o
c:
Acrylic/epoxy/SC>2
Resin
Cumene
5. minimum4
5.
4-7
-------
Version 1
Draft
Section 4—Mold and Core Making
Amount of pollutant in
component, percent
Binder system
Component
HAP present
Range
Typical
Furan hot box
Resin
Formaldehyde
2.-5.
3.
Furan nobake
Resin
Phenol
Formaldehyde
Methanol
^ v- ^
1 1 1
0 0 c\i
1.
0.1
3.
Catalyst
Methanol
0
CO
1
0
CM
27.
Furan/SC>2
Resin
Formaldehyde
Methanol
1.-4.
1.-3.
2.
2.
Oxidizer
Dimethyl phthalate
Methyl ethyl ketone
O
LO CM
I I
O O
45.
2.
Furan warm box
Resin
Formaldehyde
0.-1.
0.5
Catalyst
Methanol
45.-55.
50.
Phenolic baking
Resin
Phenol
Formaldehyde
3.-14.
0.-2.
8.
1.
Phenolic ester
nobake
Resin
Phenol
Formaldehyde
2.-8.
0.-2.
4.
0.5
Phenolic ester
cold box
Resin
Phenol
Formaldehyde
Glycol ethers
2.-8.
0.-2.
No data
4.
0.5
0.1 3
Co-reactant
Methanol
No data
27. 3
Phenolic CO2 cure
Resin
Diethylene glycol butyl ether
Ethylene glycol monophenyl ether
No data
No data
1.3
1. 3
Phenolic hot box
Resin
Phenol
Formaldehyde
2.-8.
1.-4.
5.
2.
Phenolic nobake
(acid catalyzed)
Resin
Phenol
Formaldehyde
Methanol
8.-14.
0.-2.
2.-4.
12.
0.5
3.
Catalyst
Methanol
O
CO
I
0
CM
27.
Phenolic Novolac
flake (hot coating
operations)
Resin
Phenol
1.5-8.04
5.5 5
Phenolic Novolac
liquid (warm-coating
operations)
Resin
Phenol
Formaldehyde
Methanol
1.-4.
0.-3.
0.-15.
2.
0.5
5.
Phenolic Novolac
flake (resin-coated
sand)
Resin
Catalyst
Phenol
Ammonia
1.5-8.0 4
No data
5.5 b
40.®
4-8
-------
Version 1
Draft
Section 4—Mold and Core Making
Amount of pollutant in
component, percent
Binder system
Component
HAP present
Range
Typical
Phenolic urethane
Resin
Phenol
3.-8.
6.
nobake and cold box
Formaldehyde
0.-1.
0.1
Naphthalene
0.-2.
1.
Cumene
0.-2.
0.5
Xylene
0.-1.
0.1
Coreactant
Naphthalene
0.-3.
1.
Cumene
0.-1.
0.1
Xylene
0.-1.
0.1
Biphenyl (only in cold box system)
0.-1.
0.1
MDI
60.-95.
80.
Urea formaldehyde
Resin
Formaldehyde
1.-4.
1.
1 Source: Stone, 1999, and Jonathan A. Stone, Delta Resins and Refractories, Delta-HA (private communication to
J. H. Maysilles, U.S. EPA., November 15, 1999) except where noted.
2 Only HAP that could be emitted because of incomplete reaction or nonreaction are listed.
3 No data. Concentrations estimated based on other binder systems.
4 Information supplied by Joe Fox, Ashland Chemical, Inc. Private communication to J. H. Maysilles, U.S. EPA,
August 16, 2000.
5 Information is based on Material Safety Data Sheets from foundries that use this system.
6 Assume catalyst is 99+% of hexamethylenetetramine and 40% of hexamethylenetetramine converts to ammonia.
4-9
-------
Version 1
Draft
Section 4—Mold and Core Making
Example 4-3: Estimating Mold and Core Making Emissions using Methodology Rank 3
A facility uses a phenolic urethane cold box (PUCB) system. During 2011, the facility used
400,000 lbs of Part I and 340,000 lbs of Part II and 70,000 lbs of TEA gas (> 99.9% purity).
Calculate the HAP emissions from the PUCB mold and core making system given that the
composition of the Part 1 and Part II binder chemicals are as follows (based on MSDS sheets):
Compound
Concentration (wt%)
Part I
Part II
Formaldehyde
0.5
—
Phenol
6
—
Xylene
0.2
0.1
Cumene
0.5
—
Naphthalene
1.0
2.0
1,2,4-Trimethylbenzene
1.0
—
Methylene phenylene isocyanate (MDI)
80
Biphenyl
0.2
Apply Equation 4-1 for each binder component and compound of interest using the direct
composition data (Methodology Rank 3A for Mold and Core Making). The calculation for
naphthalene follows:
Ej = (400,000 lbs x^-x-^-)+ f340,000 lbs x^-x-^-) = 972 lbs in 2011
1 V 100% 100%/ V 100% 100%/
Converting to tons per year, Enaphthalene = 972/2,000 = 0.486 tons/yr
Similar calculations are used for the other compounds to yield the following results:
Eformaldehyde 0.02 tOns/yr
Ephenoi= 0.024 tons/yr
Exyiene= 0.051 tons/yr
Etrimethylbenzene 0.18 tOns/yr
Emdi= 0.00136 tons/yr
Ebiphenyl = 0.0306 tons/yr
TEA gas catalyst emissions from a cold box system should also be calculated. If the emissions
from the catalyst gas are not controlled, then the emissions from the PUCB mold and core
making system is:
Etea = 70,000 lbs/(2,000 lbs/ton) = 35 tons/yr
If the TEA emissions are controlled with an sulfuric acid wet scrubber (pH<4), a control
efficiency of 99% can be assumed, so the TEA emissions would be 1% of the uncontrolled
case, or 0.35 tons/yr.
4-10
-------
Version 1
Draft
Section 4—Mold and Core Making
4.3 Methodology Rank 4 for Mold and Core Making
4.3.1 Pollutant Emissions from Chemical Binder Systems
It is anticipated that most facilities will not have direct routine measurement data or site-specific source
test data for their mold and core making operations and will have to use Methodology Rank 4. If specific
binder chemical usage rates are not known, but the sand usage rates are known for each type of
chemically bonded mold and core making process, then the factors in Table 4-4 can be used to estimate
emissions (Methodology Rank 4). The chemical content provided in Table 4-4 considers the average
composition of the compound in the combined binder system when multiple components are used.
Equation 4-2 is used to estimate emissions based on the information in Table 4-4.
g yN / S x X E111F X| \
1 ~~ ^x=l ^ 2,000 )
Eqn. 4-2
where:
E; = emissions for compound i, tons/yr)
Sx = quantity of sand used with binder system ""x". tons/yr
EmFxJ = emission factor for compound "i" for binder system ""x". from Table 4-4, lbs/ton
2,000 = Conversion factor, lbs/ton.
The emission factors in Table 4-4 were derived from typical binder usage rates by binder component
(from U.S. EPA, 1998), the default composition data in Table 4-3, and the percent emitted values in
Table 4-2. For more information about how the factors were developed, see Appendix C.
Table 4-4. Default Pollutant Emission Factors for Sand Binders
Binder system
Binder-to-Sand
Concentration
(lbs/ton sand)
Pollutant
Emission Factor
(lb/ton sand)
Alkyd oil
30
MDI
1.1x10"4
Acrylic/epoxy/SC>2
34
Cumene
2.6x10"2
Furan hot box
40
Formaldehyde
6.0x10"2
Furan nobake
24
Phenol
Formaldehyde
Methanol
3.4x10"4
3.4x10"4
1.22
Furan/SC>2
30
Formaldehyde
Methanol
Dimethyl phthalate
Methyl ethyl ketone
6.6x10"3
0.165
3.04
0.135
Furan warm box
32
Formaldehyde
Methanol
6.4x10"3
3.2
Phenolic baking
30
Phenol
Formaldehyde
1.2*10"2
1.5x10"2
Phenolic ester nobake
33
Phenol
Formaldehyde
2.6x10 3
3.3x10"3
4-11
-------
Version 1
Draft
Section 4—Mold and Core Making
Binder system
Binder-to-Sand
Concentration
(lbs/ton sand)
Pollutant
Emission Factor
(lb/ton sand)
Phenolic ester cold box
33
Phenol
Formaldehyde
Glycol ethers
Methanol
2.6x10 3
3.2x10 3
1.6x10"2
0.405
Phenolic CO2 cure
30
Diethylene glycol butyl ether
Ethylene glycol monophenyl ether
1.5x10"3
1.5x10"3
Phenolic hot box
30
Phenol
Formaldehyde
7.5x10"3
3.0x10"2
Phenolic nobake (acid
catalyzed)
27
Phenol
Formaldehyde
Methanol
4.4x10"3
1.8x10"3
1.44
Phenolic Novolac
flake (hot coating operations)
50
Phenol
1.4x10"2
Phenolic Novolac
liquid (warm-coating
operations)
50
Phenol
Formaldehyde
Methanol
0.20
1.3x10"2
2.5
Phenolic Novolac
flake (resin-coated sand)
50
Phenol
Ammonia
2.8x10 3
2.0
Phenolic urethane nobake
25
Phenol
Formaldehyde
Naphthalene
Cumene
Xylene
MDI
1.7x10"3
2.8x10"4
4.0x10"2
1.3x10"2
4.0x10"3
9.0x10"5
Phenolic urethane cold box
30
Phenol
Formaldehyde
Naphthalene
Cumene
Xylene
Biphenyl
MDI
2.0x10"3
3.3x10"4
2.7x10"2
8.6x10 3
2.7x10"3
1.2x10"3
1.1x10"4
Urea formaldehyde
30
Formaldehyde
6.0x10"3
4-12
-------
Version 1
Draft
Section 4—Mold and Core Making
Example 4-4: Estimating Binder System Emissions using Methodology Rank 4
A facility uses furan nobake system to make cores for its castings. The facility used 200,000
tons of sand in its core process in 2011. Calculate the 2011 annual HAP emissions from the
core making system.
Without specific chemical usage rates, the default emission factors in Table 4-4 would be used
(Methodology Rank 4). The relevant factors from Table 4-4 are as follows:
Emission Factor
Binder system
Pollutant
(lb/ton sand)
Furan nobake
Phenol
3.4x10-4
Formaldehyde
3.4x10-4
Methanol
1.22
Apply Equation 4-2 for the single binder system used at the facility. The calculation for phenol
follows:
Ephenol = (200,000 tons X = 0.034 tons/yr
Similar calculations are used for the other compounds to yield the following results:
Eformaldehyde 0.034 tOns/yr
Emethanoi= 122 tons/yr.
4.3.2 PM Emissions from Sand Handling Operations Associated with Mold and Core
Making
Table 12.3-1 of Section 12.13 of AP-42 (U.S. EPA, 1995) reports an emission factor for sand
grinding/handling in mold and core making at steel foundries of 0.54 lb PMio-FIL/ton of sand processed.
Table 12.10-7 of Section 12.10 of AP-42 (U.S. EPA, 2003) contains an emission factor of 3.6 lbs PM
(TSP)/ton of sand for uncontrolled sand handling systems and provides factors of 0.046 and 0.20 lb/ton of
sand for sand handling systems controlled by a wet scrubber and baghouse, respectively. It is unusual that
a baghouse would be less efficient than a wet scrubber, so these differences may be in PM loading rates to
the control devices. As the PM emissions from sand handling systems are expected to be similar between
iron and steel foundries, the differences in these factors were initially considered to be due to the fact that
the steel foundry emission factor was specific to mold and core-making operations and the iron foundry
emission factor considered both sand reclamation and mold and core-making operations. Baghouse catch
data from mold sand mullers and automated (Disamatic) molding lines suggested that the iron foundry
emission factor was more appropriate than the steel foundry factor when considering only mold sand
operations. Baghouse catch data for core sand mullers suggested even higher emission factors may be
appropriate for core sand handling. Nonetheless, a single set of emission factors is presented in Table 4-5
for use for sand handling operations associated with mold and core making for both iron and steel
foundries. The default captured but uncontrolled PM emission factor for mold and core making is based
on the default factor reported for iron foundries. Use of site-specific baghouse catch data to determine the
captured but uncontrolled PM emission factor for the specific mold and core making operations would be
the preferred to use of the Table 4-5 defaults. The default emission factors can be used in Equation 4-2 to
determine the PM mass emissions. For controlled sand handling operations, it is recommended that
sources use the ducted uncontrolled emission factors and adjust the emissions based on the expected
control efficiencies of the control device as provided in Table 3-4 of Section 3.1, Melting Furnaces, of
this Protocol document.
4-13
-------
Version 1
Draft
Section 4—Mold and Core Making
As with the PM emission factor for melting furnaces in Table 3-3, the PM emission factors in Table 4-5
for PMio-FIL include emissions of PM2 5-FIL. As well, the PM coarse -FIL emissions factor is the emissions
factor for PMio-FIL minus the emissions factor for PM2 5-FIL. Therefore, to calculate PM10-FIL
emissions, the emissions for PM coarse -FIL is added to the emissions for PM2 5-FIL.
Table 4-5. Default PM Emission Factors for Sand Handling Operations Associated
with Mold and Core Making
Emission
Suggested SCC
Suggested SCC for
Factor
for Iron Foundries
Steel Foundries
Emission Source
Pollutant
(lb/ton sand)
30400350
30400716
Sand grinding/handling,
PM-FIL
3.6 1
captured and ducted
_l
LL
0
CL
3.0 1
uncontrolled to atmosphere
PM2.5-FIL
2.6 1
30400350
30400716
Sand grinding/handling,
PM-FIL
1.8 2
uncaptured fugitive dust to
_l
LL
0
CL
1.5 2
atmosphere
PM2.5-FIL
1.3 2
1 Used the default factor for iron foundries, which was mid-range of the factors reported for steel foundries, for PM-
FIL. Used the size distribution ratios for steel foundries casting shakeout exhausted, prior to controls, to determine
PM10-FIL and PM2.5-FIL values.
2 Assumes 50% of ducted emissions would be released as fugitive emissions that escape to the atmosphere for
uncaptured units.
4-14
-------
Version 1
Draft
Section 4—Mold and Core Making
Example 4-5: Estimating PM Emissions from Green Sand Mold Making
A facility processes 500,000 tons of sand per year for their green sand molds. Calculate the PM
emissions for the green sand molding operations if the system is uncaptured/uncontrolled.
What are the emissions if the facility controls these emissions using a fabric filter?
For the existing uncaptured/uncontrolled system, the default emission factors for uncaptured
fugitive emissions from Table 4-5 would be used (Methodology Rank 4). The relevant factors
from Table 4-5 are as follows:
Emission Factor
Emission Source
Pollutant
(lb/ton sand)
Sand grinding/handling,
PM-FIL
1.8
uncaptured fugitive dust to
_l
LL
o
CL
1.5
atmosphere
PM2.5-FIL
1.3
Note that PM-CON for sand handling systems is 0, so PM-PRI = PM-FIL.
Apply Equation 4-2 for the green sand system used at the facility. The calculation for PM-FIL
follows:
Epm-fil = Epm-pri = (500,000 tons X 2 oo(J = 450 tons/yr
Similar calculations are used for PM10-FIL and PM2 5-FIL to yield:
EPmio-fil= 375 tons/yr
Epm2.5-fil = 325 tons/yr.
If the facility installs a capture system and fabric filter control system, the PM at the inlet to the
control device will be twice the uncaptured emissions (based on the emission factors in Table
4-5). Thus, the inlet PM rates are 650 tons/yr and 750 tons/yr for PM2 5-FIL and PM10-FIL,
respectively. Consequently, there are 100 tons/yr of PM between 2.5 and 10 nm.
The default control efficiency for a fabric filter control device is 99% for PM <2.5 pim and
99.5% for PM between 2.5 and 10 pim. It is generally assumed that the control efficiencies for
particles greater than 10 pim in diameter is 100%, so that PM-PRI = PM-FIL = PM10-FIL.
First, calculate the PM2 5-FIL emissions as follows:
Epm2 5-FiL = 650 x (1-0.99) = 6.5 tons/yr
Next, calculate the PM emissions between 2.5 and 10 nm ("coarse" PM) as follows:
EpMj'coarse" = 100 x (1-0.995) = 0.5 tons/yr
The PMio-FIL emissions equal the sum of these emissions, i.e., 7.0 tons/yr.
4-15
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
5. Pouring, Cooling, and Shakeout
The processes discussed in this section include pouring of molten metal into molds, cooling of the casts in
the molds, and shakeout of the casts from the molds. Emissions from these processes are commonly
considered together, and the collective process is referred to as pouring, cooling, and shakeout (PCS).
Emissions from pouring are primarily PM and metal fumes released as the molten metal is poured into the
molds. In sand casting operations, organic emissions are typically released as the molten metal contacts
organics or coke used in sand mold and cores. Even in green sand systems, coke is commonly added to
ensure a reducing atmosphere and prevent oxidation of the metal as it cools. As the mold cools, pyrolysis
products continue to be generated and released as the heat from the metal penetrates further into the sand
mold. As it is difficult to distinguish organics generated from pouring from those generated during
cooling, organic emissions are commonly estimated for the combined pouring/cooling process. The
emission factors presented in this section were generally developed for automated pouring, cooling lines
where the molds are expected to light off automatically or manually.
When the castings are sufficiently cooled (solidified), they are removed from the molds using mechanical
tools or vibrating grids or conveyors to knock off or shake loose the sand from the casting. As such, this
process is commonly referred to as "knock-out" or "shakeout." This process will be referred to simply as
shakeout in the Protocol document regardless of the means used to remove/loosen the sand. Emissions
from shakeout are PM associated with the loosened sand and additional organic vapor emissions. The
organic vapors include pyrolysis vapors generated during pouring/cooling that did not diffuse out of the
mold during pouring/cooling, but are released as the mold is broken apart. Organic vapors may also be
generated as organics (or coke) contained in the outer portions of the mold come in contact with the hot
casting and pyrolyze during the shakeout process.
There are five typical casting types for iron and steel foundries: sand (includes green sand molds and
chemically bonded sand molds), centrifugal, permanent, investment, and expendable pattern casting. The
primary difference in the emissions from the different casting systems is the amount of sand used in the
casting process. Centrifugal, permanent, and investment casting operations use little to no sand
(depending on whether cores are needed), so these systems do not have large PM emissions associated
with sand system shakeout, nor do they have significant organic emissions from the pyrolysis of binders
and additives commonly used in sand systems.
Expendable pattern castings use sand molds compressed about a polystyrene pattern. When molten metal
is poured into the mold, the molten metal volatilizes the polystyrene pattern and replaces it, forming a
casting of the same shape as the pattern. Vapors generated in the process escape through the sand that
surrounds first the pattern and then the casting. Castings are removed from the loose sand and then
finished in much the same manner as those made by sand casting. Emissions from expendable pattern
castings are similar to green sand systems, but have specific organic emissions generated from the
pyrolysis of the polystyrene patterns.
The emission estimation methods for PCS operations are presented in Table 5-1. Emissions from some
PCS operations may be captured and vented through well-defined stacks, especially for shakeout. For
emissions that are captured and emitted through well-defined stacks, direct measurement methods may be
used. Direct measurements may be continuous or routinely frequent (daily or weekly) in nature so that
the measurement data can be used directly to determine emissions (Methodology Rank 1 or 2). For many
pollutants, a one-time or annual source test may be performed, but the emissions are not routinely
measured. For these pollutants, the measured data can be used to determine an emission factor based on
process throughput (quantity of metal poured or sand used), and the annual emissions can be determined
using the annual material usage multiplied by the site-specific emission factor (Methodology Rank 3).
5-1
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeont
For some PCS operations, no direct measurement data will be available. In such cases, emissions are
estimated based on the type of casting system employed, the molten metal throughput, and default
emission factors provided in this section (Methodology Rank 4).
Table 5-1. Summary of Typical Hierarchy of Pouring, Cooling, and Shakeout Emission Estimates
Rank
Methodology
Description
Application
Data Requirements
1
Direct routine
concentration and flow
measurement
Pouring, cooling, and
shakeout that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate
2
Direct routine
concentration
measurement with
engineering flow
estimates
Pouring, cooling, and
shakeout that are well-
captured and have well-
defined stack emissions
Constituent concentration and operating
parameters needed for estimating flow rate
(e.g., fan curve and amp usage)
3
Site-specific emission
factor from one-time or
periodic emissions
source test data
Pouring, cooling, and
shakeout that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate and molten metal throughput during
source test; annual molten metal throughput
4
Default emission factors
All sources
Material usage rates
5.1 Methodology Ranks 1 and 2 for PCS Operations
Emissions from vented PCS operations can be directly measured at the stack or outlet of the control
device using the direct measurement methods as described in Section 3, Melting Operations, of this
Foundries Emissions Protocol document. If CEMS are available for both a pollutant concentration and
exhaust gas flow rate, the CEMS data should be used to determine the pollutant emissions as
Methodology Rank 1 for PCS (using Equation 3-1 from Section 3.1,Melting Furnaces). If a CEMS is
used to determine a pollutant's concentration, but direct-flow measurement is not available, flow rates can
often be determined using engineering estimates, such as fan amperage-to-flow correlations, to determine
the pollutant emissions as Methodology Rank 2 for PCS. Example 5-1 demonstrates how to estimate the
flow measurement using fan curves.
5-2
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
Example 5-1: Estimation of Flow Measurement Using Fan Curves
A pouring/cooling ventilation line is equipped with a VOC CEMS with a concentration output
in ppmv as methane as sampled (i.e., wet basis). The ventilation line does not have a flow
monitor, but the ventilation rate is driven by a single speed Model 420-20 fan. The fan
manufacturer's fan chart is provided below. If the VOC monitor measures 70 ppmv and the
ventilation system pressure drop averages 0.375 inch of water gauge, what is hourly emission
rate for the system?
MODEL
420-5
420-7
420-10
420-15
HP
1/2
3/4
1 -1/2
RPM
Sonesj BHP
220
245
280
305
350
STATIC PRESSURE IN INCHES OFW.G.
9353
6.5 I 0.31
10416
7.2 I 0.43
11904
8.3 0.64
12966
9.4 0.82
14879
11.9 1.24
16367
14.4 1.65
17643
0.125
Sonesl BHP~
7846
Soresl BHP
6.0 I 0.37
9104
6.7 I 0.50
10785
7.7 0.73
11959
8.8 0.94
14037
11.2 1.38
15629
13.5 1.81
16982
0.25
5075
5.5 I 0.35
7304
6.1 I 0.53
9418
7.1 0.78
10781
8.0 1.00
13050
10.1 1.46
14731
12.4 1.91
16148
Sol
Sonesl BHP
7419
6.6 0.77
9272
7.3 1.02
11925
9.4 I 1.52
03790
15103
RPM
required
Motor horsepower
size required
Use Equation 3-1 to determine the emission rate.
N
0.50
Sonesl BHP
Utislnblt
points are rial prm
10562
8.6 1.54
12701
10.7 2.05
14363
0.75
performance
9246
8.5 1.92
12028
1.00
Sonesl B1
7cT
BHP at selected performance
CFM selected
Sories at selected performance
Ej = Y ((V)n X [1 - (f„2o)n]
n=l ^
X
(Ci)n MWj
X ¦
100% MVC
x - xi xK
Given the fan model size and system pressure, the flow rate for the ventilation system is
13,791 acfm (see above). The cumulative volume for the 1 hour time period (60 minutes) is:
(V)n = 13,791 acfm x 60 min = 827,460 acf
As the concentration is measured on the same basis as the flow rate, the moisture correction
factor is not needed:[l — (fH2o)n] = 1 The system operates essentially at atmospheric
conditions (note: 1 inch water gauge = 0.0025 atm). Although not specified, the temperature is
likely to be near room temperature. As such, the pressure (— 1 and temperature
V P o /
(-r) corrections can be assumed to be 1 without introducing significant error.
VT o /
70
(Ci)n =
X 100% = 0.007%
1,000,000
Use the molecular weight of methane (16 kg/kg-mol), since the concentration is measured as
methane.
K — j = 2.2046 (to output the mass in lbs rather than tons)
0.007%
16
Lkg.
Eyoc = 827,460 x 1 x x ¦
vuij 100% 849.5
Eyoc = 2.40 pounds per hour (lb/hr)
x 1 x 1 x 2.2046
5-3
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
5.2 Methodology Rank 3 for PCS Operations
When CEMS data are not available, but data from a single source test or a limited number of source tests
are available, a site-specific emission factor can be developed and used to assess the emissions from PCS
operations for the pollutants tested, similar to the methods described in Section 3, Melting Operations
(using Equation 3-4/ Emission of metal HAP and PM from pouring and cooling operations are generally
correlated with quantity of metal poured. Similarly, metal HAP emissions from shakeout are expected to
be correlated with quantity of metal poured. PM emissions from shakeout may be correlated with the
quantity of metal poured or the quantity of sand used. Example 5-2 demonstrates how to calculate
emission factors for PM emissions from shakeout using the different activity data.
For PCS operations that are controlled with a baghouse, site-specific emission factors can be developed
based on the baghouse catch data (the measurement of the mass of PM collected over a period of time).
To determine the captured, uncontrolled PM emissions from baghouse catch data, see Example 3-4,
above. Many PCS operations may have canopy hoods or similar capture systems, so the captured,
uncontrolled emissions estimated from baghouse catch data should be corrected for the capture efficiency
of the ventilation system to account for uncaptured PM emissions.
Baghouse dust analyses can also be used to determine the metal HAP concentration of the emitted PM,
particularly for shakeout. Baghouse dust analysis is expected to be less accurate for determining metal
HAP content associated with pouring emissions as the metal fumes emitted from the baghouse may have
a different composition than the metal particles collected in the baghouse. However, the baghouse dust
composition data is preferred to the default metal HAP compositions provided in Methodology Rank 4 for
PCS Operations.
5-4
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
Example 5-2: Development of Site-Specific Emissions Factor for PM Emissions from
Shakeout
During its most recent source test, three test runs were conducted using Method 5 to determine the PM
emission rate from shakeout operations. The PM emissions rate measured during the source tests were
20.2, 25.1, and 17.6 lbs/hr for tests 1, 2, and 3, respectively. The processing rates during the three runs
were measured for metal poured/hour and sand used/hr: 16, 17, and 15 tons of metal poured/hr; 64, 74,
and 59 tons of sand used/hr, respectively.
Calculate the appropriate emission factor for each individual run for metal poured/hour and sand
used/hr, and then average the emission factors for each processing rate as follows:
For processing rate of metal poured/hr
Run 1: Emissions/metal throughput = 20.2 [lbs/hr] 16 [tons/hr] = 1.26 lbs/ton metal
Run 2: Emissions/metal throughput = 25.1 [lbs/hr] 17 [tons/hr] = 1.48 lbs/ton metal
Run 3: Emissions/metal throughput =17.6 [lbs/hr] ^ 15 [tons/hr] = 1.17 lbs/ton metal
Average: Emissions/metal throughput = (1.26 + 1.48 + 1.17)/3 = 1.30 lbs/ton metal
For processing rate of sand used/hr
Run 1: Emissions/sand use = 20.2 [lbs/hr] 64 [tons/hr] = 0.316 lbs/ton sand
Run 2: Emissions/sand use = 25.1 [lbs/hr] 74 [tons/hr] = 0.339 lbs/ton sand
Run 3: Emissions/sand use = 17.6 [lbs/hr] 59 [tons/hr] = 0.298 lbs/ton sand
Average: Emissions/sand use = (0.316 + 0.339 + 0.298)/3 = 0.318 lbs/ton sand
There are a variety of ways to determine which emission factor is most appropriate. One
method is to compare the range of the test runs compared to the three-run average. For the
metal throughput-based emissions factors, the highest single run emission factor is 14%
[100%x(1.48-1.30)/1.30] higher than the average and the lowest single run emission factor is
10% lower than the average. For the sand use-based emissions factors, the highest single run
emission factor is 6.6% [100%x(0.339-0.318)/0.318] higher than the average, and the lowest
single run emission factor is 6.3% lower than the average. The smaller range for the sand use
emission factors (as a percentage of the average) suggests that normalizing the emissions by
sand use accounts for more of the differences in the observed emissions than does throughput.
Consequently, the sand use emission factor would be preferred to the metal throughput based
emission factor in this example.
For organic emissions, the "activity" data may be the quantity of metal poured per type of mold system
used. The annual emissions can be estimated using similar methods, as described in Section 3, Melting
Operations (using Equation 3-5). Example 5-3 provides an example of developing a site-specific
emission factor for organic emissions from PCS operations and annual emissions of organics from PCS
operations.
5-5
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
Example 5-3: Development of Site-Specific Emissions Factor
A series of source tests were performed to determine the non-methane, non-ethane organic
carbon (NMNEOC) emissions from PCS operations while producing two castings using: green
sand molds with no cores and green sand molds with phenolic urethane cold box cores. The
average NMNEOC emissions rate for the test runs performed with the green sand only system
was 12 lbs/hr, and the average NMNEOC emissions rate for the test runs performed with the
green sand and chemically bonded cores was 35 lbs/hr. The average metal pouring rates for
both tests were 15 tons/hr. The facility pours 30,000 tons of gray iron per year, and 30 percent
of their castings (by weight) do not require cores. Using the site-specific data, what are the
VOC emissions from the PCS operations at this facility?
First, NMNEOC measurements are an excellent proxy for total VOC. As such, the NMNEOC
measurements can be used directly to determine the site-specific VOC emission factors for
green sand-only castings and for cored castings. The site-specific emission factor is
determined using Equation 3-4
EmF = emissions rate / activity rate
For the green sand only system:
EmFvoc.Gs = 12 lbs/hr / 15 tons/hr = 0.8 lbs/ton
For the green sand system with cores:
EmFvoc.Gs&cores = 35 lbs/hr / 15 tons/hr = 2.33 lbs/ton
To calculate the annual emissions, first determine the total quantity of metal poured for each
type of mold system. Based on the information provided, 30% of 30,000 tons/yr or
9,000 tons/yr of castings are produced using the green sand only molds. Therefore, 21,000
(30,000 - 9,000) tons/yr of castings are poured using the green sand molds with chemically
bonded cores. The total VOC emission from the PCS operations are then determined using
Equation 3-5 as follows:
EvOC.PCS = EmFvoe.GS X QgS + EmFyoe.GS&Cores X QoS&Cores
Evoc.pcs = 0.8 x 9,000 + 2.33 x 21,000 = 56,130 lbs/yr
Evoc,pcs = 56,130/2000 = 28.1 tons/yr
5.3 Methodology Rank 4 for PCS Operations
5.3.1 Organic Pollutant Emission from PCS Operations
Different foundries can have significantly different emissions from PCS depending on the type of casting
system employed. Even within sand casters, emissions are dependent on whether the molds are made
from green sand or chemically bonded mold sand and on the relative amount of cores needed for the
specific castings. The Casting Emission Reduction Program (CERP) has also demonstrated that emissions
of benzene (typically the most prevalent HAP emitted from PCS operations) from green sand molds are
strongly correlated with percent loss on ignition (percent LOI), which relates to the concentration of
seacoal in the mold sand, the mass of metal cast (at constant mass to surface area ratio), and to the surface
area of the casting (at constant cast weight) (Technikon, 2006). As PCS emissions are primarily caused
by the pyrolysis of organics contained within the mold and/or cores, the larger the contact area and the
more organics in the sand, the greater the emissions.
5-6
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeont
Technikon (2006) provides a series of total HAP emission factors for various types of molds and binder
systems. These factors, while helpful, do not provide speciation needed for a proper HAP emission
inventory. For the most part, the emissions from green sand systems with cores were essentially
equivalent to the emission of green sand systems alone plus emissions from cored systems with no
seacoal in the mold sand. Based on the data from the various CERP studies, the emission factors and
concentration profiles provided in Tables 5-2 and 5-3, respectively, were developed. See Appendix D
for more information about the development of the factors. Using these factors, Equation 5-1 is
recommended for estimating organic emissions from PCS operations.
Ei=QMs[(MSV0C x CiMs) x GSCF + (CS
voc x Q,cs
)]/2,000
Eqn. 5-1
where:
E; = Emission of pollutant "\" (tons/yr).
Qms = Quantity of metal poured into a given type of mold system, tons/yr.
MS voc = VOC emission factor for the mold system used, lb VOC/ton metal poured (from
Table 5-2).
CSvoc = VOC emission factor for the cores used, lb VOC/ton metal poured (from Table 5-2).
Cj ms = Concentration profile of pollutant ""i" in emissions from mold system used, lb
pollutant/lb VOC (from Table 5-3).
C;,cs = Concentration profile of pollutant ""i" in emissions for core system used, lb pollutant/lb
VOC (from Table 5-3).
GSCF = Green sand correction factor, unitless; for green sand systems, GSCF = percent
LOI/5.1 percent; for all other types of mold systems, GSCF = 1.
%LOI = Percent of green sand lost on ignition, weight percent using ASTM D7348 or similar
methods .
2,000 = Conversion factor, lbs/ton.
For foundries that operate several different types of mold systems, Equation 5-1 should be applied
separately for each type of mold system. The total PCS emissions for the facility would then be the sum
of the emissions for each type of mold system. The application of Equation 5-1 is illustrated in
Example 5-4.
Table 5-2. VOC Emission Factors by Casting Type for PCS Operations3'15
Type of Mold/Core System
Factor Designation
VOC Emission Factor,
(lb/ton metal poured)
Green Sand
MS voc
1.9a
Phenolic Urethane Bonded Mold Sand
MS voc
7.4b
Other Chemically Bonded Mold Sand
MS voc
4.0b
Chemically Bonded Cores (all)
CSvoc
1.6a
Mold systems without cores
CSvoc
0
Expendable Pattern Casting (Lost Foam)
MS voc
4.8C
Permanent, Centrifugal, or Investment Casting
MS voc
0.12d
a Developed from Casting Emission Reduction Program (CERP) baseline testing (CERP 1999b, 2000; Technikon
2000, 2001a, 2003).
b Developed from CERP testing of chemically bonded mold systems (Technikon, 2001b; 2001d; 2001f; 2001g;
2004).
c Based on data reported by Twarog (1991).
d Developed from CERP "background baseline" (CERP, 1999b).
5-7
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeont
Table 5-3. Default HAP Composition Profiles for PCS Operations
HAP Compound
Concentration Ratio, lb HAP per lb VOC
Green
Sand
Molds3
Phenolic
Urethane
Bonded
Molds"
Other
Chemically
Bonded
Moldsb
Cores3
Expendable
Pattern
Casting0
Permanent,
Centrifugal,
or
Investment11
Acetaldehyde
0.005
0.00075
0.018
0.0025
0
0.07
Aniline
0.0075
0.0013
0
0.035
0
0
Benzene
0.065
0.028
0.14
0.073
0.07
0.05
Cresols (total)
0.0015
0.04
0.013
0.01
0
0
N,N-Dimethylaniline
0.0025
0
0
0.0075
0
0
Ethylbenzene
0.005
0.0005
0.0005
0.001
0
0.005
Formaldehyde
0.00075
0.0025
0.015
0.0005
0
0.013
n-Hexane
0.01
0
0
0.002
0
0
Naphthalene
0.0075
0.0018
0.0025
0.0075
0.0025
0.01
Non-Naphthalene
Other POM®
0.01
0.0025
0.013
0.015
0.0025
0.005
Phenol
0.0075
0.078
0.023
0.025
0
0.0025
Propianaldehyde
0
0.00025
0.0025
0
0
0
Styrene
0.0013
0.0013
0.00025
0.0013
0.12
0
Toluene
0.045
0.005
0.02
0.018
0.023
0.025
Xylenes (total)
0.033
0.0025
0.005
0.0075
0
0.02
a Developed from CERP baseline testing (CERP, 1999b, 2000; Technikon, 2000, 2001a, 2003).
b Developed from CERP testing of chemically bonded mold systems (Technikon, 2001b, 2001 d, 2001 f, 2001 g,
2004).
c Based on data reported by Twarog (1991).
d Developed from CERP "background baseline" (CERP, 1999b).
e POM other than naphthalene, predominately methylnaphthalenes and dimethylnaphthalenes.
Example 5-4: Organic Emissions from PCS Operations
A foundry melts 54,000 tons of gray iron per year. They pour 30,000 tons/yr of metal into
green sand molds that do not have cores, 12,000 tons/yr of metal into phenolic urethane nobake
molds, and 12,000 tons/yr of metal into green sand molds that have chemically bonded cores.
The facility's green sand has an average loss on ignition of 4.5% for all green sand
applications. What are the VOC emissions and HAP emissions from the PCS operations at this
facility?
The emission for each mold system is calculated using Equation 5-1:
Ei=QMs[(MSVoc x Ci,Ms) x GSCF + (CS
voc x Q,cs )]/2,000
For the green sand-only castings, QMs = 30,000; MSVoc = 1.9 (from Table 5-2); and GSCF =
4.5%/5.1% = 0.88. When calculating VOC emissions, ClMs = 1- Since no cores are used,
CSvoc = 0 (from Table 5-2), the second term in the equation will be zero.
Evoc.GSoniy= 30,000[(1.9 x 1) x 0.88 + (0)]/2,000 = 25.1 tons/yr
5-8
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeont
Example 5-4: Organic Emissions from PCS Operations (continued)
For the green sand molds with cores, QMs = 12,000; MSVoc = 1.9 (from Table 5-2); GSCF =
4.5%/5.1 % = 0.88, C,ms = Ci,cs =1 (for VOC); and CSVOc = 1.6 (from Table 5-2).
Evoc.os+cores = 12,000[(1.9 x 1) x 0.88 + (1.6 x l)]/2,000 = 19.6 tons/yr
For the chemically bonded molds, QMs = 12,000; MSVoc = 7.4 (from Table 5-2); GSCF=1;
Cjms =1 (for VOC); and CSVoc = 0 (from Table 5-2).
Evoc.chemMoid=12,000[(7.4 x 1) x 1 + (0)]/2,000 = 44.4 tons/yr
So the total VOC emissions are the sum of the VOC emissions determined from each on the
three mold systems used at the facility.
EVoc= 25.1 + 19.6 + 44.4 = 89.1 tons/yr
To determine the benzene emissions, Cl Ms = 0.065 for the green sand mold (with or without
cores), Cl Ms = 0.028 for the phenolic urethane bonded molds, and Cl Cs =0.073 for the cored
systems (from Table 5-3). The emission calculation for benzene is as follows:
EBz,GSoniy= 30,000[(1.9 x 0.065) x 0.88 + (0)]/2,000 = 1.63 tons/yr
EBz,GS+cores= 12,000[(1.9 x 0.065) x 0.88 + (1.6 x 0.073)]/2,000 = 1.35 tons/yr
Evoc.chemMoid= 12,000[(7.4 x 0.028) x 1 + (0)]/2,000 = 1.24 tons/yr
EBz= 1.63 + 1.35 + 1.24 = 4.22 tons/yr
Emissions of other HAP are calculated in a similar fashion. The results are summarized below:
HAP Compound
Emissions (tons/yr)
Green
Sand
only
Molds
Green
Sand
with
Cores
Chemically
Bonded
Molds
Facility
Total
Acetaldehyde
0.12
0.07
0.03
0.23
Aniline
0.19
0.41
0.06
0.66
Benzene
1.61
1.34
1.22
4.22
Cresols (total)
0.04
0.11
1.78
1.93
N,N-Dimethylaniline
0.06
0.10
0
0.16
Ethylbenzene
0.12
0.06
0.02
0.21
Formaldehyde
0.02
0.01
0.11
0.14
n-Hexane
0.25
0.12
0
0.37
Naphthalene
0.19
0.15
0.08
0.42
Non-Naphthalene
Other POM®
0.25
0.24
0.11
0.60
Phenol
0.19
0.31
3.44
3.97
Propianaldhyde
0
0
0.01
0.01
Styrene
0.03
0.03
0.06
0.12
Toluene
1.12
0.61
0.22
1.97
Xylenes (total)
0.81
0.39
0.11
1.34
5-9
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
5.3.2 PM Emissions from PCS Operations
Table 5-6 of the NESHAP Iron and Steel Foundry BID (U.S. EPA, 2003) reports emission factors for
pouring, cooling, and shakeout for foundries. Table 5-4 summarizes the default emission factors PCS
operations for both iron and steel foundries. See Appendix D for more information about the
development of these factors. These emission factors can be used in Equation 5-2 to determine the PM
mass emissions. For controlled PCS operations, it is recommended that sources use the captured,
uncontrolled emission factors and adjust the emissions based on the expected control efficiencies of the
control device, as provided in Table 3-3 of Section 3.1, Melting Furnaces, of this Protocol document.
Site-specific captured/uncontrolled emission factors can be determined using baghouse catch data. If
available, these site-specific captured/uncontrolled emission factors are preferred to the default factors in
Table 5-4.
As with the PM emission factor for melting furnaces in Table 3-3, the PM emission factors in Table 5-4
for PMio-FIL include emissions of PM2 5-FIL. As well, the PM coarSe -FIL emissions factor is the emissions
factor for PMio-FIL minus the emissions factor for PM2 5-FIL. Therefore, to calculate PM10-FIL
emissions, the emissions for PM coarse -FIL is added to the emissions for PM2 5-FIL.
M X EmF;
Ei 2,000
Eqn. 5-2
where:
E; = emissions for compound i, tons/yr
M = quantity of metal poured, tons/yr
EmF, = emission factor for compound "i", from Table 5-4, lbs/ton
2,000 = Conversion factor, lbs/ton.
Table 5-4. Summary of PM Emission Factors for PCS Lines
Suggested
SCC for Iron
Foundries
Suggested
SCC for Steel
Foundries
Emission category
PM-FIL
emissions
factor,
lb/ton
metal
poured
PM10-FIL
emissions
factor,
lb/ton
metal
poured
PM2.5-FIL
emissions
factor,
lb/ton
metal
poured
PM-CON
emissions
factor,
lb/ton
metal
poured
30400320
30400708
Pouring, captured,
uncontrolled
0.087a
0.071b
0.063b
0.23c
30400325
30400713
Cooling, captured,
uncontrolled
0.29a
0.24b
0.21b
0.77c
30400331
30400709
Shakeout, captured,
uncontrolled
79.3a
65b
57b
a Derived from Table 5-6 of NESHAP Iron and Steel Foundry BID.
b Used the size distribution ratios for steel foundries casting shakeout exhausted, prior to controls, to determine
PM10-FIL and PM2.5-FIL values.
c Developed from CERP testing (Technikon, 2001a; 2001b; 2001c; 2001d; 2001e; 2001 f; 2001g). Note that the
CERP testing did not use EPA Method 202, and it is uncertain how well the CERP procedure would compare with
EPA Method 202 in quantifying condensable PM.
5-10
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeont
5.3.3 HAP Metal Emissions from PCS Operations
Most metal HAP emissions will be associated with the filterable PM emissions. Although there are some
metal HAP emissions associated with condensable PM emissions, based on the very limited data
available, there was not enough data to develop PM-CON fractions. The most accurate method for
determining the metallic HAP composition for emitted PM, particularly for cooling and shakeout, where
there are fewer metal fumes, would be to analyze the metal HAP composition of the collected baghouse
dust, if applicable., Site-specific metal chemistry should be used to estimate the metal content of the
emitted (filterable) PM for pouring emissions. In the absence of site-specific baghouse dust or metal
chemistry data, the default metal chemistries provided in Table 5-5 can be used to estimate HAP metals.
See Appendix D for more information about the development of these factors. It can be expected that for
certain metal HAP, such as mercury, the metal HAP emissions will be primarily associated with the
melting furnace, where most of these volatile HAP will be released from the metal, so there would not be
appreciable amounts of mercury or similar HAP in the emissions associated with pouring, cooling, and
shakeout. For shakeout emissions, the metal HAP content of the PM is highly dependent on the relative
amounts of sand compared to metal particles are in the emitted dust. Shakeout systems with higher PM
emissions are likely to have lower metal HAP concentrations as these high PM emitting systems are
expected to have more sand fines than metal fines as compared to lower PM emitting systems. Because
the shakeout system for which metal HAP concentrations were determined in Table 5-5 had lower PM
emissions than the average factor in Table 5-4, the combination of these default factors are expected to
result in a conservatively high metal HAP emission estimate for shakeout.
Table 5-5. Summary of HAP Content of PM from PCS Components3
Suggested
SCC for
Iron
Foundries0
Suggested
SCC for
Steel
Foundries0
CAS No.
Metal
Constituent
HAP% of
PM-FIL
Pouring
HAP% of
PM-FIL
Cooling
HAP% of
PM-FIL
Shakeout
HAP%
of PM-
FIL PCS
Total
30400320
30400708
7440-36-0
Antimony
ND
0.0097
0.0022
0.0049
30400320
30400708
7440-38-2
Arsenic
0.0046
ND
ND
0.00072
30400320
30400708
7440-43-9
Cadmium
0.011
0.019
0.014
0.016
30400320
30400708
18540-29-9
Chromium
(hexavalent)
0.0036b
0.0066b
0.0045b
0.0051b
30400320
30400708
7440-47-3
Chromium (total)
0.12
0.22
0.15
0.17
30400320
30400708
7440-48-4
Cobalt
1.77
0.050
0.074
0.33
30400320
30400708
7439-92-1
Lead
0.43
0.21
0.63
0.18
30400320
30400708
7439-96-5
Manganese
2.01
0.49
0.29
0.64
30400320
30400708
7440-02-0
Nickel
0.28
0.18
0.27
0.23
30400320
30400708
7782-49-2
Selenium
ND
0.0039
ND
0.0016
a Derived from test data from 1999 CERP Foundry Mexico Baseline Testing (CERP, 1999a).
b Assume hexavalent chromium is 3% of total chromium emissions based on Chromium hexavalent percentages
reported for Pouring, Cooling, and Shakeout for Iron and Steel Foundry SCC in Appendix D of NATA (U.S. EPA,
2011a).
c Report all emissions under pouring SCC as there is not an SCC for total PCS emissions.
ND = Non-Detect
5.3.4 Other Criteria Pollutant Emissions from PCS Operations
It can be expected that PCS emissions from chemically bonded molds will have carbon monoxide
emissions. The default emission factors are presented below in Table 5-6. See Appendix D for more
information about the development of this factor.
5-11
-------
Version 1
Draft
Section 5—Pouring, Cooling, and Shakeout
Table 5-6. Summary of Non-PM Criteria Pollutant Emission Factors
for PCS Operations3
Suggested SCC for
Iron Foundries"
Suggested SCC for
Steel Foundries"
HAP Compound
Emission Factor, lb/ton
Chemically Bonded Molds (all)
30400320
30400708
Carbon Monoxide
3.7
aDeveloped from CERP testing of chemically bonded mold systems (Technikon, 2001b, 2001 d, 2001 f, 2001 g, 2004).
b Report all emissions under pouring SCC as there is not an SCC for total PCS emissions.
5-12
-------
Version 1
Draft
Section 6—Finishing Operations
6. Finishing Operations
All castings typically undergo some type of mechanical finishing. Finishing operations begin once the
casting is removed from the mold and cooled. Hammers, band saws, abrasive cutting wheels, flame cut-
off devices, and air-carbon arc devices may be used to remove the risers, runners, and sprues of the metal
transfer system. Metal fins at the parting lines (i.e., lines on a casting corresponding to the interface
between the cope and drag of a mold) are removed with chipping hammers and grinders. Residual
refractory material and oxides are typically removed by sand blasting or steel shot blasting, which can
also be used to give the casting a uniform and more attractive surface appearance (U.S. EPA, 2002).
The cleaning of castings precedes any coating operations to ensure that the coating will adhere to the
metal. Scale, rust, oxides, oil, grease, and dirt can be chemically removed from the surface of a casting
using organic solvents, emulsifiers, pressurized water, abrasives, or alkaline agents (caustic soda, soda
ash, alkaline silicates, and phosphates). Molten salt baths are also used to clean complex interior passages
in castings (U.S. EPA, 2002).
Castings are often given a coating to inhibit oxidation, resist deterioration, or improve appearance.
Common coating operations include painting, electroplating, electroless nickel plating, hard facing, hot
dipping, thermal spraying, diffusion, conversion, porcelain enameling, and organic or fused dry-resin
coating (U.S. EPA, 2002).
The emission estimation methods for finishing operations are presented in Table 6-1. Some equipment
for finishing operations, such as shotblasting equipment, are enclosed; other equipment for finishing
operations would require a hooding or ventilation system to direct the emissions to a well-defined stack.
For emissions that are captured and emitted through well-defined stacks, direct measurement methods
may be used. Direct measurements may be continuous or routinely frequent (e.g., daily or weekly) in
nature so that the measurement data can be used directly to determine emissions (Methodology Rank 1 or
2). For some pollutants, a one-time or annual source test may be performed or, for PM, baghouse catch
quantities can be determined even though emissions are not routinely measured. For these pollutants, the
measured data can be used to determine an emission factor based on process throughput (e.g., metal
produced, tons of castings cleaned, number of castings, gallons of coating used, etc.), and the annual
emissions can be determined using the annual material usage multiplied by the site-specific emission
factor (Methodology Rank 3). However, for most finishing operations, no direct measurement data will be
available. In such cases, emissions will be estimated based on default release factors (Methodology Rank
4a) or generic emission factors (Methodology Rank 4b).
Table 6-1. Summary of Typical Hierarchy of Finishing Operations Emission Estimates
Rank
Methodology
Description
Application
Data Requirements
1
Direct routine
concentration and flow
measurement
Finishing, grinding,
cleaning, and coating
operations that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate
2
Direct routine
concentration
measurement with
engineering flow
estimates
Finishing, grinding,
cleaning, and coating
operations that are well-
captured and have well-
defined stack emissions
Constituent concentration and operating
parameters needed for estimating flow rate
(e.g., fan curve and amp usage)
6-1
-------
Version 1
Draft
Section 6—Finishing Operations
Rank
Methodology
Description
Application
Data Requirements
3
Site-specific emission
factor from one-time or
periodic emissions
source test data
Finishing, grinding,
cleaning, and coating
operations that are well-
captured and have well-
defined stack emissions
Constituent concentration and exhaust flow
rate and material usage rate (or throughput)
during source test; annual material usage
rate (or annual throughput)
4a
Chemical-specific
release factors
Organic emissions from
cleaning and coating
materials
Composition of cleaning and coating material
(from MSDS) and cleaning and coating
material annual usage quantities
4b
Default emission factors
All PM and metal HAP
sources
Material usage rates
Emissions from finishing operations include PM emissions from the mechanical finishing operations
(e.g., cutting, grinding, shot blasting), which may contain metal HAPs as well as organic emissions from
the release or volatilization of constituents in cleaning and coating materials used to clean and coat the
metal casts. The PM produced by mechanical finishing is anticipated to be mainly coarse material that
would not remain airborne. That is, uncontrolled PM produced by mechanical finishing would not
generally escape the foundry building or be transported outside the facility boundaries. The cleaning and
coating operations may generate VOC and organic HAPs from painting; coating, and solvent cleaning and
acid and metal ion mists from anodizing; plating; polishing; hot-dip coating, etching; and chemical
conversion coating.
6.1 Methodology Ranks 1 and 2 for Finishing Operations
Emissions from finishing operations can be directly measured at the stack or outlet of the control device
using the direct measurement methods described in Section 3, Melting Operations, of this Foundries
Emissions Protocol document. If CEMS are available for both a pollutant concentration and flow rate, the
CEMS data should be used to determine the pollutant emissions as Methodology Rank 1 for Finishing
Operations. If a CEMS is used to determine a pollutant's concentration, but direct-flow measurement is
not available, flow rates can often be determined using engineering estimates, such as fan amperage-to-
flow correlations, to determine the pollutant emissions as Methodology Rank 2 for Finishing Operations.
6.2 Methodology Rank 3 for Finishing Operations
When CEMS data are not available, but data from a single source test or a limited number of source tests
are available, a site-specific emission factor can be developed and used to assess the emissions from
finishing operations for the pollutants tested, similar to the methods described in Section 3, Melting
Operations. For the finishing operations, the "activity" data will be the mass of metal poured or castings
produced. Site-specific emission factors include those developed from baghouse catch data. The method
to determine the captured, uncontrolled PM emissions from baghouse catch data was provided previously
in Example 3-4. The captured, uncontrolled emissions estimated from baghouse catch data should be
corrected for the capture efficiency, particularly for sources that employ canopy hoods or similar capture
systems. Baghouse dust analyses is the preferred source of metal HAP concentration for the emitted PM.
6.3 Methodology Rank 4 for Finishing Operations
When no emission measurement data are available, pollutant emissions will need to be estimated using
default factors and either material-specific composition (Methodology Rank 4a for Finishing Operations)
or default composition profiles (Methodology Rank 4b for Finishing Operations).
6-2
-------
Version 1
Draft
Section 6—Finishing Operations
6.3.1 Organic Pollutant Emission from Finishing Operations
Organic air pollutant emissions from cleaning and coating materials will vary widely depending on the
specific chemicals used during the cleaning and coating finishing operations. As such, chemical-specific
compositional data should always be used when estimating the emissions from these operations
(Methodology Rank 4a for Finishing Operations). The quantity of each cleaning and coating material
used should be available from purchase records or direct material usage meters. Cleaning and coating
material component composition should also be available from MSDS received from the chemical
supplier (as described in Section 4.2 of this Protocol document). Together, this information provides a
direct means of determining constituent-specific usage rates. The organic emissions emitted to the
atmosphere from the cleaning and coating materials can be assumed to be 100 percent of the volatile
constituents (as listed in Table 1-1 of the Emissions Protocol document). The annual emissions for each
volatile pollutant contained in the materials used in cleaning and coating can be estimated using the
following general equation (Equation 6-1):
E| = ExN=i (Qx x
Eqn. 6-1
where:
E; = emissions for compound i, tons/yr
Qx = quantity of material "x" used at the foundry, tons/yr
Cxj = concentration of compound "i" in material "x", weight percent
6.3.2 PM Emissions from Finishing Operations
The EPA has developed PM emission factors for finishing operations, which are provided in Table 12.10-
7 of Section 12.10 of AP-42 (U.S. EPA, 2003) and Table 12.13-2 of Section 12.13 of AP-42 (U.S. EPA,
1995), and provides size distribution analysis for iron foundry sources in Table 12.10-8 of Section 12.10
of AP-42 (U.S. EPA, 2003). Data collected through the Foundry Information Collection Request (ICR)
were also used to assess the AP-42 emission factors and to refine the default emission factors for iron and
steel foundries based on the available data. Table 12.13-2 of Section 12.13 of AP-42 (U.S. EPA, 1995)
reports one emission factor for casting cleaning at steel foundries of 1.7 lb PMio-FIL/ton of metal
processed. Table 12.10-7 of Section 12.10 of AP-42 (U.S. EPA, 2003) contains a generic emission factor
of 17 lbs PM (TSP)/ton of gray iron produced for uncontrolled cleaning and finishing but indicates that
only 0.1 lbs/ton of PM is released to the atmosphere (with 99+ percent of the PM settling within the
foundry building). The discrepancy between these factors is likely to be due to the size of the castings
produced rather than the types of metal cast. Grinding/finishing of small parts can be expected to produce
more PM per ton of cast metal than grinding/finishing of larger parts. Unfortunately, no data exists to
correlate the PM emissions to the size of the parts produced. Bag house catch data collected from the
Foundry ICR (see Table C-17 in Appendix C of the Foundry BID, [U.S. EPA, 2002]) was used to
calculate separate emission factors for cutting, grinding, and shot blasting. See Appendix E for more
information about the development of the factors. The data show significant variability across different
facilities, indicating that the default PM emission factors for finishing operations are highly uncertain.
Based on the Foundry ICR data, the emission factors presented in Table 6-2 are recommended for
finishing operations at both iron and steel foundries. For finishing operations, it is assumed that there are
no condensable PM emissions so that PM-FIL emissions equal PM-PRI emissions. These emission
factors can be used in Equation 6-2 to determine the PM mass emissions for uncontrolled units. For
controlled finishing operations, it is recommended that sources use the ducted uncontrolled emission
factors and adjust the emissions based on the expected control efficiencies of the control device as
provided in Table 3-4 of Section 3.1, Melting Furnaces, of this Protocol document. When applying the
control efficiencies from Table 3-4 to the emission factors of Table 6-2, use the control efficiency
6-3
-------
Version 1
Draft
Section 6—Finishing Operations
reported for the particle size range of 0 to 2.5 nm for the fraction of PM2 5-FIL and use the control
efficiency reported for the article size range of 2.5 to 10 pim for the fraction of PM10-FIL. It can be
assumed that the PM collection efficiencies for PM greater than 10 nm in diameter are 100 percent
(provided the PM collection efficiency for the 2.5 to 10 pim is 90 percent or greater).
As with the PM emission factor for melting furnaces in Table 3-3, the PM emission factors in Table 6-2
for PM10-FIL include emissions of PM25-FIL. As well, the PM coarse -FIL emissions factor is the emissions
factor for PM10-FIL minus the emissions factor for PM2 5-FIL. Therefore, to calculate PM10-FIL
emissions, the emissions for PM coarse -FIL is added to the emissions for PM2 5-FIL.
_ MxEmFj
Ei j —
1 2,000
Eqn. 6-2
where:
E; = emissions of pollutant ""i". tons/yr
M = quantity of metal produced, tons/yr
Em F, = emission factor of pollutant "i", from Table 6-2, lbs/ton
2,000 = Conversion factor, lbs/ton.
Table 6-2. Default PM Emission Factors for Finishing Operations
Suggested SCC for Iron
Foundries
Suggested SCC for
Steel Foundries
Emission Source
Pollutant
Emission
Factor
(lb/ton
metal
produced)
30400360
30400715
30400765
Cutting, captured and
uncontrolled
PM-FIL
PM10-FIL
PM2.5-FIL
6.0a
3.0b
1.2b
30400340
30400711
Grinding, captured and
uncontrolled
PM-FIL
PM10-FIL
PM2.5-FIL
16.0a
8.0b
3.2b
Shot blasting or sand
blasting, captured
and uncontrolled
PM-FIL
PM10-FIL
PM2.5-FIL
16.0a
8.0b
3.2b
30400360
30400715
30400765
Cutting, uncaptured and
uncontrolled
PM-FIL
PM10-FIL
PM2.5-FIL
0.06c
0.054c
0.048c
30400340
30400711
Grinding, uncaptured
and uncontrolled
PM-FIL
PM10-FIL
PM2.5-FIL
pop
CO O)
000
Shot blasting or sand
blasting, uncaptured
and uncontrolled
PM-FIL
PM10-FIL
PM2 s-FIL
pop
CO ^ O)
000
a Used data collected from Foundry ICR.
b Used the size distribution ratios of for steel foundries pouring and cooling uncontrolled, to determine PM10-FIL
(50%) and PM2.5-FIL (20%) values.
6-4
-------
Version 1
Draft
Section 6—Finishing Operations
c Assumes 1 % of ducted emissions would be released as fugitive emissions that escape to the atmosphere for
uncaptured units to roughly agree with the atmospheric release factor reported for finishing operations at iron
foundries reported in AP-42 (U.S. EPA, 2003). Assumed that 90% of PM is PM-io and 80% of PM is PM2.5 as
smaller particles are more likely to escape to the atmosphere.
6.3.3 Metal HAP Emissions from Finishing Operations
The metal HAP emissions from finishing operations will be associated with the filterable PM emissions
from finishing operations. Use Equation 6-3 to determine the emissions of specific HAP metals from the
melting furnace PM emission estimates. The PM generated during cut-off and grinding is expected to be
primarily the same composition of the cast metal; the PM generated during shot blasting is expected to be
primarily sand with some contribution from the cast metal. Baghouse dust analyses suggest that the
concentrations of metals in the collected PM from shot blasting operations are approximately one-fifth
that of dust collected from cut-off or grinding operations. The factor of 5 associated with the PM
generated from shot blasting is included to account for the lower metal content of PM emissions from
shot blasting. If baghouse dust analysis is available, the factor of 5 correction to adjust the shot blasting
metal HAP emissions in Equation 6-3 should not be used.
„ _ (nj., r,,, J}., T-.TT PM'FILghotblagA %PM"FILj
Ei — ^PM-FILcutoff + PM-FILgrinciing H - J X iQQ
-------
Version 1
Draft
Section 6—Finishing Operations
Example 6-1: Estimating PM Emissions from Finishing Operations
A facility pours 35,500 tons of steel per year to produce 31,000 tons of finished castings. The
facility has a shot blasting unit that is controlled by a high-efficiency cyclone and a grinding
station that is uncaptured/uncontrolled. The facility does not have a cut-off station. Calculate
the PM emissions for the finishing operations for this facility.
For the uncaptured/uncontrolled grinding station, use the default emission factors for
uncaptured/uncontrolled emissions from Table 6-2 in Equation 6-2. For Equation 6-2, the mass
of metal produced refers to the total mass of metal cast, not just the mass of final metal
product.
The calculation for PM-FIL for grinding follows:
Epm-fil,Grinding = Epm-pri = (35,500 tons X = 2.84 tons/yr
[Note that PM-CON for finishing operations is 0, so PM-PRI = PM-FIL.]
Similar calculations are used for PM10_FIL and PM2 5-FIL for grinding to yield:
Epmio-fil,Grinding = 2.49 tons/yr
EpM2.5-FIL,Grinding = 2.31 tOns/yr
For the shot blasting unit, first determine the inlet loading to the control device using the
captured/uncontrolled emission factors from Table 6-2 in Equation 6-2 as follows:
Epm-fil,SB,uncontrolled — ^35,500 tons X 2 00q) — 284 tons/yr
EpM10-FIL,SB,uncontolled= 142 tOns/yr
EpM2.5-FIL, SB ,uncontrolled 56.8 tOns/yr.
Next, the control device efficiency for the different size fractions is selected from Table 3-4.
For a high-efficiency cyclone (centrifugal collector), the collection efficiency for PM2 5 is 80%.
For PM between 2.5 and 10 pirn, the collection efficiency is 95%.
Calculate the PM2 5-FIL emissions based on the control device efficiency using Equation 3-4 as
follows:
EpM2.5-fil,sb = 56.8 x (1-0.80) = 11.36 tons/yr
Similarly, calculate the PM emissions between 2.5 and 10 nm as follows:
EPM-"coarse",sB = (142-56.8) x (1-0.95) = 4.26 tons/yr
The PM10-FIL emissions are the sum of the PM2 5-FIL and PM "coarse" emissions. Given the
high PM10 collection efficiency, it can be assumed that the collection efficiency for PM greater
than 10 |im in diameter is effectively 100%. Therefore,
Epm-fil,sb = Epmio-fil.sb = 11-36 + 4.26 = 15.62 tons/yr.
6-6
-------
Version 1
Draft
Section 6—Finishing Operations
Example 6-2: Estimating HAP Metal Emissions from Finishing Operations
For the facility in Example 6-1, calculate the HAP metal emissions for the finishing operations.
The facility does not have site-specific metal chemistry data available and does not have a cut-
off station.
Since the facility does not have site-specific metal chemistry data available, use the default
values in Table 3-6 of this protocol document and Equation 6-3.
Metal Constituent
%of PM-FIL
Antimony
0.02
Arsenic
0.003
Beryllium
0.0001
Cadmium
0.02
Chromium (hexavalent)
0.01
Chromium (total)
0.10
Cobalt
0.0006
Lead
1.0
Manganese
3.0
Mercury
0.02
Nickel
0.2
Phosphorus
0. 3
Selenium
0.01
PM-FILsb\ %PM-FILi
Ej - ^PM-FILcutoff + PM-FILGrinding H ——J X ¦
EAntimony = (o + 2.84 + X = 0.0012 ton/yr
Similarly,
EArsenic = 0.00018 ton/yr
EBeryilium = 0.0000060 ton/yr
Ecadmmm = 0.0012 ton/yr
Echromium (hexavalent) 0.00060 tOn/yr
Echrommm (total) = 0.0060 ton/yr
Ecobait = 0.000036 ton/yr
ELead = 0.060 ton/yr
EManganese 0.18 tOn/yr
EMercury = 0.0012 ton/yr
ENickei = 0.012 ton/yr
Ephosphorus 0.018 tOll/vT
Eseiemum = 0.00060 ton/yr
6-7
-------
Version 1
Draft
Section 7—References
7. References
AFS and CISA (1998). Form R - Reporting of binder chemicals used in foundries. 2nd Edition.
Prepared by American Foundrymen's Society (AFS) and the Casting Industry Suppliers
Association (CISA).
AFS and CISA (2007). Form R - Reporting of binder chemicals used in foundries. 4th Edition. Prepared
by AFS and CISA.
Avallone, E.A., & Baumeister, T. Editors. (1978). Marks' standard handbook for mechanical engineers.
Ninth Edition. McGraw-Hill Book Company, New York, New York.
CERP (1999a). Foundry process emission factors: baseline emissions from automotive foundries in
Mexico. January 19, 1999.
CERP (1999b). Baseline testing emission results pre-production foundry. November 1, 1999.
CERP (2000). Baseline testing emission results production foundry. February 7, 2000.
Levelton Consultants Ltd. (2005). Emission monitoring and reporting strategy - Summary and
background. Prepared for Canadian Council of Ministers of the Environment and NFPRER
Monitoring and Reporting Team.
Residuals Management Technology (RMT), Inc. (1998). Technical and economical feasibility study for
control ofVOCs from phenolic urethane cold box and no bake core- and mold-making operations
in foundries. Prepared for the Ohio Cast Metal Association and American Foundrymen's Society,
Inc. April
Technikon (2000). Ashland core binder replacement. December 15, 2000.
Technikon (2001a). Production baseline airborne emission test report. March 26, 2001.
Technikon (2001b). Phenolic urethane /iron no-bake baseline emission test. March 27, 2001.
Technikon (2001c). Pre-production air emission test report. April 25, 2001.
Technikon (200 Id). Iron phenolic no-bake delta HA Techniset. April 26, 2001.
Technikon (200 le). Greensandprepared with tap and advanced oxidant enriched water. August 1, 2001.
Technikon (200If). Ashland iron/furan no-bake NBF. May 18, 2001
Technikon (200 lg). Ashland iron no bake NBE1 and 2 ester cured phenolic/iron. June 22, 2001.
Technikon (2003). Emission comparison of phenolic urethane binders with standard solvents and
napthalene-depletedsolvents. July 30, 2003.
Technikon (2004). Product test: no-bake HA international. March 2004.
Technikon (2006). CERP organic HAP emission measurements for iron foundries and their use in
development of an AFS HAP guidance document. January 2006 & August 2007.
7-1
-------
Version 1
Draft
Section 7—References
Twarog, D.L. (1991). Identification of emissions and solid wastes generatedfrom EPC process.
American Foundrymen's Society Research Report. June 4, 1991.
U.S. EPA (1993). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Appendix C.2: Procedures for the laboratory analysis of surface/bulk dust loading
samples. AP-42, Fifth Edition. Office of Air Quality Planning and Standards, Research Triangle
Park, NC.
U.S. EPA (1995). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 12.13: Steel foundries. AP-42, Fifth Edition. Office of Air Quality Planning and
Standards, Research Triangle Park, NC.
U.S. EPA (1996). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Appendix B.2: Generalized Particle Size Distributions. AP-42, Fifth Edition. Office of
Air Quality Planning and Standards, Research Triangle Park, NC. September.
U.S. EPA (1997). Procedures for preparing emission factor documents. Office of Air Quality Planning
and Standards, Research Triangle Park, NC. EPA-454/R-95-015. November.
U.S. EPA (1999a). Iron and steel foundries manual emissions testing of cupola baghouse at Waupaca
Foundry in Tell City, Indiana. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. EPA-454/R-99-017A and EPA-454/R-99-017B. June.
U.S. EPA (1999b). Iron and steel foundries manual emissions testing of cupola wet scrubber at General
Motors Corp., Saginaw, Michigan. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. EPA-454/R-99-025A and EPA-454/R-99-025B. July.
U.S. EPA (2002). National emission standards for hazardous air pollutants for iron and steel foundries—
background information for proposed standards. Office of Air Quality Planning and Standards,
Research Triangle Park, NC.
U.S. EPA (2003). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 12.10: Gray iron foundries. AP-42, Fifth Edition. Office of Air Quality Planning
and Standards, Research Triangle Park, NC.
U.S. EPA (2006a). An inventory of sources and environmental releases of dioxin-like compounds in the
U.S. for the years 1987, 1995, and 2000. Office of Research and Development, Washington, DC,
EPA/600/P-03/002F. November.
U.S. EPA (2006b). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 13.2.2: Unpaved roads. AP-42, Fifth Edition. Office of Air Quality Planning and
Standards, Research Triangle Park, NC.
U.S. EPA (2006c). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 13.2.4: Aggregate HANDIING AND STORAGE PIIEs. AP-42, Fifth Edition.
Office of Air Quality Planning and Standards, Research Triangle Park, NC.
U.S. EPA (2006d). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 13.2.5: Industrial wind erosion. AP-42, Fifth Edition. Office of Air Quality
Planning and Standards, Research Triangle Park, NC.
U.S. EPA (201 la). An overview of methods for EPA's National-Scale Air Toxics Assessment. Office of
Air Quality Planning and Standards, Research Triangle Park, NC.
7-2
-------
Version 1
Draft Section 7—References
U.S. EPA (201 lb). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 13.2.1: Paved roads. AP-42, Fifth Edition. Office of Air Quality Planning and
Standards, Research Triangle Park, NC.
7-3
-------
Version 1
Draft
Appendix A
Appendix A:
Foundry Glossary Terms
http://www.atlasfdry.com/glossary.htm
Addition Agent: Any material added to a charge of molten metal in bath or ladle to bring alloy to
specifications. A reagent added to the plating bath.
Additives: Any material added to molding sand for reasons other than bonding or improvement of bond is
considered an additive. Bonds can be of varying types: carbonaceous (sea coal, pitch, fuel oil, graphite,
gilsonite); cellulose (wood flour, cereal hulls); fines (silica flour, iron oxide, fly ash); cereals (corn flour,
dextrine, sugar); and chemical (boric acid, sulfur, ammonium compounds, diethylene glycol).
Alloy: A substance having metallic properties and composed of two or more chemical elements of which
at least one is metal. A metallic material formed by mixing two or more chemical elements. Usually
possess properties different from those of the components. As examples, Brass is an alloy of copper and
zinc and Cast Iron contains iron, carbon and silicon.
Alloy Steel: Steel containing significant quantities of alloying elements other than carbon and the
commonly accepted amounts of manganese, silicon, sulfur, and phosphorus.
Alloying: Procedure of adding elements other than those usually comprising a metal or alloy to change its
characteristics and properties.
Alloying Elements: Elements added to nonferrous and ferrous metals and alloys to change their
characteristics and properties.
Ambient Air: The surrounding air.
Ambient Temperature: Temperature of the surrounding air.
American Foundry Society: Association that provides and promotes knowledge and services that
strengthen the metal casting industry for the ultimate benefit of its customers and society.
Anodizing: Forming a conversion coating on a metal surface by electrolytic oxidation with the work
forming the anode. This process is most frequently applied to aluminum.
Antimony: One of the elements; its chemical symbol is Sb. Its formula weight is 121.76, specific gravity
6.62, and melting point 630.5°C.
Assembling (Assembly) Line: Conveyor system where molds or cores are assembled.
Bake: Heat in an oven to a low controlled temperature to remove gases or to harden a binder.
Baked Core: A core which has been heated through sufficient time and temperature to produce the desired
physical properties attainable from its oxidizing or thermal-setting binders as opposed to a green-sand
core, which is used in the moist state.
Baked Permeability: Property of a molded mass of sand heated at a temperature above 230° F until dry
and cooled to room temperature, to permit passage of gases through it; particularly those generated during
pouring of molten metal into a mold.
Baked Strength: Compressive, shear, tensile or transverse strength of a mold sand mixture when baked at
a temperature above 231°F (111°C) and then cooled to room temperature.
A-l
-------
Version 1
Draft
Appendix A
Batch: Amount or quantity of core or mold sand or other material prepared at one time.
Bed Charge: The charge of iron placed on the coke bed in a cupola.
Bed Coke: Coke placed in the cupola well to support the following iron and coke charges.
Binder: The bonding agent is a material used as an additive to mold or core sand to impart strength or
plasticity in a "green" or dry state. May be cereal, oil, clay, resin, pitch, etc.
Binder, Plastic (Resin): Synthetic resin material used to hold grains of sand together in molds or cores;
may be phenol formaldehyde or urea formaldehyde thermosetting types.
Blast: Air driven into the cupola or furnace for combustion of fuel.
Blast Cleaning: Removal of sand or oxide scale from castings by the impinging action of sand, metal shot,
or grit projected under air, water, or centrifugal pressure.
Blended Molding Sands: Naturally bonded molding sands which have been mixed or modified by the
supplier to produce desirable properties.
Blended Sand: Mixture of sands of different grain sizes, clay content, etc., to produce one possessing
characteristics more suitable for foundry use.
Blower, Core Or Mold: A machine or device using compressed air to inject sand into a core box or a
flask.
Bond: (a) A bonding substance or bonding agents - any material other than water, which, when added to
foundry sands, imparts bond strength. The overlapping of brick so as to give both longitudinal and
transverse strength, (b) Cohesive material in sand.
Bottom Pour Ladle: Ladle in which metal, usually steel, flows through a nozzle in the bottom.
Bottom Pour Mold: A mold that is gated at the bottom.
Bottom Running Or Pouring: Filling of the mold cavity from the bottom by means of gates from the
runner.
Bottom Sand: Layer of molding sand rammed into place on the doors at the bottom of a cupola.
C: Degrees Centigrade or Celsius.
Casting (verb): A process where molten metal is poured into a mold and solidification is allowed to take
place. The act of pouring metal.
Casting (noun): A metal object obtained by pouring molten metal into a mold. The metal shape, exclusive
of gates and risers, that is obtained as a result of pouring metal into a mold.
Casting Process: A forming process in which a molten metal, polymer, or other heated liquid or plastic
material is poured into a mold or onto a substrate with little or no pressure applied; the substance cools,
solidifies, and the formed object is removed.
Cavity, Mold Or Die: Impression or impressions in a mold or die that give the casting its shape.
Centrifugal Casting (verb): Process of filling molds by pouring the metal into a sand or metal mold
revolving about either its horizontal or vertical axis, or pouring the metal into a mold that subsequently is
A-2
-------
Version 1
Draft
Appendix A
revolved before solidification of the metal is complete. Molten metal is moved from the center of the
mold to the periphery by centrifugal action.
Centrifugal Casting (noun): Casting made in molds which are rotating so as to produce a centrifugal force
in the molten metal.
Chip (verb): To remove extraneous metal from a casting with hand or pneumatically operated chisels.
Chromium: Alloying element used as a carbide stabilizer.
Cleaning: Process of removing sand, surface blemishes, runners, risers, flash, surplus metal, and sand etc.,
from the exterior and interior surfaces of castings. Includes degating, tumbling, or abrasive blasting,
grinding off gate stubs, etc.
Cobalt: Blue-white metal, melting at 2,715°F (1,492°C), used in very hard alloy such as stellite, and a
binder in carbide cutting tools.
Coke: Coal derivative resulting from the distillation of bituminous coal in the absence of air. The
distillation process removes all of the volatile material from the coal so it can be used as a very intense
source of fuel in cupola melting. Source of some carbon found in iron.
Coke Bed: First layer of coke placed in the cupola. Also the coke used as the foundation in constructing a
large mold in a flask or pit.
Coke Breeze: Fines from coke screening, used in blacking mixes after grinding; also briquetted for cupola
use.
Cold Box Process: A rapid coremaking process which does not require application of heat to cure the
cores. Hardening of the cores is accomplished by chemical reaction rather than by conventional baking. A
phenolic resin is added to the sand used to make the core. This resin reacts chemically when exposed to
an accelerator, typically an active organic gas, and hardens very quickly, forming an organic bond in the
core sand. This reaction occurs at room temperature and does not require special coreboxes or equipment.
Additionally, since the bond is organic, the sand collapses readily during shakeout and can be recovered
easily from the casting.
Combustion: Chemical change as a result of the combination of the combustible constituents of the fuel
with oxygen, producing heat.
Combustion Chamber: Space in furnace where combustion of gaseous products from fuel takes place.
Combustion Efficiency: The amount of heat usefully available divided by the maximum amount which
can be liberated by combustion; usually expressed in percentage.
Continuous Tapping: A furnace or holding ladle that is made of discharge molten metal continuously
during normal operation.
Conveyor: A mechanical apparatus for carrying or transporting materials from place to place. Types
include apron, belt, chain, gravity, roller, monorail, overhead, pneumatic, vibrating, etc.
Conveyor Belt: A continuously moving belt used in an automated or semiautomatic foundry to move
materials from one station to another.
Core: A bonded sand insert placed in the mold to form an undercut or hollow section in the casting which
cannot be shaped by the pattern. A core is frequently used to create openings and various shaped cavities
A-3
-------
Version 1
Draft
Appendix A
in the casting. The shaped body of sand which forms interior of casting and also selected external
features.
Core Assembly: A complex core made from a number of cores or sections.
Core Binder: Any material used to hold the grains of core sand together.
Core Compound: A commercial mixture used as a binder in core sand.
Core Knockout Machine: A mechanical device for removing cores from castings.
Crucible: A ceramic pot or receptacle made of materials such as graphite or silicon carbide, with
relatively high thermal conductivity, bonded with clay or carbon, and used in melting metals; sometimes
applied to pots made of cast iron, steel, or wrought steel. The name derives from the cross, the Crux, with
which ancient alchemists adorned it.
Crucible Furnace: A furnace fired with coke, oil, gas, or electricity in which metals are melted in a
refractory crucible.
Cupola: A cylindrical straight shaft furnace usually lined with refractories, for melting metal in direct
contact with coke by forcing air under pressure through openings near its base. Vertical shaft furnace
lined with refractories used to produce cast iron by high temperature melting of metallic and mineral
charge materials.
Cure: To harden.
Curing Time (No Bake): That period of time needed before a sand mass reaches maximum hardness.
Cut: Defect in a casting resulting from erosion of the sand by metal flowing over the mold or cored
surface.
Cutoff Machines, Abrasive: A machine using a thin abrasive wheel and employed in cutting off gates and
risers from casting or in similar operations.
Cyclone (Centrifugal Collector): In air pollution control, a controlled descending vortex created to spiral
objectionable gases and dust to the bottom of a collector core.
Die: A metal block used in forming materials by casting, molding, stamping, threading, or extruding. A
metal form used as a permanent mold for die casting or lost wax process.
Die Casting: (a) Forcing molten metal into permanent molds, dies. Die Casting is also called Pressure
Casting. See Pressure Die Casting, (b) noun Casting resulting from die-casting process, (c) verb Pouring
molten metal under pressure into metal molds.
Direct-Arc Furnace: Electric furnace in which the material is heated directly by an arc established
between the electrodes and the work. See Dielectric Furnace.
Draw: A term used to temper, to remove pattern from mold, as an external contraction defect on surface
of mold.
Draw (verb): To remove a pattern from a mold.
Dried Sand: Sand which bas been dried by mechanical dryer prior to use in core making.
A-4
-------
Version 1
Draft
Appendix A
Dual Metal Centrifugal Casting: Centrifugal castings produced by pouring a different metal into the
rotating mold after the first metal poured.
Ductile Iron: A type of iron in which the graphite content takes spherical rather than flake form. Ductile
iron is produced by adding magnesium. The spherical form of the graphite provides greater tensile
strengths and flexibility than other types of iron. An iron/graphite composite in which the graphite exists
in spheres or nodules, allowing the material to deform rather than fracture when placed under mechanical
stress. Also called Nodular Iron. Iron in which carbon is in the form of spherical nodules.
Electric Arc Furnace: A crucible furnace that uses an electric arc, similar to an electric arc welding
operation, to melt metal.
Electrode: Compressed graphite or carbon cylinder or rod used to conduct electric current in electric arc
furnaces, arc lamps, carbon arc welding, etc.
Expendable Pattern: In investment molding, the wax or plastic pattern that is left in the mold and later
melted and burned out. This also called a disposable pattern.
Ferroalloys: Alloys consisting of certain elements combined with iron, and used to increase the amount of
such elements in ferrous metals and alloys. In some cases the ferroalloys may serve as deoxidizers.
Finish (verb): The hand work on a mold after the pattern has been withdrawn.
Flux: Any substance used to promote fusion. Also any material which reduces, oxidizes, or decomposes
impurities so that they are carried off as slags or gases.
Foundry (Foundries, plural): The act, process, or art of casting metals. The buildings and works for
casting metals.
Foundry Ladle: A vessel for holding molten metal and conveying it from cupola to the molds.
Foundry Returns: Metal in the form of sprues, gates, runners, risers and scrapped castings, with known
chemical composition that are returned to the furnace for remelting. Sometimes referred to as "revert".
Foundry Sand: Foundry sand is used in creating cores and molds used in the casting of iron, steel, copper
and aluminum products. In construction, steel and iron beams-known as girders-are used in the building
of bridges, large office buildings and some homes. Copper pipes, aluminum supports and even the
hardware and hand tools used in construction had their origins at the foundry. Foundry sand is the second
largest industrial use of sand in terms of tons consumed.
Gray Iron: Iron in which a large percentage of the carbon content is in the form of graphite flakes.
Traditionally referred to as "Cast Iron". The graphite flakes cause it to have low shock resistance, but
high damping ability. It has a gray fracture. Gray Iron is by far the oldest and most common form of cast
iron. As a result, it is assumed by many to be the only form of cast iron and the terms "cast iron" and
"gray iron" are used interchangeably. Cast iron containing graphite in flake form and typically consisting
of 2 to 4 percent carbon and 1 to 3 percent silicon. Gray iron is widely used for engine components in
automobiles and trucks.
Gray Iron Melting: The process of melting gray iron, especially as it is done in a foundry on a commercial
scale.
Green Sand: Natural sands combined with water and organic additives, such as clay, to proper
consistency for creating molds.
A-5
-------
Version 1
Draft
Appendix A
Green Sand Core: A core that is made of molding sand but not baked.
Grinding: Removing gate stubs, fins, and other projections on castings by an abrasive wheel.
Holding Furnace: Usually a small furnace for maintaining molten metal at the proper pouring
temperature, and which is supplied from a large melting unit.
Holding Ladle: Heavily lined and insulated ladle in which molten metal is placed until it can be used.
Hot Box Process: Method of making and curing cores within a heated corebox. To form and cure the core,
the corebox is heated to approximately 500 degrees F. The sand used in this process contains a catalyst
which hardens the binders in the core upon contact with the hot corebox. Complete curing while the core
is still in the box results from the residual heat in the core, eliminating the need for conventional dryers or
ovens. Frequently, cores created with the Hot Box process are shell cores.
Indirect-Arc Furnace: An AC, Alternating Current, electric-arc furnace in which the metal is not one of
the poles. An electric furnace in which the arc is struck between two horizontal electrodes, heating the
metal charge by radiation.
Induction Furnace: An AC melting furnace which utilizes the heat of electrical induction.
Investment Casting: A pattern casting process in which a wax or thermoplastic pattern is used. The
pattern is invested (surrounded) by a refractory slurry. After the mold is dry, the pattern is melted or
burned out of the mold cavity, and molten metal is poured into the resulting cavity.
Investment Casting Process: A pattern casting process in which a wax or thermoplastic pattern is used.
The pattern is invested (surrounded) by a refractory slurry. After the mold is dry, the pattern is melted or
burned out of the mold cavity, and molten metal is poured into the resulting cavity.
Investment Molding: Method of molding using a pattern of wax, plastic, or other material which is
"invested" or surrounded by a molding medium in slurry or liquid form. After the molding medium has
solidified, the pattern is removed by subjecting the mold to heat, leaving a cavity for reception of molten
metal. This is also called the lost-wax process or precision molding.
Iron: A metallic element, mp 1535°C (2795°F). Also irons that do not fall into the steel categories, such
as Gray Iron, Ductile Iron, Malleable Iron, White Iron, Ingot, and Wrought Iron.
Knock Out: To remove sand and casting from a flask.
Ladle: Metal receptacle frequently lined with refractories used for transporting and pouring molten metal.
Different types of ladles include hand bull, crane, bottom-pour, holding, teapot, shank, lip-pour.
Lead: One of the elements; its chemical symbol is Pb. Its formula weight is 207.2 and melting point is
327.5°C.
Manganese: One of the elements; its chemical symbol is Mn. Its formula weight is 54.93; specific gravity
7.2, and melting point is 1260°C. Metallic manganese is used in the nonferrous industry both as a
deoxidizing agent and as an essential constituent to improve physical properties of certain alloys.
Melting Rate: Amount of metal melted in a given period of time, usually one hour.
Mill Sale: Iron oxide scale formed on steel during hot working processes, cooled in air.
A-6
-------
Version 1
Draft
Appendix A
Mold: Normally consists of a top and bottom form, made of sand, metal, or any other investment material
which contains the cavity into which molten metal is poured to produce a casting of definite shape and
outline.
Mold Coating: Coating to prevent surface defects, i.e., metal penetration and improve casting finish.
Nickel: An element used for alloying iron and steel as well as nonferrous metals; melting point 1455°C
(2651°F). Nickel is also a base metal for many casting alloys resistant to corrosion and high temperature
oxidation. Nickel's chemical symbol is Ni. Its formula weight is 58.69, the specific gravity is 8.90, and
nickel's melting point 1,452°C.
Nobake Binder: A synthetic liquid resin sand binder that hardens completely at room temperature,
generally not requiring baking, used in the Cold Setting process.
No-Bake Process: Molds/cores produced with a resin bonded air setting sand. Also known as the air set
process because molds are left to harden under normal atmospheric conditions.
Particulate Matter: In air pollution control, solid or liquid particles, except water, visible with or without a
microscope, that make up the obvious portion of smoke.
Pattern: An original used as a form to produce duplicate pieces. Pattern dimensions are slightly enlarged
to counteract the shrinkage of the casting as it solidifies and cools in the mold. Although patterns can be
made in one piece, a complicated casting may consist of two or more parts. The pattern may be made out
of wood, plastic, metal, or other material.
Permanent Mold: A long-life mold into which metal is poured by gravity. It is used repeatedly to produce
many castings from the same mold. It is not an ingot mold.
Pouring: Filling the mold with molten metal. Transferring the molten metal from the furnace to the ladle,
ladle to ladle, or ladle into the molds.
Preheating: A general term for heating material, as a die in die casting, as a preliminary to operation, to
reduce thermal shock and prevent adherence of molten metal.
Psi: Abbreviation for pounds per square inch.
Reverberatory Furnace: Melting unit with a roof arranged to deflect the flame and heat toward the hearth
on which the metal to be melted rests.
Sand: In metal casting, a loose, granular material high in Si02, resulting from the disintegration of rock.
The name sand refers to the size of grain and not to mineral composition. Diameter of the individual
grains can vary from approximately 6 to 270 mesh. Most foundry sands are made up principally of the
mineral quartz (silica). Reason for this is that sand is plentiful, refractory, and cheap; miscellaneous sands
include zircon, olivine, chromite, CaC03, black sand (lava grains), titanium minerals and others.
Sand Mulling: A method of evenly distributing the bond around the sand grain by a rubbing action.
Scrap Metal: Metal to be remelted; includes scrapped machinery fabricated items such as rail or structural
steel and rejected castings (metal to be re-melted, castings that have to be re-melted).
Sea Coal: Term applied to finely ground bituminous coal which is mixed with sands for foundry uses.
Selenium: A metalloid melting at 220°C (428°F) added to stainless steel to improve machinability.
A-7
-------
Version 1
Draft
Appendix A
Separator: A mechanical unit which separates or grades ground materials into constituent parts, used in
the foundry to remove fines from the system sand and dust from the air.
Shakeout: The process of separating the solidified casting from the mold material. The stage in the casting
process where the sand from the mold is cleaned off of the newly formed castings through vigorous
vibration.
Shakeout Machinery: Equipment for mechanical removal of castings from molds.
Shell Process: Process in which clay-free silica sand coated with a thermosetting resin or mixed with resin
is placed on a heated metal pattern for a short period of time to form a partially hardened shell. The bulk
of the sand mixture inside the resulting shell is removed for further use. The pattern and shell are then
heated further to harden or polymerize the resin-sand mix, and the shell is removed from the pattern.
Frequently, shell cores are made using the Hot Box process.
Shot: Metallic abrasive commonly used for cleaning casting surfaces. In die-casting, it is the phase of the
die-casting cycle when molten metal is forced into the die.
Shotblasting (Shot peening): Casting cleaning process employing a metal abrasive (grit or shot) propelled
by centrifugal or air force.
Silica: Silicon dioxide, Si02, occurring in nature as quartz, opal, etc. Molding and core sands are impure
silica. The prime ingredient of sand and acid refractories.
Silica Sand: Sand with a minimum silica content of 95% used for forming casting molds.
Slag: A fused nonmetallic material used to protect molten metal from the air and to extract certain
impurities. The nonmetallic covering on molten metal resulting from the combination of impurities in the
initial charge like ash from fuel, and any silica and clay eroded from the refractory lining. It is skimmed
off prior to pouring the metal.
Split Pattern: A pattern that is parted for convenience in molding.
Sprue: A vertical passageway that takes the molten metal from the pouring basin to the runner.
Steel: An alloy of iron and carbon, containing no more than 1.74% carbon. It must be malleable at some
temperature while in the as-cast state.
Tap: To withdraw a molten charge from the melting unit.
Tap Hole: Opening in a furnace through which molten metal is tapped into the forehearth or ladle.
Temperature: Degree of warmth or coldness in relation to an arbitrary zero measured on one or more of
accepted scales, as Centigrade, Fahrenheit, etc.
Temperature, Holding: Temperature above the critical phase transformation range at which castings are
held as a part of the heat treatment cycle. The temperature maintained when metal is held in a furnace,
usually prior to pouring.
Temperature, Pouring: The temperature of the metal as it is poured into the mold.
Transfer Ladle: A ladle that may be supported on a monorail or carried in a shank and used to transfer
metal from the melting furnace to the holding furnace or from furnace to pouring ladles.
A-8
-------
Version 1
Draft
Appendix A
Trimming: Removing fins, gates, etc. from castings.
Vanadium: A white, hard, metallic element, mp 1800°C (3272°F), used as an alloy in iron and steel; a
powerful carbide stabilizer and deoxidizer.
Wax Pattern: A precise duplicate, allowing for shrinkage, of the casting and required gates, usually
formed by pouring or injecting molten wax into a die or mold. Wax molded around the parts to be welded
by a termite welding process.
Wet Scrubber (Gas Washer): In air pollution control, a liquid (usually water) spray device for collecting
pollutants in escaping foundry gases.
A-9
-------
Version 1
Draft
Appendix B
Appendix B:
Development of Emission Factors for Section 3 Melting Operations
Particulate matter (PM) test data reported in the literature, U.S. Environmental Protection Agency (EPA)
source test data, and test data reported by the industry in response to EPA's detailed information
collection request (ICR) were compiled, and emission factors based on tons of metal melted (or tons of
metal poured) were calculated (see Appendix C of the Iron and Steel Foundries BID, U.S. EPA, 2002). A
summary of the PM emission factor data for melting operations was provided in the BID (see Table 5-2,
U.S. EPA, 2002); this table, with additional rows for AP-42 factors, is presented as Table B-l in this
appendix.
The summary of data in Table B-l indicates that there is significant variability among the reported data.
Furthermore, there appear to be some discrepancies between the different sources. For example, most of
the ICR test data suggested that the PM emission factors for controlled wet scrubbers are less than those
reported in AP-42 (overall cited as U.S. EPA, 1995, in the BID, but the iron foundry chapter was updated
in 2003), but the baghouse catch data suggests that the uncontrolled emission factor for cupolas is
understated in AP-42. In general, these data were considered along with the expected emission control
efficiency for specific control devices (as reported in Table B.2-3 in Appendix B-2 of AP-42, which is
also provided as Table 3-4 in Chapter 3 of this Emission Protocol document) to develop recommended
emission factors. This appendix provides documentation of some of the specific comparisons and
analyses performed to provide the recommended emission factors provided in Chapter 3 of this Emission
Protocol document.
Table B-1. Summary of PM Emission Factors for Melting Furnace Operations
Emission category/
source of data
Basis of
reported values
Range of
emissions
factors,
lb/ton
Median
emissions
factor,
lb/ton
Average
emissions
factor,
lb/ton
Cupolas controlled with wet scrubbers
GM - Saginaw (U.S. EPA,
1999b)
4 Run EPA source test
0.038-0.21
0.110
0.117
ICR PM Tests
11 Source tests
0.090- 1.46
0.56
0.580
AP-42 (U.S. EPA, 2003)
4 Values for different scrubber types
0.8**-5.0
3.0
Cupolas controlled with fabric filters
Waupaca - Tell City
(U.S. EPA, 1999a)
2 of 3 Run EPA source test
0.010-0.017
0.014
ICR PM Tests
3 Source tests
0.030-0.082
0.077
0.063
AP-42 (U.S. EPA, 2003)
As reported
0.70
0.70
Cupolas uncontrolled (or prior to controls)
Waupaca - Tell City
(U.S. EPA, 1999a)
3 Run EPA source test
3.45-9.7
7.7
7.0
GM -Saginaw (U.S. EPA,
1999b)
4 Run EPA source test
3.6-4.9
4.1
4.3
ICR - Baghouse catch
Data for 17 cupola
8.14-64.1
24.0
26.1
B-l
-------
Version 1
Draft
Appendix B
Emission category/
source of data
Basis of
reported values
Range of
emissions
factors,
lb/ton
Median
emissions
factor,
lb/ton
Average
emissions
factor,
lb/ton
Kearney (1971)
Data for 24 cupola
7.5-66.3
21.9
30.2
AP-42 (U.S. EPA, 2003)***
As reported
13.8
EAF melting controlled with fabric filters
ICR PM Tests
4 Source tests
0.037-0.56
0.15
0.22
EAF - BID (EPA, 1980)
Data for 11 EAF
0.052-0.69
0.15
0.23
AP-42 (EPA, 2003)***
As reported
0.4
EAF melting uncontrolled
ICR PM Tests
1 Source test (3 runs)
20.2-25.9
23.9
25.7
ICR - Baghouse catch
Data for 13 EAF
3.3-29.5
8.4
11.0
Kearney (1971)
Data for 19 EAF
4.0-40.0
12.7
13.8
AP-42 (EPA, 2003)**
As reported
12.7
AP-42 (EPA, 1995)**
As reported
4.-40.
13
EAF charging & tapping uncontrolled
EAF - BID (EPA, 1980)
As reported
1.4 iron, 1.6
steel
1.4
1.6
EAF steel - BID (EPA, 1983)
As reported
1.6-2.0
1.8
Induction furnace with PM control
ICR PM Tests
5 Source tests
0.080-0.67
0.13
0.30
AP-42 (EPA, 2003)
As reported
0.20
0.20
Induction furnaces uncontrolled
ICR PM Tests
2 Source tests
0.44-8.94
4.7
ICR - Baghouse catch
Data for 8 furnaces
0.33-4.0
1.75
2.0
BCIRA (Shaw, 1982)
Data for 14 furnace tests
0.26-3.3
0.62
0.9
AP-42 (EPA, 2003)***
As reported
0.9
AP-42 (EPA, 1995)***
As reported for total filterable PM
0.1
AP-42 (EPA, 1995)***
As reported for PM10-FIL
0.09
From Table 5-2 in Iron and Steel Foundry BID (U.S. EPA, 2002) unless otherwise noted. Emissions factors selected
for estimating baseline emissions in the BID are presented in bold.
** There is a typographical error in Table 5-2 of the Iron and Steel Foundry BID (U.S. EPA, 2002); the best performing
wet scrubber type had an emission factor of 0.8 (not 0.08) lb/ton metal melted.
*** Not directly included in Table 5-2 of the Iron and Steel Foundry BID (U.S. EPA, 2002). Most of these additional
AP-42 factors were based on the same data set presented by Kearney (1971) or Shaw (1982).
B-2
-------
Version 1
Draft
Appendix B
Development of PM Filterable Emission Factors for Cupola Melting Furnaces
For cupola melting furnaces, AP-42 provides emission factors by particle size for controlled and
uncontrolled cupolas. The PM distribution for controlled units is very different than what would be
anticipated using the control device PM control efficiencies provided in Table B.2-3 in Appendix B-2 of
AP-42 (U.S. EPA, 1996); the primary control device control efficiencies are provided here in Table B-2.
Table B-3 provides a comparison of the emission factors reported in AP-42 to those calculated from the
uncontrolled emission factor on control device emission factors.
Table B-2. Typical Collection Efficiencies of Various Particulate Control Devices3 (%)
^ o
< o
Type of Collector
Collection Efficiency for:
Filterable Particle Size (pm)
0-2.5
2.5-6
6-10
001
Wet scrubber - hi-efficiency
90
95
99
002
Wet scrubber - med-efficiency
25
85
95
003
Wet scrubber - low-efficiency
20
80
90
053
Venturi scrubber
90
95
99
053
Venturi scrubber (AP >30 inches of water)b
95
98
99
053
Venturi scrubber (AP <30 inches of water)b
88
94
99
007
Centrifugal collector - hi-efficiency
80
95
95
008
Centrifugal collector - med-efficiency
50
75
85
009
Centrifugal collector - low-efficiency
10
35
50
010
Electrostatic precipitator - hi-efficiency
95
99
99.5
011
Electrostatic precipitator - med-efficiency
80
90
97
012
Electrostatic precipitator - low-efficiency
70
80
90
016
Fabric filter - high temperature (>250 °F)
99
99.5
99.5
017
Fabric filter - med temperature (180 °F
-------
Version 1
Draft
Appendix B
It is clear that using the uncontrolled emission factors along with the control device collection efficiencies
yield emission estimates that are much lower than those reported for controlled units in AP-42.
However, the calculated emission factors are slightly higher that the emissions reported for controlled
units from source test data collected in the foundries ICR. If the uncontrolled emission factor for
uncontrolled cupolas recommended in the foundries BID (from the baghouse catch data) were used, the
calculated controlled emissions for cupolas would further exceed those observed in the ICR test data.
Therefore, the uncontrolled cupola emission factor from AP-42 of 13.8 lb/ton was selected as the
recommended emission factor for uncontrolled cupolas.
PM distribution data were available for cupola and electric arc furnace (EAF) melting furnaces. A single
distribution for melting furnaces were developed from these distributions, based on the average of the
relative fraction of PM that is less than 10 nm and the fraction less than 2.5 nm. Table 12.10.9 of the
Gray Iron Foundry section (12.10) of AP-42 indicates that PMi0 emissions account for 90% of the
uncontrolled cupola furnace PM emissions; it also indicates that PMi0 emissions account for 90% of the
uncontrolled EAF PM emissions. Table 12.10.9 of the Gray Iron Foundry section (12.10) of AP-42
indicates that PM2 5 is 84.0 percent for uncontrolled cupolas and 61.6 percent (interpolated between 2.0
|im and 5.0 |im size distribution) for uncontrolled EAF. Therefore, the average distribution was
determined to be 90% for PM10 emissions and 70% for PM2 5. This average distribution was applied for
all furnace types because it provided better agreement between the ICR test data results and the calculated
emission factors by particle size.
Due to the variability of PM emission factors for cupolas with venturi scrubbers, separate venturi scrubber
control efficiencies were recommended for venturi scrubbers operating at pressure drops exceeding 30
inches of water and venturi scrubbers operating at pressure drops less than 30 inches of water using the
default venturi scrubber control efficiencies as a guide. Table B-4 shows the revised PM emission factors
for uncontrolled cupolas based on 70% of the PM being less than 2.5 nm and the separate PM control
efficiencies for venturi scrubbers. As seen in Table B-4, the calculated emission factors for baghouses did
not change appreciably. The combination of change in particle size and control efficiencies yields an
average emission factor for high-energy venturi scrubbers to agree well with the ICR test average, while
the lower-energy wet scrubber factor agrees well with the lower performing wet scrubbers from the ICR
test data and agrees reasonably well with the cupola venturi scrubber emission factor in AP-42. Together,
these observations were used to provide the emission factors and methodologies recommended in the
Protocol document for cupola melting furnaces.
Table B-4. Recommended PM Emission Factors for Melting Furnace Operations
Emission category
PM-FIL
emissions
factor, lb/ton
PM10-FIL
emissions
factor, lb/ton
PM2.5-FIL
emissions
factor, lb/ton
Cupolas uncontrolled (recommended in Protocol)
13.8
12.4
9.7
Cupolas controlled by baghouse (calculated a)
0.11
0.11
0.10
Cupolas controlled by venturi scrubber AP>30" (calculated a)
0.60
0.60
0.58
Cupolas controlled by venturi scrubber AP<30" (calculated a)
1.44
1.44
1.39
a Calculated from the recommended uncontrolled emission factors and the control efficiencies in Table B-2; assumes
100% control efficiency for PM greater the 10 urn in diameter.
Development of PM Filterable Emission Factors for Other Melting Furnaces
The emission factors for uncontrolled EAF melting as presented in Table B-lare fairly consistent. It is
not always clear if the reported melting emission factors cover all phases of the EAF melting process,
including charging and tapping. As charging and tapping emissions may have different levels of capture
and control than the melting cycle, separate emission factors were desired for these processes. Emission
B-4
-------
Version 1
Draft
Appendix B
factors for charging and tapping (1.8 lbs/ton) are available from the EAF BID (U.S. EPA, 1983). It is not
always clear if the baghouse catch data or the AP-42 emission factors include charging and tapping
emissions, but the EAF emissions factor from these data averaged 11 lb/ton. The AP-42 emission factors
for EAFs were 12.7 and 13 lb/ton for iron foundries and steel foundries, respectively (U.S. EPA, 2003
and 1995). Based on these data, a single emission factor of 12.8 lb/ton for all EAF, regardless of the
foundry type, appears reasonable. The AP-42 foundry emission factors do not indicate a charging/tapping
emission factor for EAF melting furnaces; however, the sum of the emission factor for charging and
tapping and for uncontrolled EAF melting from the baghouse catch data is essentially equal to the
emission factors reported in AP-42 (1.8 lb/ton +11 lb/ton =12.8 lb/ton). As such, this Protocol document
recommends the use of 1.8 lb/ton for EAF charging and tapping and 11 lb/ton for direct melting.
Combined, these emission factors represent the total emissions from and uncontrolled EAF and are
consistent with the overall emissions from an EAF as suggested by the AP-42 emission factors. .
While there is significant variability in the emissions data for electric induction furnaces (EIF), the more
recent source test data and baghouse catch data suggest that the AP-42 emission factor for EIF developed
from Shaw (1982) is too low. Consequently, the baghouse catch data from the ICR was used to estimate
the overall filterable PM emissions from EIF as 2 lb/ton. Because charging and tapping emissions may
have different controls than during the primiary melting cycle, separate charging and tapping emissions
factors were desired, but none were directly available. While the charging and tapping emissions for EIF
may be expected to be similar to that for EAF, the overall emission factors reported for EIF are often less
than the EAF charging and tapping emission factor. Again, it is not always clear what steps of the
melting cycle are included in the baghouse catch data or emissions test data. A 1 lb/ton emission factor
split was considered for each process (1 lb/ton for charging and tapping and 1 lb/ton for melting) based on
the average emission factor from Shaw (1982), assuming this represented the pure melting cycle only, and
the average emission factor from the baghouse catch data, assuming this represented the total melting,
charging, and tapping emissions. However, when considering the relative duration of these different
cycles, a 1 to 1 split of emissions was expected to understate the relative magnitude of the melting
emissions compared to the charging and tapping emissions. Unfortunately, no data are available to
specifically determine the appropriate proportion between EIF charging and tapping emissions and EIF
melting emissions. When considering the cycle times, a 0.5 lb/ton emission factor for EIF
charging/tapping and a 1.5 lb/ton emission factor for EIF melting was recommended for use in the
Protocol.
Reverberatory furnaces are not that prevalent for iron and steel foundries, and there are limited data for
these furnaces. The emission factor of 2.1 lb/ton for an uncontrolled reverberatory furnace from
Table 12.10-3 of the Gray Iron Foundry Section (12.10) of AP-42 was used.
As discussed previously, PM distribution data were available for cupola and EAF melting furnaces. A
single distribution for melting furnaces were developed from these distributions, based on the average of
the relative fraction of PM that is less than 10 nm and the fraction less than 2.5 nm. Table 12.10.9 of the
Gray Iron Foundry section (12.10) of AP-42 indicates that PMi0 emissions account for 90% of the
uncontrolled cupola furnace PM emissions; it also indicates that PM10 emissions account for 90% of the
uncontrolled EAF PM emissions. Table 12.10.9 of the Gray Iron Foundry section (12.10) of AP-42
indicates that PM2.5 is 84.0 percent for uncontrolled cupolas and 61.6 percent (interpolated between 2.0
|im and 5.0 |im size distribution) for uncontrolled EAF. Therefore, the average distribution was
determined to be 90% for PM10 emissions and 70% for PM2 5. This proportional distribution was applied
to all types of melting furnaces and all cycles of the melting cycle, including charging/tapping and
melting emissions.
B-5
-------
Version 1
Draft
Appendix B
Development of PM Condensable Emission Factors for Melting Furnaces
For each melt furnace, use an emission factor of 0.05 lb/ton for PM-CON. The emission factor 0.05 was
derived from PM condensable data collected from the Foundry Information Collection Request. Table B-
5 summarizes the available PM-CON data used to develop the PM-CON emission factor for melting
furnaces in the protocol document. The cupola baghouse operated by WI-35 was a special low-
temperature, high efficiency, horizontal baghouse that set the new source "MACT floor" for cupolas and
is not expected to be indicative of typical emissions from cupolas. Excluding this data point, the average
PM-CON emission factor for cupola would be 0.04 lb/ton. Given the variability in the emissions for
different units and the limited data set for some types of units, a single emission factor of 0.05 lb/ton
averaged across all of the data is recommended for all melting furnaces.
Table B-5. Summary of PM Condensable Data for Melting Furnaces
Docket Item*
ICR Return ID
Test
Control Type
Average PM-CON lb/ton
Cupola
ll-D-82
WI-35
Avg of 3 tests
Baghouse
0.00343
ll-D-46
IA-19
Feb-98
Baghouse
0.0614
ll-D-65
NC-05
Feb-00
Baghouse
0.0184
AVG
Baghouse
0.028
Electric Induction Furnace
ll-D-56
MN-07
Aug-96
Baghouse
0.053
ll-D-57
MN-12
Mar-95
Baghouse
0.0633
ll-D-57
MN-12
May-96
Baghouse
0.0203
ll-D-75
TX-11
Oct-93
Baghouse
0.086
AVG
Baghouse
0.056
Electric Arc Furnace
ll-D-44
IA-09
Aug-96
Baghouse
0.044
ll-D-76
TX-19
Jan-95
Baghouse
0.14
ll-D-55
MN-03
May-93
Baghouse
0.019
AVG
Baghouse
0.067
Average Emission Factor for Melting Furnaces
0.05
*Docket No.: EPA-HQ-OAR-2002-0034
For emissions from charging and tapping use the emission factor of 0.01 lb/ton for PM-CON. The
emission factor 0.01 was derived considering the average proportion of PM filterable emissions for
charging/tapping versus melting as developed for EAF and EIF furnaces.
Development of Metal HAP Emission Factors for Melting Furnaces
Default metal HAP concentrations for emitted PM were determined based on data collected during the
foundries ICR. For facilities that performed a multi-metals emission determination concurrent with a PM
emissions determination, the metals mass emission rate was divided by the PM mass emissions rate to
determine a metals concentration for the emitted PM. Table B-6 provides a summary of the metal HAP
emissions as a percent of filterable PM emissions determined from the ICR source tests. These average
concentrations were used as the primary source for the filterable PM metal HAP concentrations.
Tables B-7 and B-8 provides a summary of the PM emission source test conducted by the EPA. During
these tests, metal concentration and PM emissions were determined before and after the control device.
These data were used to assess the concentration of non-PM associated HAP emissions. Metal fumes that
B-6
-------
Version 1
Draft
Appendix B
remain gaseous are expected to pass through the PM control device (particularly high-temperature
baghouses) as are condensable PM. The additional uncondensed metal HAP emissions needed to yield
the measured outlet pollutant emissions were determined and used to determine a metal HAP content for
condensable PM. For example, in the GM source test, the outlet emission rate for manganese was
0.13 lb/hr greater than would be predicted based on the PM-FIL control efficiency. The PM-CON
emission rate is estimated to be 2.1 lbs/hr (using the default PM-CON factor of 0.05 and the site specific
melting rate). Therefore, for this test, it appears that manganese comprises approximately 6 percent
(0.13/2.1) of the PM-CON. Although there are very limited data by which to assess the metal HAP
content of PM-CON emissions, it is important to consider these emissions, particularly for mercury and
other volatile metals.
Table B-6. Summary of Metal HAP Emissions as a Percent of Filterable PM Emissions
Docket
Item*
Furnace Type
% Pb
% Mn
% Cd
% Cr
%Ni
%Hg
% Other
11-1-27
Cupola - BH
1.99%
1.73%
0.12%
11-1-20
EAF
8.64%
0.69%
Charlotte
Pipe
Cupola BH
0.26%
1.08%
0.06%
0.01%
0.004%
11-1-49
Cupola BH
2.04%
0.06%
0.06%
0.09%
0.57%
11-1-24
Cupola BH
2.35%
11-1-30
Cupola
0.12%
6.20%
0.001%
0.02%
0.01%
0.03%
ll-D-114
EAF
0.22%
0.07%
ll-D-100
Cupola Gray Iron
0.36%
0.01%
0.13%
0.51%
0.19%
ll-D-100
Cupola Nodular
0.19%
0.01%
0.11%
0.26%
0.12%
ll-D-50
Cupola
0.21%
1.76%
0.01%
ll-A-32
Cupola WS
1.10%
7.28%
0.001%
0.08%
0.01%
0.04%
0.01%
ll-A-30
Cupola BH
1.03%
0.59%
0.14%
0.09%
0.06%
1.49%
0.10%
ll-D-60
Cupola
3.98%
0.12%
0.02%
0.08%
ll-D-80
Cupola 1991
1.28%
0.03%
0.004%
ll-D-80
Cupola 1994
0.06%
2.08%
0.02%
0.03%
AL-11
EAF
0.11%
2.68%
0.02%
0.29%
0.74%
0.14%
0.01%
M1-13
Cupola WS
0.29%
2.18%
IN-12
Cupola WS
1.62%
3.33%
0.93%
0.03%
0.01%
0.01%
IN-12
Preheater
cyclone
0.70%
1.35%
0.20%
0.51%
IN-12
EIF
0.25%
0.74%
OH-11
Cupola WS
2.20%
II-I-63
Pouring
0.42%
2.01%
0.01%
0.01%
0.28%
1.78%
Average
1.01%
2.70%
0.13%
0.08%
0.23%
0.43%
0.23%
*Docket No.: EPA-HQ-OAR-2002-0034
B-7
-------
Version 1
Draft
Appendix B
Table B-7. Summary of Metal HAP Emissions Measured during EPA Source Test:
Cupola Baghouse Waupaca, IN (U.S. EPA, 1999a)
Pollutant
Emission Rate (Ibs/hr)
% of PM
Run 1
Run 2
Run 3
Average
Emissions prior to Baghouse Control
PM
143
471
353
322
N/A
Antimony
0.0377
0.0611
0.0555
0.0515
0.02
Arsenic
0.0083
0.0109
0.0135
0.0109
0.003
Barium
0.0625
0.1254
0.1076
0.0985
0.03
Beryllium
0.0001
0.0007
0.0004
0.0004
0.0001
Cadmium
0.0608
0.1656
0.1295
0.1186
0.04
Chromium
0.0548
0.1071
0.0878
0.0833
0.03
Cobalt
0.0015
0.0019
0.0020
0.0018
0.001
Lead
3.5108
8.7429
6.3628
6.2055
1.92
Manganese
3.6702
9.1579
7.1378
6.6553
2.06
Mercury
0.0021
0.0082
0.0026
0.0043
0.001
Nickel
0.0137
0.0197
0.0215
0.0183
0.01
Phosphorus
0.0804
0.1204
2.6324
0.9444
0.29
Selenium
0.0032
0.0056
0.0050
0.0046
0.001
Silver
0.0006
0.0024
0.0009
0.0013
0.0004
Thallium
0.0031
0.0073
0.0057
0.0054
0.002
Zinc
16.9502
48.8963
33.3442
33.0636
10.25
Emissions after Baghouse Control
PM
0.71
ME
0.45
0.58
N/A
Antimony
0.0003
0.0003
0.0003
0.0003
0.05
Arsenic
0.00004
0.0001
0.0001
0.0001
0.01
Barium
0.0030
0.0031
0.0033
0.0033
0.56
Beryllium
ND
ND
ND
ND
ND
Cadmium
0.0017
0.0001
0.0005
0.0010
0.17
Chromium
0.0005
0.0005
0.0006
0.0006
0.10
Cobalt
ND
ND
ND
ND
ND
Lead
0.0096
0.0028
0.0056
0.0068
1.17
Manganese
0.0020
0.0029
0.0053
0.0033
0.57
Mercury
0.0121
0.0086
0.0054
0.0096
1.65
Nickel
0.0004
0.0003
0.0004
0.0004
0.06
Phosphorus
0.0048
0.0046
0.0041
0.0047
0.81
Selenium
0.0002
0.0003
0.0003
0.0003
0.05
Silver
0.00002
0.00002
0.0001
0.00004
0.01
Thallium
ND
ND
ND
ND
ND
Zinc
0.0083
0.0135
0.0368
0.0187
3.20
B-8
-------
Version 1
Draft
Appendix B
Table B-8. Summary of Metal HAP Emissions Measured during EPA Source Test:
Cupola Wet Scrubber at General Motors Corporation (U.S. EPA, 1999b)
Pollutant
Emission Rate (Ibs/hr)
%of PM
Run 1
Run 2
Run 3
Run 4
Average
Emissions prior to Wet Scrubber Control
PM
139
169
186
215
177
N/A
Antimony
0.008
0.003
0.015
0.018
0.011
0.01
Arsenic
0.004
0.005
0.005
0.006
0.005
0.00
Barium
0.041
0.070
0.052
0.070
0.058
0.03
Beryllium
ND
0.00015
ND
ND
0.0002
0.00
Cadmium
0.011
0.015
0.016
0.013
0.014
0.01
Chromium
0.063
0.065
0.054
0.051
0.058
0.03
Cobalt
0.003
0.003
0.002
0.002
0.003
0.001
Lead
1.281
0.840
1.740
1.800
1.415
0.80
Manganese
8.210
9.160
7.550
8.090
8.253
4.66
Mercury
0.001
0.001
0.007
0.001
0.003
0.001
Nickel
0.004
0.003
0.003
0.002
0.003
0.002
Phosphorus
0.055
0.355
0.234
0.230
0.218
0.12
Selenium
0.003
0.004
0.003
0.003
0.003
0.002
Silver
0.001
0.002
0.002
0.001
0.002
0.001
Thallium
0.004
0.004
0.003
0.004
0.004
0.002
Zinc
10.340
10.470
13.570
17.500
12.970
7.32
Emissions after Wet Scrubber Control
PM
8.08
7.61
1.71
1.71
4.78
N/A
Antimony
1.2E-03
7.2E-04
6.1E-04
5.6E-04
7.8E-04
0.02
Arsenic
6.4E-04
3.3E-04
ND
ND
4.9E-04
0.01
Barium
5.8E-03
5.0E-03
2.4E-03
1.6E-03
3.7E-03
0.08
Beryllium
3.7E-05
3.0E-05
ND
ND
3.3E-05
0.00
Cadmium
9.3E-04
1.1E-03
2.6E-04
1.4E-04
6.1E-04
0.01
Chromium
6.9E-03
4.9E-03
1.8E-03
1.3E-03
3.7E-03
0.08
Cobalt
ND
2.8E-04
1.4E-04
ND
2.1E-04
0.004
Lead
9.0E-02
7.6E-02
3.0E-02
1.5E-02
5.3E-02
1.10
Manganese
7.0E-01
5.2E-01
1.0E-01
7.0E-02
3.5E-01
7.28
Mercury
1.1E-03
1.0E-03
3.6E-03
1.2E-03
1.7E-03
0.04
Nickel
1.1E-03
5.8E-04
4.2E-04
ND
6.9E-04
0.01
Phosphorus
3.4E-02
3.1E-03
6.4E-03
4.0E-03
1.2E-02
0.25
Selenium
6.8E-04
4.4E-04
3.3E-04
3.4E-04
4.5E-04
0.01
Silver
5.5E-04
3.8E-04
1.4E-04
1.2E-04
3.0E-04
0.01
Thallium
4.7E-04
2.5E-04
8.1E-05
1.0E-04
2.2E-04
0.005
Zinc
7.3E-01
6.1E-01
1.8E-01
1.7E-01
4.2E-01
8.82
B-9
-------
Version 1
Draft
Appendix B
Development of Chromium Hexavalent Distribution of Total Chromium
To determine the distribution of chromium hexavalent emissions the values reported in Exhibit D-l of
"An Overview of Methods for EPA's National-Scale Air Toxics Assessment" for Standard Classification
Codes from Iron and Steel melting operations were used (U.S. EPA, 2011). For melting furnaces from
iron foundries the percent chromium hexavalent emissions of total chromium emissions was reported as 3
percent. For steel foundries the percent chromium hexavalent emissions of total chromium emissions was
reported as 12 percent. For all other operations at iron and steel foundries the percent of chromium
hexavalent emissions of total chromium emissions was reported as 3 percent. Table B-9 summarizes the
percent chromium hexavalent of total chromium emissions for the applicable SCC codes for iron and steel
foundries from Exhibit D-l.
Table B-9. Percent Cr (VI) of Total Chromium Emissions for Iron and Steel Foundry SCC
Type of Foundry
Source
SCC
% Cr (VI)
Iron
Cupola
3-04-003-01
3
Iron
Reverberatory
3-04-003-02
not in table
Iron
EIF
3-04-003-03
3
Iron
EAF
3-04-003-04
3
Iron
Scrap and Charge handling, heating
3-04-003-15
3
Iron
Pouring, Cooling
3-04-003-18
3
Iron
core making baking
3-04-003-19
not in table
Iron
Magnesium treatment
3-04-003-21
not in table
Iron
Refining
3-04-003-22
3
Iron
Shakeout
3-04-003-31
3
Iron
Cleaning and finishing
3-04-003-40
3
Iron
sand handling
3-04-003-50
3
Steel
EIF
3-04-007-05
12
Steel
EAF
3-04-007-01
12
Steel
Pouring and casting
3-04-007-08
3
Steel
Sand grinding/handling in mold and core making
3-04-007-06
3
Steel
Core ovens
3-04-007-07
not in table
Steel
Casting cleaning
3-04-007-11
3
Steel
Charge handling
3-04-007-12
3
Steel
Casting cooling
3-04-007-13
3
Steel
Open hearth
3-04-007-02
not in table
Steel
Open hearth oxygen lanced
3-04-007-03
not in table
Development of Mercury Emission Factors
Automobile scrap has more potential for mercury emissions because of the mercury switches contained in
old cars. There are other sources of mercury in scrap metals, but mercury from switches found in
automobile scrap has the most potential for mercury emissions. If a facility is using automobile scrap,
they should use the upper emission factor for mercury found in Table 3-6 of the Protocol document.
Facilities not using automobile scrap should use the lower emission factor value for mercury from Table
3-6 of the Protocol document.
B-10
-------
Version 1
Draft
Appendix B
Development of SO2 Emission Factors
Source test data for S02 emissions from cupola melting furnaces were compiled as part of the impact
analyses conducted to support MACT standards for iron and steel foundries. Table B-10 summarizes the
available source test data for S02 emissions from cupola melting furnaces. While the S02 emissions are
expected to vary based on the quantity and sulfur content of the coke used, these data were not generally
known for these tests. Nonetheless, the test data presented in Table B-10 are more recent and are better
documented than the emission factors reported in AP-42 (U.S. EPA, 2003). Therefore, the average
emission factors from these source tests are recommended for use (as presented in Table 3-5 of the
Protocol document) rather than the emission factors (dependent on sulfur content) presented in AP-42
(U.S. EPA, 2003).
Table B-10. Summary of S02 Emissions for Cupola Melting Furnaces (SCC 30400301)
Plant - cupolas with baghousesa
S02 (lb/ton)
US Pipe (NJ-03) 1991 - [ll-D-58]
0.059
US Pipe (NJ-03) 1997 [ll-D-58]
0.072
Waupaca Tell City (IN-34) 1997 [ll-D-41]
0.11
Waupaca Plant 1 (Wl-01)1998 [G. Mosher e-mail]
0.115
Grede Reedsburg (WI-35) 1998 [ll-D-117]
0.18
Charlotte Pipe (NC-05) 2000 [II-I-70]
0.23
Grede Reedsburg (WI-35) 2000 [II-I-731
0.32
Average
0.155
Standard Deviation
0.094
Plant - cupolas with wet scrubbers3
S02 (lb/ton)
Atlantic States (NJ-04) [ll-D-59]
0.0006
CMI Cast Parts (MI-13) 1997 [ll-D-49]
<0.0015
Briggs & Stratton (WI-24) [ll-D-80]
0.0015
Griffin Pipe (NJ-05) 1997 [ll-D-60]
0.002
Waupaca Plant 2/3 (WI-42) 1995 [ll-D-83]
0.0023
Waupaca Plant 1 (WI-01) 1998 [G. Mosher e-mail]
0.006
Waupaca Plant 2/3 (WI-42) 1997 [ll-D-83]
0.0097
Great Lakes Casting (MI-17) 1996 [ll-D-50]
<0.011
Waupaca Plant 2/3 (WI-42) 1994 [ll-D-83]
0.011
Charlotte Pipe (NC-05) 1994 [ll-D-61]
0.015
Waupaca (WI-01) 1994 [ll-D-78]
<0.02
LaGrange Foundry (MO-05) 1993 [ll-D-100]
0.026
Wrightsville (PA-34) 1995 [Survey response for No. 688]
0.061
GM Saginaw (MI-33) 1995 [ll-D-541
0.12
Average
0.019
Standard Deviation
0.033
altems available in Docket No. EPA-HQ-OAR-2002-0034
References
Kearney, A.T. (1971). Systems analysis of emissions and emissions control in the iron foundry
industry, Vol. /-///. Prepared by A. T. Kearney and Company for EPA Air Pollution
Control Office under Contract No. CPA 22-69-106, NTIS PB 198 348-50, February.
Shaw, F. (1982). Induction furnace emissions. AFS International Cast Metals Journal. June,
pp. 10-27.
B-ll
-------
Version 1
Draft
Appendix B
U.S. EPA (1980). Electric arc furnaces in ferrous foundries - background information for proposed
standards. Office of Air Quality Planning and Standards, Research Triangle Park, NC. EPA-
450/3-80-020a, May, pp. i-x.
U.S. EPA (1983). Electric arc furnaces and argon-oxygen decarburization vessels in steel industry -
background information for proposed revisions to standards. Office of Air Quality Planning and
Standards, Research Triangle Park, NC. EPA-450/3-82-020a, July, pp. i-xvii.
U.S. EPA (1995). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 12.13: Steel foundries. AP-42, Fifth Edition. Office of Air Quality Planning and
Standards, Research Triangle Park, NC.
U.S. EPA (1996). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Appendix B.2 Generalized particle size distributions. AP-42, Fifth Edition. Office of Air
Quality Planning and Standards, Research Triangle Park, NC.
U.S. EPA (1999a). Iron and steel foundries manual emissions testing of cupola baghouse at Waupaca
Foundry in Tell City, Indiana. Office of Air Quality Planning and Standards, Research Triangle
Park, NC. EPA-454/R-99-017A and EPA-454/R-99-017B. June.
U.S. EPA (1999b). Iron and steel foundries manual emissions testing of cupola wet scrubber at General
Motors Corp., Saginaw, Michigan. Office of Air Quality Planning and Standards, Research
Triangle Park, NC. EPA-454/R-99-025A and EPA-454/R-99-025B. July.
U.S. EPA (2002). National emission standards for hazardous air pollutants for iron and steel foundries—
background information for proposed standards. Office of Air Quality Planning and Standards,
Research Triangle Park, NC.
U.S. EPA (2003). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Section 12.10 Gray iron foundries. AP-42, Fifth Edition. Office of Air Quality Planning
and Standards, Research Triangle Park, NC.
U.S. EPA (2011). An overview of methods for EPA's National-Scale Air Toxics Assessment. Office of Air
Quality Planning and Standards, Research Triangle Park, NC.
B-12
-------
Version 1
Draft Appendix C
Appendix C;
Development of Emission Factors for Section 4 Mold and Core Making
Table 4-2 of this Protocol document provides emission factors for specific chemicals contained within
specific binder system components. Table 4-3 provides typical composition of these binder system
components, as well as typical chemical use rates per ton of sand. Together, these data provide a means
to determine default emission factors for each binder system. The general equation to calculate the
default emission factor is as follows:
IN
V1 / %emittedi r\
EmFi = Z. (Qc X Ci'c X 100% ' )
C=1
where:
Em F, = Emission factor for pollutant "i" in a given binder system (lb/ton sand)
Qc = Chemical use rate for component "c" of the binder system from Table C-l (lbs/ton
sand)
Qc = Concentration of pollutant "i" in component "c" of the binder system from Table 4-3
(weight fraction)
%emittedlc = percent of pollutant "i" in component "c" of the binder system emitted to the
atmosphere (see Table 4-2).
Typical chemical use rates were summarized in Appendix B of the Iron and Steel Foundries BID (U.S.
EPA, 2002). The binder use rates used in the development of the default emission factors are provided in
Table C-l, along with the typical concentration and percent emitted as presented in Section 3 of the
Foundries Protocol. Table C-l shows the calculated chemical and component specific emission factors.
For most chemicals, the emission factors calculated in Table C-l were rounded and presented directly in
Table 4-4 of the Protocol. For a few binder systems, a given pollutant is present in more than one binder
component. For example, methanol is used in both the resin and catalyst for fiiran nobake binder systems.
The overall emission factor for methanol for fiiran nobake is the emission factor for the resin (0.252 lb/ton
sand) plus the emission factor for the catalyst (0.972 lb/ton sand) or 0.252 + 0.972 = 1.224 lb/ton sand.
Therefore, the default emission factor for methanol from furan nobake is provided as 1.22 lb/ton sand in
Table 4-4 of the Protocol.
Table C-1. Calculation of Default Emission Factors
(B)
(C)
D=AxBxC
Binder system
Component
(A)
Component
Use Rate
(lb/ton
sand)
HAP
Typical
Concen-
tration
(wt%)
Percent
emitted
(wt%)
Emission
Factor
(lb/ton
sand)
Alkyd oil
Co-reactant
13.5
Methylene phenylene
isocyanate
80
0.001
0.000108
Resin
16.5
Cobalt
ND
0
0
Resin
16.5
Lead
ND
0
0
Acrylic/Epoxy/
Resin
34
Cumene hydroperoxide
ND
0.3
0
so2
Resin
34
Cumene
5
1.5
0.0255
Furan hotbox
Resin
40
Formaldehyde
3
5
0.06
Furan nobake
Resin
16.8
Phenol
1
0.2
0.000336
C-l
-------
Version 1
Draft
Appendix C
(B)
(C)
D=AxBxC
Binder system
Component
(A)
Component
Use Rate
(lb/ton
sand)
HAP
Typical
Concen-
tration
(wt%)
Percent
emitted
(wt%)
Emission
Factor
(lb/ton
sand)
Resin
16.8
Formaldehyde
0.1
2
0.000336
Resin
16.8
Methanol
3
50
0.252
Catalyst
7.2
Methanol
27
50
0.972
Catalyst
7.2
Sulfuric acid
ND
0
0
Furan/SC>2
Resin
16.5
Formaldehyde
2
2
0.0066
Resin
16.5
Methanol
2
50
0.165
Oxidizer
13.5
Dimethyl phthalate
45
50
3.0375
Oxidizer
13.5
Methyl ethyl ketone
2
50
0.135
Furan warmbox
Resin
25.6
Formaldehyde
0.5
5
0.0064
Catalyst
6.4
Methanol
50
100
3.2
Phenolic baking
Part 1
30
Phenol
8
0.5
0.012
Part 1
30
Formaldehyde
1
5
0.015
Phenolic ester
Resin
33
Formaldehyde
0.5
2
0.0033
nobake
Resin
33
Phenol
4
0.2
0.00264
Phenolic ester
Resin
32
Formaldehyde
0.5
2
0.0032
coldbox
Resin
32
Phenol
4
0.2
0.00256
Resin
32
Glycol ethers
0.1
50
0.016
Co-reactant
3
Methanol
27
50
0.405
Phenolic CO2
cure
Resin
30
Diethylene glycol butyl
ether (112-34-5)
1
0.5
0.0015
Resin
30
Ethylene glycol
monophenyl ether (122-
99-6)
1
0.5
0.0015
Phenolic hotbox
Resin
30
Formaldehyde,
2
5
0.03
Resin
30
Phenol
5
0.5
0.0075
Phenolic nobake
Resin
18.4
Phenol
12
0.2
0.004416
(acid catalyzed)
Resin
18.4
Formaldehyde
0.5
2
0.00184
Resin
18.4
Methanol
3
50
0.276
Acid
8.6
Methanol
27
50
1.161
Acid
8.6
Sulfuric acid
ND
0
0
Phenolic
Novolac flake
(hot coating
operations)
Resin
50
Phenol
5.5
0.5
0.01375
Phenolic
Part 1
50
Phenol
2
20
0.2
Novolac liquid
(warm-coating
operations)
Part 1
50
Formaldehyde
0.5
5
0.0125
Part 1
50
Methanol
5
100
2.5
Phenolic
Resin
50
Phenol
5.5
0.1
0.00275
Novolac flake
(resin-coated
Catalyst
10
Ammonia, catalyst
(*Assume ammonia =
40*
50
2
C-2
-------
Version 1
Draft
Appendix C
(B)
(C)
D=AxBxC
Binder system
Component
(A)
Component
Use Rate
(lb/ton
sand)
HAP
Typical
Concen-
tration
(wt%)
Percent
emitted
(wt%)
Emission
Factor
(lb/ton
sand)
sand)
40% of
hexamethylenetetramine)
Phenolic
Part 1
13.75
Formaldehyde
0.1
2
0.000275
urethane nobake
Part 1
13.75
Phenol
6
0.2
0.00165
Part 1
13.75
Xylene I
0.1
16
0.0022
Part 1
13.75
Cumene
0.5
16
0.011
Part 1
13.75
Naphthalene
1
16
0.022
Part 1
13.75
1,2,4-Trimethylbenzene
16
0
Part II
11.25
Methylene phenylene
isocyanate
80
0.001
0.00009
Part II
11.25
Xylene
0.1
16
0.0018
Part II
11.25
Cumene
0.1
16
0.0018
Part II
11.25
Naphthalene
1
16
0.018
Part II
11.25
1,2,4-Trimethylbenzene
0
16
0
Phenolic
Part 1
16.5
Formaldehyde
0.1
2
0.00033
urethane
coldbox
Part 1
16.5
Phenol
6
0.2
0.00198
Part 1
16.5
Xylene
0.1
9
0.001485
Part 1
16.5
Naphthalene
1
9
0.01485
Part 1
16.5
Cumene
0.5
9
0.007425
Part 1
16.5
1,2,4-Trimethylbenzene
0
9
0
Part II
13.5
Methylene phenylene
isocyanate
80
0.001
0.000108
Part II
13.5
Xylene
0.1
9
0.001215
Part II
13.5
Naphthalene
1
9
0.01215
Part II
13.5
Cumene
0.1
9
0.001215
Part II
13.5
Biphenyl
0.1
9
0.001215
Catalyst
3
Triethyl amine or diethyl
amine (uncontrolled)
100
100
3.0
Urea
formaldehyde
Part 1
30
Formaldehyde
1
2
0.006
References
U. S. EPA (2002). National Emission Standards for Hazardous Air Pollutants for iron and steel
foundries—Background information for proposed standards. Office of Air Quality
Planning and Standards, Research Triangle Park, NC.
C-3
-------
Version 1
Draft
Appendix D
Appendix D: Development of Emission Factors
for Pouring, Cooling, and Shakeout
Development of Organic Emission Factors for PCS Operations
The organic emission factors for pouring, cooling and shakeout (PCS) were developed primarily based on
data developed for the Casting Emission Reduction Program (CERP); most of these reports were prepared
by Technikon LLC. Initial testing results for the pre-production foundry are summarized in Table D-l.
Table D-1. Summary of Baseline Emission Results from Pre-production Foundry3
Analyte
Emission Factor (lb/ton metal)
Background
Baseline
Greensand
Baseline
Core
Baseline
Greensand /
Core Baseline
Sum of VOCs
0.0312
0.4722
0.4708
0.8324
Sum of HAPs
0.0249
0.3160
0.3161
0.5424
Benzene
0.0061
0.1244
0.1389
0.2202
Toluene
0.0031
0.0836
0.0324
0.1059
Xylene (Total)
0.0026
0.0620
0.0163
0.0790
Ethylbenzene
0.0005
0.0099
0.0015
0.0115
Naphthalene
0.0011
0.0153
0.0226
0.0113
1-Methylnaphthalene
0.0002
0.0029
0.0052
0.0036
2-Methylnaphthalene
0.0005
0.0055
0.0115
0.0075
Acetaldehyde
0.0087
0.0077
0.0060
0.0096
Formaldehyde
0.0017
0.0015
0.0008
0.0027
Phenol
BDL
0.0046
0.0137
0.0026
o -Cresol
BDL
0.0022
0.0052
0.0047
Aniline
NT
NT
0.0917
0.0533
Hexane
0.0005
0.0210
0.0011
0.0181
Styrene
BDL
0.0024
0.0016
0.0053
a As reported in by CERP (1999b).
Note that "Sum of VOCs" were the sum of individual analytes and, depending on the analytes tested, may
or may not be a complete measure of all VOC. CERP also performed "baseline" emission testing at their
production foundry and at a foundry in Mexico. These facilities were greensand foundries with
chemically-bonded cores. The Mexico study (CERP, 1999a) used primarily phenolic hot box cores, while
the CERP production foundry used phenolic urethane cold box cores. Table D-2 summarizes the test
data from these three studies. The Mexico study focused on HAPs and did not analyze for or report data
for as many non-HAP VOCs as compared to the CERP production foundry test. It is unclear exactly why
the HAP emissions from the Mexico study are lower than the CERP production foundry; it is expected to
be a combination of casting size and complexity, poorer capture in the real-world foundry, and
differences in emission potential for the different core systems. Nevertheless, the benzene and total HAP
emissions from the CERP production foundry agrees well with the benzene and total HAP emissions from
the CERP pre-production foundry, indicating that the pre-production foundry results provide a reasonable
assessment of the emissions resulting during full foundry production.
D-l
-------
Version 1
Draft
Appendix D
Table D-2. Summary of Baseline Emission Results from Production Foundry
Analyte
Emission Factor (lb/ton metal)
CERP, 1999b
CERP, 2000
Technikon, 2001a
Sum of VOCs
NR
0.6768
0.735
Sum of HAPs
0.318
0.4882
0.643
Benzene
0.0639
0.2255
0.251
Toluene
0.0421
0.0715
0.065
Xylene (Total)
0.0298
0.0460
0.035
Ethylbenzene
0.0049
0.0097
0.0064
Naphthalene
0.0110
0.0365
0.021
1-Methylnaphthalene
0.0044
0.0161
0.0095
2-Methylnaphthalene
0.0063
0.0253
0.0189
Acetaldehyde
0.0613
0.0051
0.0052
Formaldehyde
0.0276
0.0018
0.0029
Phenol
0.0338
0.0554
0.060
o -Cresol
0.0149
0.0114
0.026
Aniline
BDL
NT
0.073
N,N-Dimethylaniline
NT
NT
0.042
Hexane
NT
0.0120
0.016
Styrene
0.0053
0.0079
0.0034
NR - not reported
NT - not tested
BDL - below detection limit
CERP evaluated various chemically bonded cores systems in their pre-production foundry. These cores
were made using mold sand without any organic additives (similar to the core baseline tests). Table D-3
summarizes the results of these studies.
Table D-3. Summary of Core Binder System Studies in the Pre-production Foundry
Analyte
Emission Factor (lb/ton metal)
FB-FC 1409-125-
117public.pdf
(Technikon 2000)
CM 1256 11 GSA 3
Ashland Core Binder
Repalcement.pdf
(Technikon, 2003)
Sum of VOCs
0.4274
0.3962
Total HAPs
0.386
0.368
Benzene
0.142
0.0955
Toluene
0.039
0.0201
Xylene (Total)
0.021
0.0063
Ethylbenzene
0.0030
0.0011
Naphthalene
0.0142
0.0068
1-Methylnaphthalene
0.0071
0.0023
2-Methylnaphthalene
0.0124
0.0041
D-2
-------
Version 1
Draft
Appendix D
Analyte
Emission Factor (lb/ton metal)
FB-FC 1409-125-
117public.pdf
(Technikon 2000)
CM 1256 11 GSA 3
Ashland Core Binder
Repalcement.pdf
(Technikon, 2003)
Acetaldehyde
0.0027
0.0065
Formaldehyde
0.0006
0.0015
Phenol
0.097
0.1022
o,m,p-Cresol
0.0223
0.0348
Aniline
0.0273
0.0521
N,N-Dimethylaniline
BDL
0.0121
Hexane
0.004
0.0066
Styrene
0.0053
0.0010
NR - not reported
NT - not tested
BDL - below detection limit
CERP also evaluated various chemically bonded mold systems in their pre-production foundry. These
castings did not have cores. Table D-4 summarizes the results of these studies.
Table D-4. Summary of Chemically Bonded Mold Sand Studies in the Pre-production Foundry
Analyte
Emission Factor (lb/ton metal)
DG1211-
Phenolic
Urethane No-
Bake
(Technikon,
2001b)
FP 1410 -113
Phenolic
Urethane No-
Bake
(Technikon,
2004)
DP 1256-112
Phenolic
Urethane No-
Bake
(Technikon,
2001d)
DX 1256-
1115
Furan No
Bake
(Technikon,
2001f)
DZ 1256-
1116
Ester-Cured
Phenolic No
Bake
(Technikon,
2001g)
Sum of VOCs
4.06
1.852
1.73
1.1
0.901
Total HAP
1.797
1.470
1.123
1.059
0.807
Benzene
0.299
0.2487
0.229
0.818
0.318
Toluene
0.056
0.0519
0.045
0.106
0.054
Xylene (Total)
0.031
0.035
0.0285
0.0078
0.032
Ethylbenzene
0.0061
0.0035
0.0045
0.0017
0.0019
Naphthalene
BDL
0.0138
BDL
0.0035
0.015
1-Methylnaphthalene
0.0065
0.0030
0.0035
BDL
0.046
2-Methylnaphthalene
0.012
0.0058
0.0061
0.0027
0.004
Acetaldehyde
0.0041
0.0062
0.0069
0.066
0.069
Formaldehyde
0.0205
0.0221
0.0187
0.025
0.084
Propionaldehyde
0.0007
0.0046
0.0027
0.0012
0.017
Phenol
0.942
0.4965
0.718
0.044
0.121
o,m,p-Cresol
0.500
0.6154
0.058
BDL
0.047
Aniline
0.0185
BDL
0.011
BDL
NT
N,N-Dimethylaniline
Invalid
BDL
BDL
BDL
NT
Hexane
0.0003
0.0004
0.0026
BDL
BDL
Styrene
0.0143
0.0106
0.0122
BDL
0.0008
D-3
-------
Version 1
Draft
Appendix D
BDL - below detection limit
NT - not tested
Invalid - data rejected due to validation considerations
Looking at the baseline pre-production foundry results, it appears that the greensand plus core system
emissions are approximately equal to the baseline greensand emissions and the baseline core emissions.
This observation led to a simple additive correlation considering mold and core emissions separately. In a
summary report (CERP/Technikon, 2006), strong correlations were identified for greensand mold systems
between the emissions of benzene and the casting weight (at constant surface area to volume ratios),
casting surface area at constant pour weights, and combustible level (loss on ignition equating to seacoal
level). As the emission factors are provided on a per mass of metal basis, the variability with casting
weight is already incorporated in the emission estimation methodology. There was inadequate
information by which to develop a correction factor for surface area; however, the benzene emissions
appeared to be directly related to the percent loss on ignition (%LOI). The baseline greensand studies
used molds with a %LOI of 5.1%, so a simple correction factor "greensand correction factor" was
incorporated into the calculation methodology based on baseline emissions factors. Consequently, the
following equation is recommended for estimating organic emissions from PCS operations.
Ei= Qms[(MSVOc x Q,ms) x GSCF + (CSyoc x Cj cs)]/2,000
Eqn D-l
where:
E, = Emission of pollutant "i" (tons/yr).
Qms = Quantity of metal poured into a given type of mold system, tons/yr.
MSyoc = VOC emission factor for the mold system used, lb VOC/ton metal poured (from
Table 5-2).
CSvoc = VOC emission factor for the cores used, lb VOC/ton metal poured (from Table 5-2).
CL\is = Concentration profile of pollutant "i" in emissions from mold system used, lb
pollutant/lb VOC (from Table 5-3).
C, rS = Concentration profile of pollutant "i" in emissions for core system used, lb pollutant/lb
VOC (from Table 5-3).
GSCF = Green sand correction factor, unitless; for green sand systems, GSCF = percent
LOI/5.1 percent; for all other types of mold systems, GSCF = 1.
%LOI = Percent of green sand lost on ignition, weight percent using ASTM D7348 or similar
methods .
2,000 = Conversion factor, lbs/ton.
For foundries that operate several different types of mold systems, Equation D-l should be applied
separately for each type of mold system. The total PCS emissions for the facility would then be the sum
of the emissions for each type of mold system. The default emissions factors for Equation D-l were
developed from the data presented in Tables D-l through D-4. The greensand baseline results in Table D-
1 were used directly for the greensand mold system VOC emission factor. The median value for the
phenolic urethane no-bake mold systems was selected for phenolic urethane bonded mold sand (FP Test
in Table D-4) and the average VOC emission factor from the furan no-bake and ester-cured phenolic no-
bake (Tests DX and DZ in Table D-4) were used for the "other chemically bonded mold sand" VOC
emission factor. The VOC emission factor for chemically bonded cores was selected based on the data
presented in Table D-3 with consideration that the sum of MSVoc and CSVoc would approximately equal
the baseline emissions for greensand with chemically bonded cores in the baseline pre-production foundry
(in Table D-l). The background baseline (Table D-l) was used as the default VOC emission factor for
permanent mold, centrifugal, and investment casting systems. Finally, the HAP emission factors for PCS
D-4
-------
Version 1
Draft
Appendix D
from expendable pattern casting (EPC) of iron as presented by Twarog (1991) suggest a total HAP
emission factor of approximately 1.0 lb/ton. Based on the ratio of VOC to HAP emissions for the PCS
emissions presented in Tables D-l through D-4, a default VOC emission factor of 1.2 lb/ton was
determined. The summary of recommended VOC emission factors for Equation D-l is presented in
Table D-5
Table D-5. VOC Emission Factors by Casting Type for PCS Operations3'15
Type of Mold/Core System
Factor Designation
VOC Emission Factor,
(lb/ton metal poured)
Green Sand
MS voc
1.9a
Phenolic Urethane Bonded Mold Sand
MS voc
7.4b
Other Chemically Bonded Mold Sand
MS voc
4.0b
Chemically Bonded Cores (all)
CSvoc
1.6a
Mold systems without cores
CSvoc
0
Expendable Pattern Casting (Lost Foam)
MS voc
4.8C
Permanent, Centrifugal, or Investment Casting
MS voc
0.12d
a Developed from CERP baseline testing (CERP, 1999b and 2000; Technikon 2000, 2001a, 2003).
b Developed from CERP testing of chemically bonded mold systems (Technikon, 2001b, 2001 d, 2001 f, 2001 g, 2004).
c Based on data reported by Twarog (1991).
d Developed from CERP "background baseline" (CERP, 1999b).
In a similar fashion, default chemical composition of the emitted VOC was determined for each of these
mold systems. For each test presented in Tables D-l through D-4, a ratio of the specific analyte's
emission factor to the VOC emission factor was calculated. The greensand baseline results in Table D-l
were used directly for the greensand mold system composition profile. The compositional profile for the
"other chemically bonded mold sand" systems was significantly different for the composition profile for
phenolic urethane no-bake systems. An average composition profile was developed for the phenolic
urethane no-bake molds (based on Tests DG, FP, and DP) and a separate composition profile was
developed for other chemically bonded molds (using Tests DX and DZ).
The composition profile for cores was determined by taking the average composition for the test
summarized in Table D-2 and the composition for the baseline core test in Table D-l. The emission
factors predicted using these profiles and the default VOC factors were then compared to the emission
factors for greensand/core baseline emissions from Table D-l and D-2 (excluding the Mexico data).
Application of the default VOC emission factors and composition profiles yielded combined
greensand/core emissions that were in excellent agreement for nearly all compounds. Based on this
comparison, the average proportion of phenol from core systems was adjusted down to 0.10 (rather than
0.14) to improve the overall prediction when applying Equation D-l to estimate emissions from
greensand/core systems.
The background baseline (Table D-l) was again used as the default composition profile for permanent
mold, centrifugal, and investment casting systems. Finally, the compositional profile for iron EPC was
determined based on the data presented by Twarog (1991). The iron casting emission factors were only
provided for a limited number of compounds (6 analytes, with measurable emissions for styrene, benzene,
and toluene). The concentrations of these compounds was selected so that the concentration times the
VOC emission factor yield the emission factor reported by Twarog for these three compounds. Other data
reported by Twarog indicate that naphthalene and other polycyclic organic matter (POM) is likely to be
emitted, but no direct study of POM emissions was conducted for iron castings (only aluminum castings).
As naphthalene and other POM tend to be ubiquitous pyrolysis products, 1 percent of the VOC emissions
D-5
-------
Version 1
Draft
Appendix D
from EPC was projected to be naphthalene and 1 percent was projected to be "other POM." The
summary of recommended compositional profiles are provided in Table D-6.
Table D-6. Default HAP Composition Profiles for PCS Operations
HAP Compound
Concentration Ratio, lb HAP per lb VOC
Green
Sand
Molds3
Phenolic
Urethane
Bonded
Molds"
Other
Chemically
Bonded
Moldsb
Cores3
Expendable
Pattern
Casting0
Permanent,
Centrifugal,
or
Investment11
Acetaldehyde
0.005
0.00075
0.018
0.0025
0.07
Aniline
0.0075
0.0013
0.035
Benzene
0.065
0.028
0.14
0.073
0.07
0.05
Cresols (total)
0.0015
0.04
0.013
0.01
N,N-Dimethylaniline
0.0025
0.0075
Ethylbenzene
0.005
0.0005
0.0005
0.001
0.005
Formaldehyde
0.00075
0.0025
0.015
0.0005
0.013
Hexane
0.01
0.002
Naphthalene
0.0075
0.0018
0.0025
0.0075
0.0025
0.01
Other POM®
0.01
0.0025
0.013
0.015
0.0025
0.005
Phenol
0.0075
0.078
0.023
0.025
0.0025
Propianaldehyde
0.00025
0.0025
Styrene
0.0013
0.0013
0.00025
0.0013
0.12
Toluene
0.045
0.005
0.02
0.018
0.023
0.025
Xylenes (total)
0.033
0.0025
0.005
0.0075
0.02
a Developed from CERP baseline testing (CERP, 1999b, 2000; Technikon, 2000, 2001a, 2003).
b Developed from CERP testing of chemically bonded mold systems (Technikon, 2001b, 2001 d, 2001 f, 2001 g,
2004).
c Based on data reported by Twarog (1991).
d Developed from CERP "background baseline" (CERP, 1999b).
e POM other than naphthalene, predominately methylnaphthalenes and dimethylnaphthalenes.
D-6
-------
Version 1
Draft
Appendix D
Development of PM Emission Factors for PCS Operations
For the captured, uncontrolled PM-FIL emission factors for pouring, cooling, and shakeout operations
from Table 5-4 in the Protocol document, the emission factors were adapted from the Uncontrolled
component PM emission factor, lb/ton metal melted from Table 5-6 of the Iron and Steel Foundry BID
(U.S. EPA, 2002). Table D-7 summarizes the Uncontrolled PM emission factors from Table 5-6 of the
Iron and Steel Foundry BID.
Table D-7. PM Emission Factors for PCS Lines
PCS Component
Uncontrolled component PM
emission factor, lb/ton metal
melted
Pouring
0.0873
Cooling
0.29
Shakeout
79.3
For the captured, uncontrolled PMio-FIL and PM2 5-FIL emission factors for pouring, cooling, and
shakeout operations from Table 5-4 in the Protocol document, the size distributions presented in Section
12.13 Steel Foundries: Casting Shakeout of Appendix B.l Particle Size Distribution Data and Sized
Emission Factors for Selected Sources of AP-42 were used to calculate the emission factors (U.S. EPA,
1986). Table D-8 summarizes the size distributions for PCS operations from the AP-42 Appendix B.l.
The captured, uncontrolled PM10-FIL emission factor was calculated by multiplying the captured,
uncontrolled PM-FIL emission factor for each operation by the size distribution for PMi0, 82.0%. The
captured, uncontrolled PM2 5-FIL emission factor was calculated by multiplying the captured,
uncontrolled PM-FIL emission factor for each operation by the size distribution for PM2 5, 72.2%.
Table D-8. Size Distributions for PCS Operations
Aerodynamic particle, diameter,
|im
Cumulative wt. % < stated size,
Uncontrolled
2.5
72.2
6.0
76.3
10.0
82.0
For the captured, uncontrolled PM-CON emission factors for pouring, cooling, and shakeout operations
from Table 5-4 in the Protocol document, several tests conducted by Technikon for the Casting Emission
Reduction Program (CERP) were used. The available PM-CON lb/ton values reported in the test reports
were compiled. The average of the PM-CON lb/ton values from all the tests was calculated. Table D-9
summarized the PM-CON lb/ton values reported from the Technikon test reports used in calculating the
PM-CON emission factor.
Table D-9. Size Distributions for PCS Operations
Technikon Source
2001c
2001e
2001a
2001b
2001d
2001f
2001g
PCS Combined
PM-CON lb/ton
0.2324
1.17
1.47
0.8
0.916
2.28
0.448
Average PM-CON lb/ton
1.0
D-7
-------
Version 1
Draft
Appendix D
To determine the distribution of PM-CON emissions between pouring and cooling used the same
distribution as in PM-FIL emission factors for PM-CON emission factors, 23 percent for pouring and 77
percent for cooling. For Shakeout PM-CON emissions, assumed that PM-CON emissions would be zero.
Development of HAP Content of PM for PCS Operations
For the metal HAP percent of PM-FIL from Table 5-5 of the Protocol document, the metal HAP and PM
data were used from a Casting Emission Reduction Program (CERP) testing program of a foundry in
Mexico (CERP, 1999a). Table D-10 summarizes the metal HAP and PM emission factors reported in the
test report. Calculated the percent metal HAP of PM for each of the reported metal HAPs.
Table D-10 HAP Metal and PM Emission Factors from CERP Foundry Mexico Baseline Testing
Analyte
Pouring (lb/ton
metal poured)
Cooling (lb/ton
metal poured)
Shakeout (lb/ton
metal poured)
Totals (lb/ton
metal poured)
Antimony
Non-detect
1.03E-05
2.53E-06
1.29E-05
Arsenic
1.91E-06
Non-detect
Non-detect
1.91E-06
Cadmium
4.55E-06
2.03E-05
1.67E-05
4.16E-05
Chromium
4.85E-05
2.31 E-04
1.71 E-04
4.51 E-04
Cobalt
7.39E-04
5.33E-05
8.62E-05
8.78E-04
Lead
1.79E-04
2.22E-04
7.29E-04
4.74E-04
Manganese
8.37E-04
5.21 E-04
3.39E-04
1.70E-03
Nickel
1.16E-04
1.92E-04
3.13E-04
6.20E-04
Selenium
Non-detect
4.10E-06
Non-detect
4.10E-06
Total PM
0.042
0.11
0.12
0.26
To determine the distribution of chromium hexavalent emissions the values reported in Exhibit D-l of
"An Overview of Methods for EPA's National-Scale Air Toxics Assessment" for Standard Classification
Codes from Iron and Steel pouring, cooling, and shakeout operations were used, 3 percent (U.S. EPA,
2011). Table D-l 1 summarizes the percent chromium hexavalent of total chromium emissions for the
applicable SCC codes for iron and steel foundries from Exhibit D-l.
Exhibit D-11. Chromium Speciation Table Used for the 2005 NATA
Type of Foundry
Source
SCC
% Cr (VI)
Iron
Cupola
3-04-003-01
3
Iron
Reverberatory
3-04-003-02
not in table
Iron
EIF
3-04-003-03
3
Iron
EAF
3-04-003-04
3
Iron
Scrap and Charge handling, heating
3-04-003-15
3
Iron
Pouring, Cooling
3-04-003-18
3
Iron
core making baking
3-04-003-19
not in table
Iron
Magnesium treatment
3-04-003-21
not in table
Iron
Refining
3-04-003-22
3
Iron
Shakeout
3-04-003-31
3
D-8
-------
Version 1
Draft
Appendix D
Type of Foundry
Source
see
% Cr (VI)
Iron
Cleaning and finishing
3-04-003-40
3
Iron
sand handling
3-04-003-50
3
Steel
EIF
3-04-007-05
12
Steel
EAF
3-04-007-01
12
Steel
Pouring and casting
3-04-007-08
3
Steel
Sand grinding/handling in mold and core making
3-04-007-06
3
Steel
Core ovens
3-04-007-07
not in table
Steel
Casting cleaning
3-04-007-11
3
Steel
Charge handling
3-04-007-12
3
Steel
Casting cooling
3-04-007-13
3
Steel
Open hearth
3-04-007-02
not in table
Steel
Open hearth oxygen lanced
3-04-007-03
not in table
Development of CO Emission Factor for PCS Operations Table 5-6
To determine the carbon monoxide emission factor for PCS operations in Table 5-6 of the Protocol
document used several tests conducted by Technikon for CERP. The reported CO lb/ton values were
compiled. The average of the compiled lb/ton values was calculated. The majority of the CO emissions
reported were from chemically bonded molds, there was a lack of CO data from the other mold types.
Table D-12 summarizes the Technikon carbon monoxide data used to develop the CO emission factor.
Table D-12. Summary of CO Emission Factors for PCS Operations3
Technikon Source
2001b
2004
2001d
2001f
2001g
PCS Total
CO lb/ton
4.18
0.0796
4.11
5.99
4.32
Average CO lb/ton
3.7
D-9
-------
Version 1
Draft
Appendix D
References
CERP (1999a/ Foundry process emission factors: baseline emissions from automotive foundries in
Mexico. January 19, 1999.
CERP (1999b). Baseline testing emission results pre-production foundry. November 1, 1999.
CERP (2000). Baseline testing emission results production foundry. February 7, 2000.
Technikon (2000). Ashland core binder replacement. December 15, 2000.
Technikon (2001a). Production baseline airborne emission test report. March 26, 2001.
Technikon (2001b). Phenolic urethane /iron no-bake baseline emission test. March 27, 2001.
Technikon (2001c). Pre-production air emission test report. April 25, 2001.
Technikon (200 Id). Iron phenolic no-bake Delta HA Techniset. April 26, 2001.
Technikon (2001e). Greensandprepared with tap and advanced oxidant enriched water. August 1, 2001.
Technikon (200If). Ashland iron/furan no-bake NBF. May 18, 2001
Technikon (200 lg). Ashland iron no bake NBE1 and 2 ester cured phenolic/iron. June 22, 2001.
Technikon (2003). Emission comparison of phenolic urethane binders with standard solvents and
napthalene-depletedsolvents. July 30, 2003.
Technikon (2004). Product test: no-bake HA international. March 2004.
Twarog , D.L. (1991). Identification of emissions and solid wastes generated from EPC process.
American Foundrymen's Society Research Report. June 4, 1991.
U.S. EPA (1986). Compilation of air pollutant emission factors. Volume 1: Stationary point and
area sources. Appendix B. 1: Particle size distribution data and sized emission factors for
selected sources. AP-42, Fifth Edition. Office of Air Quality Planning and Standards,
Research Triangle Park, NC.
U.S. EPA (2011). An overview of methods for EPA 's National-Scale Air Toxics Assessment.
Office of Air Quality Planning and Standards, Research Triangle Park, NC.
D-10
-------
Version 1
Draft
Appendix E
Appendix E:
Development of Emission Factors for Finishing Operations
Development of PM-FIL Captured, Uncontrolled Emission Factors for Finishing
Operations
In the development of the PM-FIL emission factors for Cutting, Grinding, and Shotblasting, captured and
uncontrolled found in Table 6-2 of the Protocol document (designated with footnote a), PM data that was
collected from the Foundry Information Collection Request (ICR) was used. The PM lb/ton data for
cutoff, grinding, and shotblasting process units that were tested separately were used. Data that came
from more than one process unit ducted together was not used. The data that had a collection period of
less than twenty hours of testing was not used in the calculations. The median emission factor for each
operation was calculated, and rounded to the nearest whole number. Table E-l summarizes the data from
the Foundry ICR used to calculate emission factors.
Table E-1. PM Data from Foundry ICR
Facility ID
Operation
Collection period, hours
Emission factor, lb/ton
358
Cutoff
5,814
3.53
358
Cutoff
5,814
3.53
159
Cutoff
2,080
5.08
227
Cutoff
64
5.47
358
Cutoff
5,814
6.67
358
Cutoff
5800
8.60
358
Cutoff
4,845
9.00
619
Cutoff
3,840
20.3
462
Grinding
2,838
1.02
519
Grinding
2,670
3.82
230
Grinding
40
4.44
389
Grinding
6,000
4.67
14
Grinding
2,000
6.15
818
Grinding
8,760
12.6
818
Grinding
8,760
12.6
529
Grinding
120
14.3
159
Grinding
2,080
15.6
140
Grinding
2,080
16.8
365
Grinding
1,353
21.0
72
Grinding
5,000
21.9
385
Grinding
5,856
23.8
363
Grinding
40
28.6
760
Grinding
80
30.8
262
Grinding
150
45.0
E-l
-------
Version 1
Draft
Appendix E
Facility ID
Operation
Collection period, hours
Emission factor, lb/ton
270
Grinding
1,200
72.7
110
Grinding
40
250
232
Shotblast
22
57.1
184
Shotblast
24
13.6
213
Shotblast
434
32.8
16
Shotblast
24
2.14
16
Shotblast
24
15.6
16
Shotblast
24
4.66
542
Shotblast
2,000
1.95
567
Shotblast
1,920
9.26
413
Shotblast
20
15.3
760
Shotblast
40
15.4
76
Shotblast
120
22.5
23
Shotblast
2,652
25.4
363
Shotblast
40
28.6
818
Shotblast
8,760
33.7
682
Shotblast
80
62.5
389
Shotblast
6,000
82.3
433
Shotblast
160
1.56
433
Shotblast
320
1.67
519
Shotblast
2,670
2.52
608
Shotblast
4,500
6.92
519
Shotblast
2,670
9.21
818
Shotblast
8,760
16.9
14
Shotblast
30
26.7
140
Shotblast
2,080
39.4
645
Shotblast
80
42.9
110
Shotblast
40
62.5
534
Shotblast
40
83.9
262
Shotblast
150
2.50
207
Shotblast
24
5.00
257
Shotblast
80
41.8
184
Shotblast
24
133
72
Shotblast
5,000
4.48
72
Shotblast
5,000
8.38
Cutoff Median Emission Factor Value
6.0
Grinding Median Emission Factor Value
16.0
Shotblast Median Emission Factor Value
16.0
E-2
-------
Version 1
Draft
Appendix E
Development of PM10-FIL and PM2.5-FIL Captured, Uncontrolled Emission Factors
for Finishing Operations
Table 12.10-9 of Section 12.10 of AP-42 (U.S. EPA, 2003) lists the particle size distributions for PM2 5
and PM10 for different foundry operations. The table does not include particle size distributions for
finishing operations. The particle size distributions for pouring, cooling operations at foundries were as a
starting point in estimating particle size distributions for finishing operations. Table E-2 summarizes the
size distribution values from Table 12.10-9 and the values used to calculate the PM10-FIL and PM25-FIL
emission factors in Table 6-2 of the Protocol document. The size distribution ratio for PMi0 and PM2 5
was applied to each PM-FIL captured, uncontrolled value for Cutting, Grinding, and Shot blasting to
calculate the PM10-FIL and PM2 5-FIL emission factors for each emission source.
Table E-2. Size Distribution of PM10-FIL and PM2.5-FIL for PCS Operations
Source
Particle Size (pm)
Table 12.10-9. Particle
Size Distribution Data
and Emission Factors for
Gray Iron Foundries from
Section 12.10 of AP-42
Cumulative Mass % <
Stated Size
Size Distribution
Ratio Used
Pouring, cooling
Uncontrolled
2.5
24.0
20
10.0
49.0
50
Development of PM Uncaptured, Uncontrolled Emission Factors for Finishing
Operations
Table 12.10-7 of Section 12.10 of AP-42 (U.S. EPA, 2003) shows that 0.1 PM lb/ton is emitted to the
atmosphere from 17 PM lb/ton emissions from uncontrolled finishing operations at a foundry. Thus,
approximately 99 percent of the finishing emissions are expected to remain (redeposit) within the foundry
building and approximately 1 percent of the emission are projected to be released to the atmosphere. The
uncaptured and uncontrolled PM-FIL values for each emission source in Table 6-2 of the Protocol
document were calculated by multiplying the captured and uncontrolled PM-FIL value for each emission
source by 1 percent.
For the PM10-FIL and PM2 5-FIL emission factors, it was assumed that 90 percent of the uncaptured PM-
FIL is PM10-FIL and 80 percent of the uncaptured PM-FIL is PM2 5-FIL as smaller particles are more
likely to escape to the atmosphere than be captured by a control device. The uncaptured and uncontrolled
PM10-FIL values for each emission source in Table 6-2 of the Protocol document were calculated by
multiplying the uncaptured and uncontrolled PM-FIL value by 90 percent. The uncaptured and
uncontrolled PM2 5-FIL values for each emission source in Table 6-2 of the Protocol document were
calculated by multiplying the uncaptured and uncontrolled PM-FIL value by 80 percent.
References
U.S. EPA (2003). Compilation of air pollutant emission factors. Volume 1: Stationary point and
area sources. Section 12.10: Gray iron foundries. AP-42, Fifth Edition. Office of Air
Quality Planning and Standards, Research Triangle Park, NC.
E-3
-------
Version 1
Draft
Appendix F
Appendix F:
Control Efficiency and Particulate Matter Size Distribution
To simplify the protocol document, the collection efficiencies for filterable particles size 2.5-6 and 6-
10 nm were condensed down to be presented as the collection efficiencies for filterable particle size 2.5-
10 nm. Table F-l provides the collection efficiency values as reported in AP-42. In general, the average
collection efficiency for PM in the 2.5-10 nm range was determined as the arithmetic average of the
collection efficiencies for the 2.5- 6 pim PM size range and the 6-10 pim PM size range.
Table F-1. Typical Collection Efficiencies of Various Particulate Control Devices3 (%)
AIRS
Codeb
Type of Collector
Collection Efficiency
Average
Collection
Efficiency
for PM
2.5-10 (jm
Filterable Particle Size (pm)
0-2.5
2.5-6
6-10
001
Wet scrubber - hi-efficiency
90
95
99
97
002
Wet scrubber - med-efficiency
25
85
95
90
003
Wet scrubber - low-efficiency
20
80
90
85
004
Gravity collector - hi-efficiency
3.6
5
6
5
005
Gravity collector - med-efficiency
2.9
4
4.8
4
006
Gravity collector - low-efficiency
1.5
3.2
3.7
3.4
007
Centrifugal collector - hi-efficiency
80
95
95
95
008
Centrifugal collector - med-efficiency
50
75
85
80
009
Centrifugal collector - low-efficiency
10
35
50
42
010
Electrostatic precipitator - hi-efficiency
95
99
99.5
99
011
Electrostatic precipitator - med-efficiency
80
90
97
93
012
Electrostatic precipitator - low-efficiency
70
80
90
85
014
Mist eliminator - high velocity >250 FPM
10
75
90
92
015
Mist eliminator - low velocity <250 FPM
5
40
75
57
016
Fabric filter - high temperature (>250 °F)
99
99.5
99.5
99.5
017
Fabric filter - med temperature (180 °F 30 inches of water)
95
98
99
98
053
Venturi scrubber (AP < 30 inches of water)
88
94
99
96
054
Process enclosed
1.5
3.2
3.7
3.4
055
Impingement plate scrubber
25
95
99
97
056
Dynamic separator (dry)
90
95
99
97
057
Dynamic separator (wet)
50
75
85
80
058
Mat or panel filter - mist collector
92
94
97
95
059
Metal fabric filter screen
10
15
20
17
061
Dust suppression by water sprays
40
65
90
77
F-l
-------
Version 1
Draft
Appendix F
AIRS
Codeb
Type of Collector
Collection Efficiency
Average
Collection
Efficiency
for PM
2.5-10 (jm
Filterable Particle Size (pm)
0-2.5
2.5-6
6-10
062
Dust suppression by chemical stabilizer or wetting agents
40
65
90
77
063
Gravel bed filter
0
5
80
42
064
Annular ring filter
80
90
97
93
071
Fluid bed dry scrubber
10
20
90
55
075
Single cyclone
10
35
50
42
076
Multiple cyclone w/o fly ash reinjection
80
95
95
95
077
Multiple cyclone w/fly ash reinjection
50
75
85
80
085
Wet cyclonic separator
50
75
85
80
086
Water curtain
10
45
90
67
a Data represent an average of actual efficiencies. Efficiencies are representative of well designed and well operated
control equipment. Site-specific factors (e. g., type of particulate being collected, varying pressure drops across
scrubbers, maintenance of equipment) will affect collection efficiencies. Efficiencies shown are intended to provide
guidance for estimating control equipment performance when source-specific data are not available. Table derived
from Table B.2-3 Typical Collection efficiencies of various particulate control devices of Appendix B.2 of AP-42,
Volume I, Fifth Edition (U.S. EPA, 1996)
b Control codes in Aerometric Information Retrieval System (AIRS), formerly National Emissions Data Systems.
References
U.S. EPA (1996). Compilation of air pollutant emission factors. Volume 1: Stationary point and area
sources. Appendix B.2: Generalized Particle Size Distributions. AP-42, Fifth Edition. Office of
Air Quality Planning and Standards, Research Triangle Park, NC. September.
F-2
-------
Version 1
Draft
Appendix G
Appendix G:
List of Suggested SCC for Iron and Steel Foundry Operations
Table G-1. Iron and Steel Foundry SCC
Source
Classification
Code
SCC Level
One
SCC Level Two
SCC Level
Three
SCC Level Four
Suggested SCC for Iron Foundry Operations
30400301
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Cupola
30400302
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Reverberatory Furnace
30400303
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Electric Induction Furnace
30400304
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Electric Arc Furnace
30400310
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Inoculation
30400314
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Scrap Metal Preheating
30400315
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Charge Handling
30400316
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Tapping
30400319
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Core Making, Baking
30400320
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Pouring/Casting
30400321
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Magnesium Treatment
30400322
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Refining
30400325
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Castings Cooling
30400331
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Casting Shakeout
30400340
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Grinding/Cleaning
30400350
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Sand Grinding/Handling
30400360
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Castings Finishing
30400370
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Shell Core Machine
30400371
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Core Machines/Other
2294000000
Mobile
Sources
Paved Roads
All Paved
Roads
Total: Fugitives
2296000000
Mobile
Sources
Unpaved Roads
All Unpaved
Roads
Total: Fugitives
G-l
-------
Version 1
Draft
Appendix G
Source
Classification
Code
SCC Level
One
SCC Level Two
SCC Level
Three
SCC Level Four
Other SCC for Iron Foundry Operations
30400305
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Annealing Operation
30400317
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Pouring Ladle
30400318
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Pouring, Cooling
30400330
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Miscellaneous Casting-
Fabricating **
30400332
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Casting Knock Out
30400333
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Shakeout Machine
30400341
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Casting Cleaning/Tumblers
30400342
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Casting Cleaning/Chippers
30400351
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Core Ovens
30400352
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Sand Grinding/Handling
30400353
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Core Ovens
30400354
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Core Ovens
30400355
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Sand Dryer
30400356
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Sand Silo
30400357
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Conveyors/Elevators
30400358
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Sand Screens
30400398
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Other Not Classified
30400399
Industrial
Processes
Secondary Metal
Production
Grey Iron
Foundries
Other Not Classified
Suggested SCC for Steel Foundry Operations
30400701
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Electric Arc Furnace
30400705
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Electric Induction Furnace
30400708
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Pouring/Casting
30400709
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Casting Shakeout
30400711
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Cleaning
30400712
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Charge Handling
G-2
-------
Version 1
Draft
Appendix G
Source
Classification
SCC Level
SCC Level
Code
One
SCC Level Two
Three
SCC Level Four
Industrial
Secondary Metal
Steel
30400713
Processes
Production
Foundries
Castings Cooling
Industrial
Secondary Metal
Steel
30400715
Processes
Production
Foundries
Finishing
Industrial
Secondary Metal
Steel
30400716
Processes
Production
Foundries
Sand Grinding/Handling
Industrial
Secondary Metal
Steel
30400730
Processes
Production
Foundries
Shell Core Machine
Industrial
Secondary Metal
Steel
30400731
Processes
Production
Foundries
Core Machines/Other
Industrial
Secondary Metal
Steel
30400741
Processes
Production
Foundries
Scrap Heating
Industrial
Secondary Metal
Steel
30400765
Processes
Production
Foundries
Billet Cutting
30400768
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Scrap Handling
Industrial
Secondary Metal
Steel
30400770
Processes
Production
Foundries
Slag Storage Pile
Mobile
All Paved
2294000000
Sources
Paved Roads
Roads
Total: Fugitives
2296000000
Mobile
Sources
Unpaved Roads
All Unpaved
Roads
Total: Fugitives
Other SCC for Steel Foundry Operations
Industrial
Secondary Metal
Steel
30400702
Processes
Production
Foundries
Open Hearth Furnace
Industrial
Secondary Metal
Steel
Open Hearth Furnace with
30400703
Processes
Production
Foundries
Oxygen Lance
Industrial
Secondary Metal
Steel
30400704
Processes
Production
Foundries
Heat Treating Furnace
Industrial
Secondary Metal
Steel
30400706
Processes
Production
Foundries
Sand Grinding/Handling
Industrial
Secondary Metal
Steel
30400707
Processes
Production
Foundries
Core Ovens
Industrial
Secondary Metal
Steel
30400710
Processes
Production
Foundries
Casting Knock Out
Industrial
Secondary Metal
Steel
30400714
Processes
Production
Foundries
Shakeout Machine
Industrial
Secondary Metal
Steel
30400717
Processes
Production
Foundries
Core Ovens
Industrial
Secondary Metal
Steel
30400718
Processes
Production
Foundries
Core Ovens
30400720
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Sand Dryer
Industrial
Secondary Metal
Steel
30400721
Processes
Production
Foundries
Sand Silo
Industrial
Secondary Metal
Steel
30400722
Processes
Production
Foundries
Muller
Industrial
Secondary Metal
Steel
30400723
Processes
Production
Foundries
Conveyors/Elevators
Industrial
Secondary Metal
Steel
30400724
Processes
Production
Foundries
Sand Screens
G-3
-------
Version 1
Draft
Appendix G
Source
Classification
SCC Level
SCC Level
Code
One
SCC Level Two
Three
SCC Level Four
Industrial
Secondary Metal
Steel
30400725
Processes
Production
Foundries
Casting Cleaning/Tumblers
Industrial
Secondary Metal
Steel
30400726
Processes
Production
Foundries
Casting Cleaning/Chippers
Industrial
Secondary Metal
Steel
Electric Arc Furnace:
30400732
Processes
Production
Foundries
Baghouse
Industrial
Secondary Metal
Steel
Electric Arc Furnace:
30400733
Processes
Production
Foundries
Baghouse Dust Handling
Industrial
Secondary Metal
Steel
30400735
Processes
Production
Foundries
Raw Material Unloading
30400736
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Conveyors/Elevators: Raw
Material
Industrial
Secondary Metal
Steel
30400737
Processes
Production
Foundries
Raw Material Silo
30400739
Industrial
Processes
Secondary Metal
Production
Steel
Foundries
Scrap Centrifugation
Industrial
Secondary Metal
Steel
Reheating Furnace: Natural
30400740
Processes
Production
Foundries
Gas
Industrial
Secondary Metal
Steel
30400742
Processes
Production
Foundries
Crucible
Industrial
Secondary Metal
Steel
30400743
Processes
Production
Foundries
Pneumatic Converter Furnace
Industrial
Secondary Metal
Steel
30400744
Processes
Production
Foundries
Ladle
Industrial
Secondary Metal
Steel
30400745
Processes
Production
Foundries
Fugitive Emissions: Furnace
Industrial
Secondary Metal
Steel
30400760
Processes
Production
Foundries
Alloy Feeding
Industrial
Secondary Metal
Steel
30400775
Processes
Production
Foundries
Slag Crushing
Industrial
Secondary Metal
Steel
30400780
Processes
Production
Foundries
Limerock Handling
Industrial
Secondary Metal
Steel
Roof Monitors - Hot Metal
30400785
Processes
Production
Foundries
Transfer
Industrial
Secondary Metal
Steel
30400799
Processes
Production
Foundries
Other Not Classified
G-4
------- |