United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-453/R-96-008a
September 1996
Air
<&EPA Hazardous Air Pollutant
Emissions from the
Production of Flexible
Polyurethane Foam ~
Basis and Purpose Document
for Proposed Standards
-------
HAZARDOUS AIR POLLUTANT
EMISSIONS FROM THE PRODUCTION
OF FLEXIBLE POLYURETHANE FOAM
Basis and Purpose Document
for Proposed Standards
Emission Standards Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
September 1996
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard. 12th Floor
Chicago, II 60604-3590
-------
DISCLAIMER
This report has been reviewed by the Emission Standards Division
of the Office of Air Quality Planning and Standards, EPA, and has
been approved for publication. Mention of trade names or
commercial products is not intended to constitute endorsement or
recommendation for use.
11
-------
ENVIRONMENTAL PROTECTION AGENCY
Hazardous Air Pollutant Emissions from the Production of Flexible
Polyurethane Foam - Basis and Purpose Document for Proposed
Standards
1. The standards regulate hazardous air pollutant emissions
from the production of flexible polyurethane foam. Only
flexible polyurethane production facilities that are part of
major sources under Section 112(d) of the Clean Air Act
(Act) will be regulated.
2. For additional information contact:
Mr. David Svendsgaard
Organic Chemicals Group
U.S. Environmental Protection Agency (MD-13)
Research Triangle Park, NC 27711
Telephone: (919) 541-2380
3. Paper copies of this document may be obtained from:
U.S. Environmental Protection Agency Library (MD-36)
Research Triangle Park, NC 27711
Telephone: (919) 541-2777
National Technical Information Service (NTIS)
5285 Port Royal Road
Springfield, VA 22161
Telephone: (703) 487-4650
111
-------
TABLE OF CONTENTS
Page
LIST OF FIGURES viii
LIST OF TABLES viii
1.0 PURPOSE OF DOCUMENT 1-1
2.0 INTRODUCTION 2-1
3.0 DESCRIPTION OF THE AFFECTED INDUSTRY 3-1
3.1 INDUSTRY DESCRIPTION 3-1
3.1.1 Slabstock Foam Facility Distribution . . . .3-2
3.1.2 Molded Foam Facility Distribution 3-2
3.2 SLABSTOCK AND MOLDED FOAM CHEMISTRY 3-5
3.2.1 Chemistry of Flexible Polvurethane Foam:
Slabstock and Molded 3-5
3.2.2 Auxiliary Blowing Agents 3-6
3.3 SLABSTOCK FOAM PRODUCTION PROCESS 3-7
3.3.1 Horizontal Maxfoam Production 3-7
3.3.2 Vertifoam Production Process Description . 3-10
3.3.3 Foam Fabrication 3-10
3.4 HAP EMISSION SOURCES FROM SLABSTOCK FOAM
PRODUCTION 3-11
3.4.1 Storage Emissions 3-11
3.4.2 Equipment Leaks from Components in HAP
Service 3-11
3.4.3 Foam Tunnel and Curing 3-11
3.4.4 Equipment Cleaning 3-12
3.4.5 Fabrication 3-12
3.5 CONTROL TECHNOLOGIES FOR SLABSTOCK HAP EMISSIONS 3-12
3.5.1 Control Technologies for ABA Emissions . . 3-12
3.5.2 Reducing Releases From Chemical Storage and
Handling, Equipment Cleaning, and Components
in HAP Service 3-13
3.5.2.1 Chemical Storage and Handling . . . 3-13
3.5.2.2 Components in HAP Service 3-13
3.5.2.3 Eguipment Cleaning 3-13
3.5.3 Reducing Releases From Fabrication/Repair
Operations: Fabrication Adhesives and
Molded Foam Repair 3-13
3.6 MOLDED FOAM PRODUCTION PROCESS 3-14
3.7 MOLDED FOAM HAP EMISSION SOURCES 3-16
3.7.1 Mixhead Flush 3-16
3.7.2 Mold Release Agents 3-16
3.7.3 Molded Foam Repair 3-17
3.8 HAP CONTROL TECHNOLOGIES FOR MOLDED FOAM .... 3-17
3.8.1 Control Technologies for Reducing Releases
from Mixhead Flushing 3-17
3.8.2 Control Technologies for Reducing Releases
of Mold Release Agent 3-17
3.9 REBOND FOAM PRODUCTION 3-18
iv
-------
TABLE OF CONTENTS
(continued)
3.9.1 Rebond Foam Process 3-18
3.9.2 Rebond Emission Sources 3-18
3.9.3 Rebond Control Techniques 3-18
4.0 SUBCATEGORIZATION OF THE LISTED SOURCE CATEGORY . . . .4-1
4.1 SUBCATEGORIZATION 4-1
4.2 PROCESS AND HAP EMISSION DESCRIPTIONS 4-1
4.2.1 Polyurethane Foam Chemistry 4-2
4.2.2 Slabstock Polvurethane Foam Production . . . 4-2
4.2.3 Molded Polvurethane Foam Production .... 4-2
4.2.4 Fabrication Operations 4-3
4.2.5 Rebond Foam Production 4-3
4.3 RATIONALE FOR SUBCATEGORIZATION 4-4
4.4 SUMMARY 4-6
5.0 BASELINE EMISSIONS 5-1
5.1 SLABSTOCK BASELINE EMISSION CALCULATIONS 5-5
5.2 MOLDED BASELINE EMISSION CALCULATIONS 5-5
5.3 REBOND BASELINE EMISSION ESTIMATES 5-6
6.0 MACT FLOORS AND REGULATORY ALTERNATIVES 6-1
6.1 CLEAN AIR ACT (CAA) REQUIREMENTS FOR MACT FLOORS . 6-1
6.2 CONSIDERATIONS IN DETERMINING MACT FLOORS 6-1
6.2.1 Subcatecrorization 6-2
6.2.2 Major Source Determination 6-2
6.2.3 Grouping of Emission Sources 6-3
6.2.4 Approach to Determining the MACT Floor . . . 6-4
6.3 MACT FLOOR CONCLUSIONS 6-5
6.3.1 Molded Foam 6-5
6.3.1.1 Storage/unloading 6-6
6.3.1.2 Mixhead flush 6-6
6.3.1.3 Mold release agents 6-7
6.3.1.4 Foam repair 6-7
6.3.1.5 In-process vessels, components in
HAP service, in-mold coatings, foam
reactant dispensing, and demolding . . . 6-8
6.3.2 Slabstock Foam 6-8
6.3.2.1 Storage/unloading 6-8
6.3.2.2 Components in HAP service 6-9
6.3.2.3 Equipment cleaning 6-9
6.3.2.4 ABA emission sources 6-10
6.3.2.4.1 Existing sources 6-10
6.3.2.4.2 New sources 6-16
6.3.3 Rebond Foam Production 6-21
6.4 SUMMARY OF MACT FLOORS 6-21
6.5 POTENTIAL LEVELS OF CONTROL 6-21
6.5.1 Molded Foam Facilities 6-21
6.5.1.1 Mixhead Flush 6-23
-------
TABLE OF CONTENTS
(continued)
6.5.1.2 Mold Release Agents and Foam Repair 6-23
6.5.1.3 Other Molded Emission Sources . . . 6-23
6.5.2 Slabstock Foam Facilities 6-23
6.5.2.1 Storage/Unloading 6-23
6.5.2.2 Components in HAP Service (Equipment
Leaks) 6-23
6.5.2.3 Equipment Cleaning 6-24
6.5.2.4 Auxiliary Blowing Agent 6-24
6.5.2.4.1 Existing sources 6-24
6.5.2.4.2 New sources 6-24
6.5.3 Rebond Foam 6-24
6.6 REGULATORY ALTERNATIVES 6-25
6.6.1 Molded Foam 6-25
6.6.2 Slabstock Foam 6-25
6.6.3 Rebond Foam 6-29
7.0 MODEL PLANTS 7-1
7.1 SLABSTOCK FOAM MODEL PLANT DESCRIPTIONS 7-2
7.2 MOLDED FOAM MODEL PLANT DESCRIPTIONS 7-6
8.0 ENVIRONMENTAL AND ENERGY IMPACTS
OF REGULATORY ALTERNATIVES 8-1
8.1 PRIMARY ENVIRONMENTAL IMPACTS 8-1
8.1.1 Primary Environmental Impacts for Molded
Foam 8-1
8.1.2. Slabstock Foam Primary Environmental
Impacts 8-2
8.1.3 Rebond Foam Primary Environmental Impacts . 8-6
8.2 SECONDARY ENVIRONMENTAL IMPACTS 8-6
8.2.1. Air Pollution Impacts 8-6
8.2.2. Water Pollution Impacts 8-7
8.2.3. Solid and Hazardous Waste Impacts 8-8
8.3 ENERGY IMPACTS 8-8
9.0 COST AND ECONOMIC IMPACTS OF REGULATORY ALTERNATIVES . . 9-1
9.1 MODEL PLANT COSTS 9-1
9.1.1 Slabstock Foam 9-1
9.1.1.1 Storage/Unloading 9-2
9.1.1.2 Equipment Cleaning 9-2
9.1.1.3 Equipment Leaks 9-4
9.1.1.4 HAP ABA emissions 9-4
9.1.1.4.1 Chemical Alternatives . . . . 9-6
9.1.1.4.2 Carbon Dioxide as an ABA . . . 9-6
9.1.1.4.3 Acetone as an ABA 9-9
9.1.1.4.4 Variable Pressure Foaming . . 9-9
9.1.1.4.5 Forced Cooling 9-9
9.1.2 Molded Foam 9-12
9.1.2.1 Mixhead Flushing 9-14
vi
-------
TABLE OF CONTENTS
(continued)
9.1.2.2 Mold Release Agents 9-14
9.1.2.3 Repair Adhesives 9-14
9.2 NATIONWIDE COST IMPACTS 9-14
9.2.1 Slabstock Foam Nationwide Cost Impacts . . 9-18
9.2.1.1 Storage/Unloading. Equipment Leaks.
and Equipment Cleaning 9-19
9.2.1.2 HAP ABA Emissions 9-19
9.2.2 Nationwide Cost Impact Summary for
Slabstock Foam 9-21
9.2.3 Molded Foam Nationwide Cost Impacts . . . 9-24
9.2.3.1 Mixhead Flush 9-24
9.2.3.2 Mold Release Agents 9-26
9.2.3.3 Repair Adhesives 9-26
9.2.4 Nationwide Cost Impact Summary for Molded
Foam 9-26
9.3 ECONOMIC IMPACTS 9-29
10.0 SELECTION OF THE STANDARDS 10-1
10.1 SUMMARY OF THE PROPOSED STANDARDS 10-1
10.1.1 Applicability and Compliance Schedule . . 10-1
10.1.2 Standards for Molded Flexible Polyurethane
Foam Production 10-2
10.1.3 Standards for Rebond Foam Production . . 10-2
10.1.4 Standards for Slabstock Flexible
Polyurethane Foam Production 10-2
10.1.4.1 Diisocyanate emissions 10-3
10.1.4.2 HAP ABA storage and equipment leak
emissions, HAP ABA emissions from the
production line, and equipment cleaning
HAP emissions 10-3
10.1.4.2.1 HAP ABA storage vessel
requirements 10-4
10.1.4.2.2 HAP ABA equipment leaks . . 10-4
10.1.4.2.3 HAP ABA Emissions from the
production line 10-5
10.1.4.2.4 Equipment cleaning HAP
emissions 10-7
10.1.4.3 Source-wide emission limitation . 10-7
10.1.5 Monitoring Requirements 10-9
10.1.5.1 Storage vessel monitoring .... 10-9
10.1.5.2 Polvol and HAP ABA monitoring at
the mixhead 10-9
10.1.5.3 Recovered HAP ABA monitoring . . . 10-10
10.1.5.4 Monitoring to determine the amount
of HAP ABA in a storage vessel .... 10-10
10.1.5.5 Monitoring to determine the amount
of HAP ABA added to a storage vessel . 10-10
10.1.6 Testing Requirements 10-11
vii
-------
TABLE OF CONTENTS
(continued)
10.1.7 Alternative Means of Emission Limitation 10-11
10.1.8 Applicability of General Provisions . . . 10-12
10.1.9 Reporting Requirements 10-12
10.1.9.1 Initial notification 10-12
10.1.9.2 Application for approval of
construction or reconstruction . . . . 10-13
10.1.9.3 Precompliance report 10-13
10.1.9.4 Notification of compliance status 10-14
10.1.9.5 Semi-annual compliance reports . . 10-14
10.1.9.6 Other reports 10-15
10.1.10 Recordkeepina Requirements 10-15
10.1.10.1 Storage vessel records 10-15
10.1.10.2 Equipment leak records 10-15
10.1.10.3 HAP ABA records 10-16
10.1.10.3 Records for area sources .... 10-16
10.2 RATIONALE FOR THE SELECTION OF THE SOURCE
CATEGORY 10-17
10.3 RATIONALE FOR THE SELECTION OF POLLUTANTS AND
EMISSION POINTS TO BE COVERED BY THE PROPOSED
STANDARDS 10-18
10.4 RATIONALE FOR THE SELECTION OF THE LEVELS OF THE
PROPOSED STANDARDS 10-19
10.4.1 Selection of the Levels of the Proposed
Standards for Existing Sources 10-19
10.4.1.1 Slabstock foam production . . . . 10-19
10.4.1.2 Molded foam production 10-20
10.4.1.3 Rebond foam production 10-21
10.4.2 Selection of the Levels of the Proposed
Standards for New Sources 10-21
10.4.2.1 Slabstock foam production .... 10-21
10.4.2.2 Molded foam production 10-21
10.4.2.3 Rebond foam production 10-22
10.5 RATIONALE FOR THE SELECTION OF THE FORMATS OF THE
PROPOSED STANDARDS 10-22
10.5.1 Molded Foam Production 10-22
10.5.2 Slabstock Foam Production 10-22
10.5.2.1 Storage vessels 10-22
10.5.2.2 Equipment leaks 10-23
10.5.2.3 Equipment cleaning 10-23
10.5.2.4 HAP ABA emissions from the
production line 10-23
10.5.2.4.1 Emission limitation format 10-23
10.5.2.4.2 Averaging time format . . . 10-25
10.5.2.5 Source-wide HAP ABA and equipment
cleaning HAP emission limitation . . . 10-26
10.5.3 Rebond Foam Production 10-27
10.6 SELECTION OF EMISSION TEST METHODS 10-27
10.7 SELECTION OF MONITORING REQUIREMENTS 10-28
viii
-------
TABLE OF CONTENTS
(continued)
10.8 SELECTION OF RECORDKEEPING AND REPORTING
REQUIREMENTS 10-29
10.8.1 Recordkeepinq Requirements 10-29
10.8.2 Reporting Requirements 10-29
10.9 MODIFICATION AND RECONSTRUCTION CONSIDERATIONS . 10-30
10.10 OPERATING PERMIT PROGRAM 10-30
IX
-------
LIST OF FIGURES
Figure 3-1
Figure 3-1
Typical Slabstock Foam Production Process . .3-8
Typical Molded Foam Production Process ... 3-15
TABLE 3-1
TABLE 3-2
TABLE 5-1
TABLE 5-2
TABLE 6-1
TABLE 6-2
TABLE 6-3
TABLE 6-4
TABLE 6-5
TABLE 6-6
TABLE 6-7
TABLE 6-8
TABLE 6-9
TABLE 6-10
TABLE 7-1
TABLE 7-2
TABLE 7-3
TABLE 8-1
TABLE 8-2
TABLE 9-1
TABLE 9-2
TABLE 9-3
TABLE 9-4
TABLE 9-5
TABLE 9-6
TABLE 9-7
LIST OF TABLES
DISTRIBUTION OF SLABSTOCK FOAM FACILITIES BY
STATE 3-3
DISTRIBUTION OF MOLDED FOAM FACILITIES
BY STATE 3-4
BASELINE HAP EMISSIONS FOR SLABSTOCK FOAM
PRODUCTION 5-2
BASELINE HAP EMISSIONS FOR MOLDED FOAM
PRODUCTION 5-3
NUMBERS OF PLANTS REPRESENTED IN FORMULATION
DATABASE 6-13
HAP ABA FORMULATIONS FROM ICR'S 6-14
BASELINE HAP ABA FORMULATIONS 6-15
MACT FLOOR PERCENTAGE REDUCTIONS 6-17
MACT FLOOR HAP ABA FORMULATION LIMITATIONS . 6-18
FLEXIBLE FOAM MACT FLOOR CONCLUSIONS .... 6-20
REGULATORY ALTERNATIVES FOR MOLDED FOAM - EXISTING
SOURCES 6-23
REGULATORY ALTERNATIVES FOR MOLDED FOAM - NEW
SOURCES 6-26
REGULATORY ALTERNATIVES FOR SLABSTOCK FOAM -
EXISTING SOURCES 6-27
REGULATORY ALTERNATIVES FOR SLABSTOCK FOAM - NEW
SOURCES 6-28
SLABSTOCK FOAM PRODUCTION MODEL PLANT
PARAMETERS 7-3
SLABSTOCK FOAM PRODUCTION MODEL PLANT FORMULATION
INFORMATION 7-5
MOLDED FOAM MODEL PLANT PARAMETERS 7-7
MOLDED FOAM REGULATORY ALTERNATIVE HAP EMISSION
REDUCTION 8-3
SLABSTOCK FOAM REGULATORY ALTERNATIVE HAP EMISSION
REDUCTION 8-5
SLABSTOCK FOAM MODEL PLANT COSTS FOR VAPOR
BALANCING 9-3
SLABSTOCK FOAM REGULATORY ALTERNATIVE 1 EQUIPMENT
LEAK MODEL PLANT IMPACTS 9-5
SLABSTOCK FOAM MODEL PLANT COSTS FOR CHEMICAL
ALTERNATIVES HAP ABA EMISSION REDUCTION . . . 9-7
SLABSTOCK FOAM MODEL PLANT COSTS FOR CARDIO HAP
ABA EMISSION REDUCTION 9-8
SLABSTOCK FOAM MODEL PLANT COSTS FOR ACETONE HAP
ABA EMISSION REDUCTION 9-10
SLABSTOCK FOAM MODEL PLANT COSTS FOR VARIABLE
PRESSURE FOAMING HAP ABA EMISSION REDUCTION 9-11
SLABSTOCK FOAM MODEL PLANT COSTS FOR ENVIROCURE
HAP ABA EMISSION REDUCTION 9-13
-------
TABLE 9-8
TABLE 9-9
TABLE 9-10
TABLE 9-11
TABLE 9-12
TABLE 9-13
TABLE 9-14
TABLE 9-15
TABLE 9-16
TABLE 9-17
LIST OF TABLES
(continued)
MOLDED FOAM MODEL PLANT COSTS FOR TECHNOLOGIES TO
REDUCE MIXHEAD FLUSHING HAP EMISSIONS ... 9-15
MOLDED FOAM MODEL PLANT COSTS FOR TECHNOLOGIES TO
REDUCE MOLD RELEASE AGENT HAP EMISSIONS . . 9-16
MOLDED FOAM MODEL PLANT COSTS FOR TECHNOLOGIES TO
REDUCE HAP EMISSIONS FROM THE USE OF FOAM REPAIR
ADHESIVES 9-17
DISTRIBUTION OF ABA EMISSION REDUCTION
TECHNOLOGIES USED TO ESTIMATE THE SLABSTOCK FOAM
NATIONWIDE REGULATORY ALTERNATIVE COSTS BY MODEL
PLANT 9-20
SLABSTOCK FOAM REGULATORY ALTERNATIVE COSTS 9-22
DISTRIBUTION OF TECHNOLOGIES USED TO ESTIMATE THE
MOLDED FOAM NATIONWIDE REGULATORY ALTERNATIVE
COSTS BY MODEL PLANT 9-25
MOLDED FOAM REGULATORY ALTERNATIVE COSTS . . 9-28
SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC IMPACTS
FOR SLABSTOCK FOAM 9-29
SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC IMPACTS
FOR MOLDED FOAM 9-30
SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC IMPACTS
FOR SLABSTOCK FOAM - SENSITIVITY ANALYSIS . 9-32
XI
-------
1.0 PURPOSE OF DOCUMENT
This Basis and Purpose Document provides background
information on, and rationale for, decisions made by the
Environmental Protection Agency (EPA) related to the proposed
standards for the reduction of hazardous air pollutants (HAP)
emitted during the production of flexible polyurethane foam.
This source category includes both molded and slabstock flexible
polyurethane foam production. This document is intended to
supplement the preamble for the proposed standards.
This document is separated into 10 chapters providing a
combination of background information and EPA rationale for
decisions made in the standards development process. Chapters 2,
3, 5, 7, 8, and 9 provide background information; Chapter 2 is an
introduction, Chapter 3 describes the affected industry,
Chapter 5 presents the baseline HAP emissions, Chapter 7 presents
model plants, Chapter 8 presents the predicted environmental
impacts associated with the regulatory alternatives, and Chapter
9 presents the predicted cost and economic impacts associated
with the regulatory alternatives. Chapters 4, 6, and 10 provide
EPA rationale for subcategorization, determination of MACT
"floors" and development of regulatory alternatives, and
rationale for the selection of the proposed standards,
respectively. Supporting information and more detailed
descriptions for each technical and rationale chapter are
contained in the memoranda referenced in this document and
contained in the project docket.
Supporting information is located in Air Docket A-95-48.
Detailed descriptions of the analyses presented in Chapters 3
through 9 are contained in a series of memoranda assembled in the
Supplementary Information Document (SID), (Docket Item number
II-A-3)-1
1-1
-------
2.0 INTRODUCTION
Section 112 of the Clean Air Act, as amended in 1990 (1990
Amendments) gives the EPA the authority to establish national
standards to reduce air emissions from sources that emit one or
more hazardous air pollutants (HAP). Section 112(b) contains a
list of HAP to be regulated by National Emission Standards for
Hazardous Air Pollutants (NESHAP), and Section 112(c) directs the
EPA to use this pollutant list to develop and publish a list of
source categories for which NESHAP will be developed. The EPA
must list all known source categories and subcategories of "major
sources" that emit one or more of the listed HAP. A major source
is defined in section 112(a) as any stationary source or group of
stationary sources located within a contiguous area and under
common control that emits, or has the potential to emit, in the
aggregate, considering controls, 10 tons per year or more of any
one HAP or 25 tons per year or more of any combination of HAP.
This list of source categories was published in the Federal
Register on July 16, 1992 (57 FR 31576), and includes the
flexible polyurethane foam source category. Flexible
polyurethane foam is used in furniture, bedding, packaging,
carpet cushioning, automobile interiors, medical supplies, and a
variety of miscellaneous other uses.
2-1
-------
3.0 DESCRIPTION OF THE AFFECTED INDUSTRY
The purpose of this chapter is to present a brief
description of the flexible polyurethane foam industry. This
chapter is arranged in several sections. Section 3.1 presents a
general description of the industry, including facility location
and other general statistics. Section 3.2 describes the foam
chemistry. Section 3.3 describes the slabstock foam production
process, including fabrication processes, followed by sections
that describe where HAP emissions occur in the slabstock process,
and control technologies for these emissions. The next sections
present the same information for the molded foam process, from
the production process to emission control. The final sections
discuss the rebond foam process, emission sources, and control
techniques. A more detailed industry description is provided in
the SID.2
3.1 INDUSTRY DESCRIPTION
The flexible polyurethane foam source category is contained
in the initial list of source categories for NESHAP under the
amended Clean Air Act of 1990. In the Environmental Protection
Agency's (EPA) initial source category listing,3 the source
category is defined as follows:
The Flexible Polyurethane Foam Production Source category
includes any facility which manufactures foam made from a
polymer containing a plurality of carbamate linkages in the
chain backbone (polyurethane).
Three types of polyurethane foam facilities appear to fit in this
category description: slabstock flexible polyurethane foam (i.e.,
slabstock foam), molded flexible polyurethane foam (i.e., molded
foam), and rebond foam. Slabstock foam is produced in large
continuous buns that are then cut into the desired size and shape
(fabricated). Slabstock foam is used in furniture, bedding,
packaging, and carpet cushioning. Molded foam is produced by
3-1
-------
"shooting" the foam mixture into a mold of the desired shape and
size. The major use of molded foam, by weight, is in automobile
interiors, but is used in many other applications such as
packaging, novelty applications, and medical supplies. Rebond
foam is made from scrap foam that is converted into a material
primarily used for carpet underlay.
The foam chemistry of the slabstock and molded segments of
the industry is analogous; however, the equipment, production
processes, emission sources, and control techniques are very
different. The rebond foam segment differs from both other
segments in these areas, as well as in the chemistry.
3.1.1 Slabstock Foam Facility Distribution
The EPA estimates that there are 78 slabstock foam
facilities in the United States. Table 3-1 shows the
distribution of foam facilities by state. Data were received
from all 78 facilities through the distribution of an Information
Collection Request (ICR) by the EPA.4 Since slabstock foam is
produced in large "buns," which must be cut into the desired
sizes and shapes, fabrication operations are sometimes co-located
with slabstock foam production facilities. Also, rebond foam
production operations, which use foam scraps as the primary
starting material to produce the foam product, are sometimes co-
located with slabstock foam production facilities.
3.1.2 Molded Foam Facility Distribution
The EPA estimates that there are 228 molded foam production
facilities in the United States. The EPA used several sources to
obtain this estimate. First, ICR responses were received from
46 molded foam facilities. Using the "Polyurethane Industry
Directory and Buyer's Guide - 1994",5 an additional 182 flexible
molded foam companies were identified based on company
descriptions. Table 3-2 presents the distribution of these 228
molded foam facilities by state.
3.1.3 Rebond Foam Facility Distribution
The EPA estimates that there are 52 rebond foam facilities
in the United States.6 Of these rebond facilities, 21 are
3-2
-------
TABLE 3-1. DISTRIBUTION OF SLABSTOCK FOAM
FACILITIES BY STATE
State
Arkansas
California
Delaware
Florida
Georgia
Illinois
Indiana
Iowa
Kansas
Kentucky
Maryland
Massachusetts
Michigan
Mississippi
Minnesota
New Jersey
New Mexico
North Carolina
Ohio
Oregon
Pennsylvania
Tennessee
Texas
Virginia
Washington
Wisconsin
Total
Number of Facilities
2
8
1
5
4
3
8
1
1
1
1
1
2
8
1
2
1
9
2
1
4
5
5
1
1
1
78
3-3
-------
TABLE 3-2. DISTRIBUTION OF MOLDED FOAM
FACILITIES BY STATE
State
Alabama
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Illinois
Indiana
Iowa
Kansas
Kentucky
Maine
Maryland
Massachusetts
Michigan
Minnesota
Number of
Facilities
3
1
1
19
6
8
2
1
4
11
7
5
3
3
1
5
4
29
7
State
Mississippi
Missouri
Montana
Nebraska
Nevada
New Hampshire
New Jersey
New York
North
Carolina
Ohio
Oregon
Pennsylvania
Rhode Island
South
Carolina
Tennessee
Virginia
Washington
West Virginia
Wisconsin
Number of
Facilities
1
7
1
1
2
2
11
9
5
23
2
14
1
2
5
4
4
1
8
TOTAL
228
3-4
-------
located at plant sites that also produce slabstock flexible
polyurethane foam. Data were obtained from the ICR responses for
these rebond operations.
3.2 SLABSTOCK AND MOLDED FOAM CHEMISTRY
This section briefly describes the chemistry involved in
producing slabstock and molded flexible polyurethane foam.
Section 3.3 then discusses the slabstock foam production process.
The primary references for these sections are the ICI
Polyurethanes Book7 and a site visit report to a foam production
facility.8
3.2.1 Chemistry of Flexible Polvurethane Foam: Slabstock and
Molded
Polyurethanes are made by reacting a polyol with a
diisocyanate. The polyol is typically a polyester or a polyether
with two or more -CH2OH functional groups, and the diisocyanate
is usually a mixture of 2,4- and 2,6- isomers of toluene
diisocyanate (TDI). TDI is the most often used diisocyanate in
slabstock foam production, while methylene diphenyl diisocyanate
(MDI) is most often chosen in molded foam production.
Polyurethane foams are made by adding water to the reaction
mixture. Once the ingredients are mixed, two main polymerization
reactions occur. Isocyanate groups react with hydroxyl groups on
the polyol to produce urethane linkages (hence the term
"polyurethane"). The other main reaction is that of the
isocyanate and water. The initial product of the reaction with
water is a substituted carbamic acid, which breaks down into an
amine and carbon dioxide (C02). The amine then reacts with
another isocyanate to yield a substituted urea linkage.
Surfactants and catalysts are also added to the mixture.
The surfactants aid in mixing incompatible components of the
reaction mixture, and also help control the size of the foam
cells by stabilizing the forming gas bubbles. Catalysts balance
the isocyanate/water and isocyanate/polyol reactions, and assist
in driving the polymerization reaction to completion.
The CO2 formed in this reaction acts as the "blowing agent"
(blowing agents will be discussed in more detail later in this
3-5
-------
section) and causes the bubbles to expand. The bubbles
eventually come into close contact, forming a network of cells
separated by thin membranes. At full foam rise, the cell
membranes are stretched to their limits and rupture, releasing
the blowing agent and leaving open cells supported by polymer
"struts."
The final polymer is composed of the urethane and urea
linkages formed in the isocyanate/polyol and isocyanate/water
reactions. The polyol-to-isocyanate urethane linkages provide
strength, and the isocyanate-to-isocyanate urea linkages give the
foam its firmness.
3.2.2 Auxiliary Blowing Agents
As noted in the previous section, one result of the
isocyanate-water reaction is the liberation of C02 gas. The
blowing action of this CO2 is termed "water-blowing," because the
C02 blowing agent is produced from the isocyanate-water reaction.
Many grades of foam can be produced using only this C02 gas as a
blowing agent.
Increasing the amount of water in a formulation generally
produces a lower-density foam, because additional CO2 blowing
agent is produced. However, there is a practical limit to the
amount of water that can be used. First, an increase in the
water level results in an increase in the number of urea linkages
in the final polymer. These linkages tend to make the polymer
stiffer because they undergo hydrogen bonding. Second, the
isocyanate-water reaction is extremely exothermic. An excessive
level of water can cause high temperatures that can scorch the
foam, or even cause the foam to ignite.
As a result, some grades of foam require the use of an
auxiliary blowing agent (ABA). The ABA is mixed with the foam
reactants as a liquid when the reactant mixture is first poured.
As the exothermic polymerization reactions raise the temperature
of the polymer mass, the ABA vaporizes, supplementing the blowing
action of C02 from the water-isocyanate reaction. The
vaporization of the ABA also serves to remove excess heat from
the foam, reducing the potential for scorching or auto-ignition.
3-6
-------
Auxiliary blowing agents are more widely used in the
production of slabstock foams than in the production of molded
foams. The amount of ABA required depends on the grade of foam
being produced and the ABA used. ABAs are most important for low
density and soft foams. In these grades, water-blowing alone
would cause problems with either overheating or with increased
foam stiffness. For low-density foams, ABAs are used in
conjunction with water-blowing to avoid overheating. In the case
of soft foams, ABAs provide blowing action without increasing the
foam's stiffness. The most common ABA used in the foam industry
is a HAP, methylene chloride (MeCl2) .
3.3 SLABSTOCK FOAM PRODUCTION PROCESS
Figure 3-1 depicts a typical slabstock foam production
process. Flexible slabstock foam is produced as a large
continuous "bun" that is later cut into sections with the desired
dimensions. There are variations in the design of the machines
that produce the foam. They may be horizontal or vertical, with
the horizontal foam line being the most common. There are
several types of horizontal foam machines found in foam
facilities. The most common system is called "Maxfoam", which is
described below. Following the description of the Maxfoam line
will be a description of the Vertifoam process, a vertical foam
production process.
3.3.1 Horizontal Maxfoam Production
From bulk chemical storage, raw ingredients are moved to
smaller feed tanks. The chemicals are pumped from the feed tanks
to the mixing head of the foam line where they are vigorously
mixed. The amount of each chemical sent to the mixing head is
carefully controlled by metering pumps. The mixture is
discharged through a mixing head into a trough where the
reactions begin to occur (i.e., "creaming" begins). From this
trough, the froth flows onto the foam tunnel. The mixture
quickly spreads evenly across the width of the tunnel.
The bottom of the tunnel consists of a series of five
adjustable fall-plates that are covered by paper. The foam
reaches its maximum height, or "full rise" about 25 feet from the
3-7
-------
-5.*:
°
_-.
2 S
J5
5 i
§.=•> *
S. S 5 S
0, g VI M
® see
S 033
CA oo m CQ
x a
» •» 2 s
II1?!
•» C t» X "
S e *> « M~S
?i s-s^s
e S ^"T e g
tn £ 0,-g, « =
^ i e 3 s*
IP! 11
SlIlJl
v CM en v m
m
CQ
-------
nozzle. The full rise time is dependent on the grade of foam,
with lower density foams rising highest and at the fastest rate.
Instead of "rising," the foam actually expands downward along the
slope of the fall plates. The sides of the foam tunnel are
vertical conveyors, covered with plastic, which move the foam
down the tunnel. The fall plates are stationary, and it is the
side plastic and the bottom paper that move the foam to the belt
conveyor portion.
The belt conveyor carrying the foam block moves at an
average speed of 15 to 20 feet per minute. Additional time on
the conveyor after full rise is required to allow the
polymerization reactions to be completed so the foam will
solidify. The side papers are then removed from the bun, and the
bun is sawed into the desired lengths. After sawing, the end of
the bun is marked with the foam grade, and the bun moves off the
belt conveyor onto a roller-type conveyor moving at a higher rate
of speed. This conveyor continues through the wall of the
pouring area, through the foam storage area, and then into the
foam curing area.
In the curing area, the buns are removed from the conveyor
with overhead cranes and placed on the floor. Typically, buns
are cured 12 to 24 hours before being moved from the curing area
to foam storage. In the storage area, buns are piled 4 or 5
high. The buns remain in the storage area until ready for
fabrication or shipping.
3.3.2 Vertifoam Production Process Description
In the Vertifoam process, the foam reaction mixture is
introduced at the bottom of a completely enclosed chamber. This
chamber is lined with paper or plastic, which is drawn upwards at
a controlled rate. The rate is dependent on the pressure in the
chamber, the foam formulation, and the rate of production. With
a controlled rate of upward pull, the rheology of the foam
reaction process, combined with the effect of gravity, ensures a
stable foaming front, and prevents the mixing of the still liquid
reacting mixture with the partially gelling foam poured a few
seconds earlier.
3-9
-------
3.3.3 Foam Fabrication
As mentioned earlier, slabstock flexible polyurethane foam
is produced in large buns which are typically 4 feet tall, 8 feet
wide, and 50 to 100 feet long. Prior to being delivered to the
furniture manufacturer or other end-user, the large buns are
"fabricated" according to the end-use. The simplest type of
fabrication is to cut the foam into the desired shape by use of
specialized saws, by hand-cutting, or other techniques. However,
many customers desire foam products that are more "finished" or
complex. To produce such products generally requires the gluing
of foam-to-foam, or foam to some other material such as cotton
batting. The most commonly used adhesives contain methyl
chloroform (a HAP). Since methyl chloroform has also been
identified as an ozone depleting substance, fabricators have been
searching for alternative adhesives. It appears that the most
popular replacements have been MeCl2-based adhesives.
Fabrication operations are sometimes co-located with foam
production operations (i.e., are on-site). Information from the
ICR's revealed that approximately 40 percent of the foam produced
is fabricated on-site. Off-site fabrication facilities fabricate
the remaining 60 percent of the foam produced.
3.4 HAP EMISSION SOURCES FROM SLABSTOCK FOAM PRODUCTION
This section will briefly discuss the HAP emission points
for slabstock foam production and fabrication. The four main
sources of HAP emissions are storage of raw materials, leaking
components in HAP service, the foam tunnel and curing area, and
equipment cleaning. Fabrication emission sources will be
discussed in this section, as well.
3.4.1 Storage Emissions
Raw HAP chemicals are received at foam facilities by
railcar, tank truck, and in drums. Emissions can occur as
working losses during the unloading of the HAP from the railcar
or tank truck. There can also be small amounts of HAP emitted
from the storage tank due to diurnal temperature or pressure
changes.
3-10
-------
3.4.2 Equipment Leaks from Components in HAP Service
There can be small amounts of HAP releases from leaking
components in HAP service. Some examples of components in HAP
service that may leak are pumps, valves, and connectors.
3.4.3 Foam Tunnel and Curing
There are two HAP that are emitted from the foam tunnel and
curing area: MeCl2 used as an ABA, and TDI. As mentioned
earlier, the TDI is a primary reactant in the polyurethane
reaction. There is a small opportunity for TDI emissions at the
point the foam mixture is initially poured on the conveyer.
However, the TDI reacts very quickly, leaving little residual TDI
to be emitted.
MeCl2 is the principal ABA used, and its role is simply to
volatilize and expand the foam. Therefore, all of the MeCl2 that
is added is eventually emitted. The MeCl2 ABA is emitted in
three primary areas. The first is in the foam tunnel, where the
increasing temperatures from the exothermic isocyanate-water
reaction cause the MeCl2 to volatilize. A significant amount of
MeCl2 blowing agent remains in the foam after the tunnel, and is
emitted when the bun is cut into sections, as well as while in
the curing area. Estimates place the general distribution of
these ABA emissions at 30-40 percent in the foam tunnel and 40-55
percent in the curing area.9
3.4.4 Equipment Cleaning
HAP are also emitted through the use of HAP cleaning
solvents. Methylene chloride is used as a cleaner to rinse
and/or soak foam machine parts such as mixheads and foam troughs
at the end of a pour. Hardened foam residue forms on the trough,
fall plates, and other equipment, and must be removed after each
produc t i on run.
3.4.5 Fabrication
The HAP emissions in fabrication operations occur due to the
use of HAP-based spray adhesives for gluing fabric to foam, or
foam to foam. Fabrication covers the broad range of die cut
parts, cut parts, and glued parts, and not all fabrication
involves gluing..
3-11
-------
3.5 CONTROL TECHNOLOGIES FOR SLABSTOCK HAP EMISSIONS
The EPA reviewed information from the ICR responses that
were received from flexible polyurethane foam producers, as well
as the information contained in other pertinent project files, to
identify potential HAP emission reduction and control
technologies. Based on their findings, the EPA created a report
entitled "Flexible Polyurethane Foam Emission Reduction
Technologies Cost Analysis".10 This report will hereafter be
referred to as the Cost Report. The technologies investigated in
the cost report will briefly be identified and discussed in this
section.
3.5.1 Control Technologies for ABA Emissions
There were several alternatives identified to either reduce
or eliminate the use of MeCl2 as an ABA in the manufacture of
slabstock flexible polyurethane foam. The technologies
identified were acetone or liquid CO2 as an ABA, foaming in a
controlled environment, forced cooling, chemical modifications,
and carbon adsorption.
3.5.2 Reducing Releases From Chemical Storage and Handling,
Equipment Cleaning, and Components in HAP Service
3.5.2.1 Chemical Storage and Handling
There were two methods identified for reducing HAP emissions
from this source: carbon canisters and a vapor balance system.
Carbon canisters control emissions by capturing the vapor
released during unloading or storage. The vent systems for the
storage tanks lead to this carbon canister that is filled with
activated carbon.
Vapor balancing is another method frequently used on TDI
storage tanks to control unloading emissions. When a TDI storage
tank is filled, the vapors are vapor-balanced. That is, the
vapors present in the tank are forced out by the incoming liquid
and are routed back to the railcar or tank truck using piping.
3.5.2.2 Components in HAP Service
Two methods for controlling equipment leaks from components
in HAP service were identified: equipment modifications and leak
detection and repair (LDAR) programs. The equipment modification
3-12
-------
identified for the slabstock industry was the use of leakless
pumps for TDI and MeCl2. Leak detection and repair programs
require the periodic monitoring of components to detect and
repair leaks.
3.5.2.3 Equipment Cleaning
Methylene chloride is used as a cleaner to rinse and/or soak
foam machine parts such as mixheads and foam troughs. The two
alternatives for eliminating these HAP emissions which were
identified were steam cleaning and non-HAP cleaners.
3.5.3 Reducing Releases From Fabrication/Repair Operations:
Fabrication Adhesives and Molded Foam Repair
HAP-based adhesives are used in both slabstock and molded
foam facilities. As the reduction alternatives are the same for
both subcategories, they will be discussed together. In
slabstock facilities, spray adhesives are used to glue fabric to
foam, or foam to foam. The main use of adhesives in molded foam
facilities is for the repair of voids and tears in the molded
pieces. Three alternatives were identified that might eliminate
HAP emissions from the use of adhesives. These are (1) hot-melt
adhesive, (2) water-based adhesives, and (3) Hydrofuse.
3.6 MOLDED FOAM PRODUCTION PROCESS
Figure 3-2 illustrates a typical molded foam production
line. The primary references used in this section were the ICI
Polyurethanes Book6 and a site visit report to a molded foam
production facility.11 The production line includes multiple
molds, with each mold consisting of top and bottom sections,
joined by hinges. The molds are mounted on a circular or oval-
shaped track. Both the molds and the track can vary broadly in
size. The molds travel around the track, and the necessary
process operations are performed at fixed stations. The
following paragraphs describe a basic molding cycle.
The first step in the molding cycle is the application of
mold release agent. This is a substance that is applied to the
mold to facilitate removal of the foam product. After the mold
release agent is applied, any special components to be molded
3-13
-------
00
"I
E
V
00
.5
•s
e
I
/ I!
CO
ra
-------
into the foam are placed in the mold. These might include
covers, springs, or reinforcing materials.
Raw materials, including polyol, diisocyanate, water,
catalyst, and surfactant are all pumped to a common mixhead in
predetermined amounts. The mixhead injects a precisely measured
"shot" of raw material into each mold. There are two types of
mixheads used in the industry, high-pressure (HP) and low-
pressure (LP). The two types of mixheads have different cleaning
requirements, resulting in a dramatic difference in overall
emissions from the process, which will be discussed in the HAP
emission section that follows.
The mold is closed and the polymerization reaction occurs,
producing a foam product that fills the mold. Most molded foams
are produced without any ABA, using only the blowing action of
C02 gas from the water-isocyanate reaction. After curing, the
molds are opened and the product is removed. The mold is then
cleaned and starts the circuit again.
Another important variety of molded foam is the integral
skin foam, also known as a self-skinning foam. An integral skin
foam is a foam with a dense, tough outer surface. The skin is
produced by overpacking the mold and using an ABA, usually
Freon-11. Unlike other types of molded foams, integral skin
foams require an ABA. The skin production is also driven by the
temperature gradient between the center of the foam mass and the
relatively cooler surface of the mold. Integral skin foams are
used in such products as steering wheels and footwear.
3.7 MOLDED FOAM HAP EMISSION SOURCES
This section will briefly discuss the HAP emission points
for molded foam production. The three main areas of HAP
emissions from molded foam production are mixhead flush, mold
release agents, and repair operations (from adhesive use).
3.7.1 Mixhead Flush
Methylene chloride emissions for flushing of LP mixheads is
the largest emission source for flexible molded foam manufacture.
With LP mixheads, the chemical streams enter the mixing chamber
at approximately 40 to 100 psi, and are blended by rotating mixer
3-15
-------
blades before being released or "shot" into the mold. Residual
materials can remain in the chamber, as well as on the blades.
This material needs to be cleaned out, either after every shot,
or after several, depending on the conditions. Flushing is
necessary because the residual froth can harden and clog the
mixhead or can interfere with the necessary precision required
regarding the volume of the foam shot.
3.7.2 Mold Release Agents
Mold release agents were another source of HAP emissions
from molded foam. Mold release agents are sprayed on the mold
surface before the foam mixture is poured into the mold, to
prevent adhesion and create a smooth surface. Traditional mold
release agents consist of a resin in a solvent carrier,
frequently methylene chloride or 1,1,1-trichloroethylene (methyl
chloroform), which are both HAP. The carrier evaporates, leaving
the resin, which prevents the foam from sticking to the mold.
3.7.3 Molded Foam Repair
Once a foam piece has been removed from the mold, it is
inspected for tears or holes. If repair is needed, scrap foam
pieces, or the original piece that stuck to the mold, are glued
to fill in the void. A HAP-based adhesive may be used for this
process, with the carrier solvent being a HAP. The emissions
occur when the solvent carrier evaporates after the adhesive is
applied.
3.8 HAP CONTROL TECHNOLOGIES FOR MOLDED FOAM
As was discussed in the HAP control technologies section for
slabstock, the EPA identified emission reduction and control
techniques for the flexible molded polyurethane foam industry.
Mixhead flush and mold repair emission reduction technologies are
identified below. The emission reduction technologies for repair
adhesives are the same as for slabstock fabrication, and were
discussed in section 3.5.3. A more complete discussion of the
HAP control technologies can be found in the industry description
memorandum.
3-16
-------
3.8.1 Control Technologies for Reducing Releases from Mixhead
Flushing
As noted in the discussion of HAP emission sources,
methylene chloride emissions for flushing of LP mixheads was the
largest emission source for flexible molded foam manufacture.
Several technologies were identified that could reduce or
eliminate this source of HAP emissions for the molded foam
producer. These technologies include non-HAP flushes, HP
mixheads, self-cleaning mixheads, and solvent recovery units.
3.8.2 Control Technologies for Reducing Releases of Mold Release
Agent
As mentioned in the earlier section on HAP emission sources
for molded foam, emissions occurred from the evaporation of the
carrier solvent from mold release agents. Alternatives being
used, or being investigated, by the industry include water- or
naphtha-based agents, and reduced-VOC solvent agents.
3.9 REBOND FOAM PRODUCTION
Another flexible foam product is rebond foam. Rebond foam
is produced at flexible foam production facilities, as well as at
stand-alone, or off-site facilities. Rebonding is a process
where scrap foam is converted into a material that is used for
carpet underlay and several other end-uses such as school bus
seats.
3.9.1 Rebond Foam Process
The scrap foam may have been generated at the facility from
its slabstock operations, or may have been shipped or bought from
other foam facilities. There is such a high demand for this
product that foam scrap is imported from overseas. The scraps
are received in "bales." The baled foam is foam "chewed" into
smaller pieces. These small pieces are loaded into a blender,
where a mixture of polyol and TDI is added. The foam and binder
mixture, and occasionally a dye, is poured into a cylindrical
mold, that is below floor level. This mold has a central core so
that there is a hole that runs the length of the cylinder.
Pressure and steam are applied to the mixture in the mold, and
3-17
-------
then the roll is taken out of the mold and allowed to cool, or
"set," for about 24 hours.
3.9.2 Rebond Emission Sources
Three HAP emission points were identified at rebond
operations. First, a small amount of TDI emissions occur at the
molding pit area where foam pieces, TDI, and polyol are subj ectecl
to pressure and steam. Also, one rebond facility reported the
use of a MeCl2-based mold release agent, and the use of MeCl2 as
an equipment cleaner.
3.9.3 Rebond Control Techniques
Since only one facility reported the use of an MeCl2-based
mold release agent and the use of MeCl2 as an equipment cleaner,
the EPA assumes that other products are available to accomplish
these same functions, without emitting HAP. Further, the EPA was
informed that the facility that originally reported the use of
these HAP products had discontinued their use.12 There were no
methods identified to control TDI emissions from rebond
operations.
3-18
-------
4.0 SUBCATEGORIZATION OF THE LISTED SOURCE CATEGORY
The purpose of this chapter is to present considerations and
conclusions regarding the subcategorization of the flexible
polyurethane foam production source category. The first section
identifies potential reasons for subcategorizing a source
category, the second section presents brief descriptions of the
operations and hazardous air pollutant (HAP) emission sources
associated with flexible polyurethane foam source production and
related processes, and the third section presents EPA's rationale
for the selection of subcategories for this industry. The final
section summarizes this subcategorization decision.
4.1 SUBCATEGORIZATION CONSIDERATIONS
Subcategories, or subsets of similar emission sources within
a source category, may be defined if technical differences in
emissions characteristics, processes, control device
applicability, or opportunities for pollution prevention exist
within the source category.13
There are three distinct production processes associated
with flexible polyurethane foam: slabstock foam production,
molded foam production, and rebond foam production. Each will be
briefly discussed below, along with slabstock foam fabrication.
4.2 PROCESS AND HAP EMISSION DESCRIPTIONS
This section briefly describes the molded, slabstock,
fabrication, and rebond processes, including the main HAP
emission points, as well as a brief discussion of flexible
polyurethane foam chemistry. More details on the foam chemistry,
production processes, HAP emissions, and control technologies can
be found in Chapter 3 of this document, and in the Industry
Description Memorandum.1
4-1
-------
4.2.1 Polyurethane Foam Chemistry
Flexible polyurethane foam is produced by mixing three major
ingredients: a polyol polymer, an isocyanate, and water. When
the polyol, diisocyanate, and water are mixed, two main
polymerization reactions occur. Isocyanate groups react with
hydroxyl groups on the polyol to produce urethane linkages (hence
the term "polyurethane"). The other main reaction is that of the
isocyanate and water, which forms a urea linkage and CO2. The
CO2 formed in this reaction acts as the "blowing agent" and
produces bubbles, causing the foam to expand to its full volume
within minutes after the ingredients are mixed and poured. The
final polymer is composed of the urethane and urea cross-linkages
formed in the isocyanate/polyol and isocyanate/water reactions.
4.2.2 Slabstock Polyurethane Foam Production
Flexible slabstock foam is produced as a large continuous
"bun" that is cut into sections with the desirable dimensions.
The major HAP emission source at slabstock facilities is from the
use of MeCl2 as an ABA. Methylene chloride's role is simply to
volatilize and expand the foam, not directly participate in the
polyurethane reaction. Therefore, all of the methylene chloride
that is added is eventually emitted. Other HAP emission
sources at slabstock production facilities include unreacted TDI
from the foam tunnel (very small amount), emissions from leaking
TDI and MeCl2 pumps, valves, and other equipment; and equipment
cleaning.
4.2.3 Molded Polyurethane Foam Production
Molded foam production uses somewhat different chemical
formulations from those used for slabstock foam production,
although the basic polyurethane foam reaction is the same.
Molded flexible foams have higher densities than the slabstock
flexible foams and, therefore, seldom use an ABA. In contrast to
the slabstock process, the molding method is an intermittent
batch process where the raw ingredients are placed in a mold and
allowed to react.
After a foam piece is removed from the mold, it is generally
trimmed and inspected for tears or holes, and any tears and/or
4-2
-------
holes are repaired. Repair operations are carried out at glue
stations, which may by equipped with local ventilation systems to
remove solvent vapors emanating from the glue.
Methylene chloride emissions for flushing of LP mixheads is
the largest emission source for flexible molded foam manufacture.
Mold release agents are another source of HAP emissions from
molded foam. Traditional mold release agents consist of a resin
in a solvent carrier, frequently methylene chloride or 1,1,1-
trichloroethylene (methyl chloroform), both HAP. If repair of
the molded piece is needed, scrap foam, or the original piece
that stuck to the mold, are glued to fill in the void. A HAP-
based adhesive may be used for this process, with the carrier
solvent being a HAP.
4.2.4 Fabrication Operations
As mentioned earlier, slabstock flexible polyurethane foam
is produced in large buns which are typically 4 feet tall, 8 feet
wide, and 50 to 100 feet long. Prior to being delivered to the
furniture manufacturer or other end-user, the large buns are
"fabricated" according to the end-use. The simplest type of
fabrication is to cut the foam into the desired shape by use of
specialized saws, by hand-cutting, or other techniques. However,
many customers desire foam products that are more "finished" or
complex. To produce such products generally requires the gluing
of foam-to-foam, or foam to some other material such as cotton
batting. The most commonly used adhesives are either methyl
chloroform based, or MeCl2-based.
4.2.5 Rebond Foam Production
Rebonding is a process where scrap foam is converted into a
foam product that is used for carpet underlay and several other
end-uses. The scrap foam is converted into small pieces, which
are loaded into a blender, and a mixture of polyol and TDI is
added. The foam and binder mixture is poured into a cylindrical
mold. Pressure and steam are applied to the mixture in the mold,
and then the roll is taken out of the mold and allowed to cool or
"set" for about 24 hours. There is the potential for very small
TDI emissions during the process, but the largest potential for
4-3
-------
HAP emissions at rebond facilities is from the use of MeCl2 as an
equipment cleaner.
4.3 RATIONALE FOR SUBCATEGORIZATION WITHIN THE FLEXIBLE
POLYURETHANE FOAM PRODUCTION SOURCE CATEGORY
As is evident from the information presented in the
paragraphs above, the only characteristic that molded and
slabstock foam share is a similar chemistry to produce a flexible
polyurethane foam product. Further, the rebond foam process is a
dissimilar process, with the only similarity being between the
final products. Therefore, the EPA concludes that the flexible
polyurethane foam production industry should be separated into
three distinct subcategories, slabstock and molded flexible
polyurethane foam production, and rebond foam production. While
the foam chemistry and final products are similar, the equipment,
emission sources, and control techniques are very different.
Molded and rebond foam are manufactured in batch-type processes,
while slabstock is made in a continuous method. The major
emission source for slabstock foam is from the use of ABA, and
there is no analogous emission point for either molded or rebond
foam. The only significant HAP emission point that the three
segments share is equipment cleaning (mixhead cleaning for
molded), and the reasons for these emissions, and the control
technologies that could be used, are very different. Therefore,
the three segments are treated as three separate subcategories
for the purpose of this rulemaking.
During its investigations, the EPA became aware of flexible
polyurethane foam fabrication operations, which are sometimes co-
located with slabstock foam production operations (i.e., are
located on-site). The EPA also obtained information related to
HAP emissions from fabrication operations in the ICR responses.
The total reported HAP emissions from fabrication operations at
the 69 facilities with on-site fabrication operations was
1,382 tons per year. Therefore, the average fabrication
operation HAP emissions was around 23 tons per year of a single
HAP.
4-4
-------
As noted earlier, the ICR responses indicated that
approximately 40 percent of the foam produced is fabricated on-
site, meaning that the majority of slabstock foam is fabricated
at sites that do not also produce slabstock foam. Given the
emissions information for on-site fabrication operations noted
above, the EPA concluded that the potential for HAP emissions
from foam fabrication is enormous. Therefore, the Agency decided
that further investigation of this segment of the foam
fabrication industry is necessary.
Due to the relationship between foam production and foam
fabrication, and since some information had already been
collected for foam fabrication (at operations co-located with
slabstock production operations), the Agency considered expanding
the flexible polyurethane foam production source category to
include foam fabrication. However, this option was rejected for
several reasons. First, the EPA was not able to determine if the
information collected for on-site facilities was representative
of the entire fabrication industry. Since the information
collected was for on-site fabrication operations owned by foam
producers, it was believed that this information may not have
been representative of smaller, independent fabricators, and it
was estimated that there could be as many as 2,000 small
independent fabricators in the U.S.14 Obtaining general
information on the fabrication industry in a timely manner was
made more difficult, since no trade organization was identified
that represents the fabrication industry. In addition, the EPA
became aware of fabrication operations using HAP-based adhesives
at rebond facilities, which were not represented in the original
data. The EPA concluded that the investigation of the foam
fabrication industry that was needed to allow the development of
comprehensive and appropriate standards could not be accomplished
in the schedule for the foam production source category.
Therefore, the EPA has listed flexible polyurethane foam
fabrication as a separate source category. The scheduled
promulgation date for the fabrication source category will be
November 15, 2000. This will allow the EPA to have time to
4-5
-------
gather more information on stand-alone fabrication facilities, as
well as to work with States and industry on identifying potential
control options specifically for fabrication. On-site and off-
site facilities and their similarities and differences can be
considered during the development of the foam fabrication
regulation.
4.4 SUMMARY
In summary, the EPA has separated the flexible polyurethane
foam production source category into three subcategories:
slabstock foam production, molded foam production, and rebond
foam production. During the analysis of the foam production
industry, the EPA became aware of foam fabrication operations
that emit HAP, and has listed flexible polyurethane foam
fabrication as a major source category. The proposed standards
do not address the foam fabrication source category, as
regulation of this new fabrication source category will take
place on a different schedule.
4-6
-------
5.0 BASELINE EMISSIONS
This chapter presents the baseline HAP emissions for the
flexible polyurethane foam production source category. As
discussed in Chapter 4, there are three subcategories of flexible
polyurethane foam: molded, slabstock and rebond. Baseline HAP
emissions for slabstock foam are presented in Table 5-1. The
same information is provided in Table 5-2 for molded foam. As
shown in these tables, the total nationwide estimated HAP
emissions are over 16,500 tons per year (15,000 Mg/yr) for the
slabstock subcategory, and almost 3,200 tons per year (2,900
Mg/yr) for molded foam. Therefore, over 19,700 tons per year
(17,950 Mg/yr) of total HAP are emitted from the source category,
including 2.5 tons per year (2.3 Mg/yr) from rebond foam
production.
As described in Section 3.4, in the manufacture of slabstock
foam, HAP are emitted from storage and unloading, equipment
leaks, ABA usage and other emission sources involved in foam
production and equipment cleaning. The use of ABA during foam
production comprises the largest portion of these emissions,
making up over 98 percent (16,250 tons per year) of total
slabstock HAP emissions.
The HAP emitted from the manufacture of slabstock foam
include MeCl2, methyl chloroform, propylene oxide, and TDI. For
the purpose of establishing baseline emissions, it was assumed
that the use of methyl chloroform as an ABA will be phased out,
and replaced by MeCl2. Propylene oxide is contained in very small
amounts as a stabilizer in MeCl2, and was not included in the
baseline emission estimates.
Section 3.7 describes the HAP emission sources for molded
foam. The major emission points are mixhead flushing, mold
release agents, and foam repair. The use of a HAP mixhead flush
5-1
-------
TABLE 5-1. BASELINE HAP EMISSIONS FOR
SLABSTOCK FOAM PRODUCTION
Emission Source
Baseline HAP Emissions
(tons/yr)
Chemical Storage/Unloading
Equipment Leaks
Foam Production
ABA
Other
Equipment Cleaning
TOTAL
17
162
16,250
9
130
16,568
5-2
-------
TABLE 5-2. BASELINE HAP EMISSIONS FOR MOLDED FOAM PRODUCTION
Emission Source
Chemical Unloading/Storage
Equipment Leaks
Day Tanks
Foam Production
Dispensing
Mixhead Flush
Mold Release Agent
Demolding
In-mold Coating
Other Production
Equipment Cleaning
Foam Repair
TOTAL
Baseline HAP
Emissions (tons/yr)
10
55
25
67
2,561
287
12
32
11
10
116
3,186
5-3
-------
is the largest emission source, making up over 80 percent
(2,561 tons per year) of the total molded HAP emissions.
The HAP emitted from the manufacture of molded foam include
MeCl2, methyl chloroform, MDI, and TDI. Small amounts of
diethanol amine (DEOA) were reported, because DEOA is used as an
additive. Methyl ethyl ketone (MEK) and toluene were reported
primarily by a few facilities as being used as a carrier for in-
mold coatings. Methanol was predominantly reported as a mixhead
flush and as an equipment cleaner. Other HAP were also reported,
but in very small quantities.
The HAP potentially emitted from rebond facilities are TDI
and MeCl2. TDI is a reactant used to adhere the foam scraps
together. MeCl2 can be used as an equipment cleaner at rebond
facilities, and MeCl2-based mold release agents were also
reported to have been used in the past.
The primary basis for baseline emission estimates for this
industry was information submitted to the Environmental
Protection Agency (EPA) by the flexible foam manufacturers in
response to information collection activities conducted under the
EPA's Section 114 authority. However, there were several
instances where the direct use of the ICR response data as the
nationwide baseline emissions was not appropriate. Two different
methods were used: the extrapolation of model plant emission
estimates for primary emission sources, and the extrapolation of
ICR emission estimates for minor emission sources. A brief
summary of the sources of the baseline emissions for each
subcategory follows. A more detailed description of the
determination of baseline emissions is provided in the SID.15
5.1 SLABSTOCK BASELINE EMISSION CALCULATIONS
For HAP ABA emissions and equipment cleaning, it was
believed that the information provided in the ICR responses
represents the nationwide emissions. For these two emission
points, the combination of model plant baseline emissions and the
estimate of the number of facilities represented by each model
plant were determined so that the nationwide emissions calculated
5-4
-------
from the model plants matched the ICR responses as closely as
possible.
For baseline HAP emissions from storage/unloading and
equipment leaks, model plant emission estimates were calculated,
then were multiplied by the number of facilities represented by
each of the model plants to obtain nationwide estimates. For the
model plants, unloading (working loss) emissions were calculated
using AP-42 emission factors for chemical storage tanks.16
Since some model plants assume storage tank controls, the
baseline HAP emissions consist of a combination of controlled and
uncontrolled emissions.
For equipment leaks, emission information was submitted in
the ICR responses for pumps and valves. However, the assumptions
made in calculating these emissions were inconsistent with
typical EPA assumptions, and no information was received for
other components in HAP service (flanges/connectors, open-ended
lines, or pressure relief valves). Therefore, model plant
emissions from components in HAP service were calculated using
assumed component counts and synthetic organic chemical
manufacturing industry (SOCMI) average emission factors.17 The
nationwide estimates were then calculated by multiplying the
model plant emissions by the number of facilities represented by
each model plant.
5.2 MOLDED BASELINE EMISSION CALCULATIONS
ICR responses were received from 46 of the estimated 228
nationwide molded foam facilities. Therefore, it was necessary
to extrapolate information from these facilities to approximate
nationwide HAP emissions. Two different methods were used to
accomplish this estimation. The first was to calculate
nationwide emissions using model plants, and the second was to
extrapolate directly from the ICR response totals. Both methods
are discussed briefly below.
Emissions from only three sources were consistently reported
by all molded foam producers. These sources were mixhead flush,
mold release agents, and adhesives used in foam repair. These
were the only three emission sources included in the molded foam
5-5
-------
model plants. The model plant emissions for these three sources
were then multiplied by the number of facilities represented by
each model plant to obtain the nationwide baseline emission
estimates.
As mentioned earlier, there were many other minor HAP
emission sources identified in the ICRs. However, the emissions
were very small, and the occurrence of the emission points was
too inconsistent to accurately create a model for these
emissions. Therefore, baseline HAP emissions were estimated by
direct extrapolation of the emission information reported in the
ICR responses. This extrapolation was based on the percentage of
facilities reporting emissions from each emission source, and an
assumption that the same percentage of the remainder of the
industry would also report comparable emissions from the same
type of source.
5.3 REBOND BASELINE EMISSION ESTIMATES
The 21 rebond facilities co-located with slabstock
production facilities reported a total of 1.0 tons per year of
HAP emissions. As noted earlier, one facility originally
reported emissions from a HAP-based mold release agent and a HAP
cleaner, but it has subsequently discontinued the use of these
products. Therefore, the total nationwide HAP emissions from the
production of rebond foam is 2.5 tons per year, which is a linear
extrapolation of the emission estimates for the 21 facilities to
the estimated 52 nationwide facilities.
5-6
-------
6.0 MACT FLOORS AND REGULATORY ALTERNATIVES
This chapter presents the results of the maximum achievable
control technology (MACT) "floor" determination for the flexible
polyurethane foam source category. Sections 6.1 through 6.4
discuss the determination of MACT floors. Following the
presentation of the MACT floors, the control options more
stringent than the MACT floors are identified. The final section
discusses the construction of regulatory alternatives, and
presents the regulatory alternatives considered for the flexible
polyurethane foam industry.
6.1 CLEAN AIR ACT (CAA) REQUIREMENTS FOR MACT FLOORS
Section 112(d) of the CAA, as amended in 1990, defines a
minimum level of control referred to as the "MACT floor," for
standards established under Section 112(d). For new sources,
emission standards "shall not be less stringent than the emission
control that is achieved in practice by the best controlled
similar source." For existing sources, the emissions standards
must be at least as stringent as either "the average emission
limitation achieved by the best performing 12 percent of the
existing sources," or "the average emission limitation achieved
by the best performing 5 sources" for categories or subcategories
with less than 30 sources. The EPA has interpreted the term
"average emission limitation" in the statute to mean a measure of
central tendency, such as the arithmetic mean, median or mode.
6.2 CONSIDERATIONS IN DETERMINING MACT FLOORS
There are several fundamental decisions that must be made
before the MACT floor can be determined. These decisions are
discussed below.
6.2.1 Subcategorization
Since a separate MACT floor must be developed for each
subcategory, the first thing that must be determined is if
6-1
-------
subcategorization of the industry is warranted. As discussed in
Chapter 4, the flexible polyurethane foam production source
category was divided into three subcategories: slabstock foam
production, molded foam production, and rebond foam production.
6.2.2 Major Source Determination
The next step was to identify the major sources in each
subcategory. The facility-wide hazardous air pollutant (HAP)
emission totals reported in the ICR responses were used to
identify the major sources within each subcategory, along with a
facility's "potential to emit" (PTE).
A facility's PTE is calculated by considering all the
emission source types at a facility, using assumptions that would
provide the maximum emissions expected for an annual period for
conditions such as highest number of operational hours, highest
HAP content, etc. Inherent limitations based on a facility's
operations can be considered, such as the production rate being
limited by storage space. However, operational practices that
reduce emissions are not considered unless they are due to a
federally enforceable requirement. For example, if some
facilities used a non-HAP solvent for cleaning while other,
similar facilities used a HAP solvent for the same purpose, in
the determination of PTE, it must be assumed that all facilities
use a HAP solvent, unless a facility is prohibited from using the
HAP solvent by a federally enforceable requirement.
Within the molded foam segment, all HP facilities reported
HAP emissions below the major source thresholds. It is not
believed that any HP facilities could be considered major sources
based on their PTE. Only a conversion from HP to LP mixheads
would increase these facility's PTE above the major source
threshold. This conversion would be expensive, as well as
impractical, requiring significant operational and equipment
changes.
Five LP facilities are major sources based on reported
emissions.3 The remaining facilities are also major sources due
to their PTE. While the reported HAP emissions were below the
major source thresholds for these facilities, no federally
6-2
-------
enforceable requirements were reported that would prohibit any
facility from switching to a HAP solvent to flush their mixhead.
Based on the available information, it would be expected that the
combination of the use of a HAP mixhead flush and maximum
potential operation would cause HAP emissions at all LP molded
facilities to be above the major source level.
Seven slabstock facilities reported actual emissions below
the major source thresholds.3 However, it is possible for these
facilities to change their product mix, or types of foam
produced, in a manner that would make their HAP emissions greater
than the major source cutoffs. There were no federally
enforceable limits identified limiting their HAP emissions.
Therefore, all slabstock production facilities were considered to
be major sources.
It was assumed that no rebond foam production process
emitted HAP above the major source levels. However, the 21
rebond processes co-located with slabstock production operations
are considered major sources, since the plant-wide emissions are
above the major source thresholds. It was assumed that the
remaining 31 rebond facilities were area sources.
6.2.3 Grouping of Emission Sources
After the subcategories and major sources within them were
identified, the groups for which separate MACT floors were to be
determined were established. Under each subcategory, individual
emission points were separated into several general emission
source types. Additional grouping decisions were then made in
the determination of MACT floors within each emission source
type. Consideration was given to the following: equipment type,
equipment size, equipment contents, stream characteristics, and
other elements that can affect the emission potential or the
ability to reduce emissions from that point.
The following are the emission source groupings chosen for
molded facilities:
• storage (including unloading emissions)
• in-process vessels (includes day and mix tanks)
• components in HAP service (e.g., pumps and valves)
6-3
-------
• in-mold coatings
• foam reactant dispensing
• mixhead cleaning
• mold release agent application
• demolding
• foam repair
For slabstock facilities the following emission source
groupings were chosen:
• storage (including unloading emissions)
• in-process vessels (includes day tanks)
• components in HAP service
• ABA related emission sources (e.g., foam tunnel,
curing, or foam storage)
• equipment cleaning
In addition, MACT floors were determined for three emission
sources at rebond operations.
• TDI emissions from rebond production
• equipment cleaning
• mold-release agent application
6.2.4 Approach to Determining the MACT Floor
MACT floors were identified for each emission source
grouping within each subcategory. For existing sources, the MACT
floor levels were established by determining some measure of
central tendency of the emission control for the top 12 percent,
or top 5, facilities in each emission source type in each
subcategory. This "average" emission limitation is expressed in
several different manners for different emission source types.
When possible, each MACT floor for existing sources was
expressed as emission limits that represent the average emission
limitation achieved by the top 12 percent. Where the MACT floor
was determined to be a technology or work practice, performance
criteria were defined that best characterized the "average" means
of HAP reduction for the top 12 percent.
For new sources, the MACT floor levels were established by
determining the emission control for the best controlled facility
in the subcategory for each emission source type. The formats of
6-4
-------
the MACT floors for new sources are consistent with those of
existing sources for each emission source type (e.g. work
practice standards, equipment specifications, etc.).
6.3 MACT FLOOR CONCLUSIONS
Based on existing EPA policy and on the information
presented in the ICRs, the EPA's conclusions on what the MACT
floors are for all emission source types identified in each
subcategory are described below. All emissions reported and used
in calculating the MACT floors were taken from the ICRs, and are
generally based on 1992 information.
6.3.1 Molded Foam
As mentioned previously, all HP molded facilities were
identified as area sources, and all LP facilities were identified
as major sources. Therefore, the MACT floors for molded foam are
based only on LP facilities. In the LP molded foam subcategory,
information was available for a total of 19 facilities, which
reported annual HAP emissions of 256 tons of HAP per year. Since
information was available for less than 30 LP facilities, the
MACT floor for each emission source type was based on the top
five performing facilities. The top five facilities were
determined on a case by case basis for each emission source type;
therefore, the same 5 facilities were not always used in the
floor determinations.
6.3.1.1 Storage/unloading
HAP emissions from storage and unloading at LP molded foam
facilities were reported to be less than 1 ton per year. Only
11 facilities provided information on their storage emissions,
and they all reported no control for these storage and/or
unloading emissions. Since all facilities reported no control,
the MACT floor for storage/unloading for both new and existing LP
molded foam facilities was concluded to be no control.
6.3.1.2 Mixhead flush
HAP emissions from mixhead cleaning occur when HAP solvents
are used to "flush" the mixhead between foam shots to remove
residual foam material. Total HAP flush emissions from this
6-5
-------
source were reported at 227 tons/year, approximately 89 percent
of the total reported HAP emissions for the LP facilities.
Unlike the other emission source types for the LP molded
subcategory, the identification of the five "best performing"
facilities for the mixhead flush emission source type was not
straightforward. The reported HAP emissions varied among
facilities from less than 1 ton per year to almost 60 tons per
year. In order to assist in the comparison of facilities,
several approaches were considered. The first was an emission
factor approach based on the reported HAP emissions per weight of
product. However, this does not provide a legitimate means of
comparison, because the number of flushes (and consequently, the
resulting HAP emissions) is dependent on the number of "pieces"
produced and not the weight of foam produced. Thus, producers of
small foam parts would have artificially high HAP emission
factors. Another approach was to simply consider the annual
reported HAP emissions. While this is more reflective of a
facility's HAP reducing activities, it does not take into account
substantial differences in the operating schedules reported, or
the size of a facility. Therefore, the annual flush emissions
were divided by the annual hours of operation. This HAP
emissions per hour of operation factor was used to identify the 5
"best performing" facilities.
None of the 5 facilities identified as the best performing
reported any control techniques in the ICR responses. In each
case, process-specific factors were responsible for the low
emissions. For example, one facility designed its molds to be
closer together than normal, reducing the frequency of flushing
needed.18 Another facility had such small pieces that the
volume of flush needed per shot was very small, resulting in
lowered total emissions.19 Therefore, the EPA's conclusion for
this emission source type is that the average emission limitation
of the best performing 5 facilities is no control. These
facilities are the best performing because of unique process
considerations that are not applicable at all LP molded foam
facilities.
6-6
-------
6.3.1.3 Mold release agents
Approximately 3 percent of the reported HAP emissions from
LP molded foam facilities was from the evaporation of the HAP
carrier solvent from mold release agents.3 When the mold release
agents are sprayed on the mold, the carrier is designed to
evaporate, leaving a waxy material on the mold to prevent the
foam from sticking.
Of the 18 facilities reporting the use of mold release
agents, 10 reported using a non-HAP based agent, resulting in
zero HAP emissions. The remaining eight facilities reported HAP
emissions, and no controls. Since the top 5 facilities all used
non-HAP based mold release agents, the floor for mold release
emissions for both new and existing sources was judged to be the
total elimination of the use of HAP-based mold release agents.
6.3.1.4 Foam repair
The main use of adhesives in molded foam facilities is for
the repair of voids and tears in the molded pieces. The
adhesives used are approximately 20 to 40 percent solids, while
the remainder consists of a solvent carrier, such as methyl
chloroform or MeCl2• The HAP emissions occur as this solvent
carrier evaporates after the adhesive is applied. There were
4.5 tons of HAP emissions reported from this source at LP molded
facilities (less than 2 percent of total reported HAP emissions).
Eight facilities did not report having any repair
operations. Four of the remaining facilities reported HAP
emissions from this activity, with no control identified. The
remaining seven facilities reported repair operations, but zero
HAP emissions. It was assumed these eight facilities were using
a non-HAP based adhesive. Since the top 5 facilities were
assumed to be using a non-HAP based adhesive, the MACT floor for
foam repair for both new and existing sources was concluded to be
the elimination of HAP-based adhesives in repair operations.
6.3.1.5 In-process vessels, components in HAP service, in-mold
coatings, foam reactant dispensing, and demolding
In the ICRs, HAP emissions were reported for each of these
emission source types. The total reported emissions for all of
6-7
-------
these combined was 25 tons/yr, which is approximately 9 percent
of the total reported molded foam HAP emissions. No control was
reported for any of these emission source types in the ICR
responses. Therefore, the floor for both new and existing
sources for these emission source types was concluded to be no
control.
6.3.2 Slabstock Foam
There are 78 slabstock foam facilities in this subcategory.
The MACT floor for each emission source type was based on the top
12 percent of the subcategory (top 10 facilities). The top
performing facilities were determined on a case-by-case basis for
each emission source type; therefore, the same facilities were
not always used in the floor determinations.
6.3.2.1 Storage/unloading
HAP emissions from storage/unloading accounted for less than
1 percent of the total reported slabstock HAP emissions.3 Of the
facilities reporting HAP ABA and/or TDI storage breathing loss
emissions, no control was reported for either type of HAP. For
HAP ABA unloading, 17 facilities reported using vapor balancing
for emission control. The remaining facilities reported no
control. For TDI unloading, 29 facilities reported using vapor
balancing. The remaining facilities reported using no control.
The EPA concluded that the floor for TDI and HAP ABA
storage/unloading at new and existing sources is vapor balancing.
6.3.2.2 Components in HAP service
HAP emissions from pumps and valves accounted for less than
one percent of the reported emissions per year. Thirty-three of
the 77 facilities reported using "canned pumps" (a type of
sealless pump for TDI). Since more than 10 plants reported
canned pumps for TDI, the new and existing source MACT floors for
TDI pumps were concluded to be "sealless" pumps. Four facilities
reported using canned pumps for MeCl2, and no other facilities
reported any type of control. A median-based approach was used
to determine the existing source floor for MeCl2 pumps. Since
"no control" was most frequent in the top 10 list, the EPA
6-8
-------
concluded the floor was to be no control for existing sources,
and sealless MeCl2 pumps for new sources.
In addition, no facility reported the control of HAP
emissions from any of the other components in HAP service
(valves, connectors, etc.). Therefore, the new and existing
source floors for all components except pumps were concluded to
be no control.
6.3.2.3 Equipment cleaning
Methylene chloride is used as a cleaner to rinse and/or soak
foam machine parts such as mixheads and foam troughs. This use
resulted in less than 1 percent of total reported slabstock HAP
emissions.3
Eight of the 73 slabstock facilities reporting equipment
cleaning operations reported the use of non-HAP cleaning methods.
The remainder used no control. A median-based approach was used
to determine the existing source floor for equipment cleaning.
Since 8 out of the top 10 facilities reported using a non-HAP
cleaning method, the EPA concluded the floor for both new and
existing sources is the use of non-HAP cleaning solvents.
6.3.2.4 ABA emission sources
Methylene chloride is the principal ABA used. The role of
the methylene chloride is simply to volatilize and expand the
foam; it does not directly participate in the polyurethane
reaction. Therefore, all of the methylene chloride that is added
to the process is eventually emitted. The use of MeCl2 as an ABA
was the largest emission source of HAP's for slabstock facilities
reported in the ICRs, making up over 80 percent of the total HAP
emissions from slabstock facilities.3 In addition, methyl
chloroform (another HAP) was used by some facilities, resulting
in over twelve percent of the total reported slabstock
emissions.3 There were several alternatives identified in the
ICRs that facilities are using to either reduce or eliminate the
use of a HAP ABA in the manufacture of flexible slabstock
polyurethane foam.
The mix of foam grades produced at a plant has a strong
influence on the amount of ABA that is used and emitted. There
6-9
-------
are many grades of foam that can be produced without any ABA,
while for some soft and light foams, ABA usage can be as high as
24 parts per hundred parts polyol (pph), or almost 300 pounds per
ton of foam produced. Industry representatives have stressed the
need to take this difference into account in determining the
floor, emphasizing that low-ABA and high-ABA grades of foam are
not interchangeable, either from a production cost standpoint,
from an emission standpoint, or from an end-use standpoint.
6.3.2.4.1 Existing sources. The EPA agrees that some
differentiation among foam grades is appropriate, and concluded
that any ABA emissions limitation should be a function of the
grades produced. Therefore, the EPA defined the MACT floor for
HAP ABA emissions by determining a MACT floor set of HAP ABA
formulation limits.
In the ICR, foam producers were asked to provide formulation
information (i.e., parts HAP ABA per 100 parts polyol, or pph)
for all foam grades produced at the facility. Therefore, the EPA
was able to separate the formulation information by grade (where
a grade is represented by its density and IFD). Foamers claimed
all formulation information confidential; therefore, the actual
formulation database and summary is not publicly available.
The foam grades were combined into density/IFD "groups."
The Polyurethane Foam Association (PFA) suggested the following
eight-group classification system:20
• IFD greater than 20 pounds, density greater than 1.4
pounds per cubic foot (pcf)
• IFD greater than 20, density from 1.15 to 1.4 pcf
• IFD greater than 20, density from 1.05 to 1.15 pcf
• IFD greater than 20, density from 0.95 to 1.05 pcf
• IFD greater than 20, density less than 0.95 pcf
• IFD from 15 to 20, at any density
• IFD from 10 to 15, at any density
• IFD less than 10, at any density
The EPA's initial attempt at determining the MACT floor HAP
ABA formulation limits was to simply calculate the average of the
lowest 12 percent of the formulations reported for each of the
6-10
-------
grade ranges suggested by the PFA. However, this approach led to
inconsistent results.21 Therefore, the approach described
below was developed and used to determine the MACT floor HAP ABA
formulation limitations.
This approach could simply be described as the determination
of a baseline, and the application of MACT floor reductions to
this baseline. The initial step in this approach was to
determine baseline formulations. As part of this analysis, the
EPA re-examined the PFA-recommended foam grade groupings.22
This analysis, which used the formulation database to examine ABA
usage/emission trends by IFD and density, concluded that the PFA-
recommended groupings needed revisions in two areas. The PFA-
recommended groupings assumed that (1) ABA usage was not
dependent on density for foam grades with IFDs less than 20
pounds, and (2) ABA usage was not dependent on IFD for IFDs
greater than 30 pounds. The EPA's analysis found both of these
assumptions to be inaccurate. Therefore, all subsequent analysis
was conducted on a 30-group grid, using the five PFA-recommended
density ranges, and six IFD ranges (0-10, 11-15, 15-20, 21-25,
26-30, and 31+ pounds).
The EPA attempted to use the overall average of formulation
information submitted for each grade group. As discussed above
for the average of the top 12 percent, the overall average
formulations often did not follow a reasonable pattern. For
instance, the average ABA level would increase with increasing
IFD, and with increasing density. The EPA primarily attributed
this problem to the low number of data points for several
density/IFD groups. Table 6-1 shows the number of plants
represented in the database for the 30 density/IFD groups.
In the development of the EPA's "Flexible Polyurethane Foam
Emission Reduction Technologies Cost Document," a representative
facility was created.23 The foam formulations used for this
representative facility were generated in a cooperative effort
between the EPA and the PFA, and the EPA believes that they
represent typical formulations for the industry. Table 6-2 shows
both the average (or range) of formulation information from the
6-11
-------
TABLE 6-1. NUMBERS OF PLANTS REPRESENTED IN
FORMULATION DATABASE
Table values
are numbers of
plants
reporting
formulation
information in
each group
0-10
11-15
j 16-20
D 21'25
26-30
31+
Densit
0-
0.95
4
5
24
15
:y ranges
0.96-
1.05
7
17
14
17
22
26
(pounds
1.06-
1.15
4
9
15
15
7
21
per cubic
1.16-
1.40
7
12
20
36
: foot)
1.41 +
<9
12
19
23
31
36
6-12
-------
ICR responses, and the representative facility formulation. From
this information, the EPA selected baseline formulation levels,
which are shown in Table 6-3, for each foam grade grouping.
The next step was to apply reductions, representing MACT
floor reductions, to these baseline formulation values. The
amount of reduction achievable also varies by foam grade.
Numerous technologies exist that allow the production of higher-
density, higher- IFD foams with significantly less, or even no,
HAP ABA. However, the amount of HAP ABA reduction that can be
achieved for low-density, low- IFD foam grades of acceptable
quality is much less. Therefore, the reductions had to be
determined on a grade-specific basis.
The problems discussed above, associated with insufficient
numbers of data points in certain density/IFD groups, would have
caused similar problems if reductions were determined for
individual groups. Therefore, the EPA combined groups until at
least 30 data points were available. For instance, the six
groups with IFDs of 20 pounds or less and densities of 1.16 pcf
or less were combined. This resulted in 41 data points in this
combined group .
For each combined group, the overall average formulation was
calculated, as well as the average formulation for the 12 percent
of the data points with the lowest HAP ABA formulations. Then
the reduction from the overall average to the top 12 percent
average was calculated. For example, if the overall average
formulation was 9 pph and the top 12 percent average was 5 pph,
the percentage reduction was calculated as follows :
PercentReduction = ~ (100) = 44 percent
(9 pph)
The results of the percentage reduction determination are
shown in Table 6-4. These percentage reductions were then
applied to the individual density/IFD group baseline formulation
levels in Table 6-2 to obtain the existing source MACT floor HAP
ABA formulation limitations, which are shown in Table 6-5. It
should be noted that this analysis resulted in equivalent HAP ABA
6-13
-------
TABLE 6-2. HAP ABA FORMULATIONS FROM ICR'S
Table values
in parts ABA
per hundred
parts polyol
0-10
11-15
F 16-20
D
21-25
26-30
31 +
Densit
0-
0.95
(21)
(16-18)
(16-18)
10
(10)
(9-10)
.y ranges
0.96-
1.05
22
(18)
19
(15)
14
(11-13)
(12-15)
8
(7)
(7-8)
(pounds pe
1.06-
1.15
(20)
(13-18)
14
(12-14)
(12)
7
(6-8)
(4-5)
r cubic i
1.16-
1.40
(10)
(10)
(7-8)
5-6
(5-6)
2
(3)
loot)
1.41 +
(2)
(2-6)
10-13
(6-8)
(4-5)
6-7
(5)
1-2
(2)
NOTES: Top numbers are the estimates provided by the
PFA and chemical suppliers in comments on the
cost report. Numbers in parentheses and
italics are from the ICR data.
6-14
-------
TABLE 6-3. BASELINE HAP ABA FORMULATIONS
Table values
in parts ABA
per hundred
parts polyol
0-10
11-15
j 16-20
D 21'25
26-30
31+
Densil
0-
0.95
21
18
16
10
9
ty ranges
0.96-
1.05
20
19
14
13
8
7
(pounds p
1.06-
1.15
20
18
14
12
7
5
>er cubic
1.16-
1.40
10
10
8
6
2
foot)
1.41 +
6
6
6
5
5
2
6-15
-------
formulation limits for more than one density/IFD group.
Therefore, the result was actually 18 density/IFD groups in the
MACT floor HAP ABA formulation limits table.
6.3.2.4.2 New sources. The new source HAP ABA MACT floor
formulation limitations were determined by examining the lowest
reported HAP ABA formulation for each of the 30 density/IFD
blocks. Formulations for foam grades only produced in small
amounts were not considered. This analysis found that the
production of many foam grades were reported with no HAP ABA.
However, the results were not always logical. For instance, all
foam grades with between 0.96 and 1.05 pcf were reported to be
produced with no HAP ABA (with the exception of IFDs less than
15 pounds), but there were no foam grades with densities between
1.06 and 1.15 pcf that were reported to be produced with no HAP
ABA. The EPA concluded that this was more a function of the
randomness of the foam grades reported, rather than the inability
to produce foams of densities between 1.06 and 1.15 pcf with no
HAP ABA. In fact, the Agency believes that if foam grades with
densities between 0.96 and 1.05 pcf can be produced with no HAP
ABA, then foam grades of corresponding IFDs with densities
greater than 1.05 pcf can also be produced with no HAP ABA. The
lack of sufficient data for the foam grades where the new source
MACT floor was not determined to be zero led the Agency to
conclude that the new source MACT floor for these grades was
equal to the existing source MACT floor, as shown in Table 6-5.
Therefore, the EPA concluded that the MACT floor HAP ABA
formulation limitations for new sources are as shown in
Table 6-6.
6-16
-------
TABLE 6-4. MACT FLOOR PERCENTAGE REDUCTIONS
Table values
are
percentage
reduction
from baseline
0-10
11-15
p 16-20
D 21-25
26-30
31+
Densi
0-
0.95
by ranges
0.96-
1.05
(pounds j
1.06-
1.15
>er cubic
1.16-
1.40
45
68
foot)
1.41+
75
100
6-17
-------
TABLE 6-5. MACT FLOOR HAP ABA FORMULATION
LIMITATIONS FOR EXISTING SLABSTOCK SOURCES
Table
values in
parts ABA
per hundred
parts
polyol
0-10
11-15
16-20
F 21-25
D
26-30
31+
Density ranges (pounds per cubic
foot)
0-
0.95
12
10
9
6
5
0.96-
1.05
1.06-
1.15
11
8
7
4
4
1.16-
1.40
6
4
3
2
1
1.41+
2
2
0
6-18
-------
TABLE 6-6.
MACT FLOOR HAP ABA FORMULATION LIMITATIONS
FOR NEW SLABSTOCK SOURCES
0-10
11-15
j 16-20
D 21-25
26-30
31 +
Densit
0-
0.95
12
10
9
6
5
;y ranges
0.96-
1.05
(pounds
1.06-
1.15
11
per cubic
1.16-
1.40
6
0
z foot)
1.41 +
6-19
-------
The HAP ABA formulation limitations will be used, in
conjunction with the following equation, to calculate the
allowable HAP ABA emissions from foam production:
(limit ±) (polyol j
where :
emissallow= Allowable emissions due to use of a HAP ABA
for a specified time period, Megagrams .
limiti = HAP ABA formulation limit for foam grade i, parts
ABA per 100 parts polyol .
polyolj_ = Amount of polyol used in the time period in the
production of foam grade i, pounds.
n = Number of foam grades produced in the time period.
6.3.3 Rebond Foam Production
Rebond is the process where scrap foam is cut up and placed
in a mold with a small amount of TDI and treated with steam.
This causes the small pieces to adhere and form a solid cylinder
of foam. This cylinder is then peeled into sheets, which is
mainly used as carpet padding.
No control for the TDI emissions was reported. Since all
facilities reported no control for TDI, and only one of the 21
rebond operations reported other HAP emissions, the EPA concluded
the floor for rebond at both new and existing sources is "no
control" for TDI emissions, and the elimination of all HAP
cleaners and mold release agents.
6.4 SUMMARY OF MACT FLOORS
Table 6-7 summarizes the MACT floor conclusions for both new
and existing source for each emission source type.
6.5 POTENTIAL LEVELS OF CONTROL
This section discusses levels of control that could be used
as the basis for developing regulatory alternatives for the
flexible polyurethane foam production industry. For each
emission source type, the potential levels of control are
6-20
-------
CO
g
I
OS1
o
O
EH
U
CQ
H
X
w
W
S
01
a
t
a.
J 01
h 0)
o
gj
m
01
o
3
0
CO
c1
•H
JJ
m
•H
X
H
i
a
b.
e
u
u
I
CO
§
M
CO
CO
M
£
U
£
I
Ch
ra 01
jj jj
o u
3 3
O O
a a
•a -o
i
rH 3 m >H C -
C rH 10 -H CO -H en
3 XH oi ra 01 c
"V, rH (X 0) --rl
oi T) oi a
o X rH (0 i k ra
JJ -H O O C 0) -H
co E S CK M rat)
ra . 2
ra rH c
01 o jj o
rH rl 0 -H
« JJ 3 JJ
0) C t3 (0
ra o o jj
0 ^ -H
i Q. g
O -H
01 M a T3 r-l
O rH
GUI ra c
rH S » 5 -2
(On JJ
14 (0 ra rl 2 3
o ao i g
ft H g £i C )H
10 Q 3 JJ O O
> H ao % b
^
§
JJ Oi
o rt
3 X
0 C
S O
T)
•0 01
(U m Q
rH ra 10 c
O (IS XI O
rl XI -H
a jj i c jj
a c & o at
go rt -H jj
3 U X JJ-H
O. (C E
O 1H JJ-H
01 ra C O -H rH
u m E
C 0)1 C -rl G
10 rH O rH O
rH (0 ft* *H -H
(0 (U rij JJ G JJ
Xt 01 CC <0 O (0
C -HrH
rl 1 ^1 -H Q 3
a oi E ra E
MX! -H -H ft
(0 Q JJ rH gO
> HO Ed Cd >w
(U
O
-H
t
0)
CQ
01
tn fti c co
C rf -H C
•H a co
•0 (5 -H
10 c 01 m
o -H rn ra
rH t) -H
G CO g
g 3 JJ JJ 01
i I 1 S I
2 co u u 2
a
•9
rH
01
a.
rH 2 rH
2.«ic,
JJ O JJ
e •o c
o c c S
o O 10 tn
•H (0
O JJ CO
C 10 rl 0)
C 4) 01
1 -H C 10
g ID O
M -H 0) rH
Q rH rH Q)
EH U 0 rl
HI
(0
(U
rH
01
rl
•a
rH
•a
c
m
ra
SH
a)
C
(0
0)
o
cu
rngi
2>M
JJ O
G
O C
o o
-H
O JJ
G 10
G m
1 -H JJ
g G
M -H 0)
QrH Oi
H Cd (0
g
1
6-21
-------
identified. For both new and existing sources, the first level
of control is the MACT "floor" level, which was discussed in the
previous sections. Levels of control more stringent than the
MACT floor level are also presented.
6.5.1 Molded Foam Facilities
HAP emissions are attributable to three predominant emission
source types at molded foam facilities: (1) HAP mixhead flushes,
(2) HAP-based mold release agents, and (3) HAP-based adhesives
for foam repair. The control options for each of these emission
source types are discussed in the following sections. As
discussed below, additional emission source types have been
identified at molded facilities; however, the combination of low
frequency of occurrence of these sources, very low emissions, and
no identified control options led the EPA to omit control options
for these emission source types from all regulatory alternatives.
6.5.1.1 Mixhead Flush
The MACT floor for this emission source type was identified
as no control for existing sources. For new sources, the MACT
floor was identified as the prohibition of the use of HAP flush
agents. There were two levels of control above the existing
source MACT floor: work practices that reduce HAP emissions, and
the prohibition of the use of HAP flush agents.
6.5.1.2 Mold Release Agents and Foam Repair
The MACT floor for both new and existing sources for these
two emission sources was identified as the total elimination of
HAP-based products. Therefore, no level of control above the
MACT floor is possible.
6.5.1.3 Other Molded Emission Sources
The MACT floors for new and existing sources were identified
as no control for the following emission source types: in-mold
coatings, storage/unloading, in-process vessels, components in
HAP service, foam froth dispensing, and demolding. No levels of
control above the MACT floor were identified for these emission
points at molded polyurethane foam facilities.
6-22
-------
6.5.2 Slabstock Foam Facilities
At slabstock foam facilities, there are four HAP emission
source types: (1) the storage/unloading of HAP compounds,
(2) leaking components in HAP service, (3) the use of a HAP as an
equipment cleaner, and (4) the use of HAP ABA'S. The levels of
control for each of these emission source types are discussed in
the following paragraphs.
6.5.2.1 Storage/Unloading
The MACT floor for tank truck/railcar unloading of both TDI
and HAP ABA at new and existing sources was identified as the use
of either a vapor balance or carbon canister system. No control
options that would result in a level of control above the MACT
floor were reported to be in use in the industry.
6.5.2.2 Components in HAP Service (Equipment Leaks)
The MACT floor for new and existing sources for this
emission source type was determined to be leakless pumps for TDI
components (except for high pressure metering pumps), and no
control for all other components in HAP service. There was one
level of control identified above the MACT floor for new and
existing sources. The level of control above the MACT floor is
an LDAR program for all other components in HAP service.
6.5.2.3 Equipment Cleaning
The MACT floor for both new and existing sources for this
emission source was identified as the total elimination of
HAP-based products. No levels of control above the MACT floor
are possible.
6.5.2.4 Auxiliary Blowing Agent
There are many "grades" of flexible polyurethane slabstock
foam produced, each with slightly different end-uses that require
slightly different foam properties. Different grades can require
varying levels of HAP ABA. The levels of control discussed below
account for this variation.
6.5.2.4.1 Existing sources. The existing source MACT floor
for HAP ABA was determined to be a limit on the amount of HAP ABA
emissions. The limit is based on the grades of foam and amount
6-23
-------
of each grade produced at a facility, and is determined using
grade-specific HAP ABA formulation limitations.
A level of control above the MACT floor identified was the
complete elimination of HAP ABA emissions. The EPA also
identified an intermediate alternative based on the combination
of extended grade ranges, the emission reduction achieved by the
best performing three facilities in each grade range group, and
the emission reduction achieved by carbon adsorption.
6.5.2.4.2 New sources. The new source MACT floor for HAP
ABA emissions was determined to be a limit on the amount of HAP
ABA emissions. The complete elimination of HAP ABA was an
alternative identified more stringent than the new source MACT
floor. In addition, there was an intermediate alternative
identified between the MACT floor and the complete elimination of
HAP ABA.
6.5.3 Rebond Foam
The MACT floor for both new and existing sources for rebond
foam was the total elimination of HAP-based products for mold
release agents and equipment cleaning. Therefore, no level of
control above the MACT floor is possible. The floor level of
control for TDI emissions was determined to be no control. No
additional control techniques were identified for TDI emissions.
6.6 REGULATORY ALTERNATIVES
The following section presents and discusses the regulatory
alternatives developed for new and existing polyurethane foam
facilities. The amount of emission reduction achieved increases
with each alternative above the floor. It is also expected that
costs will increase with each more stringent alternative.
6.6.1 Molded Foam
For molded foam production, two existing source regulatory
alternatives above the MACT floor were developed. These
alternatives are summarized in Table 6-8. Alternative 1 would
require facilities to use work practices to reduce the HAP
emissions from mixhead flushing. The work practice requirements
may take many forms, one of which may be to simply require
operators to cover the barrel used to collect the HAP mixhead
6-24
-------
O
fa
Q
W
Q
J
i
01
o
fa
CO CO
W H
> U
< O
S co
Is
CO
>H M
01 X
o w
s
J
B
w
01
CX)
I
U)
H
m
0)
i
AdhesJ
•H
(0
Pi
0)
tf
to
a
0)
tn
^
a)
a
c0
0)
H
T)
10
0)
•S
-H
g
^0)
egulato
Iternatj
/w» "^
tfrt:
c
o
-H
4J
-H
X!
-H
& -»
O
a
Qj
^j
s
O
-H
-H
Q
-H
XH ->
O
a
§
a
o
-H
rohibit
a
pj
fi
H
U
Oi
6-25
-------
flush. The other emission sources at the facility would be
required to be controlled at the MACT floor level. Alternative 2
prohibits the use of any HAP-based mixhead flush. While TDI and
MDI could still be used in the foam formulation, this alternative
would practically eliminate HAP emissions from molded foam
facilities.
The MACT floor level of control for all applicable emission
sources at new sources is the prohibition of the use of HAP-based
products. Therefore, the MACT floor Regulatory Alternative,
which is shown in Table 6-9, is the only one offered.
6.6.2 Slabstock Foam
Two existing source alternatives above the MACT floor were
developed for slabstock production. These alternatives are
presented in Table 6-10. Alternative la adds a unique LDAR
program for equipment leak emissions, and increases the level of
control for HAP ABA emissions to the intermediate emission limit.
Alternative Ib is approximately equivalent to Alternative la
in stringency, but incorporates a novel implementation approach
that would reduce the reporting, recordkeeping, and monitoring
burden on the industry. During the P-MACT process, industry was
concerned that the cost of controlling ABA emissions from
storage/unloading and equipment leaks was unreasonable, given the
relatively low emissions from these sources (around 50 tons per
year, or 0.2 percent). Under Alternative la, the amount of HAP
ABA allowed to be emitted from the entire facility would be
determined using the HAP ABA emission limit equation and
formulation limitations. This limit would then apply to the
entire facility, rather than only to the HAP ABA added at the
mixhead. In the absence of add-on control, the entire amount of
HAP ABA used is emitted. Therefore, the total amount of HAP ABA
used during the compliance time period would be compared to the
emission limit to determine compliance. Under this alternative,
the amount of ABA used could be determined using simple inventory
procedures, in lieu of more expensive LDAR techniques. This
approach would encourage the source to reduce storage and
equipment leak emissions so more ABA would be available to be
6-26
-------
CO
H
U
CO
CTi
I
m
tt)
i
Adhesd
n
•
•r|
(0
&
«
to
4J
g
Ol
rtj
0)
a
(0
0)
0
«
•O
o
g
£
to
p4
P
rH
III
TJ
(0
0
fl
•H
S
&^
0-H
4J AJ
(0 (0
H H
^ 0)
D)4J
0) H
« <;
fl
O
•H
•ohibit
M
&
cu
i
c
o
-H
4-)
-H
XI
•H
X!
0
C,
M
a.
ft
a
0
-H
• i
4-J
-H
XI
•H
X5
O
S-i
a
pt
g
j-i
0
0
rH
b
H
U
6-27
-------
o
H
I
H
E-i
n
a
0
•H
01
m
i
0)
a
•H
a
a
0
rH
o
a
2
6
-3-
H
0
U
•rl
>
M
0)
M
0i
^
W
C
ponents 1
B
0
u
M
rt 0
•ud
cog
•
4J
rH
^
0)
o>
•3
< . T(
Existing soui
MACT floor HAP
emission lim
§
•rH
4J
-H
43
-H
O
o
W
ft
0
fl
1
01 01
01 4J
0) (3
rH
-------
used in foam formulations, while giving them considerable
flexibility.
Alternative 2 would totally eliminate HAP ABA emissions, and
would require controls for TDI storage/unloading and equipment
leaks.
There are also three regulatory alternatives for new
slabstock sources. The first alternative represents the new
source MACT floor level. The second alternative adds leak
detection and repair for equipment leaks, and increases the HAP
ABA to an "intermediate" level of emission reduction. There is
also a source-wide compliance option for the new source
Regulatory Alternative 1. The third new source alternative
includes a prohibition of HAP ABA emissions. The new source
slabstock alternatives are shown in Table 6-11.
6.6.3 Rebond Foam
For the reasons stated in section 6.5.3, no regulatory
alternatives above the MACT floor level were developed for rebond
foam.
6-29
-------
HAP ABA
Emissions
a e
0) "
Et £3
Pl H
"3 S
&rl
w u
Components in
HAP Service
^ to
-H
ft
-H
A
0
ft
d,
a!
TDI pumps -
leakless
1 a
o
P -2
EH ^j
iS °
O
O
H
t.
h"i
H
U
§
New Source
Intermediate
HAP ABA
emission
limit
a
o
-H
4J
•H
,Q
-H
r*.
O
ft
CU
til
TDI pumps -
leakless
HAP ABA pumps
-unique LDAR
1 a
O
o-0
^ rt
!!
(0
r emission
CQ g
1-
CQ ll
ro ^"^
CD ^
rH ^ >,
{/* i)
" _j fll ,
CO y -H
1-1 3 "o
- r« tO
CW) n i
**-*
° 5
'S iSJ 05
flj ^* 4-J
^ \ /^
vapor balance/c
>ment cleaning -
limit for tc
i P^(
•H
1-1 ^
Q m
EH N
oS
i
cu
-
H
HAP ABA
prohibition
o
•H
4-)
-H
f)
•H
O
ft
DJ
EH
U
ra £J
ft
6-30
-------
7.0 MODEL PLANTS
This chapter presents model plants for the flexible
polyurethane foam production industry. A model plant does not
represent any single actual facility, but rather it represents a
range of facilities with similar characteristics that may be
impacted by a standard. Each model plant is characterized in
terms of facility type, size, and other parameters that affect
estimates of emissions, control costs, and secondary impacts.
The model plants discussed in this chapter were used to determine
baseline emissions, which were discussed in Chapter 5, and to
analyze cost and environmental impacts of regulatory
alternatives, which are discussed in Chapters 8 and 9,
respectively.
The molded and slabstock segments of the foam industry were
treated separately in the development of the model plants. In
both cases, model plants were developed based on available
information including data in company responses to the ICR's,
observations made during site visits, and information received
from PFA representatives, vendors, manufacturers, and foam
producers.
All 21 rebond facilities located at major source plant sites
reported emission controls that would be in compliance with the
proposed standards. It is estimated the remaining 31 rebond foam
facilities are area sources, and would not be subject to the
regulation. Therefore, since it is estimated that no rebond
facilities will be subject to the standard, no rebond model
plants were developed.
The following sections describe the slabstock and molded
foam production model plants. A complete description of the
development of the model plants is provided in the memorandum
7-1
-------
entitled "Flexible Foam Model Plants," which is contained in the
SID.24
7.1 SLABSTOCK FOAM MODEL PLANT DESCRIPTIONS
Five basic model plants were developed for the slabstock
segment of the industry, which are presented in Table 7-1. The
model plants include the following operating parameters:
production, number and size of storage vessels, number of
components in HAP service, number of production lines and line
speed. The use of a HAP (usually MeCl2) as an ABA is the primary
source of HAP emissions at a slabstock foam facility. However,
the amount of ABA needed varies considerably depending on the
"grade" of the foam being produced. Therefore, the model plants
include a breakdown of foam production by grade. It is assumed
that all model plants produce the same grades of foam, with
variation in the amount of each grade produced. The grades
produced by the model plants represent the most commonly produced
grades in the industry, and the formulations are based on ICR
responses and input from PFA representatives. The grade-specific
information is presented in Table 7-2.
The model plants also include emission levels for TDI and
MeCl2 from the components mentioned above. For each of the five
model plants, the average facility HAP ABA usage was determined
using information from the actual facilities represented by the
model plant. The amount of each grade of foam produced by the
model plant was adjusted so that the total annual HAP ABA usage
for the model plant equalled the average of the real plants. All
HAP ABA used was assumed to be emitted. Storage emissions were
calculated using AP-42 emission factors for chemical storage
tanks.16 Emissions from components in HAP service (e.g. pumps,
valves, flanges, etc.) were calculated using SOCMI average
emissions factors.25
Thirty-three percent of the slabstock facilities reported
the use of MeCl2 as a general equipment cleaner, including
facilities represented by all five model plants. However, no
correlation was found between the use. of a HAP cleaner (or the
amount of HAP used as a cleaner) and the amount of foam produced.
7-2
-------
m
m m
0 *"
s .
i „
S ^
•«• CO
.2
s:
i
^
m 00
S "
£
1
s ^
CM ca
a N
.a
0.
5 c^
_ 00
e
J2
o.
S <
Wl
| OPERATING PARAMETER
Al
O
-O
A]
<0
6
CM
°!
;2
O
»
6
QO
Ov
—
O
CO
O\
O
t~
0
Ov
6
6
1
IPoam Production Range (Ihous
tons/yr)
° \o o
—
1 — --as*. ~-ss=2
£ _ «n £2.M« <"-SSo£
f"* \« « rM O fM
2 — r* m v —
2 — = o
XJ — CM — •- _ pj fM >o CM — n m o a\
® — — — — * 2 CM « — — ^ g 0 NO
o, «, m CM M= _ «
0
CM •"• d —CM
1
1 a.
•1 « = » <•
^ A ^ ' U «
c °° ° Z -^ 2 -^
-2 S a „., 23S3 „« 25S
§ | § If |.2= || 8« =
E -I I If-lil l|-lil
I -s o al5«,|i sBsseig
t £ o -S-E JJ ^ 2 S. 5. 2^5-2SS.
o ^ °-uHS>a.a.O_K-S>B.a.O
CA ^* •* V •* ** O
E>» oo ^ u «5 r*
"^ & B ^T
> s o §5
< 2 55 Z £
- "» S
-~
- 2 §
- 2 S
._. O
— £ o»
— «n °
^
- 2 §
_. O
«-. «n ®
"• 2
s ° -
•^ •*
s §; *
«n
S 5 »n
»n O^
S ° -
ao <^>
cs 5^
o\ —
s
s
HAP ABA Used (1000 gallons
MeClj)
HAP Used as Cleaner (gallons
MeCl,)
Number of Plants Represented
Nationwide
EMISSIONS (tons/yr)
7-3
-------
2
u
CO
o
5
BJ
in 03
a
Q.
"3
•?
2 J5
1- ffi
^
C3
a.
1
S <
«T
a n
a.
1
S <
rJ 03
2 M
j«
o
•g
"
m
S
CQ
a.
•g
S <
°d!2 11° 3s
0 .»
= 3S IS^S
r»t «N M o ml «
61 C) ,3 en enl en
0 Old 5 en "" enlS
= 3S 9i-»
= 3S 3S..K
-33 s..»
= 3S s-a<
I! OPERATING PARAMETERS
•> o '!•>*"
•a £ I -a r?
w M « o O
_1 Q 0 J 2
1sH g g
«! all
g 3 g < -53
S3" J* S ™ O* CP
_ 55 tu cj Co < cu 01
£ S
s
00
eo
en
en
en
^
en
en
en
S
^
*""
a
TJ-
CO
H TOTAL HAP EMISS10
| COST PARAMETERS
o
o
o
^
en
q
ob
f2
q
oo
VO
q
cs
q
q
q
en
q
"**
1 Cost of ABA (ihousand$/yr)
o
.
o
^.
^
o
5
o
,
*
o
•*
I
1 Cost of HAP Cleaner (thousand
e
S
w
JS
Ot
S
I
Ot
W)
_r
^
•0
en
V
i
s
oo
en
00
oo"
n
w
*Q
es
N
•£
| Total Chemical Cost (thousands
oo
r-"
S
*ft
-
m
^
^
".n
^
^
K
o
—
o_
»>
S
O
a
0
i
en
00
§
en
1 Total Facility Operating Costs
(thousand$/yr)
o
o
12
•"
o
s
^*»
o
o
o
g
o
»n
«•
I
!>
O
•s
s
u
i
s-
s
7-4
-------
c
o
"a
E
o
3
O
a.
*»
o
2
o
TJ
O
E
a
o
o
o
•3
a
^
a
~
in
c
n.
a
2
(!)
oen^in^^cocoeocM
cof^co^eDN-^^ot^^inin
COCMCM»-CM»-CMCOCM ^ i-
roinco^incMcooocotnco'O-
cdor^coi^irif-cotri^-iricocd
COCMt-*-CMCMi-CMi- CMT-
OOOh-OOOt^-Of^OOO
o ^ ^» oo in co co co ^ CM ^ in co
••r^rcO'.-cMT-t-i-i- CM
ocococMoeocoocoTj-^oeo
cDi>»«*-^eMor».o«oeMeMO'*
coencnr^ococococMODcococo
inin*co'*cocMcocMOOco-r-
ooooooooooooo
t^ en O) en t^ en r^ in r^ f*1* r^ en co
inT-T-T-u5»-inc»ininir>"-cr>
ooooooooooooo
in^v— v-incM^co^tn^cM^
ooooooooooooo
ooooooooooooo
(OCOCOCMinCMCOinCOCMCOCOCM
ooooooooooooo
^•CMCMCMTrCMCOOCOCOCOCMCM
ooooooooooooo
QD *f f *f CO ^f *O O *O fO *O ^f ^f
-T—
•^ CM ^" "^ T" ^
OOlOOOOOOOOOOO
CJ ^ T" C>J rt O O O O ^~ ^r™ CM P5 CO ^ tO U>
oomoooooooooo
CO ^ ^ CM CO CM CO P3 CO ^ ^ CM CO
O3OOOO^-»-CMCOCO^lOin
£1
CM
en
^
CO
CO
*»•
to
^*
o
en
0
CD
o
o
in
o
CD
o
1~
CM
o
o
in
co n
CO CO
in in
in in
r- i^
CM CM
•» *
CM CM
CO CO
0 0
O O
en en
0 0
CO CO
o o
o o
•*»• •*
o o
CD CD
O O
^ *~
•r- •**
O O
O O
CO 1^
CM CM
CO O
N. in
t^o
CO* t—
CM i-
oco
CM CO
O CM
«* CM
O O
co in
o o
^ ^
o o
o o
•* CO
o o
CO CO
*~ ^"
o o
O CO
T™
o o
o o
CO CO
0
o
o
o
Q
0
o
0
o
o
CM
1-
o
CO
o
o
CO
o
CD
o
T-
o
o
o
CO
CM
O)
V-
0
^
^
^
CO
CO
CM
0
in
o
^
o
o
CO
0
CO
o
CD
^
o
o
en
K
in
CO
*
CO
CO
o
CM
f-
o
o
CO
CO
i
CD
o
o
in
6
o
o
o
^~
^
o
en
o
CT>
o
o
0
0
0
0
o
o
o
o
o
Tf
CO
CD
o
h*
CO
o
o
CM
o
o
*~
o
o
m
o
in
CM
in
CM
o
CM
0
A
O
CO
in
CO
o
ir)
in
UJ
in
UJ
2
u
7-5
-------
Therefore, each of the five model plants were separated into two
smaller model plants (i.e., la and Ib) , with one using MeCl2 as a
cleaner, and the other employing cleaning methods that do not use
a HAP. One-third of the nationwide facilities represented by
each of the five model plants were assigned to the smaller model
plant, which uses a HAP cleaner.
Because many of the emission reduction technologies that
will be studied for the slabstock industry will involve process
changes or other pollution prevention techniques, the model
plants include baseline operational costs, including the total
annual chemical costs, the total annual chemical costs for ABA-
blown foam, and the total annual operating costs. In addition,
baseline costs associated with the use of MeCl2 as a cleaner are
included. The baseline costs that were developed assumed the
following raw chemical costs: $0.50 per pound for polyol, and
$1.00 per pound for TDI, and $0.40 per pound for MeCl2.
7.2 MOLDED FOAM MODEL PLANT DESCRIPTIONS
While the basic processes are similar at all molded foam
facilities, there are two different types of equipment used to
pour the foam, which creates large differences in their HAP
emission potential. The two types of equipment are low pressure
(LP) and high pressure (HP) mixheads. LP mixheads require a
solvent flush between foam shots to remove residual foam, while
HP mixheads do not. Consequently, HAP emissions from LP
facilities are usually significantly higher than emissions from
HP facilities. All of the molded foam major sources were LP
facilities, and all of the HP facilities were area sources.
Four molded foam model plants were developed, one with a HP
mixhead, and three with LP mixheads. They are presented in
Table 7-3. Only one model plant was created for HP mixhead
facilities since the other parameters and HAP emissions were not
found to be linked to production. The LP model plants differ
mainly in the amount of foam produced, and the amount of HAP used
to flush. The model plants include the following operational
parameters: production, number of carrousels, operating
schedule, and type of mixhead. The model plants also include the
7-6
-------
TABLE 7-3. MOLDED FOAM MODEL PLANT PARAMETERS
Operating Parameters
Foam production range
(tons/yr)
Average foam
production (tons/yr)
Number of carrousels
Operating schedule
Type of mixheads
HAP flush amount
(55 -gal drums)
Waste MeCl2 from flush
(55 -gal drums)
HAP mold release agent
amount (gal /year)
HAP repair adhesive
amount (gal)
HAP content of
adhesive
Number of facilities
represented nationwide
Emissions (tons/yr)
Emissions from HAP
mixhead flush/plant
Emissions from HAP
mold release/plant
Emissions from HAP
adhes ive /plant
Cost Parameters
Disposal cost of waste
MeCl2
Cost of HAP-based mold
release agent
Cost of HAP-based
adhesive
HP Model
Plant
0-15,000
3,331
2
16 hrs/day
250 days/yr
HP
0
0
0
581
70%
27
0
0
2.093
0
0
$4,939
LP Model
Plant 1
0-99
26
5
8 hrs/day
250 days/yr
LP
19
2
41
0
0
109
5.14
0.15
0
$1,600
$198
0
LP Model
Plant 2
100-499
308
5
16 hrs/day
250 days/yr
LP
72
8
32
0
0
54
19.69
0.12
0
$6,400
$154
0
LP Model
Plant 3
aSOO
2718
5
16 hrs/day
250 days/yr
LP
177
18
831
66
70%
44
21.31
6.0
1.35
$14,400
$4,005
$561
7-7
-------
amounts of mixhead flush, mold release agent, and repair adhesive
used, which determine the HAP emissions for each model plant.
As with slabstock foam, many of the emission reduction
technologies that will be studied for the slabstock industry will
involve pollution prevention techniques. Therefore, the model
plants include baseline costs for disposal of waste MeCl2 from
the mixhead flush, and the cost of HAP-based mold release agent
and HAP-based adhesive.
7-8
-------
8.0 ENVIRONMENTAL AND ENERGY IMPACTS
OF REGULATORY ALTERNATIVES
This chapter presents the primary environmental impacts,
secondary environmental impacts, and energy impacts of the
existing source regulatory alternatives described in Chapter 6.
The cost and economic impacts will be discussed in Chapter 9.
There are no anticipated new source impacts for this source
category. The industry has stated that there is enough unused
capacity at existing facilities to handle any increased demand
for foam products.26 A more detailed description of the
impacts analysis is provided in the SID.27
8.1 PRIMARY ENVIRONMENTAL IMPACTS
Primary environmental impacts are the reductions of HAP
emissions that occur as a result of application of the regulatory
alternatives presented in Chapter 6. The primary emission
reductions were calculated using the model plants presented in
Chapter 7.
8.1.1 Primary Environmental Impacts for Molded Foam
Only major sources of HAP will be subject to the Flexible
Polyurethane Foam Production NESHAP. Since the HP molded model
plant and LP model plant 1 have emissions below the major source
threshold, it was assumed that the facilities represented by
these model plants would not be affected by the Flexible
Polyurethane Foam Production NESHAP. It could be maintained that
these facilities, particularly the LP facilities, have the
potential to emit major source levels of HAP. However, it was
assumed that these facilities would obtain federally enforceable
permit requirements limiting HAP emissions below major source
levels, rather than installing controls in accordance with the
NESHAP. Therefore, for molded foam, the nationwide regulatory
alternative impacts are based on LP model plants 2 and 3 only.
8-1
-------
As shown in Table 8-1, the total HAP emission reductions expected
for the molded regulatory alternatives ranges from 331 tons (300
Mg) to 2,332 tons (2,120 Mg) per year, depending on the
alternative chosen. These levels of emission reduction represent
a 10 to 73 percent reduction over the baseline emission level.
The MACT floor alternative has a reduces HAP emissions by
10 percent. The emission reduction is low because there is no
attempt to control the mixhead flush emissions, which are the
largest emission source. The MACT floor (and therefore
subsequent alternatives) calls for the total elimination of HAP-
based mold release agents and repair adhesives. The percent
emission reduction from baseline presented in Table 8-1, is not
100 percent for either of these emission sources, as HP model
plant 1, and LP model plant 1 have emissions from these sources.
As discussed earlier, these model plants were not used for
nationwide impacts, as they are assumed to represent area
sources.
Regulatory Alternative 1 adds a work practice requirement to
control mixhead flush emissions. Table 8-1 shows a nationwide
emission reduction for this source of 59 percent. As mentioned
above, the other two emission sources require total elimination
of HAP-based products. The resulting emission reduction from
nationwide baseline for Regulatory Alternative 1 is 58 percent.
Regulatory Alternative 2 requires total elimination of HAP-
mixhead flush, resulting in a 78 percent reduction for this
emission source. Again, there is not a complete elimination
because of the small amount of emissions from the area source
model plants. There is also the complete elimination of HAP-
based mold release agents and repair adhesives. The resulting
total emission reduction for Regulatory Alternative 2 is 73
percent.
8.1.2. Slabstock Foam Primary Environmental Impacts
Table 8-2 presents the primary environmental impacts for
slabstock foam. As was shown in Table 6-9, three basic
regulatory alternatives were developed for slabstock foam, with
Regulatory Alternative 1 including two implementation options.
8-2
-------
TABLE 8-1.
MOLDED FOAM REGULATORY ALTERNATIVE HAP EMISSION
REDUCTION
Regulatory
Alternative
MACT Floor
Regulatory
Alternative 1
Regulatory
Alternative 2
HAP Emission
(percent of
Mixhead
Flush
0 (0)
1,501
(59)
2,001
(78)
Reduction- tons/yr
total reduction)
Mold Total
Release Repair
Agent Adhesive
271
(94)
271
(94)
271
(94)
61 (53) 332 (10)
61 (53) 1,832
(58)
61 (53) 2,332
(73)
8-3
-------
The first option (Alternative la) consists of emission point
specific requirements, and the second (Alternative Ib) consists
of a source-wide emission limitation that allows the owner or
operator to select the emission points to control, as long as the
source-wide emission limitation is achieved. The HAP emission
reductions presented in this section for Regulatory Alternative 1
were calculated using the emission point specific requirements of
Alternative la. The EPA believes that the environmental impacts
of Alternative Ib would be analogous to those presented for
Alternative la.
There are two sets of impacts shown for Regulatory
Alternative 2. The first only takes into account the "direct"
HAP emission reductions associated with the regulatory
alternative emission requirements. However, the elimination of
the use and emissions of HAP ABA will also result in the
elimination of HAP ABA emissions from storage and equipment
leaks. The second set of impacts include these "indirect" HAP
emission reductions. The storage and equipment leak alternatives
specify separate requirements for TDI, which is a reactant in the
formation of polyurethane foam, and HAP ABA.
As shown in Table 8-2, the HAP emission reductions expected
for the slabstock regulatory alternatives range from 9,421 tons
(8,564 Mg) to 16,542 tons (15,038 Mg) per year, depending on the
alternative chosen. These reductions represent from 57 to over
99 percent reductions over the baseline HAP emissions level.
The MACT floor alternative has a total emission reduction of
57 percent. Storage and unloading emissions are controlled by
using vapor balance or carbon canisters, resulting in
approximately 88 percent reduction for this emission point.
Equipment leaks are reduced by only 1 percent because the control
of TDI emissions using sealless pumps is the only control
required at the MACT floor level.
Regulatory Alternative 1 has a total emission reduction of
69 percent. One reason for the increase over the MACT floor
reduction is that Alternative 1 reduces equipment leaks by
49 percent by requiring a unique LDAR program as well as the TDI
8-4
-------
CD
I
s
-H
,—4
-l 0)
fi-H
O 4J
•H (I)
ca C
ca M
-H cu
g 4J
0)
J3 1-1
EH 0
8-5
-------
control required at the MACT floor level. Another reason is the
increased control of HAP ABA emissions (from a 57 to a 69 percent
reduction).
Regulatory Alternative 2 has a total emission reduction of
almost 100 percent. This is the result of both direct and
indirect impacts. The prohibition of HAP ABA emissions results
in complete elimination of HAP ABA storage/unloading and HAP ABA
equipment leak emissions. HAP-based equipment cleaner emissions
were already eliminated as in the other alternatives.
8.1.3 Rebond Foam Primary Environmental Impacts
It is predicted that no rebond facilities will be affected
by the proposed regulation. Therefore, no primary environmental
impacts are anticipated from this subcategory.
8.2 SECONDARY ENVIRONMENTAL IMPACTS
While the primary impact of the regulatory alternatives is
to reduce HAP emissions, the application of control technologies
can have other positive environmental effects such as reduction
in non-HAP volatile organic compound (VOC) emissions, or a
reduction in hazardous waste. However, the environmental effects
can also be negative, such as the generation of additional
wastewater or solid waste. In this section, the secondary
impacts for both slabstock and molded foam on air pollution,
water pollution, and solid and hazardous waste are discussed. As
discussed earlier, it is predicted that no existing rebond
operations will be affected by the proposed rule. Therefore, no
secondary impacts are anticipated.
8.2.1. Air Pollution Impacts
For both slabstock and molded foam, the secondary air
pollution impacts are a potential increase in VOC emissions.
There could be an increase in the amount of VOC emitted from
slabstock foam facilities if the non-HAP cleaner required by all
the alternatives contained VOC's rather than HAP. If this is
true, HAP emissions could be replaced by VOC emissions. The same
situation could occur in molded foam with non-HAP based repair
adhesives, mold release agents, or mixhead flushes. However, the
8-6
-------
replacement products typically have low volatility, so the amount
of VOC emitted should be small.
There is also the potential for an increase in the amount of
criteria pollutants as a result of the combustion of coal, oil,
or natural gas, which are used to generate the additional energy
needed for some control equipment. This combustion results in
the emission of nitrous oxide (NOX) , carbon monoxide (CO),
particulate matter (PM), and sulfur dioxide (S02). These off-
site air impacts were not included in this analysis, although
energy impacts are discussed in Section 8.3.
8.2.2. Water Pollution Impacts
Potential water pollution impacts occur in the slabstock
foam subcategory for Regulatory Alternatives 1 and 2. The
potential impact comes from the use of carbon adsorption as one
of the control options to reduce or eliminate ABA emissions.
Once the HAP ABA is captured by the carbon bed, that carbon bed
is regenerated. The regeneration process is where there is a
potential wastewater impact. The carbon bed is regenerated by
passing steam through the bed, and condensing steam and MeCl2,
then distilling and recovering the MeCl2. The distillation
condenser bottoms can potentially contain small amounts of MeCl2
that could be introduced into the wastewater. However, the
Agency is aware of only one facility using a carbon adsorber to
recover HAP ABA, and does not expect any additional
installations.
A similar situation could occur in the solvent recovery
systems that could potentially be used to collect mixhead flush
emissions in the molded foam subcategory for Regulatory
Alternative 2. The only difference is the size of the carbon
bed, which would be smaller for the solvent recovery system.
8.2.3. Solid and Hazardous Waste Impacts
For the slabstock foam subcategory there is the potential
for a slight increase in the amount of hazardous waste due to the
control of storage and unloading emissions required by all the
alternatives. The increase would occur when the carbon canisters
used to capture the HAP ABA, or the TDI, from storing and
8-7
-------
unloading operations become saturated. These carbon canisters
would be contaminated with these HAP and would be considered a
hazardous waste. It is uncertain as to the frequency of
saturation of these canisters and subsequent need for their
disposal; however, the industry has indicated that they do not
believe it will be frequent.
For the molded foam subcategory, there is the potential for
a decrease in the amount of hazardous waste, although this
decrease is really a shift to solid waste. A small amount of
hazardous waste is generated from mixhead flushing operations
when the waste flush (MeCl2) is separated from any solids
generated during the flush operations. The MeCl2 can be
regenerated, but the solids must be disposed of as a hazardous
waste. However, by replacing the HAP-based mixhead flush with
non-HAP based flush, these same solids can most likely now be
sent to a solid waste disposal site, depending on the
contaminants.
8.3 ENERGY IMPACTS
Due to the use of several control technologies in both
slabstock and molded foam there will be some increase in the
amount of energy used by this source category. The impact will
vary depending on which control technology is chosen by each
facility, and therefore cannot be accurately calculated.
However, in the discussion of the cost impacts to follow in
Section 9.1, the increased energy usage was evaluated as part of
the annual cost for each pertinent technology.
8-8
-------
9.0 COST AND ECONOMIC IMPACTS OF REGULATORY ALTERNATIVES
This chapter presents the cost and economic impacts of the
existing source regulatory alternatives described in Chapter 6.
The primary and secondary environmental impacts, as well as the
energy impacts were discussed in Chapter 8. As was mentioned in
Chapter 8, there are no anticipated new sources for this source
category.
9.1 MODEL PLANT COSTS
The bases for the model plants are provided in Chapter 7.
The Cost Document was the basis for the majority of the model
plant costs.23 There are several situations in which all or some
portion of the model plant annual costs are negative (i.e.,
represent a cost savings). Throughout this chapter, cost savings
will be denoted in parentheses. Since it is anticipated that no
rebond facilities will be affected by the proposed regulation, no
model plant costs were developed for rebond facilities.
9.1.1 Slabstock Foam
There are five basic model plants for slabstock foam,
representing varying levels of production. Each basic model
plant is separated into facilities that use MeCl2 as an equipment
cleaner and facilities that do not. Several sources were used to
estimate costs for the slabstock foam production model plants.
Information developed for other EPA efforts, supplemented by
vendor quotes, was used for determining equipment leak and
storage tank impacts. The HAP ABA and equipment cleaning costs
were primarily derived from the Cost Document. The following
sections present information for each emission source type, as
well as a summary of their model plant costs.
9.1.1.1 Storage/Unloading
The MACT floor level of control for storage and unloading of
both TDI and HAP ABA is an equipment standard that requires
9-1
-------
either a vapor balance system to return the displaced HAP vapors
to the tank truck or railcar, or a carbon canister through which
emissions must be routed prior to being emitted to the
atmosphere. The subsequent regulatory alternatives do not
contain more stringent requirements. However, since there are no
HAP ABA emissions allowed under Regulatory Alternative 2, there
will be no storage tanks containing HAP ABA under this regulatory
alternative.
The model plant impacts are based on the installation of
vapor balance systems. The basis for estimating vapor balance
impacts was the "Background Information Document for the proposed
gasoline distribution NESHAP."28 This document asserts that
the emission reduction for vapor balance is 95 percent. The bulk
plant model costs in the referenced document include costs for
the vapor balancing of both incoming and outgoing loads. The
unloading of TDI and MeCl2 at flexible polyurethane foam
facilities is comparable to "incoming loads" at bulk plants.
The slabstock foam production model plant costs for
storage/unloading emission control are provided in Table 9-1.
There are no costs for model plants 4 and 5, because all TDI and
MeCl2 storage tanks for these model plants were assumed to be
controlled at baseline. TDI storage tanks at model plant 3 were
also assumed to be controlled.
9.1.1.2 Equipment Cleaning
The MACT floor level of control for equipment cleaning is
the complete elimination of HAP emissions. The subsequent
regulatory alternatives do not contain more stringent
requirements. While there are several alternatives available to
eliminate the use of MeCl2 or other HAP for cleaning the mixhead
and other equipment, model plant costs were developed for only
one alternative: non-HAP cleaners.
The amount of MeCl2 to clean the equipment is consistent for
all model plants. Therefore, the impacts shown below are
applicable for all model plants. More detail on these model
plant costs is provided in the regulatory alternative impacts
memorandum mentioned earlier.
9-2
-------
TABLE 9-1. SLABSTOCK FOAM MODEL PLANT COSTS
FOR VAPOR BALANCING TO CONTROL STORAGE EMISSIONS
MACT Floor and
Regulatory Alternative la
Capital Investment ($)
Annual Cost ($/yr)
Regulatory Alternative 2
Capital Investment ($)
Annual Cost ($/yr)
Costs
Model
Plant 1
$8,220
$1,573
$4,110
$873
(1994 dollars)
Model
Plant 2
$12,330
$2,402
$8,220
$1,745
Model
Plant 3
$4,110
$438
NAb
NAb
The model plant costs presented were calculated using the emission
point specific requirements of Alternative la.
Model Plant 3 was assumed to have baseline controls that would
meet the requirements of Regulatory Alternative 2. Also, Model
Plants 4 and 5 were assumed to have baseline controls that would
meet the requirements of all regulatory alternatives.
9-3
-------
Capital Cost - $0
Annual Cost - ($275)/yr
9.1.1.3 Equipment Leaks
The MACT floor level of control for equipment leaks was
determined to be sealless pumps for TDI transfer pumps. The
first regulatory alternative adds a unique LDAR program for HAP
ABA components. Since Regulatory Alternative 2 does not allow
the emission of any HAP ABA (which, in effect, prohibits the use
of MeCl2 or any other HAP as an ABA), this alternative only
contains the MACT floor requirement for TDI pumps.
For the MACT floor Regulatory Alternative, the cost would
simply be the cost of replacing existing TDI transfer pumps with
sealless pumps for model plant 1. All other model plants have
sealless TDI transfer pumps at baseline. Since there are no HAP
ABA emissions allowed under Regulatory Alternative 2, there will
be no components in HAP ABA service. The only direct equipment
leak impacts under this regulatory alternative will be the TDI
cost associated with the MACT floor Regulatory Alternative. The
model plant costs for equipment leaks are presented in Table 9-2.
9.1.1.4 HAP ABA emissions
There are three levels of control for HAP ABA emissions.
The MACT floor and first regulatory alternative levels are
emission limits based on formulation limitations. The second
regulatory alternative requires the complete elimination of HAP
ABA emissions. For each level of control, model plant impacts
were developed for several technologies. While there are
numerous technologies available to reduce HAP ABA emissions, the
effectiveness of individual technologies is widely disputed
within the foam industry. Therefore, the EPA made assumptions,
based on its knowledge of the industry, regarding the
technologies that could be used to meet each of the three HAP ABA
levels of control.
It is assumed that some technologies can be used to meet
more than one level of control. In these cases, it was assumed
that the technologies would only be used to the degree necessary
to meet the level of the regulatory alternative. In other words,
9-4
-------
OJ
I
W
I
u
M
a
H
H
4
cn
cn
H
m
JJ
CQ
o
0
in
1
Mode]
Plant
0) JJ
•0 C
I*
Si
PI
^
Mode!
Plant
N
H
01 JJ
]S
SrH
Ol
H
H
g> 4J
•o a
A
S 1
< rt!
S &
flj (0
o .
O (N
in H
0) —
> •
4J
« JJ -^
ecu
V4 i
IB 6 -^~
•O JJ 4-1 «rt-
CH ra —
cd < a)
> JJ
h >i 0 to
0 W H O
O 0 U
H JJ rH
b Id 1C rH
H 4J id
EH 3 -H 3
CJ 0) Pi pi
§« U §
H
n
•w-
in
[0
•w-
^f
in
r-
•w-
in
in
(N
H
•w-
H
0) —
> -w-
•H ^T
4J
n) 4-i
c a
M 0)
0) g
4J JJ
i-l 0)
<< 0)
s
>1 fl
M H
0
JJ rH
id m
rH 4J
3 -H
0) ft
0) (C
P5 O
O
H
00
in
o
00
in
•w-
o
00
in
•w-
0
00
in
in
cs
•w-
^^
^i
^
*\
•w-
^^
4J
(Q
0
U
rH
id
£*
<3
JJ
QJ
CN
rH 0)
3 >
O-H
rt
JJ C
rd ri
XJ 0)
jJ jJ
r-H
CQ i
JJ O
C JJ
O id
O rH
Q) CJ1
G (U
•H Pi
rH
0) "O
CQ G
rd fd
XI
(U
rd
m
in
(D
>H O
JJ CQ
JJ
OJ G
0)
CQ g
JJ 0)
04 D1
0)
i~H M
(U
T3 (U
04:
S JJ
JJ
a
-H
o
a
o
-H
CQ
CQ
•H
H
O)
c
-H
CQ
(U
JJ
rd
U
O 5H
JJ (U
4J
TJrH
Q)
CD -H
S JJ
rd
r^ £H
(U rH
JJ
-------
although variable pressure foaming (VPF) can be used to totally
eliminate the use of HAP ABA, it was assumed that at the MACT
floor level, the amount of MeCl2 allowed would still be used and
emitted.
9.1.1.4.1 Chemical Alternatives. It is assumed that the
only regulatory alternative that can be met solely by the use of
chemical alternatives is the MACT floor alternative. The capital
costs of the use of chemical alternatives are the same for all
model plants. Therefore, the capital recovery and other indirect
annual costs (which are a function of the total capital
investment) are also uniform. The other contribution to the
annual cost is the materials cost. This is the cost of the
chemical alternatives minus the cost of the MeCl2 no longer used.
The model plant chemical alternative costs are summarized in
Table 9-3.
9.1.1.4.2 Carbon Dioxide as an ABA. It is assumed that the
HAP ABA requirements for both the MACT floor and first regulatory
alternative can be met using CO2 as an ABA. While there are
other C02 systems available, the model plant costs presented for
C02 as an ABA are costs of the licensed CarDio technology. The
model plant costs for CarDio are summarized in Table 9-4.
The capital costs for the installation of the CarDio
technology are uniform for all regulatory alternatives for all
model plants. Therefore, the annual capital recovery charges are
also analogous for all situations. Similarly, the CO2 tank
rental fee is the same in all situations. The material costs
consist of the costs of the liquid CO2 minus the MeCl2 savings.
It was assumed that the total elimination of HAP ABA
(Regulatory Alternative 2) could not be achieved using only
CarDio. This was because the base formulation would need more
than 3 parts MeCl2 per hundred parts polyol to allow the
substitution of CO2 as an ABA. However, these low-ABA foam
grades could be produced using chemical alternatives. The
combination of CarDio and chemical alternatives was assumed to
achieve the complete elimination of the use of HAP ABA.
9-6
-------
g
M
2
w
CJ
[• COSTS FOR
N REDUCTION
§ °
3 W
s ^
2
w w
§<
^ P-I
° ^
SLABSTOCK ]
'ERNATIVES
t-<
m
i
W
CQ
3
EM
in
*
"• 2
IS 9
H
rH
O
•0
O
01
01
H
n n.
ii CJ
n s
0
U
jj
(3
«
Oi
H
•8 2
fi 9
H
" A
01 2?
r~ °^
J2- ^>
w -w-
in ^
p S
s I
in {£
P 5
S -
^ a.
^ H
? ^
si
m n
O) (N
r~ m
rH Tj*
n in
•w- -w-
i2
4J —
4) >i
^J *Q-
ra ^
o>
> -u
M C CD
OHO
0 O
H rH
4-> 03
H -rl 3
U ft C
9-7
-------
I
en
W
in
SS
«•
8
•w-
o
01
0\
a _.,
• a
0 S
o
4J
a
Pi
H
O
•8
i «
*H Q|
H
a
0
o
ro
01
(N
•w-
0
O
ro
Ol
(N
•w-
o
o
ro
Ol
OJ
•w-
o
0
ro
N
o
o
ro
01
OJ
•w-
, ^
4J
a
0)
g
4J
CQ
•H
id
g
0)
4J
H
*4j
s^
^
0
jj
iH
1
0
o
ro
*.
Ol
•w-
o
o
ro
01
OJ
•w-
o
0
ro
Ol
(N
•w-
0
o
ro
(N
•w-
0
o
ro
01
(N
•w-
^^.
•w-
4J
a)
g
4J
CQ
0)
^
C
H
H
ctt
-H
a
(8
u
01
in
in
ro
•w-
^^
ro
CO
ro
•w-
ro
0
•W-
in
ro
(N
•W-
in
ro
vo
•w-
^-^
^
*^
•CO-
* —
4J
CQ
0
U
rH
(C
1
M
0)
>
i> id
id u
g'g
ID 0)
43 J3
r-l O
< a
a ID
>*3 S>
H H -H
fi p, -U
43 flj
* 0 C
r-l -H h
3 Q
in
o
H
VD
*
in
(N
o
H
vo
v>
t ,
•w-
4J
c
g
4J
CO
0)
^
c
H
1-1
as
-H
a
OS
u
>
N
(N
vo
H
•w-
vo
in
H
•w-
vo
vo
vo
Ol
(N
•w-
*"*
o
vo
vo
CN
•w-
ro
vo
H
CO
•w-
^-^
£
-^
•co-
•+-*
4J
m
0
u
H
nS
3
9-8
-------
9.1.1.4.3 Acetone as an ABA. It is assumed that acetone
can be used to meet the HAP ABA requirements for all three
regulatory alternatives. The model plant costs for acetone as an
ABA are summarized in Table 9-5. As with CarDio, the capital
costs, capital recovery, and other indirect costs are the same
for all model plants and for all regulatory alternatives. The
differences in annual costs are the result of the differences in
material costs and licensing fees. The material costs were
simply the cost of the acetone minus the MeCl2 savings.
9.1.1.4.4 Variable Pressure Foaming. It is assumed that
the HAP ABA requirements for all three regulatory alternatives
can be met using VPF technology. The capital costs are
consistent across all model plants for all regulatory
alternatives. Similarly, the utilities and labor costs, capital
recovery, and other indirect annual costs are also consistent.
Therefore, the only element that changes in the VPF model plant
annual costs is the material costs. Since there is not a need
for additional chemicals or other materials, this simply
represents the cost savings from the unused MeCl2. The VPF model
plant costs are summarized in Table 9-6.
9.1.1.4.5 Forced Cooling. It is assumed that the HAP ABA
requirements for the MACT floor and the first regulatory
alternative can be met using forced cooling. While there are
numerous forced cooling systems in operation or under development
in the United States, the model plant costs presented are based
on the costs of the licensed Envirocure technology, except that
licensing fees are not included. The model plant costs for
forced cooling are provided in Table 9-7. The capital costs for
the installation of forced cooling are uniform for model plants 2
through 5 for all regulatory alternatives. The capital costs for
model plant 1 are lower, assuming that a smaller forced cooling
unit could be installed. Therefore, the annual capital recovery
charges and other indirect annual costs are analogous for all
regulatory alternatives. The utilities charges, which were
provided by the vendor, are also the same for all regulatory
alternatives.
9-9
-------
w
s
o
W
%
o
CO
EH
CO
0 S
0 O
H
EH EH
s o
53 D
| "| Q
o< w
PH
w a
Q 0
O M
S CO
CO
o w
CJ Sj
o
CO <
CO
in
i
W
§
H
in
ft
2
„
to §•
rl *
rH
rH
•8
o
en
H
01 ro
JJ ft
ra 2
0
CJ
4J
c
m
1
p^
ft
rH
a)
T)
i N
f^|
s
H
ft
O
O
O
^
cn
H
•W-
0
o
o
«.
H
•W-
0
O
0
%
(n
H
•W-
O
0
o
*~
o\
S
vo
M
•W-
in
vo
H
H
•W-
n
00
00
•w-
H
H
in
M
•w-
^.^
^
^N^
•w-
^•^
JJ
01
0
CJ
rH
I
H
o
JJ
g
o
4J
r-l
rij
s,
M
o
(0
r-l
g.
0)
0!
0
o
0
v
cn
H
•w-
o
0
o
cn
H
•w-
0
o
0
^
cn
H
•w-
0
o
o
H
•w-
o
o
o
cn
•w-
»-^
ii
JJ
0)
e
JJ
01
0)
fl
H
rH
nj
JJ
-H
ft
n)
U
cn
„
CO
•w-
"^
r)
C3
in
in
•w-
o
in
CO
•w-
0
•w-
in
in
0)
ro
•w-
^^
^
^^
•w-
•*^
jj
0
U
rH
1
s
(N
o
£
(tf
c
0)
JJ
rH
(^
^
^,1
0
JJ
id
rH
g.
o
ft
o
o
0
*•
cn
H
•W-
0
o
o
cn
H
•W-
O
0
o
••
1*
H
O
O
0
-------
in
s
A
• a
n
(0
rH
o
•0
o
X
H
f>
n £j
n
o
o
C
2
H
O
•u N
0 0|
g
o
0
o
o
0
in
..
*?
•w-
o
o
o
*.
0
o
in
^
^
•w-
0
o
o
o
o
in
•w-
0
o
o
«-
o
o
in
«.
^*
0
0
o
I.
o
o
in
^T
^^
•w-
jj
C
0)
4J
01
0)
M a
0 H
0
H rH
b (tf
4J
U ft
^ (C
a u
0
(N
H
ro
N
VD
•w-
o
(N
f^
CS
VO
•w-
o
(N
(N
in
•w-
o
o
^
01
f^
•w-
0
o
^
r-
•w-
^^
>1
^^
•w-
* —
4J
W
0
U
r-l
m
g
c
5
H
a>
>
4J
IT)
g
0)
H
s^
^_l
o
0
H
01
(D
O
0
o
o
o
in
«.
^*
*
o
o
o
V
o
o
in
^
^«
o
o
o
o
o
in
•w-
o
o
o
*
o
o
in
».
^*
•w-
o
o
o
*,
o
o
in
^T
•w-
.
•w-
j i
C
d)
g
4-1
to
0)
^
0
H
H
o
00
^*
00
0
c^
•w-
o
OJ
H
01
•w-
^^
>1
^^
•W-
"^-*
4J
CQ
0
u
rH
m
(^
^
M
0)
. 1
•rt
4J
a
g
0)
4->
iH
^J
S^
^1
o
4J
10
rH
g,
V
K
o
0
o
0
0
in
..
^
*a-
o
o
o
^
o
o
in
,.
^
•w-
0
o
o
o
o
in
•w-
o
o
o
*.
o
o
in
^
*4*
O
0
O
..
0
o
in
•*"
•w-
^^
•w-
jj
(U
g
4J
01
0)
G
H
i-l
as
jj
-H
ft
flj
U
CTl
f^
VO
in
OA
^
•w-
vo
^*
in
H
ro
in
•w-
H
ro
in
in
ro
in
•w-
vo
00
^*
vo
vo
•w-
o
*4*
^*
in
in
•w-
,^
^
•"\
•w-
*~*
jj
03
O
U
l~f
05
p*
<
9-11
-------
p
W
u
£
PH
o
M-4
CO S
0 ^
O h^
J3
P.. Q
fe W
get;
E! S3
o
U 1
w^
o 5
2 W
*>* ^rf*
l£? ^
P*^ pQ
xj ^C
fe
-^
ky ^~j
u a
8
gji
^ i_n
CO g
u
*
t***
1
a\
W
h-J
ffj
rfl
EH
in
u
^"
~ *
u s
M A
H
Q
•0
Ol
H
^-*
m <*>
.y Oj
01 ^
O
u
4J
g
^1
Oi
H
(D
"O /M
O O
Jg S
S
rH
Ol
«
O
0
o
o
o
0
•.
0
o
in
vo
H
•w-
T:
<»
>— '
u
o
u
rH
m
c
H
ID
}>
•H
S
*
r-l
*
^
o
4J
16
H
3
0)
o
0
o
o
o
0
V
CN
•W-
o
o
o
o
o
o
CN
•W-
o
o
o
o
o
o
M
•w-
o
o
0
^
o
0
o
«.
Ofl
o
o
o
o
0
o
«.
H
•w-
^^
•w-
^— '
fi
V
05
(U
H
rH
tfl
4J
-H
ft
0
n
CN
C"\
CTl
H
•w-
CO
cr\
01
o
o
^
ro
CN
H
CN
•w-
o
vo
CN
V
0\
CO
CN
•w-
0
CN
^1
in
H
•w-
^
"*-'
^
0
u
rH
(0
3
in
CN
[^
H
n
o
4.
CN
•w-
in
CN
r~-
H
n
o
CN
•w-
in
(N
t~-
H
n
o
CN
•w-
in
CN
>
•«
H
n
o
^
CN
«
m v*
•H o — '
JJ O >
«J 3 -rt 4J
*J 0) H -P
H C (D n
f< -H 4J (U
k o H
O U rH
.P It rH
(8 -0 O (0
H 0) -rl 4J
3 0 fi -H
5)H « ft
H
•U>
17
^
-— -
u
0
u
HI
ns
9-12
-------
It was assumed that the total elimination of HAP ABA
(Regulatory Alternative 2) could not be achieved using forced
cooling only. This assumption was made due to the many doubts
within the industry that low density, low IFD foams of acceptable
quality can be made using forced cooling without any ABA.
However, it was assumed that these foam grades could be produced
using the combination of forced cooling and chemical
alternatives. The capital costs, capital recovery, and other
indirect annual costs for this combination were simply the sum of
the capital costs of forced cooling and chemical alternatives.
9.1.2 Molded Foam
There are four molded foam production model plants. One of
these model plants represents larger molded foam facilities using
HP mixheads, primarily to produce automobile seats. The
remaining three model plants represent smaller producers that use
LP mixheads to produce a variety of foam products.
The molded foam production model plant costs were developed
directly from the Cost Document, except that the costs were
adjusted to the characteristics of the model plants. All
technologies, except for work practices to reduce mixhead
flushing emissions, totally eliminate HAP emissions. The
following sections present information for each emission source
type, as well as a summary of their model plant costs. More
details regarding the molded foam model plant costs can be found
in the regulatory alternative impacts memorandum mentioned
earlier.
9.1.2.1 Mixhead Flushing
Model plant impacts were developed for four technologies to
reduce or eliminate mixhead flushing emissions: work practices
for Regulatory Alternative 1; and non-HAP flushes, HP mixheads,
and self-cleaning mixheads for Regulatory Alternative 2. Costs
were developed for the work practice of an emission suppression
and solvent recovery system for the MeCl2 used to flush the
mixhead. Table 9-8 presents a summary of the model plant costs
for mixhead flushing for these technologies.
9-13
-------
f)
4J
e
a
rH
Hi*
rH
•o
o
is
CN
a «
S m
^ P* Q.
H K) ™
O __.
•o o
M "0
o\ O
Hi
ID
*> H _i
ID 0) "
0 *O I*
O 0 "ti
H
hJ
rH
a
O
rH
O
Jl
U
0)
EH
o -—
•w-m
vo
o
h.
00
•w-
0 ^-,
•w-ro
oq
CO
«.
ro
•w-
> — *
0 --,
•w-o
~~
0 O
4J ~
fl M
4J
m —
(U
m c to
3 H o
H U
fo rH
(tf rH
Pi 4-> (0
3-H S
w ft d
i nS C
c u .3
0
in in
d ro
H in
>. »
00 (N
in H
ID H
•W--W-
in t^
H 0
OJ H
CO \D
in ^"
VO H
•W--W-
in in
r>q H
H CO
CO ro
in vo
VO H
•W--W-
0 O
™
•0 ""
* 4J —
ID fl SH
•H 4J-W-
S M ^
4J
Vl fl CO
3 H 0
a u
ID rH
0» nSH
H 4-> (t
ft-H 3
JH nj (H
0>U 3
•rl
SB
00 ^^
CO H
vo ^
*• ro
in -
fs] \Q
C* H
•W--W-
00 H
oo ro
VO OJ
V k.
in >
W rH
M
•w-
00 00
oo m
W C^l
in •*
M ro
OJ -W-
•w-
0 0
*"*
•o ^
« 4J -^
» fl ^
*H jJ
& to •*— *
0)
G C to
•H H O
fi U
a) H
01 (IS H
H 4J (C
U-H 3
' ftfl
i
^1 JJ
h M " —
01 (U
0 fl to
OHO
a> a
PS rH
(0 rH
•u 4J as
0) ft fl
r* (0 fl
rH CJ ^
0
01
9-14
-------
9.1.2.2 Mold Release Agents
The MACT floor level of control for mold release agents is
the prohibition of the use of HAP-based mold release agents,
resulting in a 100 percent emission reduction. Model plant
impacts were developed for three technologies that can achieve
this level: reduced VOC mold release agents, naphtha-based mold
release agents, and water-based mold release agents. A summary
of mold release agent model plant costs is in Table 9-9.
9.1.2.3 Repair Adhesives
The MACT floor level of control for repair adhesives is also
the prohibition of the use of HAP-based adhesives, resulting in a
100 percent emission reduction. Model plant impacts were
developed for three technologies that can achieve this level:
hot-melt adhesives, water-based adhesives, and hydrofuse
adhesive. Table 9-10 provides a summary of the repair adhesive
model plant costs.
9.2 NATIONWIDE COST IMPACTS
This section presents the nationwide costs and HAP emission
reductions associated with the existing source regulatory
alternatives. The basic approach used to estimate these
nationwide impacts was to apply the model plant impacts presented
in the previous section to those facilities represented by the
model plants. As was mentioned in Chapter 7, in several
instances more than one technology could be used to achieve the
level of control required by the regulatory alternative. In
these cases, the EPA made assumptions regarding the number of
facilities represented by each model plant that would use the
various technologies.
In addition to costs of the technologies to control HAP
emissions, facilities will also incur costs for the associated
monitoring, recordkeeping and reporting (MRR). For the purposes
of this analysis, these MRR costs were estimated to be 8 percent
of the control costs. Therefore, the total control costs
calculated using the model plant costs was multiplied by
108 percent to obtain the total cost of control for each
regulatory alternative.
9-15
-------
^^
B
H
a
rH
rH
O
TJ
at
at
H
a
4J
m
O
0
H _
O
Is
32
n*4
0 N
Is
32
rH
Q rH
ii
a*
®4J
** iH
K
O
i-H
0
S
o
Q)
H
o ro
•w-og
o
h.
H
•to
O ffl
•w-ro
•to-
o o
•to in
•to-
0 0
•to -to
s~— *
4J ^»
C ^i
0) >.
C jj -w-
O w ^
0) 0)
f-3 M
OHO
o u
>rH
i ns H
•d -u ns
ffl -H 3
0 ft C
3 (« S
•O 0 <
(D
os
O
•W-ffl
in
„
S
o ro
•to- CM
CN
•
o in
t^
ro
•to-
0 0
•w-to-
•to
"""^
4J ^
c o) >i
o S^-
0>4J -to
di ra ^
0)
•8 >• -U
« C m
OHO
id u
« rH
I (IS H
<8 4-> m
J3 -H 3
•U ftC
A m 2
ftCJ i?
85
O H
•to 00
CTl
•to
O 00
•to n
•to
O 00
v>*#
•w-
o o
•to-to
w
*^^
4J —~
CH ^4
0) >i
a 4J-to
ID W —
&l JJ
C to
•OHO
0 U
BH
a * pH
(D 4J (C
1 -H S
(D Hi fl
•U CJ <
id
S
9-16
-------
H
ttl
H M
"0 .y
S
H
T| -S ^
H ^l
1 ss
01
0)
iH
O "O jj
4J O rt
a S3 5
0 JS
O d n!
,_!
Is
SHJ
H
B"
O
rH
0
U
fl>
« m
O IO
00 00
». «.
W> H
•W--W-
o o
•W--W-
0 O
•W--W-
o r-
oo n
(X> VD
•W--W-
s>
c w
0) ^*i
» g"^
> 4J -W-
•H 0} ^
a a)
« > 4J
J3 C to
•OHO
r-t
*J (C iH
iH AJ (C
t^
in
•w-
^
""""*
C ^
•H 4J -W-
o to ^
0) 0)
•C ^ xj
•8 fl 10
ft, H O
U
O r-1
m m 1-1
3 -U (8
IH -rl 3
0 ft C
ti td c
•o u 3.
^
a
o ^^
vt r*-
01
•w-
O 0
•W--W-
O 0
•l/V-W-
o —.
•w- ^
in
00
•w-
•w-
>i
fl) £ ^^
JCJ 4J -W-
•O W ^
rt! a)
r* 4^
•o fl m
O H O
a u
Id r-l
n ns H
i 4J rt
i4 -rl 3
ffl P.G
(d u S
9-17
-------
As was also mentioned in Chapter 7, only major sources of
HAP will be subject to the Flexible Polyurethane Foam Production
NESHAP. All slabstock foam facilities were considered to be
major sources. However, as was discussed, since the HP molded
model plant and the smallest LP molded model plant have emissions
below the major source thresholds, it was assumed that the
facilities represented by this model plant would not be affected
by the Flexible Polyurethane Foam Production NESHAP. Therefore,
the nationwide regulatory alternative impacts are based on LP
model plants 2 and 3. Further, it is predicted that all stand-
alone rebond facilities are area sources, and all co-located
rebond facilities are already in compliance with the proposed
requirements. Therefore, no impacts are predicted for rebond
foam production operations. The following sections briefly
discuss how the model plant cost impacts were used to estimate
nationwide impacts. Complete details on the application of the
model plant costs, and assumptions made regarding the numbers of
facilities utilizing various technologies can be found in the
regulatory alternative impacts memorandum referenced earlier in
this chapter.
9.2.1 Slabstock Foam Nationwide Cost Impacts
The following sections briefly describe the derivation of
impacts for each of the four slabstock foam production emission
sources. As was shown in Table 6-9, three basic regulatory
alternatives were developed for slabstock foam, with Regulatory
Alternative 1 including two implementation options. The first
option (Alternative la) consists of emission point specific
requirements, and the second (Alternative Ib) consists of a
source-wide emission limitation that allows the owner or operator
to select the emission points to control, as long as the source-
wide emission limitation is achieved. The model plant costs
presented in this section for Regulatory Alternative 1 were
calculated using the emission point specific requirements of
Alternative la.
9-18
-------
9.2.1.1 Storage/Unloading. Equipment Leaks, and Equipment
Cleaning
As discussed earlier, model plant costs and HAP emission
reductions were only developed for one technology for each of
these emission sources. Therefore, the nationwide regulatory
alternative costs were simply determined by multiplying the model
plant costs for each technology by the number of facilities
represented by each model plant.
9.2.1.2 HAP ABA Emissions
As was discussed in the model plant cost section, there are
a variety of technologies that reduce or eliminate the use of HAP
ABA. In estimating the nationwide regulatory alternative costs,
it was necessary to make assumptions regarding the number of
facilities that would use each available technology to comply
with the HAP ABA requirements. The distribution of technologies
used to estimate the HAP ABA slabstock foam nationwide regulatory
alternative costs by model plant is provided in Table 9-11.
For the MACT floor regulatory alternative, it was assumed
that the majority of the facilities represented by the smaller
model plants would choose to comply through the use of chemical
alternatives. It was assumed that most of the facilities not
using chemical alternatives would comply through the use of
either CarDio or forced cooling. The reduction of HAP ABA usage
through forced cooling is a technology that has already
penetrated the industry to a large degree.
For Regulatory Alternative 1, the technologies available to
meet this level were acetone, forced cooling, VPF, and CarDio.
Due to the relatively low costs, it was assumed that the majority
of smaller facilities would select CarDio. For model plants 3
and 4, where it was assumed that capital costs were less of a
problem, a smaller percentage were assumed to choose the CarDio
technology over acetone and forced cooling. For model plant 5,
it was assumed that the four technologies would be used equally.
The rationale for the Regulatory Alternative 2 assumptions
was basically the same as that for Regulatory Alternative 1
(noting of course that CarDio and forced cooling also use
9-19
-------
TABLE 9-11. DISTRIBUTION OF ABA EMISSION REDUCTION
TECHNOLOGIES USED TO ESTIMATE THE SLABSTOCK FOAM NATIONWIDE
REGULATORY ALTERNATIVE COSTS BY MODEL PLANT
Technology
MACT Floor
CarDio
Acetone
VPF
Forced cool
Chem Alts
Reg Alt I
CarDio
Acetone
VPF
Forced cool
Reg Alt II
CarDio +
Chem Alts
Acetone
VPF
Forced cool
+ Chem Alts
Number of Facilities Using the Technology
MP1 MP2 MP3 MP4 MP5
4
1
0
2
12
13
3
0
3
10
3
0
6
9
1
0
4
14
18
4
0
6
15
3
1
9
4
1
0
3
6
6
3
0
5
5
2
1
6
4
1
0
3
3
5
2
0
4
3
2
2
4
1
0
2
2
1
1
1
2
2
1
1
2
2
9-20
-------
chemical alternatives for Regulatory Alternative 2). The one
exception is that it was assumed that a few more large facilities
would choose to spend the resources to install VPF under
Regulatory Alternative 2.
9.2.2 Nationwide Cost Impact Summary for Slabstock Foam
The nationwide regulatory alternative impacts for slabstock
foam production are provided in Table 9-12. The table includes
the emission reductions that were discussed in Chapter 8, as well
as their incremental cost effectiveness. As discussed earlier,
the impacts shown for Regulatory Alternative 1 were developed
assuming compliance with the emission point specific requirements
of Alternative la. The EPA believes that the cost and economic
impacts of Alternative Ib would be slightly less than those of
Alternative la, since the owner or operator could select the most
cost-effective controls for their facility. However, for the
purpose of this analysis, the EPA assumes that the costs
presented in this section are representative of both
Alternative la and Ib.
There are two sets of impacts shown for Regulatory
Alternative 2. The first only takes into account the "direct"
HAP emission reductions associated with the regulatory
alternative emission requirements. However, as discussed
earlier, the elimination of the use and emission of HAP ABA will
also result in the elimination of HAP ABA emissions from storage
and equipment leaks. The second set of impacts include these
"indirect" HAP emission reductions.
The cost effectiveness of all three regulatory alternatives
are less than $2,000 per ton of emission reduction, with the
highest being the MACT floor level of control at $1,262 per ton.
Due to the incremental annual cost savings for the Regulatory
Alternative 1 HAP ABA requirements, the overall incremental cost
effectiveness from the MACT floor to Regulatory Alternative 1 is
negative. The incremental cost effectiveness from Regulatory
Alternative 1 to Regulatory Alternative 2 is $800 per ton,
considering the indirect emission reductions.
9-21
-------
CQ
E-i
CO
O
U
M
H
3
PH
H
j
^
pH
(Vj
O
[-J
J
D
0
W
MH
i
?»
fe
w
u
o
EH
CO
ffj
j
i— i
CQ
Q
M
0
EH
f£
r*S
^
.
C^
1
O^
1— 1
CD
i^j
C__j
rH
<0
4J
0
H
J
o 5
o 3
a
g
W
CQ
(3
O
IQ jj 0)
•H g "d
"ri
VD H
^* nH
•to- -to-
CS ^O
•to- 00 OOJ3
Ol H '
VO
[C
•to
^~*
o 01 ro VD
o in o
o co o
in in co
Ol Ol -to-
•W--W-
o vo in 01
VO Ol H ^"
Ol O 00
00 00 f-
in H -to-
in H
•W--W-
*£
^i^^
^ a
m o
fl\
•to- 4J -to-
*-* *-^ +-*
^^
•U ^ Ci CO
a ^,0 m
0) -C.-H (U
g v> ** fl
4J ^^ O (U
to 3 >
CD IB "O -rl
> 4J 0) 4J
h fl to e; u
OHO d)
0 CJ fl -to-
ro VO H
Ol ^J< H
Ol 00
CO t-
VO -to-
•to-
0 H 01 VO
o in vo 01 (u
01 01 01 in i
- ^ » -to-
H s^
i> V* 4J -to-
•H — •~"~
4J ^.^
fl) 4J Lj C3 W
G ti &o m
to CU^-H 0) rH
0) g C
< 0) IB Tf -H rH (U
> 4-> d) 4J H g
N C to rt U c8 0)
h H O 0) ^ tn
0 U C MH <0 U
4J rH O MH > C
(8 (« rH -rl W O H
i-H 4-)
•to- -to-
es ro o VO T3
in H o <
Ol f H Z
oo o X
o >
ro -w- m
•to- -to
17
^-i^^
~~- C
d CO 0
rj | I
0) -^-* O ^^ fj
> -to- 4J -to- O
•H ~— •" ^- -H
•U ^ tO
« 4J S-| C W tO
C C >, 0 to -H
h (U-C-H (I) rH g
C (OH
4J 4J ^ O tt) -U
rH tQ 3 > fl 4J
> 4J <1> 4J rH g 0)
C* H O 0) ^ iH -H
O U C! C C •
i-H 4-1 (C CO
3 -H 3 tQ 4J £
6) ft C -H CQ 4J
« rt 5 g 0 -H
« O < W CJ !£
00 rH -^ Ol H
in in I-H oo o
rH ^f dj VO 00
-4J -to-to-
vo O
H 4J
M-i IM
1 1
44 M-I
1 1
H
•* tn
H I
> W
H '
a
o
-H
4J
U
T3
(U
PH tO
to
C O rH
o a «
-H a
01 -H rH (D
-H .U rH g
ra g u n5 d
u m H o H
O rf|
C P< to
0) rt O
K W u
9-22
-------
CN
H
1
CTi
CD
rH
42
OS
EH
SH
O
4H
CQ
CD
4-1
O
53
4-1
O
0
fc
CD
SH
CU
u
•H
(-]
^
»
ra
4->
ra
O
u
Cn
S3
-H
Jj
rH
o
ft
CU
SH •
ra
73 4-1
53 m
OS O
u
CJIrH
53 O
•H rH
ft4J
CU S3
cu o
jy U
73
>H rH
O OS
U 4->
m
-H Cn
J_) £|
0-H
CU >
MH OS
14H W
CU
4-1
4-1 m
ra o
o u
D
•H
•MH
-H
U
CU
ft
ra
4-1
S3
-H
o
ft
S3
o
-H
ffl
CQ
-H
g
CU
CU
43
4-1
Cn
S3
•H
ra
£j
73
CD
OS
rH
rj
u •
rH OS
OS H
u
d)
CD 4-1
S OS
73 rH
CD CD
4-1 4-1
S3 rH
CD S3
u cu
OS £
ft CU
C* Cj
•H -H
p
cu D1
43 CU
CU
43
4-1
SH 4H
0 0
4J
as m
rH 4J
3 S3
Cn CD
CU £
SH cu
j_i
CU -H
43 3
4-> cr
CD
O
O
ra 43
4J 4-1
53
CD O
E 4->
CU
(. 1 . r
-H OS
&CJ
-H
CD 4-1
rH 53
CD
CD 73
43 -H
4J
CD
CD SH
U OS
53
-H CD
ra u
f-i
CQ 3
4-1 0
u ra
OS
ft S3
£ o
•H -H
CQ
rH CQ
OS -H
4-1 £
53 CD •
CD CD
£ 0) >
0) 43 -H
SH 4-> 4-)
U OS
53 SH 53
•H O rH
4H CD
0 4J
S3 CD rH
> OS
CD -H
SH 4J ra
o$ as 3
53 0
CD rH-H
SH cu >
CU 4-1 CU
43 rH rH
EH OS ft
ra
-H
ra
O
u
rH
OS
3
53
53
OS
rH
OS
4-1
S3
CU
CU
rH
o
S3
-H
CU
ra
3
OS
U
CD
•°
73
CU
4J
OS
rH
3
U
rH
OS
u
Jj
o
S3
ra
ra
cu
S3
cu
j>
-H
4J
U
CD
M— 1 *
H [ ^^
CD 4-1
•H
4-1 4->
m S3
O OS
tJ1
rH
OS CD
4-1 >
S3-H
CD 4-1
£ OS
CD Cn
r< CD
U S3
M OS
ernative
4-1
rH
OS
rXl
rH
0
4-1
OS
rH
;3
Cn
CD
rH
43
4-1
-H
^
73
CU
4-1
OS
-H
U
O
ra
m
0$
ra
53
0
•H
4-1
u
73
CU
rH
S3
O
-H
m •
ra CD
-H U
£ rH
CD 3
o
4-i ra
u
Q) S3
SH O
-H -H
73 ra
S3 ra
-H -H
£
o cu
S3
ra
Q) -H
SH43
OS 4-1
0) SH
rH 0
CD MH
43
EH CN
are no
for HAP
crementa
S3
CD CD -H
rH >
CU -H CD
43 4-143
4-1 OS 4-1
S3
- SH -
t£ (U 0)
CQ 4-) SH
ft
-H ra SH ft
£ 4-> CD OS
CD S3 ra
CD 4-1
CD £ E ft ra
-H ft £ CD
4-) -H O 43
OS 3 U 4-1
S3 D1
SH CD SH SH
CD O O
4-> SH M-4
•H o ra
OS rH CQ
Cn CD ra
>i 53 ra CD
SH -H CQ S3
O 73 CD CD
4-1 OS > >
OS O -H
rH rH CD 4-1
3 S3 Cn u
Cn 3 as CD
CD \ SH 14H
SH CD O MH
(314-1 CD
CD as ra
U SH 4->
S3 O < ra
-H 4-i eg o
co ra 5j u
9-23
-------
9.2.3 Molded Foam Nationwide Cost Impacts
The distribution of technologies used to estimate the molded
foam nationwide regulatory alternative costs by model plant is
provided in Table 9-13.
9.2.3.1 Mixhead Flush
Regulatory Alternative 1 requires work practices to reduce
mixhead flush emissions. As discussed earlier, the EPA developed
model plant costs for four technologies: one for Regulatory
Alternative 1 (solvent recovery) and three for Regulatory
Alternative 2 (non-HAP flushes, HP mixheads, and self-cleaning
mixheads). While the three technologies that totally eliminate
mixhead flush emissions could also be used to comply with
Regulatory Alternative 1, it was assumed that no facility would
totally eliminate HAP mixhead flushes to comply with the work
practice standard. Therefore, the cost for HAP mixhead flushes
for Regulatory Alternative 1 are entirely based on the
application of the solvent recovery model plant impacts to the 98
facilities represented by LP model plants 2 and 3.
Regulatory Alternative 2 prohibits the use of HAP mixhead
flushes. As noted above, model plant costs were developed for
three technologies that could be used to meet this requirement.
To determine nationwide costs, self-cleaning mixheads were not
considered an option, because they have significant limitations
and are currently not in use in the industry. It was assumed
that 90 percent of LP model plant 2 facilities (49 facilities)
and 80 percent of LP model plant 3 facilities (35 facilities)
would utilize non-HAP flushes, and the remainder would install HP
mixheads.
9.2.3.2 Mold Release Agents
The emission limitation for mold release agents was the
prohibition of HAP-based mold release agents for all three
regulatory alternatives. As discussed earlier, model plant costs
were developed for three technologies that meet this level:
naphtha-based release agents, reduced-VOC release agents, and
water-based release agents. It was assumed that the majority of
facilities would choose either water-based or reduced-VOC release
9-24
-------
TABLE 9-13. DISTRIBUTION OF TECHNOLOGIES USED TO ESTIMATE
THE MOLDED FOAM NATIONWIDE REGULATORY ALTERNATIVE COSTS
BY MODEL PLANT
Number of Facilities Using the
Technology
Emission Source/Technology Low-Pressure Low-Pressure
Model Plant 2 Model Plant 3
Mixhead Flush
Reg Alt I
Solvent recovery 54 44
Reg Alt II
Non-HAP flush 54 35
HP mixheads 0 9
Mold Release Agents
MACT Floor
Reduced VOC
agents
Naphtha -based
agents
Water-based
agents
18
18
18
15
14
15
Repair Adhesives
MACT Floor
Hot -melt
adhesives
Water-based
adhesives
N/A
N/A
22
22
9-25
-------
agents. For each model plant it was assumed that 1 percent would
use naphtha-based agents, and the remainder would be split evenly
between the other two options.
9.2.3.3 Repair Adhesives
The emission limitation for repair adhesives was the
prohibition of the use of HAP-based adhesives for all three
regulatory alternatives. There were three technologies for which
model plant costs were developed that meet this level: hot-melt
adhesives, water-based adhesives, and hydrofuse. Hydrofuse was
not considered in the regulatory cost analysis, because there are
no known facilities in this industry using this technology.
Since no other information was available regarding industry
preference, it was assumed that 50 percent of the facilities
would use water-based adhesives, and 50 percent would use hot-
melt adhesives.
9.2.4 Nationwide Cost Impact Summary for Molded Foam
The nationwide regulatory alternative impacts for molded
foam production are provided in Table 9-14. The table includes
the emission reductions that were discussed in Chapter 7, in
order to show the overall cost effectiveness of the regulatory
alternatives, as well as their incremental cost effectiveness.
9-26
-------
TABLE 9-14.
NATIONWIDE MOLDED FOAM REGULATORY
ALTERNATIVE COSTS
Impacts by Emission Source
Regulatory Alternative
MACT Floor
Capital Investment ( $ )
Annual Cost ($/yr)
Emission Reduction
(tons/yr)
Cost Effectiveness
($/ton)
Regulatory Alternative 1
Capital Investment ($)
Annual Cost ($/yr)
Emission Reduction
(tons/yr)
Cost Effectiveness
($/ton)
Overall
Incremental
Regulatory Alternative 2
Capital Investment ($)
Annual Cost ($/yr)
Emission Reduction
(tons/yr)
Cost Effectiveness
($/ton)
Overall
Incremental
Mixhead
Flush
$0
$0
0
$0
$4,630,500
($123,767)
1,501
_b
c
$5,923,125
$563,867
2,001
$282
$1,375
Mold
Release
Agent
$0
$153,965
270
$569
$0
$153,965
270
$569
NAd
$0
$153,965
270
$569
NAd
Repair
Adhesive
$149,688
$41,492
61
$693
$149,688
$41,942
61
$693
NAd
$149,688
$41,942
61
$693
NAd
Total
$149,688
$195,907
331
$592
$4,780,188
$72,140
1,832
$39
_b
$6,072,813
$759,775
2,332
$326
$1,375
Annual costs include monitoring, recordkeeping, and reporting costs,
which were assumed to be 7.6 percent of the total control costs.
Cost effectiveness not calculated because net annual cost is a
negative quantity (cost savings).
Incremental cost effectiveness not calculated because incremental
annual cost is a negative quantity.
There are no incremental impacts since the requirements of the
regulatory alternative for the emission source are identical to the
requirements of the previous alternative.
9-27
-------
9.3 ECONOMIC IMPACTS
An economic impact analysis of the proposed regulatory-
options was prepared to evaluate primary and secondary impacts on
(1) the slabstock and molded foam sectors of the flexible
polyurethane foam industry, (2) consumers, and (3) society.
As shown in Table 9-15 for the slabstock foam sector of the
industry, the total annualized social cost (in 1994 dollars) of
the each of the regulatory alternatives is $11.9 million for the
MACT floor, $7.18 million for Regulatory Alternative 1 (the
proposed standard), and $10.9 million for Regulatory
Alternative 2. Market price is estimated to increase by 2.28
percent, 2.20 percent, and 3.82 percent under each of the
regulatory alternatives respectively. Corresponding decreases in
market output are estimated to be 1.12 percent, 1.08 percent, and
1.86 percent respectively. Employment losses are estimated to be
1.12, 1.09, and 1.86 percent (i.e., 99, 96, and 164 jobs)
respectively. These economic factors indicate that Regulatory
Alternative 1 has the lowest cost estimate and lower economic
impacts than the MACT floor.
As shown in Table 9-16 for the molded foam sector, impacts
on price and output are estimated to be smaller than those
predicted for the slabstock market. The total annualized social
cost (in 1994 dollars) of the each of the molded foam regulatory
alternatives is $0.19 million for the MACT floor, $0.06 million
for Regulatory Alternative 1, and $0.71 million for Regulatory
Alternative 2 (the proposed standard). Price is estimated to
increase by 0.07 percent, 0.84 percent, and 1.14 percent under
each of the regulatory alternatives. Corresponding decreases in
market output are estimated as 0.04 percent, 0.42 percent, and
0.56 percent respectively. Employment losses in the molded
sector are estimated to range from 0.04 to 0.67 percent (2 to 37
jobs) .
Given the predicted changes in market price and output, the
industry will experience increases in the value of shipments
(i.e., industry profits) under all regulatory scenarios, because
estimated price increases more than offset the lower production
9-28
-------
TABLE 9-15.
SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC
IMPACTS FOR SLABSTOCK FOAM
Price Increase
(percent)
Production Loss
(percent)
Employment Loss (jobs
lost)
Emission Reduction
(percent)3
Total Economic Costs
(MM$, 1994) b
Economic Cost
Effectiveness ($/ton)c
Plant Closures
MACT Floor
2.28
1.12
99
57.6
11.86
1213
2
Regulatory
Alternative 1
2.22
1.09
96
70.4
7.46
624
3
Regulatory
Alternative 2
3.82
1.86
164
100
10.92
644
4
a Includes emission reductions achieved from control, plus
reductions attributable to closures.
b Incorporates material cost savings, but does not include
opportunity cost of lost foam quality from formulation changes.
c Differs from engineering cost effectiveness due to dynamics in the
market (price and quantity produced changes) resulting from
regulation implementation.
9-29
-------
TABLE 9-16. SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC
IMPACTS FOR MOLDED FOAM
Price Increase
(percent)
Production Loss
(percent)
Employment Loss (jobs
lost)
Emission Reduction
(percent)3
Total Economic Costs
(MM$, 1994)
Economic Cost
Effectiveness ($/ton)b
Plant Closures
MACT Floor
0.07
0.04
2
11.2
0.19
568
0
Regulatory
Alternative 1
0.84
0.42
23
62.3
0.06
30
3
Regulatory
Alternative 2
1.36
0.67
37
78.6
0.71
305
0
a Includes emission reductions achieved from control, plus
reductions attributable to closures.
b Differs from engineering cost effectiveness due to dynamics in the
market (price and quantity produced changes) resulting from
regulation implementation.
9-30
-------
volumes. Since no significant export or import markets exist for
the industry (due to prohibitive transportation costs), no
impacts on foreign trade are expected.
The analysis also predicts the number of plant closures that
may result from the imposition of compliance costs on a facility.
For the analysis, a worst-case assumption is adopted that the
facilities with the highest emission control costs are the least
efficient producers in the market. Actual plant closures will be
less than that predicted if plants with the highest emission
control costs are not the least efficient producers in the
industry. In addition, the outcome of predicted closures is
sensitive to the wide variety of emission control technologies
assigned to the model plants. If the control technology assigned
to the representative model plant is different than that which
would be chosen by an actual facility, the analysis could
overestimate the number of predicted plant closures. Therefore,
a sensitivity analysis was performed to test the outcome of
closures based on the assignment of control technology to model
plants. Economic impacts of the sensitivity analysis are
provided in Table 9-17. For the slabstock sector, plant closures
are estimated to range from 1 to 2 facilities for the MACT floor,
1 to 3 facilities for Regulatory Alternative 1 and 1 to 4
facilities for Regulatory Alternative 2. For the molded foam
sector, closures are estimated to be zero for the MACT floor, 3
for Regulatory Alternative 1, and zero for Regulatory Alternative
2 (a sensitivity analysis was not performed for the molded foam
production subcategory). Given the significant amount of
restructuring currently occurring in the industry (mergers, buy-
outs, and shut-downs), the number of facility closures that will
result from the proposed regulation is likely to be minimal.
9-31
-------
TABLE 9-17. SUMMARY OF REGULATORY ALTERNATIVE ECONOMIC
IMPACTS FOR SLABSTOCK FOAM - SENSITIVITY ANALYSIS
Price Increase
(percent)
Production Loss
(percent)
Employment Loss (jobs
lost)
Emission Reduction
(per cent )a
Total Economic Costs
(MM$, 1994)b
Economic Cost
Effectiveness ($/ton)c
Plant Closures
MACT Floor
2.28
1.12
99
57.6
11.75
1202
1
Regulatory
Alternative 1
2.22
1.09
96
70.4
7.31
612
1
Regulatory
Alternative 2
2.49
1.22
108
100
10.11
596
1
a Includes emission reductions achieved from control, plus
reductions attributable to closures.
b Incorporates material cost savings, but does not include
opportunity cost of lost foam quality from formulation changes.
c Differs from engineering cost effectiveness due to dynamics in the
market (price and quantity produced changes) resulting from
regulation implementation.
9-32
-------
10.0 SELECTION OF THE STANDARDS
The purpose of this chapter is to provide the rationale for
the selection of the standards for the flexible polyurethane foam
production source category. In order to provide background for
the subsequent discussions, the first section of this chapter is
a summary of the proposed rule. This is followed by a discussion
of the rationale for the selection of various aspects of the
standards, including the source category and pollutants to be
regulated, the level and format of the standards, and the
compliance, reporting, and recordkeeping provisions.
10.1 SUMMARY OF THE PROPOSED STANDARDS
This section provides a summary of the proposed regulation.
The full regulatory text is available in Docket No. A-95-48,
directly from the EPA, or from the Technology Transfer Network
(TTN) on the EPA's electronic bulletin boards. More information
on how to obtain a copy of the proposed regulation is provided in
the preamble for the proposed standards.
10.1.1 Applicability and Compliance Schedule
The proposed standards would regulate HAP emissions from
facilities that produce slabstock, molded, or rebond flexible
polyurethane foam, provided that a facility is a major source or
is located at a plant site that is a major source. Flexible
polyurethane foam processes meeting one of three criteria are
exempted from the regulation: (1) a process located at a plant
site where the plant site is limited by a federally enforceable
limit to emissions less than 10 tons per year of any single HAP
and less than 25 tons per year of all HAP; (2) a process
exclusively dedicated to the fabrication of flexible polyurethane
foam; and (3) a research and development process.
10-1
-------
Processes subject to the proposed regulation would be
required to comply within three years of the effective date of
the regulation.
10.1.2 Standards for Molded Flexible Polvurethane Foam
Production
At molded foam facilities subject to the proposed rule,
emissions from three emission sources are covered by the proposed
rule: mixhead flushing, mold release agent usage, and the use of
adhesives to repair molded foam. For each of these emission
sources, the proposed rule prohibits the use of HAP or HAP-based
products at the new and existing sources. Other than the initial
notification and notification of compliance, there are no
associated monitoring, reporting, or recordkeeping requirements
for molded foam producers.
10.1.3 Standards for Rebond Foam Production
The proposed regulation prohibits the use of HAP-based
cleaners or mold release agents in the production of rebond foam
at new and existing sources. Other than the initial notification
and notification of compliance, there are no associated
monitoring, reporting, or recordkeeping requirements for rebond
foam producers.
10.1.4 Standards for Slabstock Flexible Polyurethane Foam
Production
At slabstock foam facilities subject to the proposed rule,
emissions from four emission sources are covered by the proposed
rule: storage vessels, equipment leaks, HAP ABA use, and
equipment cleaning. The requirements are separated into two
basic categories corresponding to the two major uses of HAP in
the slabstock process: (1) diisocyanate used as a reactant in
the foam process, and (2) HAP ABA. The diisocyanate use'd in the
production of slabstock foam is almost always TDI, and the HAP
ABA used is almost always MeCl2.
10.1.4.1 Diisocvanate emissions
Emissions of diisocyanate from storage vessels and equipment
leaks are covered by the proposed standards. For new and
existing sources, there are two compliance options for storage
10-2
-------
vessels. The vessel can be equipped with a vapor return line
that returns vapors displaced during storage vessel filling to
the tank truck or railcar. The second option is to equip the
storage vessel with a system in which displaced vapors are routed
through a carbon adsorption system prior to being discharged to
the atmosphere. Storage vessels equipped with carbon adsorption
systems must monitor the outlet of the carbon system to detect
breakthrough.
Transfer pumps in diisocyanate service must be either
sealless pumps, or submerged pump systems that are visually
monitored weekly to detect leaks. Any transfer pump leaks
detected must be repaired within 15 calendar days. Diisocyanate
leaks for other components in diisocyanate service (valves,
connectors, and pressure-relief valves) detected by visual,
audible, olfactory, or any other detection method must be
repaired within 15 calendar days, as well.
10.1.4.2 HAP ABA storage and equipment leak emissions, HAP
ABA emissions from the production line, and equipment cleaning
HAP emissions
HAP ABA emissions from three emission sources - storage
vessels, equipment leaks, and the production line - are covered
by the proposed regulation. In addition, HAP emissions from
equipment cleaning are covered.
The proposed regulation requires that owners or operators
comply with requirements for each of the four types of emission
points (HAP ABA emissions from storage vessels, equipment leaks,
and the production line, and HAP emissions from equipment
cleaning).
However, since methylene chloride is the primary HAP used as
an ABA and as an equipment cleaner, the proposed rule allows
flexibility in compliance with the HAP ABA and equipment cleaning
provisions. As an alternative to the emission point specific
limitations, the owner or operator can elect to comply with a
source-wide emission limitation. Owners or operators selecting
the source-wide emission limitation must maintain the combined
emissions from all of these sources below the required level.
10-3
-------
While this option is slightly more stringent than the emission
point specific limitations, the EPA believes the flexibility it
provides will prove to be beneficial for sources selecting this
alternative. The emission point specific limitations are
described below in sections 10.1.4.2.1 through 10.1.4.2.4.
Owners or operators selecting the source-wide emission limitation
must maintain the combined emissions from all of these sources
below the required level. This option is described below in
section 10.1.4.2.5.
10.1.4.2.1 HAP ABA storage vessel requirements. The
requirements for HAP ABA storage vessels are identical to the
diisocyanate storage vessel requirements discussed above in
section 10.1.4.1. Storage vessels can be equipped with either a
vapor return line to the tank truck or railcar, or a carbon
adsorption system. The requirements for new and existing sources
are identical.
10.1.4.2.2 HAP ABA equipment leaks. The proposed standards
contain requirements for pumps, valves, connectors, pressure-
relief devices, and open-ended valves or lines in HAP ABA
service.
Pumps and valves must be monitored quarterly for leaks using
EPA Method 21, where a leak is defined as an instrument reading
of 10,000 parts per million (ppm) or greater. Leaks must be
repaired within 15 calendar days after their detection.
Alternatively, leakless pumps can be used. Valves that are
designated as unsafe-to-monitor must be monitored as frequently
as possible, and difficult-to-monitor valves must be monitored
once per year.
Connectors must be monitored annually, unless the connector
has been opened or the seal broken. In these cases, the
connector must be monitored within 3 months after being returned
to HAP ABA service. As with the other components, a leak is
defined as an instrument reading of 10,000 ppm or greater, and a
leak must be repaired within 15 calendar days. Connectors can
also be designated as unsafe-to-monitor, in which case they must
be monitored as frequently as possible.
10-4
-------
Pressure-relief devices must be monitored using Method 21 if
evidence of a potential leak is found by visual, audible,
olfactory, or any other detection method. If a leak is found
(10,000 ppm), it must be repaired within 15 calendar days.
Each open-ended valve or line in HAP ABA service must be
equipped with a cap, blind flange, plug, or a second valve.
10.1.4.2.3 HAP ABA Emissions from the production line.
Compliance with the proposed provisions for HAP ABA emissions
from the production line is determined by comparing actual HAP
ABA emissions to an allowable emission level for a 12-month
period. Compliance would be determined each month for the
previous consecutive 12-month period.
The proposed regulation recognizes the variability in HAP
ABA emissions for different grades of foam, where a grade of foam
is determined by its density and IFD. Therefore, the allowable
emission level is dependent on the mix of foam grades produced
during the 12-month compliance period. The nucleus of the HAP
ABA emission limitation provisions is the HAP ABA formulation
limitation equation, which determines an allowable amount of HAP
ABA for each grade of foam. This equation is:
ABAUmit = -0.25 (IFD) -19.1(_) - 16 .2 (DEN) -7.56(-) -36.5
where:
ABAlimit = HAP ABA formulation limitation, parts HAP ABA
allowed per hundred parts polyol (pph).
IFD = Internal force deflection, pounds.
DEN = Density, pounds per cubic foot.
Therefore, for each foam grade produced during the 12-month
period, the owner or operator must determine the HAP ABA
formulation limitation.
For new sources, the equation is used to determine the HAP
ABA formulation limitation for a limited number of grades.
However, for foam grades with a density greater than 0.95 pounds
per cubic foot and an IFD greater than 15 pounds, and foam grades
10-5
-------
with a density greater than 1.40 pounds per cubic foot, the
formulation limitation is automatically set to zero.
The allowable HAP ABA emissions for a consecutive 12 -month
period are calculated as the sum of allowable monthly HAP ABA
emissions for each of the individual 12 months in the period.
Allowable HAP ABA emissions for each individual month are
calculated using the following equation.
emiss
emi SS allow, month
where:
emissalloWfinonth = Allowable HAP ABA emissions from the
slabstock foam production process for the
month, pounds.
Number of slabstock foam production lines.
Number of foam grades produced in the month on
foam production line j.
= HAP ABA formulation limit for foam grade i, parts
HAP ABA per 100 parts polyol.
= Amount of polyol used in the month in the
production of foam grade i on foam production line
j, pounds.
The amount of polyol used is a key component of this
determination, and it must be determined by continuously
monitoring the amount of polyol added to the slabstock foam
production line at the mixhead. The monitoring requirements are
discussed in section 10.1.5.2.
Actual HAP ABA emissions are determined by continuously
monitoring the HAP ABA added to the slabstock foam production
line at the mixhead. The allowable monitoring methods for HAP
ABA are exactly the same as for polyol.
The proposed regulation also contains provisions to allow
for the use of HAP ABA recovery devices. If a recovery device is
used, the actual HAP emissions are the difference between the
uncontrolled HAP ABA emissions and the HAP ABA recovered. The
10-6
-------
uncontrolled HAP ABA emissions are determined by monitoring the
HAP ABA added to the slabstock foam production line at the
mixhead, as discussed above. The amount of HAP ABA recovered is
required to be monitored, as discussed in section 10.1.5.3.
As an alternative to the rolling annual compliance approach,
owners or operators can elect to take a monthly compliance
approach. If this approach is selected, actual and allowable
emissions are determined as discussed above. However, compliance
is determined by comparing allowable and actual emissions for
each month, rather than for the 12 previous months. This
alternative is allowed because it is more stringent than the
rolling annual average approach.
10.1.4.2.4 Equipment cleaning HAP emissions. Affected
sources complying with the emission point specific limitations
are prohibited from using a HAP, or a HAP-based product, as an
equipment cleaner. There are no associated monitoring,
reporting, or recordkeeping requirements.
10.1.4.3 Source-wide emission limitation
This alternative allows the owner or operator to choose
which of the HAP ABA emission sources to control. In other
words, an owner or operator could choose not to control HAP ABA
storage vessels and equipment leaks, and achieve a slightly
higher HAP ABA emission reduction from the production line.
Alternatively, an owner or operator could choose to control
emissions from equipment leaks and storage to "save" as much HAP
ABA as possible for use in the production line. In addition,
under the source-wide alternative, a facility could utilize a HAP
equipment cleaner, as long as the HAP used as the equipment
cleaner is the same chemical as the HAP ABA. However, the
equipment cleaning HAP emissions must be offset by emission
reductions from one of the HAP ABA emission sources.
An owner or operator electing to comply with the source-wide
emission limitation for HAP ABA and equipment cleaning determines
compliance by comparing actual emissions from the three HAP ABA
emission sources and from equipment cleaning with an allowable
10-7
-------
emissions level. Compliance is determined each month for the
previous 12-month period.
The allowable emissions level is determined using the same
procedures discussed above in section 10.1.4.2.3. Therefore, the
total HAP ABA and equipment cleaning HAP emissions allowed under
this alternative are equivalent to the allowed HAP ABA emissions
from the production line if the emission point specific
alternatives were selected.
The actual HAP ABA and equipment cleaning emissions are
determined by performing a material balance at the HAP ABA
storage vessel, using the following equation:
PW3aetual = X (STi.begin ~ STifSDd + ADDj
where:
PWEactual = Actual source-wide HAP ABA and equipment
cleaning HAP emissions for a month,
pounds/month.
STi, begin = Amount of HAP ABA in storage tank i at the
beginning of the month, pounds.
STi, end = Amount of HAP ABA in storage tank i at the end of
the month, pounds.
ADDi = Amount of HAP ABA added to storage tank i during
the month, pounds.
n = Number of HAP ABA storage vessels.
Weekly monitoring of the level of HAP ABA in the storage vessels
is required, thus providing the beginning and end of month
amounts to be used in the above equation. In addition, the
amount of each HAP ABA delivery must be determined. The
requirements for the monitoring of HAP ABA storage vessel levels
and the amount of HAP ABA added during each delivery is discussed
later in this section. Emission reductions achieved by recovery
devices can be accounted for by monitoring the amount of HAP ABA
recovered, as described above in section 10.1.3.2.3.
10-8
-------
As with the emission point specific limitation for HAP ABA
from the production line, the source-wide emission limitation
includes a monthly compliance alternative.
10.1.5 Monitoring Requirements
The proposed regulation contains monitoring requirements for
five situations: (1) storage vessels complying using carbon
adsorption systems, (2) polyol and HAP ABA added to the
production line at the mixhead, (3) recovered HAP ABA when a
recovery device is used, (4) the amount of HAP ABA in a storage
vessel, and (5) the amount of HAP ABA added to a storage vessel..
10.1.5.1 Storage vessel monitoring
Storage vessels equipped with carbon adsorption systems must
monitor either the concentration of HAP or the concentration of
total organic compounds (TOC) at the exit of the adsorption
system. Measurements of HAP or TOC concentration must be made
using Method 18 or 25A of Appendix A of 40 CFR 60. Outlet
concentration measurements must be made monthly (or each time the
vessel is filled, if filling occurs less frequently than
monthly), or the owner or operator can install a monitoring
system that continuously monitors HAP or TOC concentrations
during vessel filling.
10.1.5.2 Polvol and HAP ABA monitoring at the mixhead
All slabstock facilities must monitor the amount of polyol
added to the slabstock foam production line at the mixhead to
allow the calculation of allowable emissions. The regulation
contains two options for continuously monitoring the polyol
added: (1) a device installed and operated to continuously
monitor and record pump revolutions per minute, or (2) a flow
rate monitoring device installed and operated to measure the
amount of polyol added at the mixhead. Either of these devices
must be calibrated weekly, and must have an accuracy to within
+/- 2 percent. The owner or operator can develop an alternative
monitoring program to monitor the amount of polyol added at the
mixhead. In addition, if an owner or operator elects to comply
with the emission point specific limitations, the amount of HAP
ABA added to the slabstock foam production line at the mixhead
10-9
-------
must be monitored. The requirements for monitoring the amount of
HAP ABA added are exactly the same as discussed above for polyol,
10.1.5.3 Recovered HAP ABA monitoring
The final monitoring requirement is for slabstock facilities
using a recovery device to reduce HAP ABA emissions. The amount
of HAP ABA recovered is determined by using a device that
continuously monitors the cumulative amount of HAP ABA recovered
by the recovery device. This device must be installed,
calibrated, maintained, and operated according to the
manufacturer's specifications, and must be certified by the
manufacturer to be accurate to within +/- 2.0 percent.
10.1.5.4 Monitoring to determine the amount of HAP ABA in a
storage vessel The amount of HAP ABA in a storage vessel must
be determined by monitoring the HAP ABA level in the storage
vessel using a device that has been certified by its manufacturer
to be at least 99 percent accurate, that has either a digital or
printed output, and that is calibrated at least once a year. The
level of HAP ABA in each storage vessel must be measured and
recorded at least once per week.
10.1.5.5 Monitoring to determine the amount of HAP ABA
added to a storage vessel The amount of HAP ABA added to a
storage vessel during a delivery must be determined using any one
of three options. The first option requires that the volume of
HAP ABA added to the storage vessel be determined by monitoring
the flow rate using a device with an accuracy of 98 percent or
greater, and which is calibrated at least once every six months.
The second option allows the owner or operator to calculate the
weight of HAP ABA added by determining the difference between the
full weight of the transfer vehicle prior to unloading into the
storage vessel and the empty weight of the transfer vehicle after
unloading has been completed. The third option of determining
the amount of HAP ABA added to a storage vessel allows the owner
or operator to develop an alternative monitoring program. The
alternative monitoring program must include, at a minimum, a
description of the parameter to be monitored to determine the
amount of the addition, a description of how the results of the
10-10
-------
monitoring will be recorded and converted into the amount of HAP
ABA added, data demonstrating the accuracy of the monitoring
measurements, and procedures for ensuring that the accuracy of
the monitoring measurements is maintained.
10.1.6 Testing Requirements
There are two instances where the use of test methods is
required. First, for slabstock owners or operators complying
with the source specific HAP ABA equipment leak requirements,
testing must be conducted using Method 21 of 40 CFR part 60,
subpart A.
Second, all slabstock processes must test each grade of foam
to verify the IFD and density, as these are integral inputs into
the equation to determine the HAP ABA formulation limitation.
The proposed rule requires these parameters to be determined
using ASTM D3574, using a sample of foam cut from the center of
the foam bun. The maximum sample size for which the IFD and
density is determined shall not be larger than 24 inches by 24
inches by 4 inches.
10.1.7 Alternative Means of Emission Limitation
The proposed regulation also contains provisions to allow an
owner or operator to request approval to use an alternative means
of emission limitation. The request, which may be submitted in
the precompliance report for existing sources, the application
for construction or reconstruction for new sources, or at any
other time after the initial compliance, must include a complete
description of the alternative means of emission limitation and
documentation demonstrating equivalency with the requirements in
the regulation. The owner or operator can begin using the
alternative means of emission limitation upon approval of the
request by the Administrator.
10.1.8 Applicability of General Provisions
The General Provisions for Part 63; 40 CFR 63, Subpart A;
create the technical and administrative framework for
implementing national emission standards established under
section 112 of the Clean Air Act. The General Provisions
establish baseline applicable requirements for activities such as
10-11
-------
performance testing, monitoring, notifications, and recordkeeping
and reporting, and they also implement statutory provisions such
as compliance dates for new and existing sources and
preconstruction review requirements. The General Provisions
apply to all sources that are affected by Part 63 standards,
including the proposed standard for flexible polyurethane foam
production. However, certain requirements in the General
Provisions may be overridden in individual standards. The
proposed regulation contains a table outlining the sections of
the General Provisions that are applicable to subpart III, and
outlining the General Provisions' sections that are being
overridden or not incorporated.
10.1.9 Reporting Requirements
The proposed regulation requires the submittal of six types
of reports: (1) initial notification, (2) application for
approval of construction or reconstruction, (3) precompliance
report, (4) notification of compliance status, (5) semi-annual
compliance reports, and (6) other reports. These reports are
briefly described below in sections 10.1.9.1 through 10.1.9.6.
10.1.9.1 Initial notification
Each owner or operator of a flexible polyurethane foam
production process must submit an initial notification to the
Administrator within 120 days after promulgation of the rule.
This initial notification must contain an identification of the
facility that is subject to the regulation, the name and address
of the owner or operator of the subject facility, and a brief
description of the process.
10.1.9.2 Application for approval of construction or
reconstruction
Owners or operators constructing a new flexible polyurethane
foam production process, or reconstructing an existing process,
must submit an application for approval of construction or
reconstruction. This application must contain identification
information such as location, owner/operator, and the anticipated
completion and start-up dates. The application must also contain
a description of the planned process and how compliance will be
10-12
-------
achieved. The application must be submitted as soon as
practicable before the construction or reconstruction is planned
to commence. A permit application can take the place of this
report.
10.1.9.3 Precompliance report
One year before the compliance date, each slabstock owner or
operator must submit a precotnpliance report. This report must
contain notification of whether compliance will be achieved using
the emission point specific HAP ABA and equipment cleaning
emission limitation or the source-wide emission limitation.
The report must also indicate if either of the following
compliance options are going to be utilized:
• If compliance will be achieved on a monthly basis for either
the emission point specific limitation for HAP ABA emissions
from the production line or the source-wide emission
limitation.
• If a recovery device will be used to reduce HAP ABA
emissions.
This report must also contain a description of how the
amount of polyol and HAP ABA (if required) added at the mixhead
will be monitored. If the owner or operator is developing an
alternative monitoring plan, the plan must be submitted with the
precompliance report. In addition, slabstock flexible
polyurethane processes using a recovery device to reduce HAP ABA
emissions must include a description of the HAP ABA monitoring
and recordkeeping program to determine the amount of HAP ABA
recovered in the precompliance report.
10.1.9.4 Notification of compliance status
Each slabstock owner or operator must submit a notification
of compliance status report 180 days after the compliance date.
This report must contain notification of the compliance status of
diisocyanate storage vessels and diisocyanate transfer pumps. In
addition, for processes complying with the emission point
specific limitations for HAP ABA, this report must contain
compliance information for HAP ABA storage vessels and equipment
in HAP ABA service. Molded and rebond affected sources must
10-13
-------
submit a statement that compliance is being achieved with the
standards.
A flexible polyurethane foam or rebond foam process that is
committing to an enforceable limit to maintain emissions below
major source levels must submit an affidavit stating that annual
HAP emissions will not exceed the major source levels in the
notification of compliance status. This affidavit must be signed
by the owner, operator, or other responsible individual.
10.1.9.5 Semi-annual compliance reports
Each slabstock owner or operator must submit semi-annual
compliance reports. For processes complying with the rolling
annual compliance provisions (for either the emission point
specific HAP ABA limitations or the source-wide emission
limitation), the report must contain the allowable and actual HAP
ABA emissions (or allowable and actual HAP ABA and equipment
cleaning HAP emissions) for each of the 12-month periods ending
on each of the six months in the reporting period. For processes
complying with the monthly compliance alternative, the report
must contain the allowable and actual HAP ABA emissions (or
allowable and actual HAP ABA and equipment cleaning HAP
emissions) for each of the six months in the reporting period.
10.1.9.6 Other reports
A slabstock owner or operator must provide a report to the
Administrator indicating the intent to change the selected
compliance alternative (emission point specific limitations or
source-wide emission limitation). This report must be submitted
at least 180 days prior to the change.
Similarly, the intent to switch the compliance method
(rolling annual or monthly) must be reported. This report must
be submitted at least 12 months prior to the change.
10.1.10 Recordkeepinq Requirements
Records must be recorded in a form suitable and readily
available for expeditious inspection and review, and must be kept.
for a period of 5 years. At a minimum, the most recent 2 years
of data must be retained on-site.
10-14
-------
Records are required for storage vessels, equipment leaks,
and HAP ABA. If the owner or operator complies with the source-
wide emission limitation, no records are required for HAP ABA
storage vessels or equipment in HAP ABA service.
10.1.10.1 Storage vessel records
All slabstock processes must maintain records listing all
diisocyanate storage vessels and the type of control utilized to
comply with the regulation. For the storage vessels complying
through the use of a carbon adsorption system, the records must
include the design parameters of the system and the monitoring
records.
10.1.10.2 Equipment leak records
All slabstock processes must maintain a list of components
in diisocyanate service, and a description of the control
utilized for each transfer pump. If the process is complying
with the emission point specific limitations, then records
listing each component in HAP ABA service must also be
maintained.
When a leak, as defined in the proposed rule, is detected
for any component, the component must be marked with a readily
visible identification until the leak is repaired. For valves,
the identification must remain until 2 successive months have
passed where no leak is detected. Records must be kept
specifying when the leak was detected and when it was repaired.
10.1.10.3 HAP ABA records
All slabstock processes must keep records integral to the
calculation of allowable emissions. These include a daily log of
foam runs, and daily records of the amount of polyol added at the
mixhead for each grade of foam, and the results of the density
and IFD testing for each grade. Monthly, a cumulative record
must be maintained listing the foam grades produced during the
month, along with the total amount of polyol used for each foam
grade, and the corresponding allowable HAP ABA (or HAP ABA and
equipment cleaning) emissions level. If complying on an annual
rolling basis, the allowable HAP ABA (or HAP ABA and equipment
10-15
-------
cleaning) emissions level for the previous 12 consecutive months
must also be recorded each month.
For processes complying with the emission point specific HAP
ABA emission limitation, records must be kept regarding the
amount of HAP ABA added at the mixhead each day. In addition,
there must also be a cumulative HAP ABA usage record for each
month, and a cumulative record for the previous 12 consecutive
months (if complying on an annual rolling basis).
For processes complying with the source-wide emission
limitation, monthly records must be kept regarding the actual HAP
ABA and equipment cleaning emissions, as measured at the storage
vessel. If complying on an annual rolling basis, monthly records
must be kept of the actual cumulative HAP ABA and equipment
cleaning emissions for the previous 12 months.
If a process uses a recovery device to reduce HAP ABA
emissions, records must be kept regarding the amount of HAP ABA
recovered. In addition, records of all required calibrations
must be maintained.
10.1.10.3 Records for area sources
A flexible polyurethane foam or rebond foam process that is
committing to an enforceable limit to maintain emissions below
major source levels must keep records documenting HAP emissions.
These records can consist of basic inventory records and
engineering calculations.
10.2 RATIONALE FOR THE SELECTION OF THE SOURCE CATEGORY
The source category selected for the development of this
proposed rule was listed in the source category list published on
July 16, 1992 (57 FR 31576). The way in which source categories
or subcategories are defined is important, because it dictates
the basis upon which the MACT floor is to be determined. The
definition of the source category or subcategory describes the
"pool" of facilities that can be used to define the MACT floor.
This means that the MACT floor must be determined on the same
basis upon which the source category is defined. The definition
of the source category or subcategory is also important in that
it limits the scope of emissions averaging: collocated emission
10-16
-------
points cannot be averaged unless they belong to the same affected
source.
As discussed in Chapter 4, the flexible polyurethane foam
production source category was separated into three subcategories
due to differences in the processes, the manner in which HAP
emissions occur, characteristics of the emission points, and the
applicability of various control technologies. These
subcategories are slabstock and molded flexible polyurethane foam
production, and rebond foam production. Information gathered
during the development of this proposed rule indicated that
facilities in these subcategories are major sources, or are
located at major source plant sites. Therefore, the EPA selected
the slabstock, molded, and rebond subcategories to be covered by
the proposed regulation.
In defining the affected source for the regulation, the EPA
had two options. One option was to define an affected source on
a subcategory basis. The second option was to define the
affected source to include both molded and slabstock foam
processes located at the same plant site. Due to the technical
differences between molded and slabstock foam noted above and in
Chapter 4, and due to the fact that few slabstock and molded foam
processes are located at the same plant site, the EPA defined the
affected source on a subcategory basis. This definition is
consistent with the MACT floor analyses.
10.3 RATIONALE FOR THE SELECTION OF POLLUTANTS AND EMISSION
POINTS TO BE COVERED BY THE PROPOSED STANDARDS
At slabstock foam production processes, significant
emissions of four HAP were reported: methylene chloride, methyl
chloroform, TDI, and propylene oxide. There were also emissions
of other HAP reported in small quantities (less than one-tenth of
one percent of the total HAP emissions). The four primary HAP
were reported to be emitted from four sources: chemical
unloading/storage, equipment leaks, the foam production tunnel
and curing area, and from equipment cleaning. The proposed
regulation covers HAP emissions from all of these emission
points, with one exception. There are no proposed requirements
10-17
-------
for emissions of TDI from the foam production line. While TDI
emissions were reported from this source, the MACT floor was
defined as no control. Further, since there were no emission
controls identified for TDI emissions from the foam production
line, the Agency could not evaluate control levels more stringent
than the floor.
At molded foam facilities, emissions of over 10 HAP were
reported from 13 emission sources. However, around 90 percent of
the total emissions were from three sources: mixhead flushing,
mold release agent usage, and the use of adhesives to repair
foam. The proposed regulation prohibits HAP emissions from each
of these three sources. For each of the remaining 10 emission
sources, the MACT floor was defined as no control, and no
additional control techniques were identified. Therefore, the
EPA concluded that no control would be required for the emissions
from these 10 sources in the proposed rule. It was further noted
that these emission sources were sporadically reported,
suggesting that the reporting of HAP emissions from these sources
was often due to unique facility considerations.
Three emission sources were identified at rebond foam
production facilities: (1) TDI emissions from the rebond foam
production, (2) the application of HAP-based mold release agents,
and (3) the use of a HAP as an equipment cleaner. No control
options were identified for TDI emissions. The proposed
regulation covers the remaining two emission sources.
10.4 RATIONALE FOR THE SELECTION OF THE LEVELS OF THE PROPOSED
STANDARDS
The approach for determining the MACT floors and creating
regulatory alternatives is discussed in Chapter 6. This section
presents the rationale for the selection of the level of the
proposed standards for new and existing sources.
10.4.1 Selection of the Levels of the Proposed Standards for
Existing Sources
The discussion of the rationale for the selection of the
levels of the proposed standards for existing sources in this
section is separated by the three subcategories.
10-18
-------
10.4.1.1 Slabstock foam production
As discussed in Chapter 6, there were three regulatory
alternatives considered for slabstock foam production. The level
of control for equipment leaks and HAP ABA from the production
line increases with each alternative. The cost effectiveness
values of the three alternatives shown in Table 9-12 [$l,262/ton
($l,150/Mg) for the MACT floor, $640/ton ($580/Mg) for Regulatory
Alternative 1, and $696/ton ($630/Mg) for Regulatory
Alternative 2] are all within the range considered to be
reasonable by the EPA. The incremental cost effectiveness in
going from the MACT floor to Regulatory Alternative 1 is
negative, making the MACT floor alternative an inferior
alternative. The incremental cost effectiveness in going from
Regulatory Alternative 1 to 2 is $800/ton ($725/Mg). This
incremental cost effectiveness, in conjunction with the minimal
non-air environmental and energy impacts discussed in Chapter 7,
could lead to the selection of Regulatory Alternative 2.
However, there were several non-quantifiable issues associated
with the total elimination of the use HAP ABA at slabstock foam
facilities that were considered by the Agency in the selection of
the regulatory alternative to be proposed.
First, there is substantial concern within the industry that
a complete range of foams of acceptable market quality can be
produced without any HAP ABA. The manufacturers of the HAP ABA
reduction technologies maintain that the foam quality does not
suffer with the use of these technologies. However, the use of
these technologies in the total absence of the use of HAP ABA,
while still producing a complete product line, is very limited.
A second consideration is the issue of plant safety. Use of
a flammable solvent (acetone) as an ABA, and dangers associated
with the extremely exothermic nature of the foam polymerization
reaction, create potential fire hazards. While the designs of
these systems include safeguards against potential hazards,
industry representatives have consistently expressed reservations
about their use, due to these potential hazards.
10-19
-------
Therefore, the EPA selected Regulatory Alternative 1 as the
basis for the proposed regulation for slabstock flexible
polyurethane foam.
10.4.1.2 Molded foam production
As was shown in Chapter 6, three regulatory alternatives
were developed and considered for the molded foam subcategory,
with the stringency of the mixhead flush control level being the
only difference between the alternatives. As was shown in
Table 9-14, the highest overall cost effectiveness for any
alternative was the MACT floor alternative at $590 per ton of HAP
emission reduction ($536/Mg). Therefore, the cost effectiveness
values of all three alternatives per ton were in the range
considered to be reasonable by the Agency. The incremental cost
effectiveness from Regulatory Alternative 1 to 2 ($l,375/ton or
$l,250/Mg) was also in the reasonable range. Considering these
cost impacts, as well as the non-air environmental and energy-
impacts discussed in Chapter 7, the EPA judged that the
Regulatory Alternative 2 level of control was reasonable.
Therefore, the EPA selected this regulatory alternative as the
level of the proposed standards for molded foam production.
10.4.1.3 Rebond foam production
Since the MACT floor alternative for rebond includes the
complete elimination of HAP emissions from the use of mold
release agents and equipment cleaners, and no control options for
TDI were identified, only one regulatory alternative was
developed for rebond foam production. As discussed in Chapters 8
and 9, it is anticipated that there will be no impact on the
rebond industry from this regulatory alternative, which was
selected for proposal by the EPA.
10.4.2 Selection of the Levels of the Proposed Standards for New
Sources
The 1990 Amendments require that standards be set for new
sources that are no less stringent than the level represented by
the best controlled similar source, which is referred to as the
new source MACT floor. The EPA constructed the new source
regulatory alternatives for both subcategories by including in
10-20
-------
them the best level of control identified for each emission
source type within the subcategory.
10.4.2.1 Slabstock foam production
There were two new source regulatory alternatives for the
slabstock foam production subcategory. The first was equivalent
to the new source MACT floor, and the second includes the
complete elimination of HAP ABA emissions. Since it was
estimated that there will be no new sources in this subcategory,
an impacts analysis was not conducted. However, due to the
obstacles discussed above, the EPA selected the new source MACT
floor alterative and not the complete elimination of HAP ABA
alternative (Regulatory Alternative 1).
10.4.2.2 Molded foam production
Since the new source MACT floor is the complete elimination
of HAP emissions from the three covered emission sources, the
possibility for more stringent regulatory alternatives does not
exist. Therefore, the EPA selected the new source MACT floor
alternative for molded foam production.
10.4.2.3 Rebond foam production
Since the new source MACT floor for rebond includes the
complete elimination of HAP emissions from the use of mold
release agents and equipment cleaners, and no control options for
TDI were identified, this was the new source regulatory
alternative selected for rebond foam production.
10.5 RATIONALE FOR THE SELECTION OF THE FORMATS OF THE PROPOSED
STANDARDS
In general, the formats selected for the proposed regulation
are unique to this rulemaking. The EPA did not identify State or
local regulations for this industry from which an established
format could be emulated. The format of the MACT floor
determinations, presented in Chapter 6, determined the format of
the proposed regulation for most emission sources.
10.5.1 Molded Foam Production
The proposed requirements for molded foam production require
the complete elimination of the use (and emission) of HAP mixhead
flushing agents, HAP-based mold release agents, and HAP-based
10-21
-------
adhesives to repair foam. Given these requirements, the only
format considered was a prohibition of the use of these HAP-basecl
products.
10.5.2 Slabstock Foam Production
The format of the proposed requirements for slabstock foam
production varies, depending on compliance alternatives selected
by the owner or operator. The initial choice is whether to
comply with emission point specific limitations and a source-wide
emission limitation. Sections 10.5.2.1 through 10.5.2.3 describe
the proposed requirements.
10.5.2.1 Storage vessels
The format for the proposed slabstock storage vessel
requirements is an equipment standard. Owners or operators can
either utilize a vapor return line that returns displaced vapors
back to the railcar or tank truck, or can install a carbon
canister system through which the displaced vapor is vented. The
primary reason for the selection of this format was that it was
consistent with the data used to determine the MACT floor.
Typically, the level of detail reported was a brief description
of the storage vessel control technique utilized. No facility
reported emission tests or other data to allow the quantification
of the control effectiveness. Therefore, the EPA selected the
equipment standard format.
10.5.2.2 Equipment leaks
The format for the proposed requirements for diisocyanate
transfer pumps is an equipment standard, which is based on the
format of the data provided by the industry. The format for the
proposed requirements for other components in diisocyanate and
HAP ABA service is based on existing federal regulations,.
specifically 40 CFR 63, subpart H (National Emission Standards
for Organic Hazardous Air Pollutants for Equipment Leaks) and 40
CFR 60, Subpart WV (Standards of Performance for Equipment Leaks
of VOC in the Synthetic Organic Chemicals Manufacturing
Industry). However, since slabstock flexible polyurethane
production facilities have significantly fewer components than
10-22
-------
the industries covered by these subparts, the proposed standards
for equipment leaks are much simpler.
10.5.2.3 Equipment cleaning
The proposed emission point specific requirements for
equipment cleaning prohibit the use of HAP-based products to
clean slabstock foam equipment. Therefore, the only format
considered was a prohibition of the use of HAP-based cleaners.
10.5.2.4 HAP ABA emissions from the production line
10.5.2.4.1 Emission limitation format. The EPA considered
several formats for limiting HAP ABA emissions from the
production line. The key considerations in the selection of the
format for HAP ABA emissions were: (1) the format needed to
recognize the varying emission potential for different grades of
slabstock foam, (2) the format needed to incorporate current
industry practices to the extent possible, and (3) the format
needed to allow flexibility to allow for the use of varying types
of control or pollution prevention methods. The format for HAP
ABA can be somewhat simplified by the fact that 100 percent of
the HAP ABA used is emitted, unless an add-on control device is
used (the Agency is aware of only one facility in the United
States using add-on control to reduce HAP ABA emissions).
The first format considered was an emission factor, where
the allowable HAP ABA emissions level was based on the total
amount of foam produced. However, this format failed two of the
three key considerations listed above. First, basing the
allowable emissions on the total foam produced does not recognize
varying product mixes, and the resulting varying need to use HAP
ABA in the formulations. Second, there is not a common unit of
measure used in the foam industry for the amount of foam
produced. Some facilities measure production by board-feet,
while others measure production by foam weight, both of which
would be difficult to monitor. A more logical method would be to
base HAP ABA usage/emissions on the amount of polyol used, since
foamers almost always establish formulations on this basis. This
type of approach would not present a bias toward any specific
technology, thus satisfying consideration criteria number 3.
10-23
-------
The second format considered was a HAP ABA usage/emission
limitation based on the amount of raw materials used (i.e.,
polyol) for each foam grade. This satisfies consideration
criteria number 2 by corresponding with existing practices in the
industry- By considering the different grades in the calculation
of the allowable emissions, it would also account for the varying
HAP ABA potential for different grades. Finally, it would allow
the use of a variety of pollution prevention techniques. A
slight modification would be necessary to account for the
emission reduction achieved by add-on control.
The final format considered was a percent reduction
requirement, where compliance was based on the reduction from an
uncontrolled level. This format could also satisfy all three
considerations. However, this format would necessitate the
creation of a standard uncontrolled emissions/formulation level
for every foam grade produced in the United States, from which
the uncontrolled HAP ABA emissions could be calculated. The
regulation would potentially need to contain uncontrolled
emission/formulation limitations for hundreds of grades of foam.
This format would need the same information as the second format,
but would require an extra calculation step. Therefore, while it
satisfies all three considerations, the EPA concluded that this
format would result in more cumbersome requirements. In
conclusion, the EPA selected the second format discussed above
for the HAP ABA emissions from the production line.
10.5.2.4.2 Averaging time format. Another consideration
for the HAP ABA emission limitation was the averaging time for
compliance. The Agency believes that the determination of
compliance must occur no less frequently than monthly. The EPA
considered two averaging time formats: (1) compliance determined
monthly for the previous 12 months (i.e., a rolling annual
compliance determination), and (2) compliance determined for each
individual month.
Seasonal variations in production are common for the
slabstock industry. The rolling annual compliance format would
allow for these seasonal variations. The Agency determined that
10-24
-------
the rolling annual compliance format was most appropriate for
this industry.
However, while the slabstock industry was interested in the
flexibility of the rolling annual compliance format, they were
particularly concerned about enforcement implications.
Specifically, the concerns were (1) that a facility could be
fined for 365 days of noncompliance for each violation, and
(2) that a particularly abnormal month could cause violations for
several subsequent months.
Due to seasonal variations, the EPA selected the 12-month
rolling averaging period. However, the monthly averaging period
is still being allowed, because the EPA considers it to be more
stringent than the 12-month rolling averaging period. The EPA is
specifically requesting comments from State and local agencies,
as well as the industry, on the burdens caused by the inclusion
of this choice in the proposed regulation.
10.5.2.5 Source-wide HAP ABA and equipment cleaning HAP
emission limitation
The EPA had to determine if this alternative format for HAP
ABA and equipment cleaning emissions was viable. The EPA was
very interested in this option, as it would provide considerable
flexibility to the regulated industry. While this option could
be considered a form of emissions averaging, it is substantially
more simple and straightforward than recent emissions averaging
programs proposed and promulgated by the EPA. This is due to
several facts, including: (1) only one HAP would be allowed to
be included in the source-wide limitation, (2) the vast majority
of the HAP is used for one purpose (HAP ABA), and (3) all HAP
that is used is emitted (in the absence of add-on control).
The EPA had one predominant concern regarding the
feasibility of this alternative format. The point of compliance
for the emission point specific HAP ABA emission limitation is at
the mixhead, where the HAP ABA is added to the formulation. At
this point, the HAP ABA is accurately metered as it is added.
The point of compliance for the source-wide limitation would be
the HAP ABA storage vessel, where a monthly material balance
10-25
-------
would be performed to determine the amount of HAP ABA and HAP
equipment cleaner used/emitted. The Agency is concerned about
the accuracy of the common methods used to determine the amount
of HAP ABA in the storage vessel (such as sight glasses), and the
methods to determine the amount of HAP ABA delivered by tank
truck or railcar (often simply based on delivery receipts). The
proposed regulation contains requirements for the validation of
storage tank inventory measurements. However, this strays from
the common practices in the industry, and the accuracy of these
measurements is still somewhat in question.
The EPA concluded that the desire to provide flexibility to
the industry was greater than the concern over the accuracy of
these inventory measurement techniques. Therefore, the EPA
elected to include the source-wide emission limitation format in
the proposed regulation. However, the EPA is specifically
requesting comments on the storage vessel inventory requirements,,
the availability of technologies to meet these requirements, and
the accuracy of these technologies.
The discussion of the selected emission limitation and
averaging time formats for the emission point specific HAP ABA
limitation also applies to the source-wide emission limitation.
10.5.3 Rebond Foam Production
The proposed requirements for rebond foam production requires
the complete elimination of the use (and emission) of HAP-based
mold release agents and HAP equipment cleaners. Given these
requirements, the only format considered was a prohibition of the
use of these HAP-based products.
10.6 SELECTION OF EMISSION TEST METHODS
There are two instances in the proposed regulation where
testing is required. When complying with the emission point
specific equipment leak provisions, testing for leaks is required
using Method 21. This is the test method required by all EPA
equipment leak programs, and the Agency believes it is
appropriate for this regulation.
The second test method cited in the proposed regulation is
the method to determine the density and IFD of a foam sample.
10-26
-------
The proposed regulation incorporates ASTM Method D3574 for this
testing. This method is the method already used in the industry,
and the EPA concluded that it was appropriate to incorporate it
into this rule.
For the other emission sources, performance testing is not
required. As mentioned above, the amount of HAP ABA used is
equivalent to the amount emitted. Therefore, the need for
expensive performance testing is unnecessary, as only the amount
used needs to be determined. The proposed format for storage
vessels is an equipment standard, meaning that compliance is
achieved by the installation and maintenance of the proper
equipment. Finally, no test is needed to verify that HAP
equipment cleaners are not being used.
10.7 SELECTION OF MONITORING REQUIREMENTS
As discussed in section 10.1.5, the proposed regulation
contains monitoring requirements for five situations:
(1) monitoring of storage vessel carbon adsorption systems for
breakthrough, (2) polyol and HAP ABA monitoring at the mixhead,
(3) recovered HAP ABA monitoring, (4) the amount of HAP ABA in a
storage vessel, and (5) the amount of HAP ABA added to a storage
vessel.
The monitoring of storage vessel carbon adsorption systems
for breakthrough is necessary to ensure that the system has
retained its effectiveness in reducing diisocyanate or HAP ABA
emissions. The selected monitoring requirements are common
requirements for nonregenerable systems.
The monitoring of polyol at the mixhead is an integral part
of the calculation of allowable HAP ABA (or HAP ABA and equipment
cleaning HAP) emissions. For slabstock processes complying with
the emission point specific HAP ABA emission limitation, the
monitoring of the amount of HAP ABA added at the mixhead is used
to determine the actual HAP ABA emissions. For both these
situations, the selected monitoring requirements are based on
common monitoring techniques used in the slabstock industry.
For facilities using a recovery device, the amount of HAP
ABA recovered must be monitored to adjust the actual HAP ABA
10-27
-------
emissions to account for the effects of the recovery device. The
monitoring requirements for recovered HAP ABA were selected to
ensure accuracy in the measurement of the HAP ABA recovered,
while also providing flexibility to the slabstock process in how
the specific monitoring will take place.
Affected slabstock sources complying with the source-wide
emission limitation for HAP ABA storage and equipment leak
emissions, HAP ABA emissions from the production line, and
equipment cleaning HAP emissions, must perform a monthly material
balance around the HAP ABA storage vessel to determine source-
wide HAP ABA emissions. The data needed to perform this balance
are the amount of HAP ABA in the storage vessel at the beginning
of the month, the amount of HAP ABA added to the storage vessel
during the month, and the amount of HAP ABA in the storage vessel
at the end of the month. The proposed regulation requires that
the amount of HAP ABA in the storage vessel be monitored weekly.
While only the beginning and end of the month amounts are used in
the source-wide monthly emission estimation equation, the EPA
believes that weekly monitoring is necessary to verify the
accuracy of the beginning- and end-of-month amounts. The amount
of HAP ABA added to the storage vessel during each delivery is
also required to be monitored.
10.8 SELECTION OF RECORDKEEPING AND REPORTING REQUIREMENTS
10.8.1 Recordkeepina Requirements
The reporting and recordkeeping requirements for the
proposed regulation were summarized in section 10.1.9 and
10.1.10. The HAP ABA recordkeeping requirements were designed to
correlate, to the extent possible, with the existing
recordkeeping requirements in the slabstock industry. The Agency
believes that detailed records of polyol and HAP ABA usage are
already maintained in a manner similar to the recordkeeping
requirements in the proposed regulation.
For the equipment standard provisions of the proposed
regulation, records of the installation of the equipment are
required to document compliance. If a carbon adsorption system
is used for a storage vessel, records must be kept of the
10-28
-------
required monitoring results, as well as records of when the
carbon or carbon system is replaced. These recordkeeping
requirements were selected to demonstrate compliance.
Similarly, records to demonstrate compliance with the
equipment leak provisions are required. Records of the detection
and repair of leaking components must be maintained.
10.8.2 Reporting Requirements
As discussed in section 10.1.9, the proposed regulation
requires the submittal of six types of reports. The purpose of
the initial notification, the application for approval of
construction or reconstruction, and the precompliance report are
to inform the Administrator of information prior to the
compliance date. The notification of compliance status and semi-
annual compliance reports are to report the source's success in
complying with the regulation.
10.9 MODIFICATION AND RECONSTRUCTION CONSIDERATIONS
The anticipated extensive use of pollution prevention
process modifications to comply with this regulation is cause for
special consideration. The EPA does not believe that it would be
appropriate for a process modification undertaken to meet the
existing source standards to trigger the new source standards
because the modification meets the definition of reconstruction
in subpart A. Therefore, the proposed regulation defines a
reconstructed source as a process that meets the definition of
reconstruction, except that process modifications designed to
reduce the use and emission of HAP ABA are not considered part of
reconstruction.
10.10 OPERATING PERMIT PROGRAM
Under Title V of the 1990 Amendments, all HAP-emitting
facilities subject to this rule will be required to obtain an
operating permit. Oftentimes, emission limits, monitoring, and
reporting and recordkeeping requirements are scattered among
numerous provisions of State implementation plans (SIP's) or
Federal regulations. As discussed in the proposed rule for the
operating permit program published on May 10, 1991 (58 FR 21712),
this new permit program would include, in a single document, all
10-29
-------
of the requirements that pertain to a single source. Once a
State's permit program has been approved, each facility
containing that source within that State must apply for and
obtain an operating permit. If the State wherein the source is
located does not have an approved permitting program, the owner
or operator of a source must submit the application under the
General Provisions of 40 CFR part 63.
10-30
-------
11.0 REFERENCES
1. Hazardous Air Pollutant Emissions from the Production of
Flexible Polyurethane Foam. Supplementary Information
Document for Proposed Standards. U.S. Environmental
Protection Agency. Research Triangle Park, North Carolina.
Publication Number EPA-453/R-96-009a. July 1996.
2. Memorandum. Williams, A., EC/R Incorporated, to
Svendsgaard, D., U.S. Environmental Protection Agency.
Flexible Polyurethane Foam Industry Description.
June 17, 1996.
3. Documentation for Developing the Initial Source Category
List. U.S. Environmental Protection Agency. Research
Triangle Park, NC. Publication No. EPA-450-3-91-030.
4. Memorandum. Norwood, P., Williams, A., and Battye, W., EC/R
Incorporated to Svendsgaard, D., U.S. Environmental
Protection Agency. Summary of Flexible Polyurethane Foam
Information Collection Requests. January 24, 1994.
5. The Polyurethane Industry Directory and Buyer's Guide -
1994. Larson Publishing. Saco, Maine.
6. Telecon. Norwood, P., EC/R Incorporated, with Oler, B.,
Carpet Cushion Council. May 30, 1996. U.S. Population of
Rebond Foam Production Facilities.
7. Woods, George. The ICI Polyurethanes Book. John Wiley and
Sons. 1987.
8. Memorandum. Williams, A., and Norwood, P., EC/R
Incorporated, to Svendsgaard, D., U.S. Environmental
Protection Agency. Site Visit Report - Hickory Springs
Manufacturing Company, Hickory, North Carolina.
January 14, 1996.
9. Polyurethane Foam Association. Flexible Polyurethane Foam
(Slabstock) Assessment of Manufacturing Emission Issues and
Control Technology. May 17, 1993. p. 10.
10. Flexible Polyurethane Foam Emission Reduction Technologies
Cost Analysis. U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina. Publication Number
EPA-453/D-95-011. June 1996.
11. Memorandum. Williams, A., and Norwood, P., EC/R
Incorporated, to Svendsgaard, D., U.S. Environmental
11-1
-------
Protection Agency. Site Visit Report - Foamex
International, Morristown, Tennessee. January 18, 1996.
12. Memorandum. Williams, A., EC/R Incorporated, to
Svendsgaard, D., U.S. Environmental Protection Agency.
Summary of November 28 1995, meeting between EPA and Ohio
Decorative Products. May 31, 1996.
13. Federal Register. Initial List of Categories of Sources
Under Section 112(c)(1) of the Clean Air Act Amendments of
1990. Vol. 57, No. 137.
14. Memorandum. Seaman, J., EC/R Incorporated, to Svendsgaard,
D., U.S. Environmental Protection Agency. Summary of the
August 23, 1995 Meeting between the EPA and Industry
Representatives to Introduce the Development of the Flexible
Polyurethane Foam Fabrication NESHAP. September 15, 1995.
15. Memorandum. Williams, A., EC/R Incorporated, to
Svendsgaard, D., U.S. Environmental Protection Agency.
Baseline Emissions for the Flexible Polyurethane Foam
Production Industry. June 17, 1996.
16. Compilation of Air Pollutant Emission Factors. Volume I:
Stationary Point and Area Sources. 5th Edition. U.S.
Environmental Protection Agency. January 1995. Section 7.
17. Protocol for Equipment Leak Emission Estimates. U.S.
Environmental Protection Agency. Publication No.
EPA-453/R-93-026. Table 2-10.
18. Telecon. Williams, A., EC/R Incorporated, with Kern, A.,
Kern Foam Products. February 21, 1995. Emissions from
mixhead flush.
19. Telecon. Williams, A., EC/R Incorporated, with Bogden, W.,
EAR Specialty Composites, February 22, 1995. Emissions from
mixhead flush.
20. Letter, from Sullivan, D., Hickory Springs Manufacturing
Company, to Svendsgaard, D., U.S. Environmental Protection
Agency. February 24, 1994. Suggested PFA sub-
categorization strategy for slabstock foam.
21. Memorandum, from Norwood, P., EC/R Incorporated, to
Svendsgaard, D., U.S. Environmental Protection Agency. June
14, 1995. Summary of May 23, 1995, Flexible Polyurethane
Foam Presumptive MACT Roundtable Meeting.
22. Memorandum from Norwood, P. EC/R Incorporated, to
Svendsgaard, D. U.S. Environmental Protection Agency.
Status of Flexible Polyurethane Foam Auxiliary Blowing Agent
MACT Floor Determinations. June 22, 1995.
11-2
-------
23. Flexible Polyurethane Foam Emission Reduction Technologies
Cost Analysis. U.S. Environmental Protection Agency.
Research Triangle Park, NC. Publication No. EPA-453/R-95-
011. September 1996.
24. Memorandum, from Williams, A., EC/R Inc., to Svendsgaard,
D., EPA:ESD:OCG. May 31, 1995. Flexible Polyurethane Foam
Model Plants.
25. Protocol for Equipment Leak Emission Estimates. U.S.
Environmental Protection Agency, Research Triangle Park,
North Carolina. Publication No. EPA-453/R-93-026. Table
2-10. June 1993.
26. Memorandum. Seaman, J., EC/R Incorporated, to Svendsgaard,
D., U.S. Environmental Protection Agency. Summary of the
March 7, 1996 Meeting Between the EPA and Polyurethane Foam
Association Representatives. April 22, 1996.
27. Memorandum. Norwood, P., and Williams, A., EC/R
Incorporated, to Svendsgaard, D., U.S. Environmental
Protection Agency. Estimated Impacts of Regulatory
Alternative for the Flexible Polyurethane Foam Production
Industry. June 17, 1996.
28. Gasoline Distribution Industry (Stage I) - Background
Information for Proposed Standards. U.S. Environmental
Protection Agency. Research Triangle Park, North Carolina.
Publication No. EPA-453/R-94-002a. January 1994. pps. 4-34
and 7-17.
11-3
-------
TECHNICAL REPORT DATA
(Please read Instructions on reverse before completing)
1. REPORT NO.
EPA-453/R-96-008a
3. RECIPIENT'S ACCESSIONNO.
4 TITLE AND SUBTITLE
Hazardous Air Pollutant Emissions from the Production of
Flexible Polyurethane Foam - Basis and Purpose Document for
Proposed Standards
5. REPORT DATE
September 1996
6. PERFORMING ORGANIZATION CODE
7 AUTHOR(S)
8. PERFORMING ORGANTZATIONREPORT NO.
9 PERFORMING ORGANIZATIONNAME AND ADDRESS
Emission Standards Division (Mail Drop 13)
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANTNO.
68-D6-0008
12 SPONSORING AGENCYNAME AND ADDRESS
Director
Office of Air Quality Planning and Standards
Office of Air and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/200/04
15 SUPPLEMENTARY NOTES
16. ABSTRACT
This document provides background information on, and rationale for, decisions made by the EPA
related to the proposed standards for the reduction of HAP emitted during the production of flexible
polyurethane foam.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATT Field/Group
Air Pollution
Hazardous air pollutants
Emission reduction
Flexible Polyurethane Foam
Hazardous air pollutants
18 DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Report)
Unclassified
21. NO. OF PAGES
156
20. SECURITY CLASS (Page)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
-------
U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Floor
Chicago, IL 60604-3590
------- |