United States
Environmental Protection
Agency
Air
Office Of Air Quality
Planning And Standards
Research Triangle Park, NC 27711
November 2002
FINAL REPORT
Economic Impact Analysis of the Plywood
and Composite Wood Products NESHAP
Final Report
-------
-------
Select List of Acronyms and Abbreviations
BID - Background Information Document
CAA - Clean Air Act
CO - Carbon Monoxide
C/S- Cost to Sales Ratio
EFB - Electrified Filter Beds
EO - Executive Order
EPA - Environmental Protection Agency
EWP - Engineered Wood Products
HAP - Hazardous Air Pollutant
HB - Hardboard
ICR - Information Collection Request
lb - Pound
LSL - Laminated Strand Lumber
LVL - Laminated Veneer Lumber
MACT - Maximum Achievable Control Technology
MDF - Medium Density Fiber
NAAQS - National Ambient Air Quality Standards
NAICS - North American Industrial Classification System
NESHAP - National Emission Standards for Hazardous Air Pollutants
NOx- Nitrogen Oxides
NPR - Notice of Proposed Rulemaking
NSPS - New Source Performance Standards
NSR - New Source Review
OEM- Original Equipment Manufacturers
OMB - Office of Management and Budget
O&M - Operation and Maintenance
OSB- Oriented Strandboard
ODT - Oven Dry Tons
PB - Particleboard
P/E - Partial Equilibrium
PM - Particulate Matter
PSL - Parallel Strand Lumber
ppbdv - Parts Per Billion, dry volume
ppm - Parts Per Million
PRA - Paperwork Reduction Act of 1995
PTE - Permanent Total Enclosure
RCO- Regenerative Catalytic Oxidizer
RTO - Regenerative Thermal Oxidizer
RIA - Regulatory Impact Analysis
RFA - Regulatory Flexibility Act
R/S - Return to Sales Ratio
SAB - Science Advisory Board
SB A - Small Business Administration
vi
-------
SBREFA - Small Business Regulatory Enforcement Fairness Act of 1996
SIC - Standard Industrial Classification
SOA - Secondary Organic Aerosols
S02 - Sulfur Dioxide
SPV - Softwood Plywood Veneer
TAC - Total Annualized Cost
THC - Total Hydrocarbon
tpd - Tons Per Day
tpy - Tons Per Year
UMRA - Unfunded Mandates Reform Act
VOS - Value of Shipments
VOCs - Volatile Organic Compounds
WESP - Wet Electrostatic Precipitator
vii
-------
This page intentionally left blank.
viii
-------
Table of Contents
Executive Summary ES-1
1 Introduction 1-1
1.1 Scope and Purpose of the Report 1-1
1.2 Need for Regulatory Action 1-1
1.3 Requirements for the Economic Impact Analysis 1-2
1.4 Other Federal Programs 1-5
1.5 Organization of the Economic Impact Analysis 1-5
1.6 References 1-6
2 Profile of the Plywood and Composite Wood Industries 2-1
2.1 Introduction 2-1
2.2 The Supply Side 2-3
2.3 The Demand Side 2-20
2.4 Industry Organization 2-29
2.5 Market 2-42
2.6 References 2-60
3 Regulatory Alternatives, Emissions, Emission Reductions, and Control and Administrative
Costs
3.1 Regulatory Alternatives 3-1
3.2 Emissions and Emission Reductions 3-15
3.3 Control Equipment and Costs 3-26
3.4 Testing, Monitoring, Reporting, and Recordkeeping Costs 3-37
3.5 References 3-40
4 Economic Impact Analysis 4-1
4.1 Results in Brief 4-1
4.2 Introduction 4-1
4.3 Economic Impact Analysis Inputs 4-2
4.4 Economic Impact Analysis Methodology 4-4
4.5 Economic Impact Analysis Results 4-6
4.6 Analysis of Economic Impacts on Engineered Wood Products Sector 4-15
4.7 References 4-22
5 Small Business Impacts 5-1
5.1 Results in Brief 5-1
5.2 Introduction 5-1
5.3 Screening Analysis Data Sources 5-2
5.4 Screening Analysis Methodology 5-2
5.5 Screening Analysis Assumptions 5-3
5.6 Screening Analysis Results 5-3
5.7 Screening Analysis Conclusions 5-10
5.8 EIA Results for Small Businesses 5-1
5.9 References 5-12
-------
Appendix A Size Determination for Individual Firms A-l
Appendix B EIA Methodology B-l
B.l Introduction B-l
B.2 Background B-l
B.3 Market Segments and Demand Compounds B-3
B.4 Data Inputs B-4
B.5 Assumptions B-4
B.6 Analytical Approach B-5
B.7 Notation B-5
B.8 The Supply Side of the Market B-6
B.9 The Demand Side of the Market B-ll
B.10 Market Equilibrium B-l8
B.ll Facility Closures B-19
B.12 Social Cost of the Regulation B-20
B.13 Computational Tools B-26
B.14 Sensitivity Analyses B-26
B.15 EIA Results B-27
B.16 References B-28
List of Exhibits
Exhibit 2-1: SIC & NAICS Codes for the Plywood and Wood Composites Industries 2-2
Exhibit 2-2: Other Primary SIC Codes for the Plywood and Wood Composite Industries 2-3
Exhibit 2-3: SIC and NAICS Codes and Products 2-12
Exhibit 2-4: Specialization and Coverage Ratios, 1982 - 1997 2-13
Exhibit 2-5: Summary of Annual Costs and Shipments, 1992 -1997 2-14
Exhibit 2-6: Materials Consumed By Kind for Softwood Plywood and Veneer, 1997 2-17
Exhibit 2-7: Materials Consumed by Kind for Reconstituted Wood Products, 1997 2-18
Exhibit 2-8: Industry Outputs, by SIC Code 2-21
Exhibit 2-9: MDF Shipments by Downstream Market, 1997 2-23
Exhibit 2-10: Particleboard Shipments by Downstream Market, 1997 2-24
Exhibit 2-11: Housing Market Indicators, 1988 - 1997 2-25
Exhibit 2-12: Trade for Household Furniture (SIC 251), 1989 - 1996 2-25
Exhibit 2-13: Use of Wood and Non-wood Products in Residential Construction
1976 - 1995 2-27
Exhibit 2-14: Demand Elasticities 2-29
Exhibit 2-15: Concentration Ratios by SIC Code, 1982-1992 2-30
Exhibit 2-16: Facilities with Compliance Costs 2-31
Exhibit 2-17: Full Production Capacity Utilization Rates, Fourth Quarters, 1992 - 1997 2-33
Exhibit 2-18a: 1998 Employment at Facilities with Compliance Costs 2-34
Exhibit 2-18b: 1998 Employment at Facilities with Compliance Costs 2-34
Exhibit 2-19: Number of Mills, Average Capacity and Utilization, 1977 - 1997 2-36
Exhibit 2-20: Summary of Capital Expenditures, 1992 - 1997 2-37
Exhibit 2-21: Size Distribution of Firms Owning Affected Facilities 2-38
Exhibit 2-22: Types of Firm Ownership for Lumber and Wood Products (SIC 24), 1992 2-39
Exhibit 2-23: Indicators of Financial Condition, 1995-1997 2-41
Exhibit 2-24: Trade Balance and Selected Statistics 2-44
Exhibit 2-25: Production, Trade and Consumption Volumes for Selected Products (1988-1997) .... 2-45
Exhibit 2-26: 1997 U.S. Wood Products Imports by Region and Major Trading Partner 2-48
-------
Exhibit 2-27: 1997 U.S. Wood Product Exports by Region and Major Trading Partner 2-50
Exhibit 2-28: Lumber and Wood Products Producer Price Index, 1988-1997 2-51
Exhibit 2-29: Producer Price Indices of Plywood and Wood Composite Products 2-52
Exhibit 2-30: F.O.B. Prices of Southern plywood, OSB, and Particleboard 2-53
Exhibit 2-31: APA Forecasted Structural Panel Production and Demand 2-56
Exhibit 2-32: APA Actual and Forecasted Structural Panel Capacity and Production 2-57
Exhibit 3-1: Illustration of Total HAP Calculation for an Emission Source 3-5
Exhibit 3-2: Summary of MACT for PCWP Process Units at New and Existing Sources 3-7
Exhibit 3-3: Cost-Effectiveness Analysis Of Beyond-The-Floor Control Options 3-9
Exhibit 3-4: Illustration Of Total HAP Calculation For An Emission Source 3-17
Exhibit 3-5: Uncontrolled and Baseline HAP Emissions Estimates 3-21
Exhibit 3-6: Speciated Nationwide Uncontrolled HAP Emissions by Product 3-22
Exhibit 3-7: Speciated Nationwide Baseline HAP Emissions by Product 3-23
Exhibit 3-8: Estimated Number of Major Sources By Product 3-24
Exhibit 3-9: Estimated Nationwide Reduction in Total HAP and THC 3-25
Exhibit 3-10: Press Enclosure Exhaust Flow Rates and Capital Costs 3-33
Exhibit 3-11: Control Equipment Costed for Process Units with Controlled MACT Floor 3-34
Exhibit 3-12: Default Flow Rates 3-36
Exhibit 3-13: Estimated Nationwide Control Costs for the PCWP Industry 3-38
Exhibit 3-14: Dollars (In Total Annualized Costs) Per Ton Of HAP And THC Reduced 3-39
Exhibit 4-1: Baseline Characterization of Plywood and Composite Wood Markets: 1997 4-3
Exhibit 4-2. Market-Level Impacts of the Proposed NESHAP 4-7
Exhibit 4-3. Industry-Level Impacts of the Proposed NESHAP 4-10
Exhibit 4-4: Distribution of Industry-Level Impacts of Proposed NESHAP:
Affected and Unaffected Producers 4-12
Exhibit 4-5: Distribution of Social Costs Associated with the Proposed NESHAP 4-13
Exhibit 4-6: Primary Uses and Substitutes for LSL 4-16
Exhibit 4-7: Characteristics of LSL Plants 4-17
Exhibit 4-8: Primary Uses and Substitutes for PSL 4-17
Exhibit 4-9: Characteristics of PSL Plants 4-18
Exhibit 4-10: Retail Prices of GL and PSL Beams Delivered to Los Angeles 4-18
Exhibit 5-1: Net Profit Margins by Product Type 5-3
Exhibit 5-2: Affected Firms by Size 5-4
Exhibit 5-3: Affected Firms by Process Type 5-5
Exhibit 5-4: Affected Firms with C/S Ratios of 3 Percent or Greater 5-6
Exhibit 5-5: Affected Firms with C/S Ratios of 1 Percent or Greater 5-7
Exhibit 5-6: C/S to R/S Comparison for Firms with C/S of One Percent or Greater 5-8
Exhibit 5-7: Economic Impacts on Small Businesses Associated with
Projected Market Adjustments 5-11
List of Figures
Figure 2-1: Plywood and Veneer Production 2-6
Figure 2-2: Softwood Plywood and Veneer Value of Shipments and Production Costs,
1992- 1997 2-15
Figure 2-3: Reconstituted Wood Products Value of Shipments and Production Costs,
1992- 1997 2-16
Figure 2-4: Materials Consumed by Softwood Plywood and Veneer Products, 1997 2-17
Figure 2-5: Materials Consumed by Reconstituted Wood Product Producers, 1997 2-19
Figure 2-6 Industry Outputs of Softwood Plywood and Veneer Industry 2-22
-------
Figure 2-7: Industry Outputs of Reconstituted Wood Products Industry 2-23
Figure 2-8: Plywood and Wood Composite Facility Locations 2-32
Figure 2-9: Full Production Capacity Utilization, Fourth Quarters, 1992-1997 2-33
Figure 2-10: Value of Product Shipments, 1989-1995 2-46
Figure 2-11: Apparent Consumption, 1989-1995 2-47
Figure 2-12: APA Projected Housing Starts (000) 2-55
Figure 3-1: Variation in RTO Purchased Equipment Cost with Flow Rate 3-27
Figure 3-2: Relationship between RTO Electricity Consumption and Flow Rate 3-28
Figure 3-3: Relationship between RTO Natural Gas Consumption and Flow Rate 3-28
Figure 3-4: Variation in RTO Total Capital Investment with Flow 3-30
Figure 3-5: Variation in RTO Total Annualized Cost with Flow 3-30
Figure 4-1: Supply Curves for Affected Facilities 4-5
Figure 4-2: Market Equilibrium Without and With Regulation 4-6
-------
EXECUTIVE SUMMARY
EPA is proposing a rule to reduce hazardous air pollutant (HAPs) emissions from existing and
new plywood and composite wood products facilities that are major sources. This rule, scheduled for
proposal during 2002, is a National Emission Standards for Hazardous Air Pollutants (NESHAP), and
will reduce HAP emissions by requiring affected plywood and composite wood products facilities to meet
a level of emissions reductions needed to meet the Maximum Achievable Control Technology (MACT)
floor for these sources. This MACT floor level of control is the minimum level these sources must meet
to comply with the proposed rule. The major HAPs whose emissions will be reduced are formaldehyde,
acetaldehyde, acrolein, methanol, phenol, and propionaldehyde. The proposed rule will also lead to
emission reductions of other pollutants such as volatile organic compounds (VOC), particulate matter
(PM10), carbon monoxide (CO), and emission increases in nitrogen oxides (NOx) due to the application of
incineration-based controls. Increased electricity use due to application of controls will also lead to
general increases in the levels of sulfur dioxide (S02) and NOx emitted from electric utilities.
This proposed rule allows an affected source to use a production-based compliance option,
defined in units of mass of pollutant per unit of production, or any of six control system compliance
options if an affected source is equipped with an add-on control system. As explained in the Federal
Register proposal notice, the options entail HAP reductions of 90 percent or limiting the concentration of
HAPs in the exhaust from the control system. In addition, an affected source may choose to comply with
an emissions averaging option that allows the sources to not control or under-control some process units
while controlling other affected process units.
The proposed rule is expected to reduce HAP emissions by 11,000 tons per year in the third year
after its issuance. The rule is also expected to reduce VOC emissions, measured as total hydrocarbon,
by 27,000 tons per year, PM10 emissions by 13,000 tons per year, and CO emissions by 11,000 tons per
year in the third year. The rule is expected to increase NOx emissions at affected sources by 2,000 tons
per year in the third year. The increased electricity required to operate the control systems is also
expected to increase NOx and S02 emissions at electricity generating utilities by 2,000 and 4,000 tons,
respectively. The compliance costs, which include the costs of control and monitoring, recordkeeping
and reporting requirements, are estimated at $142 million (1999 dollars).
As shown in this economic impact analysis, the total social costs, which account for the
behavioral response of consumers and producers to higher pollution control costs, are estimated at $134.2
million (1999 dollars). The other impacts associated with these costs include price increases nationally of
0.9 to 2.5 percent for products affected by this rule, and a reduction in output of only 0.1 to 0.7 percent
nationally for the affected industries. An analysis of small business impacts shows that there are 17 small
firms affected, with 10 of them having annual compliance costs of 1 percent or greater than their sales,
and 3 of these having annual compliance costs of 3 percent or greater than their sales. The Agency has
certified that there is no significant impact on a substantial number of small entities (SISNOSE)
associated with this proposed rule. Also, an analysis of the energy impacts associated with this proposed
rule indicates that there is no significant adverse effect on supply, distribution, or use of energy from
implementation of the proposal.
ES-1
-------
1 INTRODUCTION
Under the authority of Section 112(d) of the Clean Air Act as amended in 1990, the U.S.
Environmental Protection Agency (EPA or the Agency) is proposing a regulation requiring facilities that
manufacture plywood and composite wood products to reduce their emissions of hazardous air pollutants
(HAPs). This regulation, a National Emission Standard for Hazardous Air Pollutants (NESHAP), will
apply to major sources of HAPs in this industry. This economic impact analysis (EIA) presents the
supporting documentation and analyses developed by the Agency that describe and quantify the expected
impacts of the proposed Plywood and Composite Wood Products NESHAP.
1.1 Scope and Purpose of the Report
The proposed NESHAP will require the manufacturers of plywood and composite wood products
to install additional pollution controls to reduce their emissions of HAPs to the air. The purpose of this
EIA is to present the results of the Agency's evaluation of the cost, economic impacts, and benefits from
compliance with the requirements of the proposed NESHAP.
The proposed NESHAP will apply to all new and existing major sources of HAPs that
manufacture plywood and composite wood products. These sources emit HAPs associated with heating
of wood and related to their use of resins, adhesives, and additives in the pressing and drying stages of the
production process. The EPA estimates that there are 447 facilities that produce plywood and composite
wood products. Of these, the EPA determined that 223 facilities are major sources of HAPs.
1.2 Need for Regulatory Action
The purpose of this NESHAP is to protect public health by reducing emissions of HAP from
plywood and composite wood products facilities. The authority for doing this lies in Section 112 of the
Clean Air Act (CAA), which requires EPA to list categories and subcategories of major and area sources
of HAP and to establish NESHAP for the listed source categories and subcategories. The plywood and
composite wood products source category was originally listed as the plywood and particleboard source
category on July 16, 1992 (57 FR 31576). The name of the source category was changed to plywood and
composite wood products on November 18, 1999 (64 FR 63025) to more accurately reflect the types of
manufacturing facilities covered by the source category. A major source of HAP is defined as any
stationary source source or group of stationary sources within a continuous area and under common
control that emits or has the potential to emit, considering controls, in the aggregate, 9.1 Megagrams
(Mg)/year (10 tons/yr) or more of any single HAP or 22.7 Mg/year or more (25 tons/yr) of multiple HAP.
Section 112 of the CAA requires EPA to establish NESHAP for the control of HAP from both
existing and new sources. The CAA requires the NESHAP to reflect the maximum degree of reduction in
emissions of HAP that is achievable. This level of control is commonly referred to as the maximum
achievable control technology (MACT).
The MACT floor is the minimum level of control allowed for NESHAP and is defined under
section 112 (d) (3) of the CAA. In essence, the MACT floor ensures that the standard is set at a level that
assures all major sources achieve the control level that is at least as stringent as that already achieved by
the better-controlled and lower-emitting sources in each source category or subcategory. For new
sources, the MACT floor cannot be less stringent than the emission control that is achieved in practice by
the best-controlled similar source. The MACT standards for existing sources can be less stringent than
standards for new sources, but they cannot be less stringent than the average emission limitation achieved
by the best-performing 12 percent of existing sources in the category or subcategory (or, the best-
performing 5 sources for categories or subcategories with fewer than 30 sources.)
1-1
-------
In the course of rule development, we may also consider control options that are more stringent
than the floor. EPA may establish standards more stringent than the floor based on the consideration of
cost of achieving the emissions reductions, any non-air quality health and environmental impacts, and
energy requirements.
1.3 Requirements for this Economic Impact Analysis
This section describes various legislative and executive requirements that govern the analytical
requirements for Federal rulemakings, and describes how each analytical requirement is addressed in this
RIA.
1.3.1 Executive Order 12866
Under Executive Order 12866 (58 FR 51735, October 4, 1993) as amended by Executive Order
13258 (67 FR 9385, February 28, 2002), the EPA must determine whether the regulatory action is
"significant" and therefore subject to review by the Office of Management and Budget (OMB) and the
requirements of the Executive Order. The Executive Order defines "significant regulatory action" as one
that is likely to result in a rule that may:
1) Have an annual effect on the economy of $100 million or more or adversely affect in a material way
the economy, a sector of the economy, productivity, competition, jobs, the environment, public health or
safety, or state, local, or tribal governments or communities;
2) Create a serious inconsistency or otherwise interfere with an action taken or planned by another
agency;
3) Materially alter the budgetary impact of entitlements, grants, user fees, or loan programs, or the rights
and obligation of recipients thereof;
4) Raise novel legal or policy issues arising out of legal mandates, the President's priorities, or the
principles set forth in the Executive Order.
Pursuant to the terms of Executive Order 12866 as amended by Executive Order 13258, it has been
determined that this rule is a "significant regulatory action" because the annual costs of complying with
the rule are expected to exceed $100 million. Consequently, this action was submitted to OMB for
review under Executive Order 12866 as amended by Executive Order 13258.
1.3.2 Regulatory Flexibility Act and Small Business Regulatory Enforcement Fairness Act of 1996
The Regulatory Flexibility Act (RFA) of 1980 (PL 96-354) generally requires that agencies
conduct a screening analysis to determine whether a regulation adopted through notice-and-comment
rulemaking will have a significant impact on a substantial number of small entities (SISNOSE), including
small businesses, governments, and organizations. If a regulation will have such an impact, agencies
must prepare an Initial Regulatory Flexibility Analysis, and comply with a number of procedural
requirements to solicit and consider flexible regulatory options that minimize adverse economic impacts
on small entities. Agencies must then prepare a Final Regulatory Flexibility Analysis that provides an
analysis of the effect on small entities from consideration of flexible regulatory options. The RFA's
analytical and procedural requirements were strengthened by the Small Business Regulatory Enforcement
Fairness Act (SBREFA) of 1996 to include the formation of a panel if a proposed rule was determined to
have a SISNOSE. This panel would be made up of representatives of the EPA, the Small Business
Administration (SBA), and OMB.
1-2
-------
For reasons explained more fully in Chapter 5 of this economic impact analysis for the proposed
rule, EPA has determined that there is no SISNOSE for this rule. While there are some impacts to some
small firms, these impacts are not sufficient for a SISNOSE. Therefore, the EPA has not prepared an
Initial Regulatory Flexibility Analysis for this proposed rule.
The RFA and SBREFA require the use of definitions of "small entities," including small
businesses, governments, and organizations such as non-profits, published by the SBA.1 Screening
analyses of economic impacts presented in Chapter 5 of this report examine potential impacts on small
entities.
1.3.3 Unfunded Mandates Reform Act of 1995
The Unfunded Mandates Reform Act (UMRA) of 1995 (PL-4) was enacted to focus attention on
federal mandates that require other governments and private parties to expend resources without federal
funding, to ensure that Congress considers those costs before imposing mandates, and to encourage
federal financial assistance for intergovernmental mandates. The Act establishes a number of procedural
requirements. The Congressional Budget Office is required to inform Congressional committees about
the presence of federal mandates in legislation, and must estimate the total direct costs of mandates in a
bill in any of the first five years of a mandate, if the total exceeds $50 million for intergovernmental
mandates and $100 million for private-sector mandates.
Section 202 of UMRA directs agencies to provide a qualitative and quantitative assessment (or a
"written statement") of the anticipated costs and benefits of a Federal mandate that results in annual
expenditures of $100 million or more. The assessment should include costs and benefits to State, local,
and tribal governments and the private sector, and identify any disproportionate budgetary impacts.
Section 205 of the Act requires agencies to identify and consider alternatives, including the least costly,
most cost-effective, or least burdensome alternative that achieves the objectives of the rule.
Since this proposed rule may cause a mandate to the private sector of more than $100 million,
EPA did provide an analysis of the impacts of this rule on State and local governments to support
compliance with Section 202 of UMRA. A summary of this analysis is in Chapter 4 of this EIA. In
short, no government entity is affected by this proposed rule - only businesses.
1.3.4 Paperwork Reduction Act of 1995
The Paperwork Reduction Act of 1995 (PRA) requires Federal agencies to be responsible and
publicly accountable for reducing the burden of Federal paperwork on the public. EPA has submitted an
OMB-83I form, along with a supporting statement, to the OMB in compliance with the PRA. The OMB-
831 and the supporting statement explains the need for additional information collection requirements and
provides respondent burden estimates for additional paperwork requirements to State and local
governments associated with this proposed rule.
1.3.5 Executive Order 12898
Executive Order 12898, "Federal Actions to Address Environmental Justice in Minority
Populations and Low-Income Populations," requires Federal agencies to consider the impact of programs,
policies, and activities on minority populations and low-income populations. Disproportionate adverse
impacts on these populations should be avoided to the extent possible. According to EPA guidance,
1 Where appropriate, agencies can propose and justify alternative definitions of "small entity." This RIA
and the screening analysis for small entities rely on the SBA definitions.
1-3
-------
agencies are to assess whether minority or low-income populations face risk or exposure to hazards that is
significant (as defined by the National Environmental Policy Act) and that "appreciably exceeds or is
likely to appreciably exceed the risk or rate to the general population or other appropriate comparison
group." (EPA, 1996). This guidance outlines EPA's Environmental Justice Strategy and discusses
environmental justice issues, concerns, and goals identified by EPA and environmental justice advocates
in relation to regulatory actions. The proposed plywood and composite wood products rule is expected
to provide health and welfare benefits to populations around the United States, regardless of race or
income.
1.3.6 Executive Order 13045
Executive Order 13045, "Protection of Children from Environmental Health Risks and Safety
Risks," directs Federal agencies developing health and safety standards to include an evaluation of the
health and safety effects of the regulations on children. Regulatory actions covered under the Executive
Order include rulemakings that are economically significant under Executive Order 12866 as amended by
Executive Order 13258, and that concern an environmental health risk or safety risk that the agency has
reason to believe may disproportionately affect children. EPA has developed internal guidelines for
implementing E.O. 13045 (EPA, 1998).
The proposed plywood and composite wood products rule is a "significant economic action,"
because the annual costs are expected to exceed $100 million. Exposure to the HAPs whose emissions
will be reduced by this rule are known to affect the health of children and other sensitive populations.
However, this proposed rule is not expected to have a disproportionate impact on children.
1.3.7 Executive Order 13211
Executive Order 13211, "Actions Concerning Regulations That Significantly Affect Energy
Supply, Distribution, or Use," was published in the Federal Register on May 22, 2001 (66 FR 28355).
This executive order requires Federal Agencies to weigh and consider the effect of regulations on supply,
distribution, and use of energy. To comply with this executive order, Federal Agencies are to prepare
and submit a "Statement of Energy Effects" for "significant energy actions." The executive order defines
"significant energy action" as the following:
1) an action that is a significant regulatory action under Executive Order 12866 or any successor order,
and
2) is likely to have a significant adverse effect on the supply, distribution, or use of energy; or
3) that is designated by the Administrator of the Office of Information and Regulatory Affairs as a
significant energy action.
An analysis of the effects of this proposed rule on supply, distribution, and use of energy is
summarized in Chapter 4.
1.4 Other Federal Programs
The only other federal program that may have an effect on these sources is the wood building
products surface coating NESHAP, a rulemaking scheduled to be proposed later in 2002. However, the
overlap of coverage of these rules is expected to be minimal. The wood furniture manufacturing
operations NESHAP, a rule signed in December 1995, may apply to some facilities that will be affected
1-4
-------
by the proposed plywood and composite wood products rule, but there are no overlapping requirements
for individual process units.
1.5 Organization of the Economic Impact Analysis
This report includes eight chapters that present a description of the industry, the costs associated
with the regulatory control options associated with the proposed NESHAP, results of the economic
impact analysis, a summary of impacts on small businesses, a listing of the qualitative benefits associated
with both the HAP and non-HAP emission reductions, and results of the monetized benefits analysis.
Chapter 2 profiles the plywood and composite wood products industries.
Chapter 3 summarizes the approach to estimating the costs of the proposed NESHAP, presents
the results of the cost analysis, and provide the emissions reductions for the proposed alternative.
Chapter 4 summarizes the approach to performing the economic impact analysis of the proposed
NESHAP and presents the results of the analysis. An analysis of impacts on energy distribution,
supply, or use is also in this chapter.
Chapter 5 includes the results of the analyses of the proposed NESHAP's impact on small
businesses.
Throughout this report, a distinction is made between "affected" and "unaffected" facilities and
firms. Affected facilities are those that will incur compliance costs (control and monitoring,
recordkeeping, and reporting ) to comply with the proposed rule. In general, unaffected facilities and
firms have no compliance costs. However, of the group of unaffected facilities, 51 of these will incur
costs associated with monitoring, reporting, and record keeping (MRR). MRR costs are estimated to be
$25,194 per year. The distinction between affected and unaffected facilities and firms will be noted
throughout the document.
1.6 References
Federal Register, 1993. Executive Order 12866, Regulatory Planning and Review. Vol. 58, October 4,
1993, pg. 51735.
U.S. Environmental Protection Agency, 1996. Guidance for Providing Environmental Justice Concerns
in EPA's NEPA Compliance Analyses (Review Draft). Office of Federal Activities, Washington, D.C.,
July 12, 1996.
U.S. Environmental Protection Agency, 1996. Memorandum from Trovato and Kelly to Assistant
Administrators. Subject: "Implementation of Executive Order 13045, Protection of Children from
Environmental Health and Safety Risks." April 21, 1998.
Federal Register, 2001. Executive Order 1321 1. Actions Concerning Regulations That Significantly
Affect Energy Supply, Distribution, or Use. Vol. 66, May 22, 2001, pg. 28355.
Federal Register, 2002. Executive Order 13258, Amending Executive Order 12866 - Regulatory Planning
and Review. Vol.67 , February 28, 2002, pg. 9385.
1-5
-------
2 PROFILE OF THE PLYWOOD AND COMPOSITE WOOD
PRODUCTS INDUSTRIES
2.1 Introduction
Through a 1998 information collection request (ICR), the EPA identified plants potentially
impacted by the proposed NESHAP. This profile presents information on several industries that comprise
the plywood and composite wood source category because they will be impacted by the regulation in
some way. These industries fall into three categories based on their Standard Industrial Classification
(SIC) or North American Industry Classification System (NAICS) classifications.
Softwood plywood and veneer
Reconstituted wood products
Structural wood members
The industries are represented by the three SIC codes and four NAICS codes presented in
Exhibit 2-1. The NAICS codes replaced SIC codes in federal statistical data beginning in 1997. The SIC
code for Structural Wood Members, Not Elsewhere Classified (n.e.c.) was divided into two NAICS codes
for Engineered Wood Members and Truss Manufacturing. The ICR surveyed 416 potentially impacted
facilities (EPA, 1998), and an additional 15 facilities were identified that either did not respond to the
survey or have commenced operation since the date of the survey. The Agency determined that of these
431 facilities, 223 were impacted facilities, owned by 52 firms.
EPA expects this rule to primarily impact certain facilities engaged in the manufacturing of
softwood plywood, reconstituted wood products, and structural wood members. Exhibit 2-1 shows, for
each of the three industry categories, the number of facilities EPA expects will experience compliance
costs as a result of this MACT standard and the total number of facilities. The total estimated capital
costs associated with the new MACT standard are $479 million. The annualized costs for affected
facilities are $138 million on an annual basis, including monitoring, reporting, and record keeping costs
(in 1999 dollars). Some unaffected facilities will also have monitoring, reporting, and record keeping
costs of approximately $4 million per year. Therefore, the total annualized compliance costs are $142
million (1999 dollars).
Including costs associated with monitoring, reporting, and record keeping requirements, EPA
expects 88 softwood plywood and veneer facilities to experience approximately 22 percent of the costs,
38 oriented strandboard facilities to experience approximately 18 percent of the costs, 82 other wood
composite (including medium density fiber (MDF), particle board (PB), and hardboard (HB)) to
experience approximately 58 percent of costs, and engineered wood product facilities to bear the
remaining 2 percent. Most of the discussions contained in this profile will emphasize the softwood
plywood and reconstituted wood products industries because facilities in these industries will experience
the greatest impacts associated with the new MACT standard. A discussion of the affected EWP facilities
is presented in Section 4.4 of this chapter.
2-1
-------
Exhibit 2-1: SIC & NAICS Codes for the Plywood and Composite Wood Industries
SIC Code
SIC Description
NAICS Code
NAICS
Description
Impacted
Facilities*
Total
Facilities in
Category
2436
Softwood Veneer and
Plywood
321212
Softwood Veneer
and Plywood
66
155
2493
Reconstituted Wood
Products
321219
Reconstituted
Wood Products
Total: 97
317
OSB: 23
PB/MDF: 56
HB: 18
2439
Structural Wood
Members, Not
Elsewhere Classified
321213
Engineered Wood
Members (Except
Truss)
3
53
321214
Truss
Manufacturing
0
992
* Does not include number of facilities with MRR costs only.
Sources: MRI (1999), U.S. Environmental Protection Agency (1998), Dun & Bradstreet (1999a), U.S. Department of
Commerce (1999a).
Producers of plywood and composite wood products also engage in additional manufacturing
activities including furniture and wholesale timber production. In some cases, their primary SIC code2
may be one other than those listed in Exhibit 2-1. The facilities with a primary SIC codes other than for
plywood and wood composite manufacturers are shown in Exhibit 2-2. The operations related to these
other SIC codes are unlikely to be affected by the MACT standard. In addition, the number of facilities
identified as potentially affected by this rule relative to the total number of establishments in all categories
is extremely small (under one percent for all categories). Therefore, this profile focuses on the SIC and
NAICS listed in Exhibit 2-1. In particular, the profile will focus on the softwood plywood and veneer and
reconstituted wood products industries. All facilities that are impacted by the MACT standard are
included in these analyses, regardless of their primary SIC or NAICS code.
2See section 2.4.3.1 for a description of how primary SIC codes were assigned to the surveyed facilities.
2-2
-------
Exhibit 2-2: Other Primary SIC and NAICS Codes for the Plywood and Wood Composite Industries
SIC
Description
NAICS
NAICS Title
Facilities
in ICR
Impacted
Facilities
Total
Facilities in
Category
2421
Sawmills
and
Planning
Mills,
General
321113
321912
321918
321999
Sawmills
Cut Stock, Resawing Lumber, & Planning
Other Millwork (including Flooring)
All Other Miscellaneous Wood Product
Manufacturing
32
13
5,815
2426
Hardwood
Dimension
and Flooring
Mills
321113
321912
321918
387215
Sawmills
Cut Stock, Resawing Lumber, & Planning
Other Millwork (including Flooring)
Showcase, Partition, Shelving, and Locker
Manufacturing
5
0
833
2448
Wood
Pallets and
Skids
321920
Wood Container and Pallet Manufacturing
1
0
1,929
2499
Wood
Products,
Not
Elsewhere
Classified
321920
333414
339999
321999
Wood Container and Pallet Manufacturing
Heating Equipment Manufacturing
All Other Miscellaneous Manufacturing
All Other Miscellaneous Wood Product
Manufacturing
4
0
2,760
251
1
Wood
Househol
d
Furniture,
Except
Upholster
ed
33712
2
33721
5
Non-upholstered Wood Household
Furniture Manufacturing
Showcase, Partition, Shelving, and
Locker Manufacturing
13
0
2,785
Sources: MRI (1999), U.S. Environmental Protection Agency (1998), Dun & Bradstreet (1999a), U.S.
Department of Commerce (1999a).
Section 2.2 of this chapter describes the supply side of the affected industries and characterizes the
production process, the products concerned, and the costs of production. Section 2.3 examines the
demand side of the affected industries, product uses, and consumers. Section 2.4 characterizes the
facilities and firms that comprise the industry, their organization, and their financial conditions. Finally,
Section 2.5 describes the markets and discusses domestic production and consumption, international
trade, and prices.
2.2 The Supply Side
The following section contains information concerning the supply of plywood and composite
wood products. This section describes the production processes of each of the aforementioned industries.
It then presents the products, by-products, and co-products of each industry. Lastly, the costs of
production for each of the three industries are presented. Factors, such as industry shipments, costs of
materials, fuels and electricity, payroll, capital expenditures, and materials consumed are all examined.
2-3
-------
2.2.1 Production Process
This section discusses three categories of plywood and wood composites production: plywood
and veneer; particleboard, strand and fiber composites; and structural wood members. The construction
of plywood, consists basically of combining an odd number of layers of veneer, with each layer having
one or more plies. Hardwood plywood is generally made by applying a hardwood veneer to the face and
back of a softwood plywood, MDF, or particleboard panel. The differences between the hardwood and
softwood processes occur because of different inputs and markets. Particleboard, oriented strandboard,
fiberboard, and hardboard are all processed similarly. These three types of reconstituted wood products
are manufactured by combining fragmented pieces of wood and wood fiber into a cohesive mat of wood
particles, fibers, and strands. Structural wood members are the products of multiple manufacturing
techniques. This section describes the production of glue-laminated timber and the three types of
structural composite lumber: laminated veneer lumber, parallel strand lumber, and laminated strand
lumber.
2.2.1.1 General Considerations for Plywood and Wood Composites Manufacturing
Release of hazardous air pollutants (HAPs) is primarily associated with drying and pressing
processes in the manufacturing of plywood and wood composites. Coating processes are intrinsically
related to the manufacturing process and result in further emissions through drying and pressing.
Conventional wood composites are generally made with a thermosetting or heat-curing resin or adhesive
that holds wood fiber together. Commonly used resin-binder systems include phenol-formaldehyde, urea-
formaldehyde, melamine-formaldehyde, and propionaldehyde. A number of additives are used in the
manufacturing of wood composites as well. Most notably, wax is used to provide finished products with
resistance to water penetration. Other additives include preservatives, fire retardants, and impregnating
resins.
While there is a broad range of plywood and wood composites and many applications for such
products, this section of the profile groups the production processes of these products into three general
categories: plywood and veneer; particle board, strand and fiber composites; and structural wood
members. Further descriptions of the production processes for each of these categories are provided in
this section.
2.2.1.2 Plywood and Veneer3
Construction of plywood relies on combining an odd number of layers of veneer. Layers consist
of one or more than one ply with the wood grain running in the same direction. Outside plies are called
faces or face and back plies, while the inner plies are called cores or centers. Layers may vary in number,
thickness, species, and grade of wood. To distinguish the number of plies (individual sheets of veneer in
a panel) from the number of layers (number of times the grain orientation changes), panels are sometimes
described as three-ply, three-layer, or four-ply, three-layer.
As described above, veneer is one of the main components of plywood. Most softwood plants
produce plywood veneer for their own use. Of facilities reporting drying of veneer, 86 percent of the
veneer produced was used for in-facility plywood production. Only approximately 7 percent of the
facilities in the ICR survey produced veneer solely for outside sales and non-internal plywood use (EPA,
1998).
3The descriptions contained in this section rely primarily on U.S. EPA's Lumber and Wood Products Sector
Notebook (1995).
2-4
-------
The general processes for making softwood includes: log debarking, log steaming and/or soaking,
veneer cutting, veneer drying, veneer preparation, glue application, pressing, panel trimming, and panel
sanding. Softwood plywood is generally made with relatively thick faces (1/10 inch and thicker) and
with exterior or intermediate glue. This glue provides protection in construction and industrial uses
where moderate delays in providing weather protection might be expected or conditions of high humidity
and water leakage may exist. Figure 2-1 below presents a diagram of the plywood production process.
Logs delivered to a plant are sorted, then debarked and cut into peeler blocks. Almost all
hardwood and many softwood blocks are heated prior to peeling the veneer to soften the wood. The
peeler blocks are heated by steaming, soaking in hot water, spraying with hot water, or combinations of
these methods. Heated blocks are then conveyed to a veneer lathe. The block, gripped at either end and
rotated at high speed, is fed against a stationary knife parallel to its length. Veneer is peeled from the
block in continuous, uniform sheets. Depending on its intended use, veneer may range in thickness from
1/16 to 3/16 (1.6mm to 4.8mm) for softwood and much thinner for hardwood and decorative plywood
uses (Youngquist, 1999). Slicing methods are also used to produce hardwood decorative veneers
generally in thicknesses of 1/24 inch and thinner.
After peeling, the continuous sheets of veneer are transported by conveyor to a clipping station
where it is clipped. In softwood mills and some hardwood mills, high-speed clippers automatically chop
the veneer ribbons to usable widths and defects are removed. In many hardwood mills, clipping may be
done manually to obtain the maximum amount of clear material. Wet clipped veneer is then dried.
Proper drying is necessary to ensure moisture content is low enough for adhesives to be effective.
Dryers
Two types of dryers are used in softwood veneer mills: roller resistant dryers, heated by forced
air; and "platen" dryers, heated by steam. In older roller dryers, also still widely used for hardwood
veneer, air is circulated through a zone parallel to the veneer. Most plants built in recent years use jet
dryers (also called impingement dryers) that direct a current of air, at a velocity of 2,000 to 4,000 feet per
minute, through small tubes on the surface of the veneer. Veneer dryers may be heated indirectly with
steam, generated by a separate boiler, which is circulated through internal coils in contact with dryer air.
Dryers may also be heated directly by the combustion gases of a gas- or wood-fired burner. The gas-fired
burner is located inside the dryer, whereas combustion gases from a wood-fired burner are mixed with
recirculating dryer air in a blend box outside the dryer and then transported into the dryer. Veneer dryers
tend to release organic aerosols, gaseous organic compounds, and small amounts of wood fiber into the
atmosphere. Once dried, veneer is sorted and graded for particular uses.
2-5
-------
Figure 2-1:
Flow Diagram of Veneer and Plywood Production
Source; Estimating Chemical Releases from Presswood and Laminated Wood Products Manufacturing. U.S. EPA, Office of
Pesticides and Toxic Substances, March 1988.
Note: Many veneer and plywood plants are dry.
Source: U.S. EPA (1995).
2-6
-------
Adhesives
Plywood manufacturing begins with the veneer sent to a lay-up area for adhesive application.
Various adhesive application systems are used including hard rolls, sponge rolls, curtain coaters, sprayers,
and foam extruders. The most common application for softwood plywood is an air or airless spray system,
which generally uses a fixed-head applicator capable of a 10-foot wide spray at a nozzle pressure of 300
pounds per square inch (psi). The phenol-formaldehyde (PF) adhesives typical in softwood plywood
manufacturing is made from resins synthesized in regional plants and shipped to individual plywood
mills. At the mills, the resins are combined with extenders, fillers, catalysts, and caustic to modify the
viscosity of the adhesive. This glue mixing has several additional effects: allowing the adhesive to be
compatible with the glue application method (curtain, roll, spray, foam); allowing for better adhesive
distribution; increasing the cure rate; and lowering cost.
Presses
Following the application of glue, the panels must be pressed. The purpose of the press is to bring
the veneers into close contact so that the glue layer is very thin. At this point, resin is heated to the
temperature required for the glue to bond. Most plywood plants first use a cold press at lower pressure
prior to final pressing in the hot press. This allows the wet adhesive to "tack" the veneers together,
permits easier loading of the hot-press, and prevents shifting of the veneers during loading. Pressing is
usually performed in multi-opening presses, which can produce 20 to 40 4x8-foot panels in each two- to
seven-minute pressing cycle.
Finishing
After pressing, stationary circular saws trim up to one inch from each side of the pressed plywood
to produce square-edged sheets. Approximately 20 percent of annual softwood plywood production is
then sanded. As sheets move through enclosed automatic sanders, pneumatic collectors above and below
the plywood continuously remove the sander dust. Sawdust in trimming operations is also removed by
pneumatic collectors. The plywood trim and sawdust are burned as fuel or sold to reconstituted panel
plants.
2.2.1.3 Particle, Strand, and Fiber Composites4
This group of products falls into the SIC or NAICS code category of reconstituted wood
products. The impacted facilities in this category manufacture the following products (MRI, 1999).
Medium density fiberboard
Oriented stand board
Particleboard
Hardboard
All particle, strand and fiber composites are processed in similar ways. Raw material for
particleboard, oriented strandboard (OSB), fiberboard, and hardboard is obtained by flaking or chipping
wood. The general process then includes wood drying, adhesive application, and forming a mat of wood
particles, fibers, or strands. The mat is then pressed in a platen-type press under heat and pressure until
the adhesive is cured. The bonded panel is finally cooled and further processed into specified width,
4The descriptions in this section rely primarily on Chapter 10 of the USDA's Forest Products Laboratory
Wood Handbook (Youngquist, 1999).
2-7
-------
length, and surface qualities. Specific details regarding the production processes for different products
are provided below.
Particleboard
Generally, particleboard is produced by mechanically reducing wood materials into small
particles, applying adhesive to the particles, and consolidating a loose mat with heat and pressure into a
panel product. Particleboard is typically made in three layers with the faces consisting of finer material
and the core using coarser material. Particleboard can also be made from a variety of agricultural
residues, including kenaf core, jute stick, cereal straw, and rice husks depending on the region. EPA does
not expect facilities that produce particleboard made from agricultural residues, also called agriboard, to
experience compliance cost impacts associated with the new MACT standard. EPA expects only one
facility that produces molded particleboard to experience compliance cost impacts (MRI, 1999).
The raw materials, or "furnish," that are used to manufacture reconstituted wood products can be
either green or dry wood residues. Green residues include planer shavings from green lumber and green
sawdust. Dry process residues include shavings from planing kiln-dried lumber, sawdust, sander dust,
and plywood trim. The wood residues are ground into particles of varying sizes using flakers, mechanical
refiners, and hammermills, and are then classified according to their physical properties.
After classification, the furnish is dried to a low moisture content (two to seven percent) to allow
for moisture that will be gained by the adding of resins and other additives during blending. Most dryers
currently in operation in particle and fiber composite manufacturing plants use large volumes of air to
convey material of varied size through one or more passes within the dryer. Rotating drum dryers
requiring one to three passes of the furnish are most common. The use of triple-pass dryers predominates
in the United States. Dryer temperatures may be as high as 1,100 - 1,200° F with a wet furnish.
However, dry planer shavings require that dryer temperatures be no higher than 500° F because the
ignition point of dry wood is 446° F. Many dryers are directly heated by dry fuel suspension burners.
Others are heated by burning oil or natural gas. Direct-fired rotary drum dryers release emissions such as
wood dust, combustion products, fly ash, and organic compounds evaporated from the extractable portion
of the wood. Steam-heated and natural gas-fired dryers will have no fly ash.
The furnish is then blended with synthetic adhesives, wax, and other additives distributed via
spray nozzles, simple tubes, or atomizers. Resin may be added as received (usually as an aqueous
solution), or mixed with water, wax emulsion, catalyst, or other additives. Waxes are added to impart
water repellency and dimensional stability to the boards upon wetting. Particles for particleboard are
mixed with the additive in short retention time blenders, through which the furnish passes in seconds.
The furnish and resin mixture is then formed into mats using a dry process. This procedure uses air or a
mechanical system to distribute the furnish onto a moving caul (tray), belt, or screen. Particleboard mats
are often formed of layers of different sized particles, with the larger particles in the core, and the finer
particles on the outside of the board. The mats are hot pressed to increase their density and to cure the
resin. Most plants use multi-opening platen presses. Though more popular in Europe, the continuous
press is currently being used in particleboard plants in the United States.
Primary finishing steps for all reconstituted wood panels include cooling or hot stacking, grading,
trimming/cutting, and sanding. Cooling is important for UF-resin-cured boards since the resin degrades at
high temperatures after curing. Boards bonded using PF resins may be hot-stacked to provide additional
curing time. Secondary finishing steps include filling, painting, laminating, and edge finishing. The vast
majority of manufacturers do not apply secondary finishes to their panels; panels are finished primarily by
end-users such as cabinet and furniture manufacturers. Panels are also finished by laminators who then
sell the finished panels to furniture and cabinet manufacturers.
2-8
-------
Oriented Strandboard (OSB)
OSB is an engineered structural-use panel manufactured from thin wood strands bonded together
with waterproof resin under heat and pressure. OSB manufacturing begins with debarked logs usually
heated in soaking ponds sliced into wood strands typically measuring 4.5 to 6 inches long (114 to
152mm). Green strands are stored in wet bins and then dried in a traditional triple-pass dryer, a single-
pass dryer, a combination triple-pass/single-pass dryer, or a three-section dryer. A recent advance in
drying technology is a continuous chain dryer, in which strands are laid between two chain mats so the
strands are held in place as they move through the dryer.
After drying, blending and mat formation take place, blending of strands with adhesive and wax
takes place in separate rotating blenders for face and core strands. Different resin formulations are
typically used for face and core layers. Face resins may be liquid or powdered phenolics, while core
resins may be phenolics or isocyantes. Mat formers take on a number of configurations to align strands
along the length and width of the panel. Oriented layers of strands are dropped sequentially (face, core,
face, for example), each by a different forming head. The mat is then transported by conveyer belt to the
press. Hot pressing involves the compression of the loose layered mat of oriented strands under heat and
pressure to cure the resin. Most plants utilize multi-opening presses that can form as many as sixteen 12-
by 24-ft (3.7- by 7.3m) panels simultaneously. Recent development of a continuous press for OSB can
consolidate the oriented and layer mat in 3 to 5 minutes.
Fiber Composites
Fiber composites include hardboard, medium-density fiberboard (MDF), fiberboard, and
insulation board. In order to make fibers for these composites, bonds between the wood fibers must be
broken. This is generally done through refining of the material, which involves grinding or shearing of
the material into wood fibers as it is forced between rotating disks. Refining can be augmented by water
soaking, steam cooking (digesting), or chemical treatments as well.
Fiber composites are classified by density and can involve either a wet process or a dry process.
High and medium density boards, such as hardboard and MDF, apply a dry process. Wet processes can
be used for high-density hardboard and low-density insulation board (fiberboard). Dry process involves
adhesive-coated fibers that are dried in a tube dryer and air-laid into a mat for pressing.
Wet processes differ from the dry processes. This process involves the utilization of water as a
distributing medium for fibers in a mat. Further differences lie in the lack of additional binding agents in
some wet processes. The technology is very much like paper manufacturing in this pulp-based aspect.
Natural bonding in the wood fibers occurs in this process. Refining in this process relies on developing
material that can achieve this binding with a degree of "freeness" for removal from mats. The wet
process involves a continuously moving mesh screen, onto which pulp flows. Water is drawn off through
the screen and through a series of press rolls. The wet fiber mats are dried in a conveyor-type dryer as
they move to the press. Wet process hardboard is then pressed in multi-open presses heated by steam.
Fiberboard is not pressed.
Manufacturers use several treatments alone or together to increase dimensional stability and
mechanical performance of both wet and dry process hardboards. Heat treatment exposes pressed
fiberboard to dry heat, reducing water absorption and improving fiber bonding. Tempering is the heat
treatment of pressed boards preceded by the addition of oil. Humidification is the addition of water to
bring board moisture content into equilibrium with the air.
2-9
-------
2.2.1.4 Structural Wood Members5
Structural wood members, such as glue-laminated timbers and structural composite timber, are
manufactured using a number of methods. Glue-laminated timber, or glulam, is an engineered product
formed with two or more layers of lumber glued together in which the grain of all layers, called
laminations, is oriented parallel to the length of the lumber. Glulam products also include lumber glued
to panel products, such I-joists and box beams. Structural composite lumber consists of small pieces of
wood glued together into sizes common for solid-sawn lumber.
Glue-Laminated Timber (Glulam)
Glulam is a material that is made from suitably selected and prepared pieces of wood, either
straight or curved, with the grain of all pieces essentially parallel to the longitudinal axis of the member.
The manufacturing process for glulam involves four major steps: (1) drying and grading, (2) end jointing,
(3) face bonding, and (4) finishing and fabricating.
Structural Composite Lumber
The are three major types of structural composite lumber: laminated veneer lumber, parallel
strand lumber, and laminated strand lumber. Each is described in more detail below, however, the general
manufacturing process for these composites is similar.
Laminated veneer lumber (LVL) is manufactured by laminating veneer with all plies parallel to
the length. This process utilizes veneer 1/8 to 1/10 inches. (3.2 to 2.5 mm) thick, which are hot pressed
with phenol-formaldehyde adhesive to form lumber of 8 to 60 feet (2.4 to 18.3 m) in length. The veneer
used for LVL must be carefully selected to achieve the proper design characteristics. Ultrasonic testing is
often used to sort veneer required for LVL. Once the veneer has been selected, end jointing occurs
followed by adhesive application and continuous pressing.
Parallel strand lumber (PSL) is a composite of wood strand elements with wood fibers primarily
oriented along the length of the member. PSL is manufactured using veneer about 1/8 inch (3 mm) thick,
which is then clipped into 3/4 inch (19 mm) wide strands. The process can utilize waste material from a
plywood or LVL operation. Strands are coated with a waterproof structural adhesive, and oriented using
special equipment to ensure proper placement and distribution. The pressing operation results in
densification of the material. Adhesives are cured using microwave technology. As with LVL, the
continuous pressing method is used.
Laminated strand lumber (LSL) is produced using an extension of the technology used to produce
oriented strandboard structural panels. LSL uses longer strands than those commonly used in OSB
manufacturing. LSL is pressed into a billet several inches thick in a steam-injection press, as opposed to
an OSB panel pressed in a multi-opening platen press. The product also requires a greater degree of
alignment of the strands at higher pressures, which result in increased densification.
2.2.2 Products, By-Products, and Co-Products
Exhibit 2-3 presents products, corresponding SIC and NAICS codes, and product examples of the
plywood and composite wood products industry.
5The descriptions in this section rely primarily on Chapter 11 of the USDA's Forest Products Laboratory
Wood Handbook (Moody and Liu, 1999).
2-10
-------
The plywood and composite wood products industries have unique manufacturing processes in
their use of waste wood products as an input for additional products. Planer shavings, sawdust, edgings,
and other wood by-products are inputs to many wood composites. Structural wood members were
developed in response to the increasing demand for high quality lumber when it became difficult to obtain
this type of lumber from forest resources. Therefore, many of the by- and co-products from one process
may be used in another.
Exhibit 2-3: SIC and NAICS Codes and Products
Product Description
SIC
NAICS
Example Products
Softwood Veneer and
Plywood
2436
321212
Panels, softwood plywood
Plywood, softwood
Softwood plywood composites
Softwood veneer or plywood
Veneer mills, softwood
Reconstituted Wood
Products
2493
321219
Board, bagasse
Flakeboard
Hardboard
Insulating siding, broad-mitse
Insulation board, cellular fiber or hard pressed
Lath, fiber
Medium density fiberboard (MDF)
Particleboard
Reconstituted wood panels
Strandboard, oriented
Wafer-board
Wall tile, fiberboard
Wallboard, wood fiber
Structural Wood
Members, Not
Elsewhere Classified
2439
321213
Arches, glue-laminated or pre-engineered wood
Fabricated structural wood members
Finger joint lumber manufacturing
I-joists, wood
Laminated structural wood members
Laminated veneer lumber
Parallel strand lumber
Structural wood members (except trusses)
321214
Floor trusses, wood, glue-laminated or pre-engineered
Roof trusses, wood, glue-laminated or pre-engineered
Source: U.S. Department of Labor, OSHA (no date).
Exhibit 2-4 provides ratios of specialization and coverage (product mix) calculated by the U.S.
Census Bureau for the last three Censuses of Manufacturers. The Census assigns a "primary" SIC code to
each establishment which corresponds to the SIC code for the largest (by value) single type of product
shipped by the establishment. The products shipped from that establishment that are classified in the
same industry as the establishment are considered "primary," and all other products shipped by the
establishment are considered "secondary." The Census then calculates various measures to illustrate the
product mix between primary and secondary products in each industry. The specialization ratio
represents the ratio of total primary product shipments to total product shipments for all establishments
classified in the industry. The coverage ratio represents the ratio of primary products shipped by the
2-11
-------
establishments classified in the industry to the total shipments of these products shipped by all
establishments classified in all industries.
As Exhibit 2-4 illustrates, all three industries have specialization ratios well above 80 percent and
coverage ratios above 90 percent. This implies that most establishments with these SIC codes are highly
specialized, and that most product shipments of each type originate in establishments with these SIC
codes. Therefore, the Census data on these SIC and NAICS industries provide information on the
primary production of facilities engaged in plywood and wood composite manufacturing. These ratios
have been stable overtime.
Exhibit 2-4: Specialization and Coverage Ratios, 1982 -1997
SIC
NAICS
Description
1982
1987
1992
1997
2436
321212
Softwood Veneer and Plywood
Primary products specialization ratio
87
87
84
88
Coverage ratio
96
95
94
95
2493 1321219
Reconstituted Wood Products
Primary products specialization ratio
96
97
96
97
Coverage ratio
97
95
95
97
2439 |321213
Structural Wood Members, N.E. C./Engineered Wood Members
Primary products specialization ratio
96
97
96
95
Coverage ratio
95
97
97
96
2439
321214
Structural Wood Members, N.E. C./Truss Manufacturing
Primary products specialization ratio
96
97
96
96
Coverage ratio
95
97
97
94
Source: U.S. Department of Commerce (1999a and 1995b).
2.2.3 Costs of Production
Exhibit 2-5 provides information on the overall value of shipments (VOS) and production costs (a
component of operating expenses) by SIC code as reported by the Bureau of the Census. Typical of many
intermediate goods, the cost of materials is the largest portion of production costs, with payroll
constituting 15-20 percent of VOS. In particular, timber supply plays a large role in industry costs. In
this decade, reductions in public timber supply, especially reductions in National Forest timber harvests,
combined with the economy's continued demands for wood has led to substantial increases in the cost of
timber (Spelter, 1997).
2-12
-------
Exhibit 2-5: Summary of Annual Costs and Shipments, 1992 -1997
(Thousands of 1997 Dollars)
1992
1993
1994
1995
1996
1997
% Change
Softwood Veneer and Plywood (SIC 2436, NAICS 321212)
Industry Shipments
7,321,64 ld
6,400,683j
6,755,571d
7,725,037d
6,525,702d
5,748,047d
-2l.5°A
Cost of Materials
4,169,048j
3,671,638j
4,097,921d
4,736,984d
4,330,167d
3,795,985d
-8.9%
Fuels & Electricity
220,039j
178,592j
178,601d
183,507d
176,759d
161,239d
-26.7%
Payroll
1,112,158j
897,839j
883,819d
l,047,092d
l,006,792d
912,613d
-17.9%
Ratio of Costs to
Shipments
75%
74%
76%
77%
84%
85%
Reconstituted Wood Products (SIC 2493, NAICS 321219)
Industry Shipments
5,350,565j
4,951,902j
5,517,234d
5,827,821d
5,561,099d
5,278,809d
-1.3%
Cost of Materials
2,400,670j
2,144,060j
2,342,362d
2,582,565d
2,697,471d
2,633,139d
9.1%
Fuels & Electricity
327,706j
250,8^
268,934d
316,876d
321,390d
350,950d
7.1°/
Payroll
825,718j
699,627j
707,179d
810,753d
855,237d
798,767d
-3.3%
Ratio of Costs to
Shipments
66%
62%
60%
64%
70%
72%
Structural wood members (SIC 2439, NAICS 321213 and 321214)
Industry Shipments
3,367,525j
3,281,578j
4,295,002d
4,739,339d
5,096,809d
5,112,873d
51.8°/
Cost of Materials
l,958,576j
l,966,635j
2,584,765d
2,863,098d
3,154,297d
3,007,103d
53.5°/
Fuels & Electricity
35,486j
33,406j
34,585d
39,595d
42,621d
42,090d
18.6°/
Payroll
692,377j
604,180j
740,318d
867,510d
947,403d
954,694d
37.9%
Ratio of Costs to
Shipments
80%
79%
78%
80%
81%
78%
All dollars adjusted to 1997 using Producer Price Index for Lumber and Wood Products (SIC 24).
Source: U.S. Derailment of Census (1999aY
From these data, one can estimate the sector-wide ratio of production costs to VOS. The ratio of
costs (materials, fuels and electricity, and payroll) to the VOS has been increasing over the 1992 to 1997
period for softwood plywood and veneer and reconstituted wood products. The data in Exhibit 2-5 show
that 1997 cost to shipment ratios range between 72 percent (reconstituted wood products) and 85 percent
(softwood veneer and plywood). This measure indicates the proportion of the revenues received for the
goods produced that are associated with production expenses (materials, fuel and electricity, and payroll)
Figures 2-2 and 2-3 present cost and VOS data for the softwood plywood and reconstituted wood
products industries, respectively.
2-13
-------
Figure 2-2: Softwood Plywood and Veneer Value of Shipments and Production Costs, 1992 - 1997
8,000,000
o
E 6,000,000
tfl
I—
SS
£ 4,000,000
0
V)
~o
1 2,000,000
3
o
1993 1994 1995 1996
Year
HVOS - Total Costs
~ Payroll
¦ Fuels & Electricity
~ Cost of Materials
Source: U.S. Department of Commerce (1999a).
Note: Total costs in this figure is the sum of payroll, fuels & electricity, and materials costs.
Figure 2-3: Reconstituted Wood Products Value of Shipments and Production Costs, 1992 -1997
6,000,000
O
O
5,000,000
£
rc
4,000,000
o
Q
H—
O
3,000,000
~o
c
TO
2,000,000
V)
o
£
H
1,000,000
1992
1993
1994 1995
Year
1996
1997
¦ VOS - Total Costs
~ Payroll
¦ Fuels & Electricity
~ Cost of Materials
Source: U.S. Department of Commerce (1999a).
Note: Total costs in this figure is the sum of payroll, fuels & electricity, and materials costs.
2-14
-------
The cost to shipment ratio does not reflect other operating expenses such as non-payroll
employment expenses, taxes, interest, or depreciation. Nor does it indicate whether the expenses are of a
variable or fixed nature. However, it does provide an approximate measure of how much cash, at a gross
level, the industries are generating to cover all operating expenses, use for capital investment, and provide
a return to owners. While this measure is somewhat crude, it indicates that the impacts of the rule may
potentially be more significant for the softwood plywood and veneer industry than for reconstituted wood
products.
Exhibit 2-6 and Exhibit 2-7 provide information on materials consumed by kind in 1997 for the
three sectors. In softwood plywood and veneer manufacturing, 81.6 percent of material costs result from
timber and veneer purchasing. Glues and adhesives represent 5.7 percent of the material costs in the
softwood plywood and veneer industry.
Exhibit 2-6: Materials Consumed By Kind for
Softwood Plywood and Veneer, 1997
Materials Consumed
I Delivered Cost I
i ($1,000)* i
% of Total
Materials
Stumpage cost (cost of timber, excluding land,
cut and consumed at same establishment)
! 346,8541
9.4%
Hardwood logs and bolts
I 64,6171
1.7%
Softwood logs and bolts
1 2,218,800!
60.0%
Hardwood veneer
1 27,3551
0.7%
Softwood veneer
1 363,5831
9.8%
Glues and adhesives
I 210,1051
5.7%
All other materials
! 471,717!
12.7%
TOTAL
! 3,703,031!
100%
* Excludes costs of resales and contract work.
Source: U.S. Department of Commerce (1999a).
Figure 2-4 shows the percentage materials consumed by kind by the softwood plywood and veneer
industry in 1997.
2-15
-------
Figure 2-4: Materials Consumed by Softwood Plywood and Veneer Products, 1997
Softwood logs &
Source: U.S. Department of Commerce (1999a).
Exhibit 2-7: Materials Consumed by Kind for Reconstituted Wood Products, 1997
Material Consumed
Delivered Cost ($1,000)*
% of Total Materials
Logs and bolts
80,891
3.2%
Pulpwood
400,579
15.7%
Chips, slabs, edgings, sawdust, and other
wood waste, and planer shavings
399,446
15.7%
Hardboard, MDF, and particleboard
346,052
13.6%
Paints, varnishes, lacquers, stains, shellacs,
enamels, and allied products
69,488
2.7%
Adhesives and resins
548,553
21.5%
Petroleum wax
61,173
2.4%
Vinyl and paper overlays
101,405
4.0%
All other materials, components parts,
containers and supplies
538,183
21.2%
TOTAL
2,545,770
100%
* Excludes costs of resales and contract work.
Source: U.S. Denartment of Commerce Q999aV
As with the plywood industry, timber products are the largest portion of costs for the
reconstituted wood product industry. Logs, pulpwood, wood materials, and other wood products account
for a combined 48.2 percent of material costs. Unlike plywood and veneer manufacturing, reconstituted
wood products have higher material costs for adhesives and resins, compromising 21.5 percent of costs.
Figure 2-5 shows the percentage of materials consumed by kind by the reconstituted wood products
industry for 1997.
2-16
-------
Figure 2-5: Materials Consumed by Reconstituted Wood Product Producers, 1997
All other materials,
components' parts,
containers & supplies
21.2%
Logs & bolts
3.2%
Pulpwood
15.7%
Vinyl & paper o\erlays
4.0%
Petroleum wax
2.4%
Chips, slabs, edgngs,
sawdust, & other wood
waste, & planner
shaungs
15.7%
Achesiws & resins
21.5%
Paints, \amishes,
lacquers, stains,
shellacs, enamels, &
allied products
2.7%
Hardboard, MDF, &
parti cleboard
13.6%
Source: U.S. Department of Commerce (1999a).
Wood costs for plywood and wood composite manufacturing vary according to plant location,
wood species, and facility efficiency. While there may be considerable variability in wood prices across
regions, the last decade has seen substantial increase in wood prices across all regions. Wood use
efficiency depends on wood species used, log temperature, speed of cutting, board compaction, and other
process variables. Next to wood, adhesives and wax play an important role in industry costs, especially
for the production of reconstituted wood products such as OSB, particleboard, and MDF (Spelter et al,
1997).
In 1995, sixteen percent of the output from the adhesive and sealant industry, SIC 2891, went to
the wood products market. As such, a MACT standard that greatly reduces the demand for adhesives and
sealants (or coatings) could potentially have a significant impact in the adhesive and sealant industry (Abt
Associates Inc., 1997). The response on the part of the softwood plywood and veneer and reconstituted
wood products industries will depend on the final requirements of the MACT standard and the
attractiveness of comparable resin, adhesive and sealant products that do not contribute to HAP
emissions. There will be many constraints on the ability of the impacted industries to switch away from
current adhesives, as their products generally must meet certain requirements related to building codes.
These properties are discussed in the next section.
2.3 The Demand Side
The following section contains information concerning the demand for plywood and veneer,
reconstituted wood products, and structural wood members. The characteristics of plywood and wood
composites are examined first, highlighting the numerous uses of these types of wood products. The
consumers and users of plywood and wood composites are then examined, specifically analyzing the
distribution of consumption. Substitution possibilities are addressed, looking at both wood and non-wood
2-17
-------
options. Lastly, the elasticities of demand of the plywood and composite wood products industries are
discussed.
2.3.1 Product Characteristics
Plywood and composite wood products provide a more stable product over solid wood by
reducing the variations between wood species, among trees of the same species, and even between wood
from the same tree. Unlike solid wood which is evaluated at a cellular level, composite wood is evaluated
at fiber, particle, flake, or veneer level. Properties of products can be changed by combining,
reorganizing, or stratifying these different elements. Control of the size of particles used in producing
composite wood products provides the chief means by which materials can be produced with
predetermined properties (Youngquist, 1999).
Strength is a crucial factor in determining the applicability of plywood and wood composites to
structural and other manufacturing uses. Stiffness and strength properties of a wood product depend
primarily on the constituents from which these products are made. The basic wood elements can be made
in a great variety of sizes and shapes, and may utilize any number of wood species. Plywood can be
manufactured from over 70 species of wood. The choices available for wood composites is almost
unlimited. Types of adhesives and bonding-agents also play an important role in the strength of a
composite wood product.
Durability will also determine the market for wood composites. Panels used for exterior
applications will have a fully waterproof bond and are designed for permanent exposure to weather and
moisture. Interior panels may lack the waterproof bond and be manufactured with glue products designed
for interior use.
Depending on the wood composite, a range of sizes and thicknesses are available. The range of
structural applications for which these products are used requires production of several standardized sizes
as well as custom-made pieces. Sizes and thicknesses will depend on the type of wood composite product
and the market for which it is primarily produced.
Wood panels and other composite wood structures are subject to performance-type standards as
outlined by various industry organizations. A number of organizations including American Plywood
Association - The Engineered Wood Association, Composite Panel Association (CPA), American
Hardboard Association, and others monitor products produced by their member firms to assure high-
quality production and industry conformity with testing and performance standards.
2.3.2 Consumers and Uses
Exhibit 2-8 shows industry output by SIC code. Output of plywood and veneer goes mainly to
the construction sector, primarily to the residential housing and repair industries. Almost one third of
plywood goes to the manufacturing sector, part of which is used as an input for other plywood
production, and part of which goes for furniture and other durable goods manufacturing. The "Other"
category is made up of foreign trade, inventory change, and wholesale trade. The outputs for
reconstituted wood products, including particleboard, are more evenly split between construction and
manufacturing, The "Other" category for reconstituted wood products is made up of sales to state and
local government, foreign trade, and services (Gale Business Resources, 1999).
2-18
-------
Exhibit 2-8: Consumption of Industry Outputs, by SIC Code
SIC
SIC Description
Construction
Manufacturing
Other
2436
Softwood veneer and plywood
63.5%
27.9%
8.6%
2493
Reconstituted wood products
45.7%
47.6%
6.7%
2439
Structural wood members
94.8%
0.6%
4.6%
Source: Gale Business Resources (1999).
The major use of structural panel products is for construction activities. Panel products include those
products such as plywood, OSB, particleboard, and others formed as a panel. These products may be
used for floor systems, exterior walls, roofing, and exterior siding. Figure 2-6 shows the industry outputs
by percentage for the softwood plywood and veneer industry.
Figure 2-6 Industry Outputs of Softwood Plywood and Veneer Industry
Source: Gale Business Resources (1999).
2-19
-------
Figure 2-7 shows the industry outputs by percentage for the reconstituted wood products industry.
Figure 2-7: Industry Outputs of Reconstituted Wood Products Industry
Manufacturing
Source: Gale Business Resources (1999).
MDF and particleboard are two products of the reconstituted wood products industry. Exhibits 2-9 and 2-
10 below show the downstream uses of MDF and particleboard in 1997. For each of the products, about
20 percent of the output is used for household furniture, and the remainder is used for construction,
shelving, cabinetry and other customized applications.
Exhibit 2-9: MDF Shinments bv Downstream Market. 1997
Downstream Use
Million ft2
Percent
Household Furniture
247.8
19%
Custom Laminators
208.6
16%
Stocking Distributors
286.9
22%
Kitchen and Bath
65.2
5%
Molding
130.4
10%
Millwork
65.2
5%
Partitions and fixtures
65.2
5%
All Other
182.6
14%
Other (n.e.c.)
52.2
4%
Total
1,304.0
100%
Source: Composite Panel Association (1998).
2-20
-------
Exhibit 2-10: Particleboard Shipments by Downstream Market, 1997
Downstream Use
Million ft2
Percent
Household Furniture
! 889.0
20%
Custom Laminators
! 711.2
16%
Stocking Distributors
! 755.7
17%
Kitchen and Bath
1 711.2
16%
Flooring Products
! 400.1
9%
Office Furniture
! 266.7
6%
Door Core
! 177.8
4%
All Other
1 400.1
9%
Other (n.e.c.)
! 133.4
3%
Total
! 4,445.2
100%
Source: Composite Panel Association (1998).
Construction Activities
Over sixty percent of the softwood plywood and veneer industry output and approximately 50
percent of the reconstituted wood products industry output goes to the construction sector, primarily to
the construction, remodeling and repair of single and multiple family dwellings. The majority of the
work performed by the construction sector is associated with single family dwellings, and the largest
share of their costs is associated with materials such as wood-based materials. As Exhibit 2-11 shows,
housing starts have been quite strong since 1996 and are expected to continue through at least this year.
Housing start activity is closely linked to general economic conditions, employment, income, and interest
rates. Renovation and remodeling expenditures have declined in real terms, as would be expected.
Generally, more renovation and remodeling takes place during periods when fewer new houses are being
constructed (U.S. Department of Commerce, 1995a).
Exhibit 2-11: Housing Market Indicators. 1988
-1997
Year
1 New Housing Units
I (thousand)
Renovation and Remodeling 1 Renovation and Remodeling
Expenditures Expenditures
(million current $) (million 1992 $)
1988
I 1,706
101,117!
110,874
1989
1 1,574
100,891 I
106,425
1990
I 1,381
106,773 j
109,175
1991
I 1,185
97,528 1
98,813
1992
1 1,411
103,734 j
103,734
1993
1 1,542
108,304 j
104,339
1994
I 1,761
115,0301
106,411
1995
| 1,694
111,683 j
99,362
1996
1 1,838
114,919!
99,756
1997
1 1,828
118,423 j
99,431
Source: Howard (19991
Because economic conditions can vary between regions in the U.S., the impact of housing starts
on demand for wood-based construction materials can vary. This regional variation is further amplified
by differing local preferences, housing codes, and availability of specific wood-based products.
2-21
-------
Wood Furniture Industry
The wood furniture industry produces output for a high value added market. Exhibit 2-12 below
shows the value of shipments from the household furniture sector. Wood household furniture is a portion
of this sector. Domestic shipments and apparent consumption of household wood furniture have
experienced modest growth since 1989, indicating that the shipments from the softwood plywood and
veneer and reconstituted wood products industries to the furniture sector has had limited experience for
growth.
Exhibit 2-12: Trade for Household Furniture (SIC 251), 1989 -1996
(Millions of 1997 Dollars)*
j 1989
1990
1991 j 1992 j 1993
1994
1995
1996
%
Change
Value of product 1
shipments j 23,056
22,477
21,5211 21,949 | 22,82
24,038
24,355
na
6
Value of imports I 3,301
3,200
3,117 I 3,368 ! 3,72
4,201
4,586
5,047
53
Value of exports j 565
884
1,0911 1,2521 1,29
1,385
1,361
1,342
237
Apparent Consumption j 25,792
24,793
23,547 ! 24,065 1 25,24
26,854
27,580
na
7
* Values adjusted to 1997 dollars using PPI for Furniture and Household Durables
Source: U.S. Department of Commerce (1999a).
Wood furniture manufacturers constitute a large portion of the demand, 20 to 30 percent, of the wood-
based products other than structural panels and structural members. Much of the growth in retail demand
is being met by imports. This translates into a large lost opportunity for domestic furniture manufacturers,
as well as for their suppliers, including the industries that are the subject of this profile. The potential
causes for this increase in imports are lower material and labor costs in exporting countries, and declining
availability of timber products to domestic producers (CINTRAFOR, 1999 and Dirks, 1991).
The 1992 Census of Manufacturers showed that 21 percent of the delivered cost of materials in
the manufacture of wood household furniture is associated with plywood and composite wood products.
As a result, significant price changes in the cost of plywood and composite wood products have the
potential to affect production costs of wood household furniture. As the demand for wood household
furniture is highly elastic with respect to price (see discussion in section 2.3.4), increased input costs
could affect both the demand for wood household furniture and for plywood and wood composites
supplied to furniture manufacturers.
2.3.3 Substitution Possibilities
The basic substitution in these industries is between different wood products, although non-wood
substitutes exist as well for some applications. Composite wood products were originally manufactured
in response to the growing demand for wood products as the availability of larger sized timber declined.
As new wood composites products were developed, they further replaced sawn lumber and other types of
wood products. Plywood and veneer production lost market share during the late 1980s and early 1990s
to new products that are categorized as reconstituted wood products, largely as a result of several
challenges: legislation protecting federal timber lands; recession in the early 1990s; price increases and
instability; and supply shortages. To provide an indication of the structural uses of wood panel products
and substitutes, Exhibit 2-13 outlines the use of various products in new single-family and multi-family
residential construction in the United States.
2-22
-------
Exhibit 2-13: Use of Wood and Non-wood Products in Residential Construction
1976 - 1995
Incidence of Use (%)
Single-family houses
Multi-family houses
Application
1976 ! 1988
1995
1976 ! 1988 ! 1995
Floor Sheathing
Lumber
1:
5
-
2;
6!
Structural Panels
511
56
55
511
52!
54
Softwood Plywood
51!
48
31
51!
46!
24
OSB
0!
9
24
0;
7!
3C
Nonstructural Panels
12!
9
9
10;
9!
7
Lightweight Concrete
0!
0
0
5|
7!
3
Concrete Slab
30!
30
35
32!
26!
36
Exterior Wall Sheathing
Lumber
-! 2!
-; -!
Structural Panels
16!
33
52
17;
40!
43
Softwood Plywood
16!
26
19
17;
28!
1C
OSB
0!
7
33
0;
12!
33
Fiberboard
34!
13
6
32;
11!
5
Foamed Plastic
7!
22
29
2;
18!
34
Foil-faced kraft
-!
17
3
0;
13!
1
Gypsum, other
18!
8
2
18;
15!
8
None
25!
5
8
31;
5!
9
Roof Sh eath ing
Lumber
14!
6
1
11;
2!
1
Structural Panels
85!
91
98
87|
94!
94
Softwood Plywood
84!
70
37
87;
78!
li
OSB
1!
21
61
1;
16!
li
Other
1!
3
0
2;
4!
5
Exterior Siding
Lumber
10!
12
7
9;
16!
2
Structural Panels
22!
23
9
32;
15!
4
Softwood Plywood
22!
23
4
32;
15!
2
OSB
-! -; 5
0;
-!
2
Hardboard
16!
16
6
7;
11!
5
Sfon-wood
52!
49
77
49;
58!
89
Vinyl
14!
15
29
12;
14!
41
Masonry, stucco
38!
34
48
37;
44!
4£
Other
0!
0
1
3; -!
Source: Spelter etal. (1997).
Structural wood panels hold the majority of the market share for floor, exterior wall, and roof sheathing in
single and multi-family housing construction. The major substitution effect in this market has occurred
between OSB and softwood plywood, with OSB capturing much of the market from softwood plywood
by 1995. Much of the trade-off between softwood plywood and OSB is due to lower cost for OSB.
However, questions of exterior durability with OSB have led many builders to continue plywood use
despite higher initial costs.
2-23
-------
Fiberboard has also seen reduction in market share for exterior wall systems due to increases in
OSB use. Non-wood products, mainly masonry, have captured 77 percent of the market for exterior
siding, greatly reducing the market share of structural panels in this market. Other major substitutes
include concrete slab for floor sheathing and foamed plastic, which gained major shares of the exterior
wall sheathing market from wood-based structural panels.
2.3.4 Demand Elasticities
The price elasticity of demand is the percentage change in the quantity of product demanded by
consumers divided by the percentage change in price. Demand curves slope downward, signifying a
negative response (less demand) to an increase in price. If demand is elastic (an absolute value of greater
than one) a small price increase will lead to a relatively large decrease in demand. Conversely, if demand
is inelastic with respect to price, or an absolute value less than one, the quantity demanded will change
very little relative to a change in price.
For the purposes of performing an economic analysis, short-term price elasticities are relevant as
impacts of the regulation fall directly on the entities owning facilities faced with compliance
responsibilities. In appropriating compliance costs to facilities impacted by this rule, the economic
analysis assumes that these facilities have a fixed capital stock in the short term. This method allows an
evaluation of the severity of impacts using static measures of profit and loss. This "non-behavioral"
approach differs from other behavioral approaches that take into account adjustments made by producers,
such as changing input mixes, that can generally affect the market environment in which they operate
over the longer term.
In the case of plywood and reconstituted wood production that is going to the construction
industry, the overall price elasticity of demand for these products is relatively inelastic. This is because
the wood product component of construction is fixed once the decision to construct has been made. The
other factors that contribute to the inelastic nature of demand for structural wood panels include local
building codes, home buyer and home owner preferences, and building industry investment in the training
and infrastructure required to construct with wood panels as opposed to a substitute.
The demand for each individual type of product may differ, depending on several factors,
including the product's own-price elasticity, the availability and price of other wood based and non-wood
products with comparable characteristics, and the availability and price of imported products. Cross price
elasticities are often difficult to identify or estimate. However, if available, cross price elasticities of
substitutes and imports might be considered when developing an approach to the economic analysis. For
example, analysis of the softwood plywood market may incorporate the cross-price elasticity of OSB, a
major substitute for plywood. When analyzing the OSB market, the converse would also be true. Even if
such cross price elasticities were available, other considerations would also determine whether the
economic analysis incorporates the market substitution dynamic.
We examined several recent and historical studies of price elasticities of demand. Most of these
studies were concerned with the softwood lumber sector, most likely due to the limited availability of
relevant price and consumption information at a disaggregated product level. Our review focused on the
1996 study by Joseph Buongiorno, a forestry economist, who noted that previous econometric studies of
the wood products sector have produced estimates of demand elasticities for softwood lumber, a product
with similar demand drivers, inputs, input costs, and uses, between zero and -0.96. Buongiorno also
reported that other studies have estimated the cross elasticity of lumber with respect to the price of
plywood to be between 0.5 and zero. Buongiorno developed a model using a price-endogenous linear
6The majority of studies reviewed estimated price elasticity of demand as being between -0.15 and -0.4.
2-24
-------
programming system (PELPS) that endeavored to address the entire wood products market using a system
dynamics approach. The results of this model included short-term price elasticities of demand for wood-
based products, as shown in Exhibit 2-14.
Exhibit 2-14: Demand Elasticities
Product
Price Elasticity of Demand
Plywood
-0.16
Fiberboard
-0.10
Particleboard
-0.27
Source: Buongiorno (1996).
Buongiorno's results provide the basis for imputing price elasticities for the other products that
are the subject of this MACT standard. In addition, further review of identified studies may produce
information useful in the final determination of appropriate elasticities for use in the economic analysis of
the impacts of a MACT standard on the softwood plywood and reconstituted wood products industries.
In the case of softwood plywood and reconstituted wood production going to the furniture
industry, the price elasticity of demand is highly elastic. This is because the price elasticity of demand for
wood furniture is highly elastic itself and the softwood plywood and reconstituted wood component of
production costs for wood furniture is also quite high, over twenty percent. The EPA's study of the
economic impacts of alternative NESHAPS on the wood furniture industry estimated the price elasticity
of demand for wood furniture as -3.477 (U.S. EPA, 1992). This result forms the basis for a derived price
elasticity of demand for use in the economic analysis of the impacts of the MACT standard.
2.4 Industry Organization
The following section contains information pertaining to the organization of the plywood and
veneer, composite wood , and structural wood members industries. This section will provide the basis
for understanding the following.
The industry structure
The characteristics of the manufacturing facilities
The characteristics of the firms that own the manufacturing facilities
A detailed examination of these three topics is essential, as it provides the basis for much of the
approach to estimating economic impacts of the MACT standard. In addition, this section also provides
detailed information about facilities and firms that are important inputs to the analysis itself as well as to
analysis of how the MACT standard might affect firms of different sizes.
2.4.1 In dustry Structure
Exhibit 2-15 shows concentration ratios by SIC code for the three census years, 1982, 1987, and
1992. The m-firm concentration ratios are equal to the sum of the market shares for the largest m number
of firms in the industry. A market is generally considered highly concentrated if the 4-firm concentration
ratio is greater than 50 percent. Exhibit 2-15 also shows the Herfindahl-Hirschman (HH) index, which is
an alternative measure of concentration equal to the sum of the squares of the market shares for the 50
largest firms in the industry. The higher the index, the more concentrated the industry is at the top. The
2-25
-------
U.S. Justice Department uses 1,000 as a benchmark for the presence of market concentration, where any
industry with a Herfindahl-Hirschman index less than 1,000 is considered to be unconcentrated (Arnold,
1989).
Exhibit 2-15: Concentration Ratios by SIC Code, 1982-1992*
1 Number
! of
Percent of value of industry shipments shipped
by the largest (in terms of shipment value)
Herfindahl
Year jCompanie
is in
; Industry
4 L 8 • 20 50
„ . ;Companie;„
Companies! ^ ; Companies ; Companies
Hirschman
Index**
Softwood Veneer and Plywood (SIC 2436)
1982! 135!
41! 561
74 1
92
619
1987! 131!
381 561
74 1
93
571
1992! 123!
47 i 66 i
82 1
96
797
Reconstituted Wood Products (SIC 2493)
1982!
N/A
1987! 158!
48 i 65 i
82 1
95
743
1992! 193!
501 661
81 i
94
765
Structural wood members (SIC 2439)
1982! 649!
151 221
35 1
50
104
1987! 831!
13 | 18 |
30 1
44
92
1992! 830!
19| 25 1
34j
46
166
*The latest year for which data is currently available.
**The index is based on the 50 largest companies in each SIC code.
Source: U.S. Department of Commerce (1992).
The concentration ratios presented in Exhibit 2-15 show very little evidence of market
concentration in the plywood and composite wood products industries. Four-firm concentration ratios for
the three sectors are below 50 with the exception of reconstituted wood products (classified as "General"
in the ICR survey) which is 50. The HH indices for all SIC codes are well below the benchmark of 1000.
While concentration appears to have increased in general between 1982 and 1992, there is no clear trend
as all appear to have been less concentrated in 1987.
2.4.2 Manufacturing Plants
Through an ICR, the U.S. Environmental Protection Agency identified plants potentially affected
by this rule. EPA categorized the surveyed facilities according to their production processes and
developed estimates of compliance costs for each facility. Exhibit 2-16 below presents information on the
number of potentially impacted facilities, and their corresponding primary SIC code. The exhibit also
shows the percent of potentially impacted facilities as a percent of total facilities for each SIC.
2-26
-------
Exhibit 2-16: Facilities with Compliance Cost Impacts
Facilities
SIC Code! Description
Impacted*
Total in SIC
% of Total
2436 | Softwood Veneer and
i Plywood
66
155
42.6%
i Reconstituted Wood
2493 j Products
Total
97
317
30.6%
OSB
23
PB/MDF
56
HB
18
2439 | Structural Wood Members
3
53
5.7%
* Does not include number of facilities with MRR costs only.
Note: Percentages represent survey facilities' share of total facilities in the category.
Sources: U.S. Environmental Protection Agency (1998), U.S. Department of Commerce (1999a), MRI
(1999).
2.4.2.1 Location
Nationally, facilities that produce softwood plywood and reconstituted wood products are
clustered in distinct geographic regions of the South, Pacific Northwest, and the upper Mid-West of the
U.S. Based on the 1997 Census of Manufacturers, the softwood plywood and veneer facilities have the
highest employment in Oregon, Washington and Louisiana. The Census showed that reconstituted wood
product facilities had the highest employment in Oregon, California, North Carolina, Texas, and
Michigan (source: U.S. Department of Commerce, 1999a).
Figure 2-87 is a map of locations of impacted and total ICR facilities as identified by EPA (MRI,
1999, EPA, 1998). For this figure, all types of facilities are combined. The map shows the state-by-state
distribution of the potentially impacted facilities relative to the total ICR facilities in the state. The states
with the greatest number of potentially impacted facilities are Oregon (36), Louisiana (16), Georgia (8),
Mississippi (7), Virginia (10), Texas (8), and North Carolina (7). Major producing states where impacted
facilities constitute a significant portion of all facilities in the state include Louisiana (66 percent), Oregon
(57 percent), Washington (77 percent), Georgia (38 percent) and Texas (44 percent).
7 Map developed based on original survey database dated July 23, 1999.
2-27
-------
| | States
Impacted Facilities
rn °
1 - 9
10 - 18
H 19 - 27
Figure 2-8: Plywood and Wood Composite Facility Locations
(Potentially Impacted Facilities and Total ICR Facilities by State)
Sources: U.S. Enviromnental Protection Agency (1998), MRI (1999)
Oof 2
2-28
-------
2.4.2.2 Production Capacity and Utilization
Exhibit 2-17 shows the capacity utilization rates by SIC code and for all manufacturing industries
for 1992 through 1997. The rates for softwood plywood and veneer, and reconstituted wood products are
significantly higher than the average for all lumber and wood products and for all industries. Capacity
utilization for structural wood members is below industry averages but has increased over the 1992 - 1997
period.
Exhibit 2-17: Full Production Capacity Utilization Rates, Fourth Quarters, 1992 - 1997
SIC
SIC Description
1992
1993
1994
1995
1996
1997
Change
2436
Softwood Veneer and Plywood
87
92
95
95
86
84
-3.4%
2493
Reconstituted Wood Products
87
92
92
88
86
82
-5.7%
2439
Structural Wood Members
65
66
66
74
77
72
10.8%
24
All Lumber and Wood Products
80
81
80
77
78
75
-6.3%
2000-3999
All Manufacturing Industries
77
78
80
76
76
75
-2.3%
Source: U.S. Department of Commerce (1997).
Figure 2-9 presents the capacity utilization rates of softwood plywood and veneer and reconstituted wood
products from 1992-1997.
Figure 2-9: Full Production Capacity Utilization, Fourth Quarters, 1992-1997
re
1 I
Softwood, plywood &
veneer
¦ Reconstituted wood
products
Source: U.S. Department of Commerce (1997).
The capacity utilization for softwood plywood and veneer, and reconstituted wood peaked in 1994,
consistent with utilization peaks for all manufacturing industries. Interestingly, utilization rates for
reconstituted wood product facilities declined, while softwood plywood and veneer was unchanged in
2-29
-------
1995, the year that shows the highest value of shipments for all (see Exhibit 2-17). This may be
explained, in part, by capacity expansions driven by the increased capital expenditures by softwood
plywood and veneer producers in 1994 and subsequent years.
The ICR provided further information on capacity utilization. A sample of general facilities
responding to questions regarding their production processes reported production and capacity. From this
data, capacity utilization for general facilities was 78 percent, slightly below the figures in Exhibit 2-17
2.4.2.3 Employment
Exhibit 2-18 provides information on employment at the softwood plywood veneer and
reconstituted wood products facilities responding to the ICR in 1998.
Exhibit 2-18a: 1998 Employment at Facilities with Expected Compliance Cost Impacts
Softwood Plywood and
Veneer
Oriented Strandboard
Medium Density
Fiberboard/
Particleboard
Number of
Employees
Facilities 1 % of All
in Size I Impacted
Category | Facilities
Facilities 1 % of All
in Size I Impacted
Category | Facilities
Facilities in 1 % of All
Size 1 Impacted
Category | Facilities
Not reporting
21 3.0%
o i o.o%
o i o.o%
<50
0! 0.0%
0! 0.0%
1 ! 1.8%
50 to 99
01 0.0%
0 i 0.0%
121 21.4%
100 to 249
18; 27.3%
21 1 91.3%
301 53.6%
250 to 499
341 51.5%
1 1 4.3%
21 3.6%
500 to 999
11! 16.7%
1 ! 4.3%
8! 14.3%
1,000 to 1,499
1 1 1.5%
01 0.0%
21 3.6%
>1,500
01 0.0%
01 0.0%
1 I 1.8%
TOTAL
661 100%
23 I 100%
561 100%
Sources: U.S. Environmental Protection Aeencv (1998N). MRI (1999).
Exhibit 2-18b: 1998 Employment at Facilities with Expected Compliance Cost Impacts
Hardboard
Engineered Wood
Products
Total Facilities
Number of
Employees
Facilities 1 % of All
in Size 1 Impacted
Category ; Facilities
Facilities 1 % of All
in Size 1 Impacted
Category ; Facilities
Facilities in 1 % of All
Size 1 Impacted
Category ; Facilities
Not reporting
0! 0.0%
0! 0.0%
21 1.2%
<50
01 0.0%
01 0.0%
1 1 0.6%
50 to 99
1 i 5.6%
01 0.0%
13 1 8.0%
100 to 249
81 44.4%
1 i 33.3%
771 47.2%
250 to 499
4! 22.2%
2! 66.7%
41 ! 25.1%
500 to 999
51 27.8%
01 0.0%
251 15.3%
1,000 to 1,499
01 0.0%
01 0.0%
3 1 1.8%
>1,500
01 0.0%
o i 0.0%
1 1 0.6%
TOTAL
18! 100%
3 ! 100%
166! 100%
Sources: U.S. Environmental Protection Agency (1998), MRI (1999).
2-30
-------
Potentially impacted facilities engaged in the production of plywood and composite wood
products tend to be small- to medium-sized. Just over half of the facilities reported having less than 250
employees. Softwood plywood producers tend to have larger facilities, while facilities producing
reconstituted wood products tend to be smaller.
2.4.2.4 Facility Population Trends
Plant age may be of particular significance to potential regulatory impacts. Older plants may be
less efficient as compared with newer plants utilizing technological improvements in production
efficiency. One example mentioned earlier is the development of the continuous press, enabling recently
constructed plants to produce more panel products in less time than older manufacturers. Newer plants
may utilize better volatile organic compound emission control technologies and have adapted their
processes to meet indoor air quality requirements.
While specific age information for all facilities is not available, an analysis performed by Spelter
et al. (Spelter, 1997) provides insights into the changing nature of plywood and composite wood facilities
over time. In their analysis, they traced the number of mills, average mill capacity, and capacity
utilization over the course of 20 or more years. The analysis does not present information on specific
plant closures and openings over time, but presents the total number of operating mills, which reflects the
net change resulting from both closures and openings. Exhibit 2-19 provides information on the results of
the analysis for selected years from 1977 to 1997, using census years to provide some comparison to
overall industry figures presented elsewhere in this chapter.
2-31
-------
Exhibit 2-19: Number of Mills, Average Capacity and Utilization, 1977 -1997
1977 I
1982 |
1987 !
1992 j
1997
% Change
Softwood Plywood
Number of Mills
62!
69 1
58 |
56 |
57
-8%
Average Mill Capacity (1000 m3)
1101
1381
180!
201 j
215
95%
Capacity Utilization
97!
79 |
"!
95 !
91
0%
Oriented Strandboard
Number of Mills
8!
21 !
39I
44!
66
125%
Average Mill Capacity (1000 m3)
88 |
115 1
148;
I87 j
259
194%,
Capacity Utilization
441
90 |
99 j
84
91%
Particleboard
Number of Mills
54!
43 I
44!
45 j
45
-17%
Average Mill Capacity (1000 m3)
137!
151 !
168!
181 j
1%
43%
Capacity Utilization
86 |
87 1
89 j
89 j
91
13%
Medium-density Fiberboard
Number of Mills
12!
13 !
17!
17 !
26
117%
Average Mill Capacity (1000 m3)
95!
105 I
122!
141 j
151
59%
Capacity Utilization
69!
66!
87I
91 !
86
25%
Laminated Veneer Lumber
Number of Mills
2 j
61
12 |
17
750%
Average Mill Capacity (million m3)
0.078 j
0.075 |
0.063 j
0.085
9%
Capacity Utilization
73 j
60!
75 j
93
27%
Engineered Joists
Number of Mills
12 !
12I
18!
35
192%
Average Mill Capacity (million meters)
3 |
41
51
9
200%
Capacity Utilization
69 j
73!
90j
58
-16%
information not available for some years. For softwood plywood, particleboard, and MDF, 1997
figures are from 1996. For particleboard, 1984 figures are used for 1982.
Source: Spelter etal. (1997).
Average facility capacity has shown substantial increases over the last twenty years for all
product groups. While the number of softwood plywood facilities declined by 8 percent between 1977
and 1997, the average mill capacity increased substantially, nearly 100 percent. Particleboard has
experienced some capacity growth while the number of plants has declined.
The OSB industry has shown the largest increase in per facility capacity, 194 percent, along with
large net additions of facilities. Most notably, there were nine more OSB plants than softwood plywood
plants in 1997, whereas in 1977 plywood plants outnumbered OSB plants nearly 8 to one. Recent facility
additions for OSB and MDF show these sectors have newer facilities, while the softwood plywood and
particleboard industries are generally composed of older facilities.
2-32
-------
A review of recent capital investment trends provides some insights into the facility population
trends of the softwood plywood and reconstituted wood products industries. Exhibit 2-20 shows capital
expenditures by industry sector. Capital expenditures have seen substantial overall increases in the last
five years for all three sectors, indicating increasing investment, particularly in the reconstituted wood
product and structural wood members sectors. However, investment by the softwood plywood and
veneer and reconstituted wood products sectors declined sharply from 1996 to 1997. This trend indicates
the connection between declining capital expenditures and the sharp increase in products costs' share of
the value of shipments (as shown in Exhibit 2-5) that began after 1995. If such conditions in the baseline
continue into the future, it is possible that certain firms may experience difficulty accessing capital to
cover these costs in addition to compliance costs associated with the MACT standard.
Exhibit 2-20: Summary of Capital Expenditures, 1992 -1997
(Thousands of 1997 Dollars)
1992
1993
1994
1995
1996
19971 % Change
Softwood Plywood & Veneer
110,125j
128,490^
159,685^
192,090^
212,277^
168,142} 52.7%
Reconstituted Wood Products
159,330
185,452
353,665
367,057
583,659
329,744! 107.0%
Structural Wood Members*
47,420
70,659
220,523
143,523
108,889
138,880; 192.9%
All dollars adjusted to 1997 using GDP Deflator.
* 1997 figure is sum of capital expenditures for NAICS 321213 and 321214.
Source: U.S. Department of Commerce (1999a).
For softwood plywood, the level of capital investment constitutes only 3 percent of the industry's
total value of shipments. With the number of mills in decline and average mill capacity growing, it
appears that the majority of capital expenditures made by the softwood plywood industry occur at
existing plants. This conclusion is supported by U.S. Industry & Trade Outlook '99, which reported that
only one new softwood plywood facility has opened in the last 10 years (U.S. Department of Commerce,
International Trade Administration, 1999).
Conversely, results of the growing capital investments made by the reconstituted wood products
industry can be observed in the large increases in the number of OSB and MDF plants, and the rising
average plant capacities of reconstituted wood products producers. As a group, these producers invested
6 percent of the value of shipments in 1997, twice the investment rate of the softwood plywood
producers. For example, in September of 1999, Willamette Industries announced that it will build an $85
million particleboard plant in South Carolina. The plant will have a capacity of 210 million square feet
per year and will be in operation in late 2001.
2.4.3 Firm Characteristics
Several factors will likely be of importance in determining the distribution of impacts generated
by the proposed MACT standard on companies. Size may play a role in a company's ability to absorb an
increase in compliance costs. Ownership is a second factor that may play a role. Because firms have
different legal and financial guidelines based on ownership, their approaches to complying with the
MACT standard may vary. Vertical and horizontal integration, or lack there of, in plywood and
composite wood product firms may affect the manner in which they absorb the potential costs of the
MACT standard. Lastly, the overall financial condition of the plywood and composite wood industries is
assessed, attempting to determine the industry's ability to withstand adverse conditions.
2-33
-------
2.4.3.1 Size Distribution
Firm size is likely to be a factor in the distribution of the impacts of the proposed MACT on
companies. Under the Regulatory Flexibility Act (RFA) and its 1996 amendment, SBREFA, SBA
definitions are used to designate which businesses are considered to be small. The SBA has set size
standards under the NAICS system, using various thresholds for the number of employees or revenues.
In determining the size of a company, the SBA treats a facility that has a substantial portion of its assets
and/or liabilities shared with a parent company as part of that company. In this analysis, the company's
primary NAICS code is used to determine the appropriate SBA threshold.
Exhibit 2-21 provides information on firm size for plywood and wood composite firms owning
facilities with expected compliance cost impacts. In the ICR, facilities were asked to provide information
on employment size for domestic parent firms. Many facilities did not report information on the ultimate
domestic parent. For this reason, information on ultimate domestic parent primary SIC and NAICS code
and employment size were obtained from Dun and Bradstreet's DUNS Database. Exhibit 2-21 shows the
number of firms and the facilities owned by the firms in the first two data columns. In the absence of Dun
& Bradstreet information on the owner, the facility's primary SIC and NAICS code from Dun &
Bradstreet was used to determine the appropriate SBA threshold. Based on this SIC code, facility
employment information from the ICR was used to make a size determination. Exhibit 2-21 shows the
number of firms and the facilities owned by the firms in the third and fourth data columns. In the absence
of facility primary SIC code from DUNS, the standard for lumber and wood products (all SIC 24 codes)
of 500 employees was used as the threshold. A full list of the facilities and their size determination is
provided in the economic impact analysis for this proposed rule.
Exhibit 2-21: Size Distribution of Firms Owning Facilities with Expected Compliance Cost Impacts
Size
SIC Based on
DUNS
SIC Based on ICR
Other Sources
Total
Firms i Facilities
! Owned by
1 Firms*
Firms
F acilities
Owned by
Firms*
Firms i Facilities
i Owned by
i Firms*
Fi mis
%
Facilities
Owned by
Firms*
%
Small
8 1 10
5
5
6 ! 7
19
35.2%
22
8.4%
Large
29 1 231
4
8
2 ! 2
35
64.8%
241
91.6%
Total
37 1 241
9
13
8 ! 9
54
100%
263
100%
* Includes all facilities reported, impacted and non-impacted.
Sources: U.S. Environmental Protection Agency (1998), MRI (1999). SBA Size Standards from SBA website:
http://www.sba.gov/regulations/siccodes/.
While over 35 percent of firms in the industry are considered small, 91 percent of facilities are
owned by large firms. Given the concentration ratios presented in Exhibit 2-15, there does not appear to
be any significant market power to these larger firms. However, the ability of larger firms to deal with
compliance costs, as compared to smaller firms, may have impacts on the industry organization.
The larger parent firms have both impacted and non-impacted facilities. Firms such as Georgia-
Pacific (43 ICR facilities), Louisiana-Pacific (32 ICR facilities), Willamette Industries (23 ICR facilities),
Columbia Forest Products (13 ICR facilities), Weyerhaeuser (19 ICR facilities), and Boise-Cascade (12
ICR facilities) may be able to make trade-offs between facilities and shift production to more efficient
facilities in response to compliance costs associated with the MACT standard.
2-34
-------
2.4.3.2 Ownership
The form of firm ownership has a set of legal and financial characteristics that may influence a
firm's regulatory compliance alternatives. The legal form of ownership impacts the cost of capital,
availability of capital, and effective tax rate faced by the firm. Debt-equity issues for these firm types will
play a role in financing capital-intensive controls. Firm ownership may generally be one of three types.
Sole proprietorships (companies with a private single-owner)
Partnerships (non-corporate firms with more than one owner)
Corporations (publically or privately owned companies formed through incorporation)
Exhibit 2-22 provides information on ownership type for the lumber and wood products industry.
While specific information by 4-digit SIC or 5-digit NAICS is not available, the table provides a general
sense of ownership types in the industry, assuming that ownership structure for the three industries
profiled is similar to that of the overall lumber and wood products industry.
Exhibit 2-22: Types of Firm Ownership for Lumber and Wood Products (SIC 24/NAICS 321), 1992
Corporation
Sole
1 Proprietorship
Partnerships
Other/
Unknown
Single-Facility Firms 1
1,291
14,909
Multi-Facility Firms j
17,617
61
All Firms 1
18,908
j 10,447
2,336
2,187
Source: U.S. Dept. of Commerce (1992).
Over ninety percent of single facility wood and lumber products firms are owned by sole
proprietorships, partnerships, or some other/unknown entity. Nearly all multi-facility firms are owned by
corporations. Just over half of all lumber and wood products firms are a corporation, while the remainder
are sole proprietorships (30 percent), partnerships (7 percent), or other (6 percent). These data support the
conclusion that single-facility firms owned by sole proprietors are more likely to be classified as small
businesses, while multi-facility firms owned by corporations are more likely to be classified as large
businesses.
2.4.3.3 Vertical and Horizontal Integration
The data presented in Section 2.2 on concentration and specialization ratios for the plywood and
composite wood industries, combined with the information on establishment size and ownership type
demonstrate that the majority of firms in the three industries examined in this profile are predominantly
not, or minimally, vertically or horizontally integrated. However, there are several exceptions to this
conclusion. The six largest firms that own multiple facilities are for the most part both vertically and
horizontally integrated. These firms, described in more detail below, are large multi-billion dollar
concerns that are vertically integrated through their ownership of timberland, their production facilities,
and their involvement in product distribution. Their horizontal integration is attributed to their other
product lines, generally pulp and paper.
Georgia-Pacific, a large, horizontally and vertically integrated firm, manufactures and distributes
building products, pulp and paper, and resins. The company's wood product line includes wood panels,
plywood, and hardboard. It also produces lumber, gypsum products, chemicals, and packaging. Georgia-
Pacific grows and sells timber, and participates in several other activities related to forestry management.
Its 1998 net sales revenues exceeded $13 billion, and it has 45,000 employees at 400 locations. Its
building products division reported record profits during the second quarter of 1999. It currently has
2-35
-------
plans to build an OSB plant in Arkansas and recently merged with Unisource, a major distributor of
imaging paper and supply systems (Financial Times, 1999b; PRNewsWire, 1999b).
Louisiana-Pacific is principally a manufacturer of building products, but also produces pulp and
building insulation, and owns almost one million acres of timberland. Its sales of structural lumber,
industrial panels, and exterior building products made up nearly 75 percent of the company's revenues,
which reached $2.3 billion in 1998. The company manufactures OSB, I-joists, LVL, MDF, fiberboard,
particleboard, hardboard, softwood plywood and hardwood veneer. Louisiana-Pacific has been involved
in a series of mergers and acquisitions that include Le Goupe Forex of Canada, Evans Forest Products,
and ABT Building Products (Louisiana-Pacific, 1999; Financial Times, 1999c).
Willamette Industries, a forest products manufacturing company, has three main lines of business:
brown paper, white paper, and building products. The building products division manufactures plywood,
lumber, particleboard, MDF, OSB, LVL and I-joists, among others. Approximately one third of the
company's $3.7 billion in total revenue is from its building materials segment. Most of Willamette's
recent merger and acquisition activity has been with firms in France and Mexico. It also owns plants in
Ireland and 1.8 million acres of timberland in the U.S. (Financial Times, 1999d; PRNewsWire, 1999c,
1998, 1997).
Columbia Forest Products describes itself as North America's largest manufacturer of hardwood
veneer, and laminated products. They sell their products through a network of wholesale distributors,
mass merchandisers and major original equipment manufacturers (OEMs). Their products include
decorative, interior veneers and panels used in high-end cabinetry, fine furniture, architectural millwork
and commercial fixtures. Columbia Forest Products is an employee-owned company with 13 plants in the
U.S. and four in Canada (Columbia Forest Products, 1999).
Weyerhaeuser is an integrated international forest products company. It is involved in growing
and harvesting timber, and the manufacturing and distributing of several categories of forest products.
Among its wood products are plywood, OSB, and wood composites. The company bills itself as the
world's largest private owner of saleable softwood timber and the country's largest producer of softwood
lumber and pulp. In addition, it is the top U.S. exporter of forest products. The company has
approximately 36,000 U.S. and Canadian employees and sales of $11 billion, ten percent of which comes
from exports (Weyerhaeuser, 1999).
Boise Cascade, an integrated international paper and forest products company, manufactures and
distributes paper and wood products, distributes office products and building materials, and owns and
manages over 2 million acres of timberland. Its building products include lumber, plywood,
particleboard, veneer, and engineered wood products. Sales of these products constitute 27 percent of the
company's $6.2 billion annual revenue (Financial Times, 1999a; PRNewsWire, 1999a).
2.4.3.4 Financial Condition
The financial condition of an industry's firms will affect the incidence of impacts of the costs
associated with complying with a new MACT standard. While information necessary to determine which
specific firms might experience adverse impacts is not available, one can examine industry-wide
indicators of financial condition. Each year, Dun & Bradstreet (D&B) publishes Industry Norms & Key
Business Ratios, which reports certain financial ratios for a sample of firms in the industry. This section
focuses on measures of profitability and solvency.
2-36
-------
Profitability Ratios
The return on sales ratio, also known as the net profit margin, is an indicator of a firm's ability to
withstand adverse conditions such as falling prices, rising costs, and declining sales, and is calculated by
dividing net profit after taxes by annual net sales.
Return on assets is calculated by dividing a firm's net profit after taxes by its total assets. This
ratio is a key indicator of both profitability and operating efficiency by comparing operating profits to the
assets available to earn a return. According to Dun & Bradstreet, companies that use their assets
efficiently will have a relatively higher return on assets than those firms that do not use their assets
efficiently.
The return on equity shows the profitability of the company's operations to owners, after income
taxes, and is calculated by dividing net profit after taxes by net worth. According to Dun & Bradstreet,
this ratio is looked to as a 'final criterion' of profitability, and a ratio of at least 10 is regarded as desirable
for providing dividends plus funds for future growth.
Solvency Ratios
The current ratio is calculated by dividing a firm's current assets by its current liabilities. This is
a measure of liquidity that gauges the ability of a company to cover its short-term liabilities. The standard
guideline for financial health is 2. The quick ratio is slightly different than the current ratio, because it
does not include inventories, advances on inventories, marketable securities, or notes receivables. The
quick ratio measures the protection afforded creditors in cash or near-cash assets. Any time this ratio is 1
or greater, the firm is said to be in a liquid condition.
Exhibit 2-23 shows various measures of the financial condition of the plywood and composite
wood industry over the period 1995 to 1997. The trends shown in Exhibit 2-23 confirm that the softwood
plywood and reconstituted wood products industries have experienced a profit squeeze due to increasing
costs and falling prices in recent years.
Exhibit 2-23: Indicators of Financial Condition, 1995-1997*
Indicator
Softwood Plywood
and Veneer
Reconstituted Wood
Products
Structural
Wood Members
1995 I 1996 I 1997
1995 ! 1996 ! 1997
1998
Return on Sales
5.81 3.6 j 1.7
3.8! 3.1! 3.5
5.0
Return on Assets
15.7! 13.5! 6.0
7.8 ! 5.9 ! 3.5
13.0
Return on Equity
28.7! 22.9! 8.7
15.2! 10.0! 5.7
NA
Current Ratio
3.2! 2.61 2.7
2.81 2.71 1.7
2.3
Quick Ratio
1.1! 1.31 1.2
1.8! 1.2! 1.1
1.3
*Includes 1998 data for Structural Wood Members, the only data reported for this sector.
Source: Dun & Bradstreet (1999). Indicator values are based on median values of the industrial sample.
For SIC 2436, there were 14 establishments in the sample in 1995, 15 in 1996, and 11 in 1997. For SIC 2493,
there were 28 establishments in the sample in 1995, 30 in 1996, and 31 in 1997. For SIC 2439, there were 135
establishments in 1998.
The softwood plywood and veneer industry has not maintained its relatively strong degree of
financial health, with many of its profitability indicators significantly lower in 1997 than in 1995. In
particular, the softwood plywood and veneer industry experienced 60 to 70 percent declines in all three
profitability ratios. The falling profitability of this industry is now at a level that indicates the presence of
firms that are not using their assets efficiently, are not providing the cash needed for future growth, and
2-37
-------
may more acutely experience adverse conditions associated with MACT standard compliance costs. The
currently low net profit margin is indicative of an industry that is experiencing increasing production
costs as a percentage of its value of shipments and falling capacity utilization (Exhibits 2-5 and 2-18).
The reconstituted wood products industry also saw fairly dramatic decreases in its financial
indicators over the time period shown, resulting in a relatively low return on assets and return on equity,
as well as a current ratio lower than generally considered healthy. These indicators are consistent with
recent trends in the industry associated with increases in production costs relative to the value of
shipments (Exhibit 2-5), rapid expansion of production capacity (Exhibit 2-20) and competitive pressures
on prices from overseas producers. This industry also includes firms that are not using their assets
efficiently or providing the cash needed for future growth. The reconstituted wood products industry's
profit margin is also somewhat low, but typical of all firms in the lumber and wood products sector (Dun
and Bradstreet, 1999b).
In the fall 1999 issue of Engineered Wood Products Journal, industry analyst Evadna Lynn
discussed investor response to the industry's current financial performance (APA, 1999c). Lynn
attributes several recent trends to stockholder pressure for improved financial performance.
Separating timber assets
Corporate restructuring
Cost control through consolidation
These trends have contributed to a dynamic market structure in recent years. By selling or
otherwise spinning off timber assets, forest products companies are converting them to cash and
improving financial performance. Restructuring activities have focused on gaining higher returns from
core business activities through the closure or divestiture of less profitable facilities or products. Some of
the divested facilities, particularly plywood mills, have been reopened by new owners as sawmills. The
industry has seen several major corporate mergers and acquisitions in the late 1990s, including:
Weyerhaeuser and MacMillan Bloedel, International Paper and Union Camp, and Louisiana Pacific and
Le Groupe Forex (of Canada). Most post-merger cost reductions are gained from streamlining operations,
including closure of production facilities (APA, 1999c; International Paper, 1998).
2.5 Markets
This chapter discusses general market conditions for the plywood and composite wood products
industries. In particular, this chapter discusses market structure, provide background on current market
volumes, prices, and international trade. It also presents information on future market volumes, prices
and international trade. The purpose of this chapter is to describe the current status of the industry and to
support the development and implementation of the economic impact analysis that is summarized in this
RIA.
2.5.1 Market Structure
Based on the data, background and analyses reviewed while preparing this industry profile, it is
reasonable to conclude that these industries exhibit clear signs of a competitive market for the products
that are the subject of this MACT standard. There are several reasons for this conclusion. First, as
discussed in section 2.4.2, the plywood and composite wood products industries are unconcentrated.
There is little concentration of market power evidenced by each separate industrial category having a 4-
firm concentration ratio of 50 or below (often well below) and HH indices below the 1000 benchmark.
Next, the output of several of the production sectors are substitutes for each other, putting competitive
pressures on suppliers. There are also competitive pressures from alternative products, either traditional
sawn lumber or non-wood materials. This chapter will focus on other factors of the competitive nature of
2-38
-------
these industries. For the most part, the markets for these goods also experience competitive pressures by
the presence of imported products. Finally, several industry experts have observed trends where prices of
the products respond negatively to the presence of excess capacity. The remainder of this chapter will
provide additional details related to these observations on industry competitiveness.
2.5.2 Market Volumes
This section will present a discussion of market consumption and production volumes for the
three industrial sectors examined in this study. For the most part, this discussion will rely on the data
contained in Exhibit 2-24 and Exhibit 2-25. Exhibit 2-24 shows the value of product shipments by
product class for the period 1989 to 1995s as reported by the International Trade Administration of the
U.S. Department of Commerce. Note that value of shipments data for Structural Wood Members is not
available for inclusion in this table. Exhibit 2-25 shows the physical volume of output produced, traded
and consumed between 1988 and 1997 for selected products as reported by Spelter et al. in their 1997
statistical report. International trade is discussed later in the section.
Exhibit 2-24: Trade Balance and Selected Statistics, Thousands of 1997 Dollars
1989 I 1990 | 1991 j 1992 j 1993 j 1994
j %
1995 ! Change
Softwood Veneer and Plywood (SIC 2436, NAICS 321212)
Value of product shipments
7,125 | 6,887 j 6,185 j 6,422 j 5,643 j 5,885
6,671 | -6%
Value of imports
81 j 69 1 55 1 79 1 82 1 100
111 I 37%
Value of exports
452 ! 509 j 428 j 452 j 391 j 333
375 ! -17%
Trade Surplus (Deficit)
371 I 440 j 373 j 372 j 310 j 234
263 ! -29%
Apparent Consumption
6,755 | 6,447 j 5,812 j 6,050 j 5,333 j 5,651
6,407 | -5%
Ratio of Imports to Consumption
0.01 j 0.01 j 0.01 j 0.01 j 0.02 j 0.02
0.02 | 45%
Ratio of Export to Product Shipments
0.06 ! 0.07 j 0.07 ! 0.07 j 0.07 j 0.06
0.06! -11%
Ratio of Imports to Exports
0.18! 0.14! 0.13! 0.18! 0.21! 0.30
0.30 ! 65%
Reconstituted Wood Products (SIC 2493, NAICS 321219)
Value of product shipments
5,013 | 4,761 | 4,743 j 5,359 j 4,940 j 5,511
5,772 | 15%
Value of imports
461j 409 j 364 j 540 j 616 j 861
1,080 | 134%
Value of exports
261! 334 j 350 j 328 j 271j 301
345 ! 32%
Trade Surplus (Deficit)
(200) ! (75) j (14) ! (212) j (345) j (560)
(735) ! 268%
Apparent Consumption
5,213 | 4,836 j 4,757 j 5,572 j 5,285 j 6,070
6,507 | 25%
Ratio of Imports to Consumption
0.09! 0.08 1 0.08! 0.10! 0.12 j 0.14
0.17 | 88%
Ratio of Export to Product Shipments
0.05 ! 0.07 j 0.07 ! 0.06 j 0.05 j 0.05
0.06 ! 15%
Ratio of Imports to Exports
1.76 ! 1.22 j 1.04 ! 1.65 j 2.27 j 2.86
3.13 ! 77%
Source: U.S. Department of Commerce, International Trade Administration (1998).
81995 is the latest year for which data is available.
2-39
-------
Exhibit 2-25: Production, Trade and Consumption Volumes for Selected Products (1988-1997)
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
%
Change
Softwood Plywood (M ft3, 3/8 in basis)
Product shipments
22,089j
21,385j
20,919j
18,652j
19,332j
19,315j
19,368j
19,367j
19,181d
17,963d
-19%
Imports
96
49
38
28
47
41
47
60
85
104
8%
Exports
1,0cm
1,442j
1,613j
l,322j
l,442j
l,409j
1,21 lj
l,267j
l,248d
l,548d
54%
Apparent Consumption
21,181
19,991
19,344
17,358
17,937
17,946
18,474
18,160
18,018
16,519
-22%
Other Structural Panels (M ft3, 3/8 in basis)
Product shipments
4,604j
5,1°5j
5,418j
5,613j
6,653j
7,002j
7,486j
7,903j
9,314d
10,534d
129%
Imports
815
U11,
1,313j
988
l,572j
2,163j
2,588j
3,214d
4,414d
5,272d
547%
Exports
57
49
60
78
82
157
167
193%*
Apparent Consumption
5,416
6,213
6,728
6,544
8,176
9,105
9,995
11,036
13,572
15,639
189%
Particleboard/Medium Density Fiberboard (M ft3, 3/4 in basis)
Product shipments
4,768j
4,828j
4,856j
4,730j
5,046j
5,402j
5,793j
5,307d
5,705d
5,916d
24%
Imports
l,634j
425
363
293
405
572
775
840
814
963
-41%
Exports
163
333
373
369
394
318
297
319
154
188
15%
Apparent Consumption
6,239
4,920
4,746
4,654
5,057
5,656
6,271
5,828
6,365
6,691
7%
Hardboard (M ft3, 1/8 in basis)
Product shipments
5,118j
5,196j
5,025j
4,895j
5,273j
5,248j
5,206j
4,930d
5,280d
4,501d
-12%
Imports
633
718
689
571
571
639
U19
1,152j
U83d
l,306d
106%
Exports
322
427
552
606
836
917
l,190j
l,377d
l,426d
l,259d
291%
Apparent Consumption
5,429
5,487
5,162
4,860
5,008
4,970
5,135
4,705
5,037
4,548
-16%
Source: Spelter etal. (1997).
* since 1991
2-40
-------
2.5.2.1 Domestic Production
As Exhibit 2-24 shows, the value of shipments (representing production) of softwood plywood
and veneer was slightly lower in 1995 than it was in 1989. During the period, production reached its
lowest level in 1993 and then began to climb, in response to meeting demand from rising expenditures for
renovation and remodeling and new housing starts. The value of shipments of reconstituted wood
products rose 15 percent between 1989 and 1995, linked to the underlying growth in the construction
sector and the growth in market share of structural panel products over softwood plywood.
Figure 2-10 compares the value of product shipments of softwood plywood and veneer to
reconstituted wood products from 1989-1995.
Figure 2-10: Value of Product Shipments, 1989-1995
o
o
C3
O
O
C3
££
S2
re
o
Q
N.
0>
0>
~ Softwood plywood &
veneer
¦ Reconstituted wood
products
1994
Year
1995
Source: U.S. Department of Commerce, International Trade Administration (1998).
Trends in product shipments by volume (Exhibit 2-25) have been mixed for this group of
industries. A statistical report produced by the U.S. Forest Service's Forest Products Laboratory (Spelter
et al., 1997) focused on production of softwood plywood, Other Structural Panels (OSB and waferboard),
particleboard and MDF as a group, and hardboard. Production by the Other Structural Panels category
experienced high growth during the period, with 1997 production almost 130 percent greater than it was
in 1988. Most of this increase can be attributed to the rapid increase in OSB's share of the structural
panel market in recent years. Particleboard and MDF production grew a moderate 24 percent, while
production of softwood plywood and hardboard declined by 19 percent and 12 percent respectively.
Historically, softwood plywood production made a continuous steady climb through the late 1980's. At
that point, the product began losing market share to OSB and production leveled off. This trend was
accompanied by a certain amount of mill attrition (Spelter et al., 1997).
2-41
-------
2.5.2.2 Domestic Consumption
Domestic, or apparent, consumption is the sum of domestic production and imports, less exports.
The dollar value of apparent consumption (Exhibit 2-25) for softwood plywood and veneer was slightly
lower in 1995 than it was in 1989. During the period, demand for softwood plywood and veneer dropped
slightly in the early 1990s and reached its lowest level in 1993 and then began to climb. The value of
domestic consumption of reconstituted wood products followed a similar pattern, increasing by 25
percent overall between 1989 and 1995. Drivers of consumption trends described here are the same as
those presented in the previous section on production (increased demand for renovation, remodeling and
new housing starts).
Figure 2-11 compares the apparent consumption of softwood plywood and veneer to reconstituted
wood products from 1989-1995.
Figure 2-11: Apparent Consumption, 1989-1995
7,000
6,000
5,000
o
o
o
o"
o
o
£ 4,000
(fl
= 3,000
o
Q
% 2,000
1,000
1989 1990 1991 19^qq^^^®
1993 1994 1095
Year
~ Softwood plywood &
veneer
¦ Reconstituted wood
products
Source: U.S. Department of Commerce, International Trade Administration (1998)
Further examination of consumption volumes (Exhibit 2-25) shows the following trends for
softwood plywood, other structural panels, particleboard and MDF as a group, and hardboard.
By volume, apparent consumption of softwood plywood fell by over 20 percent in the last 10
years.
At the same time, consumption of other structural panels increased by almost 200 percent.
Particleboard and MDF were consumed at a slightly higher level in 1997 than they were in 1988,
following a decline that ended in 1992.
Hardboard consumption has fluctuated during the same 10 years, with a 16 percent decline from
1988.
2-42
-------
Demand for softwood plywood and OSB combined experienced an annual average growth rate of 2-3
percent from 1970 to 1996 (Spelter et al., 1997). Most of this demand was met by increased production
of OSB by both domestic and imported producers.
2.5.2.3 International Trade
Imports
Import value trends during the 1989-1995 period (Exhibit 2-24) show the constant dollar value of
softwood plywood and veneer imports grew by 37 percent, particularly during the later years when the
price for the commodity was rising rapidly and supplies of timber were declining. The ratio of imports to
consumption of softwood plywood and veneer, while only 0.02, grew by 45 percent. The trade surplus
for softwood plywood and veneer fell by 37 percent. Imports of reconstituted wood products more than
doubled from 1989 to 1995 and the value of imports' share of consumption grew by almost 90 percent
and the trade deficit nearly quadrupled.
Looking at import volumes (Exhibit 2-25) for softwood plywood, other structural panels,
particleboard and MDF as a group, and hardboard, imports have made the biggest gains in the other
structural panel category, taking advantage of the overall growth in demand for those products. Imports
now supply over a third of the other structural panel market. Imports of hardboard have also grown, more
than doubling in volume since 1988. There was a slight increase in imports of softwood plywood over
the 10 years, and a decline of 40 percent in imports of particleboard and MDF. Exhibit 2-26 shows U.S.
imports of by major region and trading partner.
Exhibit 2-26: 1997 U.S. Wood Products Imports by Region and Ma.jo
r Trading Partner
Trade Areas
| Value*
j (Smillions)
Share
NAFTA
I 8,1281
85.1
Latin America
1 541 !
5.7
Western Europe
! 234!
2.5
Japan/Chinese Economic Areas
1 351
0.4
Other Asia
! 458!
4.8
Rest of World
1 150!
1.6
World Total
| 9,554 j
100.0
Top 5 Countries j
Canada
I 7,991 j
83.6
Indonesia
1 340!
3.6
Brazil
! 303 j
3.2
Mexico
1 1371
1.4
Chile
! 108!
1.1
*Includes Sawmills (SIC 2421), Softwood Plywood and Veneer (SIC 2436), Reconstituted
Wood Products (2435), and Hardwood Plywood and Veneer (SIC 2435).
Source: U.S. Department of Commerce, International Trade Administration (1999).
2-43
-------
Exhibit 2-26 shows that a vast majority, 85.1 percent, of U.S. imported wood products
originated in the North American Free Trade Agreement (NAFTA) trade zone, of which only 1.5 percent
originates in Mexico. The U.S. is also importing a significantly greater value of wood products than it is
exporting. In 1997 the U.S. exported about $3,683 million of wood products while it imported $9,554
million.
Imports of softwood plywood and veneer grew by 24 percent from 1996 to 1997. Seventy-seven
percent of U.S. softwood plywood and veneer imports are from Canada. This growth is consistent with
the strong demand for softwood plywood and veneer during this period. The overall penetration of
imports into the U.S. market is quite small (2 percent), which is attributed to the efficiency and low costs
of U.S. softwood plywood and veneer producers (U.S. Department of Commerce, International Trade
Administration, 1999).
Imports of reconstituted wood products grew by seven percent from 1996 to 1997. Seventy-
eight percent of U.S. reconstituted wood products imports are from Canada. The overall penetration of
imports into the U.S. market is significant (18 percent), which is attributed to recent capacity additions
by Canadian reconstituted wood products producers (U.S. Department of Commerce, International Trade
Administration, 1999).
Exports
Export trends during the 1989-1995 period (Exhibit 2-24) show the value of softwood plywood
and veneer exports fell by 17 percent, particularly during the later years when the price for the
commodity was rising rapidly and supplies of timber were declining. Economic crises in several Asian
economies and the falling value of the Canadian dollar relative to the U.S. dollar played a role in this
trend. The ratio of exports to value of shipments of softwood plywood and veneer fell by 11 percent.
Exports of reconstituted wood products grew by 32 percent from 1989 to 1995 and the proportion of
exports to shipments grew by almost 15 percent.
Export volumes (Exhibit 2-25) of hardboard quadrupled between 1988 to 1997, and constitute a
significant portion of the total shipments from this industry. Exports of softwood plywood grew by 50
percent, and have become an increasingly important part of the sector's overall production. While total
exports of other structural panels grew significantly, this market still remains a small portion of
production. Exports of particleboard and MDF grew significantly through 1992 but have dropped
steadily in recent years and are now just 15 percent higher than they were seven years ago. Exhibit 2-27
shows U.S. exports by major region and trading partner.
2-44
-------
Exhibit 2-27: 1997 U.S. Wood Product Exports by Region and Major Trading Partner
Trade Areas
| Value* |
| (Smillions) j Share
NAFTA
| 1,001 1
27.5
Latin America
! 203 !
5.6
Western Europe
| 1,230 !
33.8
Japan/Chinese Economic Areas
1 837 1
23.0
Other Asia
! 205 !
5.6
Rest of World
I 161 I
4.4
World Total
! 3,638 !
100.0
Top 5 Countries 1
Canada
! 800 1
22.0
Japan
1 636 1
17.5
Germany
! 292 !
8.0
United Kingdom
1 244 1
6.7
Mexico
i 202 i
5.5
*Includes Sawmills (SIC 2421), Softwood Plywood and Veneer (SIC 2436), Reconstituted
Wood Products (SIC 2493), and Hardwood Plywood and Veneer (SIC 2435).
Source: U.S. Department of Commerce, International Trade Administration (1999).
By region, the U.S. exports its largest share (33.8 percent) of wood products to Western Europe.
However, no single country in Europe imports the most significant share of U.S. wood products. Canada
imports the largest share, 22 percent, due to two reasons. First, Canada's economy has strengthened.
Second, on January 1, 1998 Canada completed its final stage of tariff removal as directed under the U.S.-
Canada Free Trade Agreement. For the two aforementioned reasons, U.S. wood product exports to
Canada increased 21 percent to $800 million in 1997 (U.S. Department of Commerce, International
Trade Administration, 1999).
Continued growth in U.S. exports of wood products is dependent on an Asian economic revival,
particularly in Japan's economy. In 1996, prior to the economic crisis, Japan was the largest importer of
U.S. wood products. By 1997, Japan's share of U.S. wood product exports fell to 17 percent, a 24
percent decrease from the previous year. To further exacerbate the problem, U.S. exports to Japan are
expected to decline an additional 30 percent in 1998 and 1999. Japan has undertaken several steps to
revitalize its economy, such as the implementation of the Enhanced Initiative on Deregulation and
Competition Policy. However, an increase in the Japanese consumption tax from 3 percent to 5 percent
in 1996 is believed to have canceled out the potential gains of the Policy, resulting in the expected
continuing decline in Japanese demand for U.S. plywood and wood products (U.S. Department of
Commerce, International Trade Administration, 1999).
In 1997, exports of softwood plywood and veneer accounted for about 10.6 percent of wood
product exports from the U.S. This was a 24 percent increase from the previous year, raising the total
value of softwood plywood and veneer exports to $392 million, the highest level in eight years. Exports
to the United Kingdom, Canada, and Germany, the top three importers of U.S. softwood plywood and
veneer, experienced strong gains in 1997. A healthy European market has increased the demand for
softwood plywood and veneer. In particular, the construction sector throughout Europe has seen an
2-45
-------
increase in activity. However, the recent strong performance of softwood plywood and veneer is not
expected to continue due to an increasing international acceptance of OSB, and increasing competition
from Canada, Brazil, and Indonesia (U.S. Department of Commerce, International Trade Administration,
1999).
Reconstituted wood products accounted for about 9.75 percent of U.S. wood product exports in
1997. Both the value and volume of reconstituted wood product exports increased by 15 percent from
the previous year. Canada, the United Kingdom, Mexico, and Japan are the largest export markets for
U.S. reconstituted wood. The continuing increase in exports is mainly attributable to a growing
international acceptance of OSB. Exports are expected to continue to grow in the upcoming years, but at
a slower rate than they did in 1997 (U.S. Department of Commerce, International Trade Administration,
1999).
2.5.3 Prices
An index of the change in producer prices for lumber and wood products is shown below in
Exhibit 2-28 This index was compiled by the Bureau of Labor Statistics.
Exhibit 2-28: Lumber and Wood Products Producer Price Index, 1988-1997
(1982 = 100)
1988
1989
1990
1991= 1992= 1993= 1994
1995
1996
1997 88-97
Lumber
and wood
products
(SIC 24)
Change
from
Previous
year
122.1
125.7
2.9%
124.6
-0.9%
124.9
0.2%
144.7
15.9%
183.4
26.7%
188.4
2.7%
173.4
-8.0%
179.8
3.7%
194.5
8.2%
59.3%
Source: U.S. Bureau of Labor Statistics (1999).
The biggest annual price increases for lumber and wood products occurred in 1992 and 1993 and
the overall price increase between 1988 and 1997 was nearly 60 percent. Another source, the Forest
Products Laboratory (FPL), that is part of the U.S. Department of Agriculture, provides a statistical
report with disaggregated price indices presented in Exhibit 2-29. Note that the base year of the BLS
index is 1982 while the base year for the FPL data is 1992.
2-46
-------
Exhibit 2-29: Producer Price Indices of Plywood and Wood Composite Products
(1992 =100)
Year
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
88-97
Softwood
Plywood
74.2
84.5
81.4
82.2
100.0
115.4
120.3
128.0
118.3
119.3
Change from
Previous year
0
0
0
0
0
0
0
0
0
1
Particleboard
103.4
106.0
96.7
96.5
100.0
114.8
128.5
128.4
123.3
117.6
Change from
Previous year
0
0
0
0
0
0
0
0
0
0
Hardboard
100.8
100.9
98.6
96.7
100.0
106.5
109.1
113.2
115.8
119.0
Change from
Previous year
Source: Howard (1
999).
0
0
0
0
0
0
0
0
0
0
Softwood plywood experienced the biggest price increase, 61 percent over the 1988 to 1997
period, with volatile price changes within the period with the biggest annual increases came in 1992 and
1993. Overall prices for particleboard rose 14 percent, but the large price increases in 1993 and 1994
have been offset by price declines in the last three years presented. Hardboard prices grew by 18
percent, with mostly steady annual price increases from 1994 on.
The market conditions and the factors that affect softwood plywood prices, supply and demand
are somewhat analogous to those that affect prices for softwood lumber. For example, the cost of timber
and transportation, foreign supply and demand, inventory levels as well as construction-driven demand
are factors that affect market prices for softwood lumber, as well as softwood plywood and other
structural panels.
A recent study produced by WEFA (Wharton Economic Forecast Associates) on trends in the
softwood lumber market provides some clues about the future of the three industries examined here.
Softwood lumber prices have climbed steadily since November of 1998. This climb included some
higher than expected price increases in the early summer of this year. The WEFA report cites strong
domestic demand related to housing construction as one underlying cause of the price increases in
softwood lumber. Current price conditions are partially explained by the expectation that housing
demand has peaked while remaining strong, exports to Asia will increase as those economies recover,
and imports from Canada will decrease.
For the most part, the WEFA report indicates that the construction industry has responded to
climbing prices by switching to "just-in-time" buying of products. Buyers are hoping that prices will
begin falling and are postponing inventory build-up during this period of climbing prices. Another
short-run factor affecting prices during the second quarter of this year was a constraint on truck and rail
transportation availability. WEFA concludes that the market has reached equilibrium for the moment,
although this could change if inventories increase at the same time that construction-driven demand
levels off or falls (WEFA, 1999).
Exhibit 2-30 presents the industry-reported free on board (f.o.b.) prices of southern plywood,
OSB and particleboard from 1989 to 1996. These are the product prices prior to shipping costs and
distributor mark-ups. On an adjusted basis, these prices reflect the trends demonstrated in the previous
exhibit, with large price increases during early 1992, falling back to or below 1989 levels by 1996.
2-47
-------
Exhibit 2-30:
F.O.B. Prices of Southern plywood, OSB, and Particleboard
($ per cubic meter)
Year 1 Southern plywood 1 OSB Particleboard
I As
j Reported
Adjusted ! As
$1997 ! Reported
Adjusted ! As
$1997 ! Reported
Adjusted
$1997
1989! 184
229! 166
206! 129
160
1990! 168
200! 124
148! 122
145
1991! 175
201! 144
165! 120
138
1992! 226
252! 208
232! 129
144
1993! 257
279! 227
247! 152
165
1994! 274
291! 252
268! 171
182
1995! 267
277! 242
251! 173
180
1996! 231
235! 184
187! 165
168
89-96!
2.8%!
-10.1%!
4.5%
Prices adjusted by the GDP deflator.
Source: Spelter, etal. (1997).
Softwood Plywood
Long-term price trend data presented in the report "Review of the Wood Panel Sector in the
U.S." showed a fairly stable price pattern for softwood plywood between 1977 and 1991. At that point,
prices increased steadily from 1992 to their peak in 1994. Prices declined over 15 percent from 1994 to
1996. The report authors observe that with softwood plywood prices at their current high levels,
producers will have a difficult time competing against the newer, more cost effective OSB producers.
However, the authors note that softwood plywood producers may be able to hang on to market share and
justify the higher prices by differentiating their product as a premium construction material (Spelter et
al., 1997).
Oriented Strand Board
The "Review of the Wood Panel Sector in the U.S." report presents OSB price data over time
that shows a 27 percent decline in price during 1995 and 1996, after a continuous trend of price increases
since 1977. The report's authors attribute this weakening to a rapid increase in capacity that contributed
to an increase in production, putting downward pressure on prices. Due to the ability of users to
substitute plywood for OSB, these low OSB prices have only added to the growing market share enjoyed
by OSB. Falling prices have cut into the net revenues of OSB producers, after a period from 1992 to
1995 where the industry enjoyed excellent cost/price margins, drawing more investment to OSB
production capacity (Spelter et al., 1997).
Particleboard
Particleboard price data from 1984 to 1992 presented in the report, "Review of the Wood Panel
Sector in the U.S." show some variation within a relatively small range, with a substantial price increase
in the years 1993 to 1995, declining slightly in 1996. The price trend for particleboard from 1977 to
1996 is very similar to that of plywood. One reason for this similarity is the close relationship of
particleboard input costs to the plywood manufacturing industry. About 25 percent of industry
production cost is for wood inputs, which are primarily made up of wastes from lumber and plywood
production (Spelter et al., 1997).
2-48
-------
Medium Density Fiberboard (MDF)
Producer-reported MDF prices were $235 per ton in September of 1996 and declined by 15
percent to $205 per ton as of April, 1997. Despite this drop, there continues to be a price gap between
MDF and less costly particleboard, although increasingly narrow. The price drop was attributed to MDF
production capacity expansions that resulted in an increase in supply, putting pressure on the profits of
MDF producers (Spelter et al., 1997).
Structural Wood Members
Producer-reported prices for I-joists reach a high in 1994 and have been declining since that
time. Recent price conditions have made I-joists more competitive with traditional 2" by 10" lumber on
an installed cost basis, typically for floor framing applications. In particular, I-joists are price
competitive with lumber when lumber prices are high. However, precise estimates of market prices are
difficult to obtain. The authors found that prices varied depending on whether the product was being
sold under a brand name, on sale, or under a volume discount. Laminated veneer lumber, presented in
the Review at $550/m3 f.o.b., is generally more expensive than 2" by 10" lumber, and is used mostly for
structural applications or as an input to I-joists (Spelter et al., 1997).
2.5.4 Market Forecasts
Production and Consumption
A study published by WEFA in the summer of 1999 examined housing starts and concluded that
housing starts will decline throughout 1999, resulting in a decline in lumber demand (WEFA, 1999).
However, housing starts continue to remain strong well into 1999, keeping demand for lumber and other
wood products for construction strong as well. The WEFA study also noted that another factor affecting
demand for softwood lumber is interest rates and concluded that rising interest rates could have a
dampening effect on demand. Higher interest rates will not only affect the affordability of new homes,
but also will curtail purchases of existing homes and mortgage refinancing activity, both major sources
of demand for materials used in home remodeling. Based on the relationship between housing starts,
purchases of existing homes, and remodeling and renovation (the construction-based demand for
plywood and other products examined in this profile), this decline in demand can be expected to affect
the plywood and wood composite industries, as 60 to 70 percent of their output goes to the construction
sector. WEFA expects the industry to experience most of this decline in 2000 (WEFA, 1999).
The most recent wood products market outlook published by APA - The Engineered Wood
Association (APA) shows U.S. housing starts exceeding expectations in 1999 (APA, 1999d). Similar to
the WEFA study, the forecast expects higher interest rates in the future to play a role in reducing future
housing demand in the period from 2000 to 2002. The report also forecasts the same trends for
residential improvements and repairs, but notes the long-term outlook for remodeling to be good as
home ownership increases. Figure 2-12 below provides information from the APA on U.S. housing
starts. The APA forecast also reports the industrial outlook is good for other wood-consuming sectors.
The APA expects demand for furniture and fixtures to remain healthy, but not at peak levels as existing
home sales will be declining from the current peak rates. Nonresidential construction is forecasted to
peak in 1999 and 2000 with declines in 2001 and 2002. Increased school construction will be a driving
factor in the upward trend for nonresidential construction (APA, 1999d).
2-49
-------
Figure 2-12: APA Projected Housing Starts (000s)
2000
1500
1000
500
0
~ US Single Family ¦ US Multi-Family
J
/
/
il
/
/
/
/
/
1997 1998 1999 2000 2001 2002
Source: APA (1999d).
In addition to providing overall forecasts for the market demand, the APA outlook includes
detailed forecasts of the demand for and production of structural panels, specifically softwood plywood
and OSB.9 These forecasts, summarized in Exhibit 2-31, show the demand from each of the major
markets for structural panels, in order of their share of market demand: new residential construction,
remodeling, industrial uses including furniture and materials, nonresidential construction, and foreign
demand. The industrial use category will have the largest domestic demand increase over the forecast
period, 8 percent. Foreign demand shows significant increase of 78 percent. However, reductions in
U.S. production as imports gain a large market share point to increased pressure from imports.
9While the report does not specify whether the forecast is exclusively for softwood plywood or includes
hardwood plywood, it is assumed to cover softwood plywood only, as hardwood plywood is typically not used for
structural panels.
2-50
-------
Exhibit 2-31: APA Forecasted Structural Panel Production and Demand
(million sq. ft. 3/8" basis)
1999
2000
2001
2002
% Change
New Residential
18,415j
17,715j
17,585j
18,435j
0.00
Remodeling
7,440j
7,440j
7,475j
7,550j
1%
Industrial/Other
6,575j
6,720j
6,875j
7,085j
8%
Nonresidential
3,800j
3,800j
3,735j
3,670j
-3%
Domestic Demand
36,230j
35,675j
35,670j
36,740j
1%
Foreign Demand
990
l,275j
l,705j
l,760j
78%
Total Demand
37,220.00j
36,950.00j
37,375.00j
38,500.00j
3%
Imports (Canada only)
(7,345)j
(7,400)j
(8,300)j
(9,330)j
27%
Total Domestic Production
29,875.00j
29,550.00j
29,075.00j
29,170.00j
-2%
Plywood
18,135j
17,450j
16,575j
16,295j
-10%
OSB
11,740
12,100
12,500
12,875
10%
Source: APA ri999dV
The APA forecasts for panel capacity and production provide additional insight into substitution
between softwood plywood and OSB. Exhibit 2-32 below shows these projected trends. Softwood
plywood shows significant decreases in capacity (down 24 percent) and production (down 16 percent)
from 1992 to 2002. Meanwhile, OSB has shown significant increases in capacity and production and is
projected to continue to capture the market for structural panels. The relatively constant capacity
utilization in the plywood sector with significant decreases in production supports the forecast of
expected plant closures in the future, while the opposite is true for OSB with expected increases in the
number of facilities.
2-51
-------
Exhibit 2-32: APA Actual and Forecasted Structural Panel Capacity and Production
(million Sq Ft, 3/8" Basis)
Plywood
OSB
Capacity
Production
Utilization
Capacity
Production
Utilization
1992
23,700
19,332
82%
7,040
6,653
95%
1993
23,300
19,315
83%
7,560
7,002
93%
1994
21,875
19,638
90%
7,920
7,486
95%
1995
22,070
19,367
88%
8,830
7,903
90%
1996
21,150
19,181
91%
11,285
9,314
83%
1997
19,275
17,965
93%
11,575
10,534
91%
1998
19,075
17,776
93%
12,050
11,227
93%
1999
19,275
18.135
94%
12,250
11.740
96%
2000
18.835
17.450
93%
13.120
12.100
92%
2001
18.260
16.575
9.1%
13.725
12.500
91%
2002
18.010
16.295
90%
14.380
12.875
90%
% Change
-0.24
-0.16
1.04
0.94
Shaded areas represent estimated values.
Source: APA (1999d).
The spring edition of the APA's on-line Engineered Wood Journal reports that the expectation of
overall production of structural panels in 1999 would be roughly the same as it was in 1998 (APA,
1999b). However, the long term prospects for the softwood plywood and veneer sector indicates that the
industry is in for a difficult time. APA members are bracing for a battle to preserve market share, a
particularly challenging goal in the face of expected declines in housing starts. Further, the APA's
spring journal focuses on the multiple pressures on its market share. A primary source of pressure is
from the expanding sentiment that wood products are not environmentally sensitive. They are concerned
that environmental advocacy groups are becoming increasingly successful at convincing major
corporations that the use of wood products should be curtailed in order to preserve trees and forested
land (APA, 1999b).
Shipments of reconstituted wood products are expected to increase 4 percent in 1998 and 1999
according to the U.S. Industry and Trade Outlook 1999. Strong demand in the furniture market has
proved beneficial to particleboard, MDF, and hardboard producers. For reconstituted wood products, the
forecast predicts an increase in growth of 3.3 percent per year from 1998 to 2003 as furniture markets
and residential construction remain healthy (U.S. Department of Commerce, International Trade
Administration, 1999).
In their article, "A Look at the Road Ahead for Structural Panels, " authors Spelter and
McKeever compare the situation of the OSB industry in 1996 to that of the MDF industry in the 1970's.
The MDF industry experienced a major upheaval in the 1970's when an economic slump hit the U.S.
right when the industry had added a significant amount of capacity. In this article, Spelter looks at
whether the OSB industry is in danger of experiencing the same process. While the OSB industry's
capacity additions reflect those of the MDF industry, the economic conditions in the late 1990's lead the
2-52
-------
author to conclude that the OSB industry conditions probably will not lead to closures like those
experienced by the MDF sector in the seventies. However, Spelter does not expect that the OSB
producers will continue to enjoy the gains in market share they have experienced over the last 10 to 15
years. He cites the near 100 percent market share held by OSB in the northeast and the Midwest as the
peak of growth opportunity in those markets. Further, the market share split in the south and west may
have stabilized due to the entrenchment of softwood plywood in those areas (Spelter and McKeever,
1996).
At the same time, manufacturers of substitutes for wood-based construction such as steel, cement
and plastic, are aggressively pushing their products hard on the construction industry using the argument
that their products are environmentally friendly, and have advantages in the areas of price stability,
quality, and performance. In-roads by these competing non-wood substitutes are expected to continue as
overall costs for wood-based products continue to climb and the underlying price advantage that wood-
products have traditionally held is undermined. Other concerns expressed the Engineered Wood Journal
include having adequate supply of timber in the long run to meet producers' needs (APA, 1999b).
International Trade
The hope for the plywood and composite wood products export markets is that declining
domestic prices and economic recovery, particularly of the Asian economies, will boost the demand for
U.S. produced wood-based products. This is of particular importance to the softwood plywood
industry, as they are currently exporting approximately 10 percent of their production. Another
international driver of demand for domestically produced wood-based building products is the effect of
regulatory changes in countries such as Japan and Korea to promote wood-based housing construction.
WEFA attributes most of the increases in exports from North America during 1999 to the U.S. rather
than Canada. Continued growth in this market is limited by expected falling housing starts in Japan
(WEFA, 1999). Any future changes in the U.S.-Canadian exchange rate will likely have a positive effect
on exports in the short-term (in the next 2-3 years), as will any modifications to tariff structures in place
for U.S. exports.
The APA outlook includes an international forecast that projects a positive outlook for wood
product exports from 2000 to 2002. This projection is based on expectation that the markets in Europe,
Mexico, South America, and Japan will pick up in 2000, causing a weaker dollar and better overall
climate for exports as (APA, 1999d). The strength of the dollar in 1999 placed U.S. wood products at a
disadvantage in world markets, but APA projects significant increases in exports from 2000 to 2002 (see
Exhibit 2-31 for structural panel export forecast). The 1999 fall edition of the APA's Engineered Wood
Journal pointed to continuing pressures on U.S. exports coming from recent increases in European
production capacity as posing a sizeable challenge to structural wood panel products in the U.S. (APA,
1999c).
The U.S. Industry & Trade Outlook notes growth in European markets and removal of tariff
barriers throughout the world as contributing to modest growth in the wood products industry. At the
same time, the report cautions that economic conditions in Asia, especially Japan, may be of some
concern. While OSB is making significant strides in residential construction in Japan and elsewhere, an
Asian recession could threaten this progress. Softwood plywood is still considered the material of
choice in many markets unfamiliar with OSB. Nontraditional markets such as South America, eastern
Europe, and China could provide significant opportunity for growth in the wood products industry,
especially softwood plywood (U.S. Department of Commerce, International Trade Administration,
1999).
2-53
-------
Prices
The recently published WEFA report on softwood lumber forecasts a 5 percent increase in the
price index for those products during the third quarter of 1999 from the previous quarter. Because of the
expected leveling off or decline in construction, prices are expected to decline during the year's fourth
quarter. Based on WEFA's forecast, overall annual prices in 1999 are expected to be about 8 percent
higher than they were in 1998. Year 2000 prices are forecast to rise only marginally over 1999.
The "Review of the Wood Panel Sector in the U.S. and Canada" presents a forecast for structural
wood panels (softwood plywood and OSB combined). In the 1997 report, Spelter and his co-authors
assume that the long run average annual growth in demand for softwood plywood and OSB combined
will continue at the historical 3 percent rate. Using that assumption, these industries will have excess
capacity until 2001, when capacity utilization reaches 95 percent (Spelter et al., 1997).
This forecast concluded that current and planned production capacity will exceed demand until
2001. This excess capacity will continue to put downward pressure on prices, a trend that began in
1996. The report authors expect that this price pressure will result in a market correction, requiring both
the plywood and the OSB sectors to adjust capacity through the closure of some high cost, low
productivity plants.
2-54
-------
2.6 References
Abt Associates Inc. 1997. Market Profile of the Adhesive Industry. Prepared for Wendy Hoffman,
Regulatory Impacts Branch, Office of Pollution Prevention and Toxics, U.S. Environmental
Protection Agency, 401 M Street, SW, Washington, DC 20460.
American Plywood Association - The Engineered Wood Association. 1999a. Web page information on
product characteristics and manufacturing. Tacoma, WA: APA - The Engineered Wood
Association, http://www.apawood.org/
American Plywood Association - The Engineered Wood Association. 1999b. Engineered Wood
Journal Online. Tacoma, WA: APA - The Engineered Wood Association. Spring.
http://www.apawood.org/news/journal/spring99.html
American Plywood Association - The Engineered Wood Association. 1999c. Engineered Wood Journal
Online. Tacoma, WA: APA - The Engineered Wood Association. Fall.
http://www.apawood.org/news/journal/fall99.html
American Plywood Association - The Engineered Wood Association. 1999d. Market Outlook. Lumber,
Structural Panels, and Engineered Wood Products: 2000 - 2002. Tacoma, WA: APA - The
Engineered Wood Association.
American Plywood Association - The Engineered Wood Association. 1998. Engineered Wood Journal
Online. Tacoma, WA: APA - The Engineered Wood Association. Fall,
http: //www .apawood .org/ne ws/j ournal/fall9 8 .html
American Plywood Association - The Engineered Wood Association. 1997. Design/Construction
Guide: Residential & Commercial. Tacoma, WA: APA — The Engineered Wood Association.
Bentley, Jerome T., C. C. Lawson, and D. Crocker (Mathtech, Inc.) 1992. Economic Impacts of
Alternative NESHAPS on the Wood Furniture Industry. Prepared for U.S. Environmental
Protection Agency, Cost and Economic Impact Section, Emissions Standards Division, office of
Air Quality Planning and Standards. September.
Buongiorno, Joseph. 1996. "Forest sector modeling: a synthesis of econometrics, mathematical
programming, and system dynamics methods. International Journal of Foresting, 12, pp. 329-
343.
CINTRAFOR. 1999. "Asian Crisis Hits PNW: Export Revenues Down 27% Between 1998-1998."
Fact Sheet #29. January. http://www.cintrafor.org/FS29.htm
Columbia Forest Products. 1999. "About our Company." Portland, OR.
www. columbiafore stproducts. com
Composite Panel Association. 1998. 1997 Annual Shipments Report & Downstream Market Survey:
U.S./Canada Particleboard & Medium Density Fiberboard Industry. Gaithersburg, MD:
Composite Panel Association, http://www.pbmdf.com/pubs/shipments97.html
Dirks, John. 1991. Perspectives on Wood Furniture Production, Marketing and Trade: A Survey of
Research Results. CINTRAFOR Special Paper #1. http://www.cintrafor.org/SP01.htm
2-55
-------
Dun & Bradstreet. 1999a. D-U-N-S Number Database. Murray Hill, NJ: Dun & Bradstreet. May 24.
http: //www. dnb. com/dunsno/dunsno .htm
Dun & Bradstreet. 1999b. Industry Norms & Key Business Ratios 1998-1999, Information for SIC
Codes 2435,2436, 2439, and 2493.
Eastin, Ivan, C. Lane, T. Waggener, R. Fight, J Barbour. 1996. An Assessment of the Market for
Softwood Clearwood Lumber Products. CINTRAFOR Working Paper No. 59.
http: //www .cintrafor. org/WP5 9 .htm
Financial Times Limited. 1999a. "Boise Cascade Corporation." Company Briefs/Company Financials.
http://www.ft.com
Financial Times Limited. 1999b. "Georgia-Pacific Corporation." Company Briefs/Company
Financials, http://www.ft.com
Financial Times Limited. 1999c. "Louisiana-Pacific Corporation." Company Briefs/Company
Financials, http://www.ft.com
Financial Times Limited. 1999d. "Willamette Industries, Inc." Company Briefs/Company Financials.
http://www.ft.com
Gale Business Resources. 1999. Online business database. Essays by industry SIC code downloaded
from http://www.gale.com.
Howard, James L. 1999. U.S. Timber Production, Trade Consumption, and Price Statistics 1965-1997.
General Technical Report FPL-GTR-116. Madison, WI: U.S. Department of Agriculture, Forest
Service, Forest Products Laboratory, 76p. July.
International Paper. 1998. "International Paper, Union Camp Corporation Announce Merger." Release,
November 24.
http://www.internationalpaper.com/ca/page/press_releases.cgi?fiincs=links&ID=493957&back=
archive
Louisiana-Pacific Corporation. 1999.
"Headquarters: Product Lines Overview" http://www.lpcorp.com/hq/products.html
"Louisiana-Pacific Corp. Successfully Completes Tender Offer for ABT Building Products
Corp." Release No.: 112-2-9. February 24.
http://www.lpcorp.com/pressrel/19990224_gen_pr.html
"Louisiana-Pacific to Acquire Evans Forest Products, Ltd." Release No.: 138-8-9. August 24.
http://www.lpcorp.com/pressrel/19990824_gen_pr.html
"Louisiana-Pacific Corp. Successfully Completes Tender Offer for Le Groupe Forex." Release
No.: 140-9-9. September 10. http://www.lpcorp.com/pressrel/19990910_gen_pr.html
McKeever, Davis B. 1997. Engineered Wood Products: A Response to the Changing Timber Resource.
Pacific Rim Wood Market Report, November.
Midwest Research Institute. 1999. Memorandum from Becky Nicholson and Melissa Icenhour to Mary
Tom Kissell and Larry Sorrels, U.S. EPA: "Preliminary facility-specific cost estimates for
implementation of the plywood and composite wood products NESHAP." October 20.
2-56
-------
Moody, Russell C. and J. Y. Liu. 1999. "Chapter 11: Glue Structural Members," Wood
handbook—Wood as an engineering material. Gen. Tech. Rep. FPL-GTR-113. Madison, WI:
U.S. Department of Agriculture, Forest Service, Forest Products Laboratory.
PRNewsWire. 1999a.
"Boise Cascade to Acquire Furman." June 15.
"Boise Cascade Reports Sharply Increased Second-Quarter Earnings." July 15.
http: //www .prnewswire. com/gh/cnoc/comp/109525 .html
PRNewsWire. 1999b.
"Georgia-Pacific Group Reports Fourth Quarter and 1998 Results." January 26.
"Georgia-Pacific to Construct Oriented Strand Board Plant in Arkansas." February 23.
"Georgia-Pacific Group Reports Improved First Quarter Earnings and All-Time Quarterly
Record for Building Products." April 22.
"Georgia-Pacific Completes Merger with Unisource Worldwide, Appoints Key Leadership; Sets
Target for Resumption of Share Repurchases." July 6.
http://www.prnewswire.com
PRNewsWire. 1999c.
"Willamette Industries Reports Fourth Quarter Earnings." January 20.
"Willamette Industries Completes Darbo Acquisition." June 22.
"Willamette Industries to Build Particleboard Plant; Name New CFO." August 5
http://www.prnewswire.com/comp/971763 .html
PRNewsWire. 1998. "Willamette Industries Finalizes Purchase of French Plant." March 4.
http: //www .prnewswire. com/gh/cnoc/comp/109525 .html
PR NewsWire. 1997. "Willamette Industries to Acquire Mexico City Box Plant." December 15.
http://www.prnewswire.com/comp/971763 .html
Spelter, Henry. 1998. Substitution, Timber Processing, March.
Spelter, Henry, D. McKeever, and I. Durbak. 1997. Review of Wood-Based Panel Sector in United
States and Canada, General Technical Report FPL-GTR-99. Madison, WI: U.S. Department of
Agriculture, Forest Service, Forest Products Laboratory. 45p.
Spelter, Henry. 1996. "Emerging Nonwood Building Materials in Residential Construction." Forest
Products Journal, 46(7/8).
Spelter, Henry and T. McKeever. 1996. "A Look at the Road Ahead for Structural Panels." C. C. Crow
Publications, Inc., 11(5). October.
Spelter, Henry. 1995. "Capacity Changes in U.S. Particleboard, Southern Pine Plywood, and Oriented
Strandboard Industries." Canadian Journal of Forest Resources, 25: 614-620.
U.S. Bureau of Labor Statistics. 1999. Website, http://stats.bls.gov/ppihome.htm
2-57
-------
U. S. Department of Commerce. 1999a. 1997 Census of Manufactures, Industry Series:
Hardwood Veneer and Plywood, EC97M-3212A
Softwood Veneer and Plywood, EC97M-3212B
Engineered Wood Member (Except Truss) Manufacturing, 1997 Economic Census,
Manufacturing Industry Series, EC97M-3212C
Truss Manufacturing, EC97M-3212D
Reconstituted Wood Products Manufacturing, EC97M-3212E
U. S. Department of Commerce. 1999b. 1997 Census of Manufactures, Industry Series:
Adhesive Manufacturing, EC97M-3255B
Wood Office Furniture Manufacturing, EC97M-3372A
U.S. Department of Commerce. 1997. Current Industrial Reports, Survey of Plant Capacity, 1997.
U.S. Department of Commerce. 1995a. 1992 Census of Construction Industries, Industry Series:
General Contractors — Single-Family Houses, Industry 1521. CC92-I-1. Washington, D.C.:
U.S. Government Printing Office. June.
U. S. Department of Commerce. 1995b. 1992 Census of Manufactures, Industry Series, Millwork,
Plywood, and Structural Wood Members, Not Elsewhere Classified, Industries 2431, 2434,
2435, 2436, and 2439. MC92-I-24B. Washington, D.C.: U.S. Government Printing Office.
U. S. Department of Commerce. 1995c. 1992 Census of Manufactures, Industry Series, Household
Furniture, Industries 2511, 2512, 2514, 2515, 2517, and 2519. MC92-I-25A. Washington,
D.C.: U.S. Government Printing Office.
U. S. Department of Commerce. 1992. Census of Manufactures, Subject Series: General Summary, MC-
92-SC-l.
U.S. Department of Commerce. No date. Census of Manufactures:
Industry 2435, Hardwood Veneer and Plywood
Industry 2436, Softwood Veneer and Plywood
Industry 2439, Structural Wood Members, N.E.C.
Industry 2493, Reconstituted Wood Products
U.S. Department of Commerce, International Trade Administration. 1999. U.S. Industry &Trade
Outlook.
U.S. Department of Commerce, International Trade Administration. 1998. "Outlook Trend Tables."
http://www.ita.doc.gov/industry/otea/usito98/tables.htm.
U.S. Department of Labor, Occupational Safety and Health Administration. No date. SIC descriptions.
http://www.osha.gov/oshstats/sicser.html.
U.S. Environmental Protection Agency. 1998. Industry Specific Information Collection Request (ICR)
for the Development of Plywood and ParticleboardMaximum Achievable Control Technology
(MACT) Standards.
U.S. Environmental Protection Agency. 1995. Lumber and Wood Products, Office of Compliance
Sector Notebook Project.
2-58
-------
U.S. Environmental Protection Agency. 1992. "Economic Impacts of the Alternative NESHAPS on the
Wood Furniture Industry."
U.S. Small Business Administration. 1999. Small Business Size Standards Matched to SIC Codes,
http://www.sba.gov/regulations/siccodes/siccodes.html.
WEFA. 1999. Informational Web site on softwood lumber,
http: //www .we fa. com/news/storie s/99/lumber. cfm
Weyerhaeuser. 1999a. "Weyerhaeuser Completes Acquisition of MacMillan Bloedel; Haskayne to Join
Weyerhaeuser Board of Directors. Release. November 1.
http://www.weyerhaeuser.com/news/mb/default.asp
Weyerhaeuser. 1999b. "Facts About Weyerhaeuser." Tacoma, WA: Weyerhaeuser Company,
http: //www .weyerhaeuser. com/facts/wey er .htm
Willamette Industries, Inc. 1999. "Willamette Industries to Build $85 Million Facility in South
Carolina." Release, http://www.wii.com/newsexp.htm
Youngquist, John A. 1999. "Chapter 10: Wood-Based Composite and Panel Products." Wood
handbook—Wood as an engineering material. Gen. Tech. Rep. FPL-GTR-113. Madison, WI:
U.S. Department of Agriculture, Forest Service, Forest Products Laboratory.
2-59
-------
3 REGULATORY ALTERNATIVES, EMISSIONS, EMISSION
REDUCTIONS, AND CONTROL AND ADMINISTRATIVE COSTS
3.1 Regulatory Alternatives
3.1.1 Background
This chapter provides background information on the regulatory alternatives, emissions,
reductions expected from implementation of the proposed rule, and the costs associated with the
regulatory alternatives.
Section 112 of the Clean Air Act (CAA) requires that EPA establish NESHAP for the control of
hazardous air pollutants (HAP) from both new and existing major sources. A major source of HAP is
defined as any stationary source or group of stationary sources within a contiguous area and under
common control that emits or has the potential to emit, considering controls, in the aggregate, 10 tons per
year or more of any single HAP or 25 tons per year of combined HAP. The CAA requires the NESHAP
to reflect the maximum degree of reduction in emissions of HAP that is achievable. This level of control
is commonly referred to as the maximum achievable control technology, or MACT. The MACT floor is
the minimum control level allowed for NESHAP and is defined in Section 112 (d) (3) of the CAA.
The requirements for new sources are potentially more stringent than those for existing sources.
For new sources, Section 112(d)(3) of the CAA requires EPA to set standards for each category or
subcategory that are at least as stringent as "the emission control that is achieved in practice by the best
controlled similar source." For existing sources, Section 112(d)(3) requires the HAP standards to be no
less stringent than "the average emission limitation achieved by the best-performing 12 percent of the
existing sources" for source categories or subcategories with at least 30 sources and "the average emission
limitation achieved by the best-performing five sources" for source categories or subcategories with fewer
than 30 sources.1
In a previous rulemaking, the EPA promulgated a final rule (59 FR 29196) that presented the
Agency's interpretation of the statutory language regarding the basis of the MACT floor.2 The EPA's
interpretation of the "average emission limitation" is that it means a measure of central tendency, such as
the median. If the median is used when there are at least 30 sources, then the emission level achievable
by the source and its control system that is at the bottom of the top 6 percent of the best-performing
sources (i.e., the 94th percentile) then becomes the MACT floor. For example, assume that there are 100
sources, and HAP emissions from approximately 15 of these sources (15 percent nationwide) are
controlled using thermal oxidizers and the HAP emissions from the remainder of the sources are
uncontrolled. In this example, the 94th percentile is represented by the control system applied to the
source ranked at number 6 (6/100 = 6 percent). However, in this example, the same type of add-on
control technology used by the source at the 94th percentile (thermal oxidizer) is used by sources ranked
below the 94th percentile. Assuming that there are no significant design or operational differences
between the different thermal oxidizers that would affect their performance, all 15 sources equipped with
thermal oxidizers would be considered representative of the MACT floor. Thus, when determining the
performance level of the MACT floor technology, EPA would evaluate the available data for any and all
of the sources equipped with thermal oxidizers.
When there are less than 30 sources, the emission level achievable by the source and its control
system that is the median of the 5 sources represents the MACT floor. For example, if there are 10
sources nationwide and the emissions from 2 of these sources are controlled with thermal oxidizers and
the emissions from the remaining 8 are uncontrolled, then the MACT floor is "no emission reduction"
3-1
-------
because the top 5 sources include the 2 that are controlled, plus 3 that are uncontrolled. In this example,
the median source (the source ranked "number 3") is uncontrolled.
3.1.2 Control Technologies and Practices in MACT Floor Determination
Control systems in use the PCWP industry include add-on control systems and incineration of
process exhaust in an onsite combustion unit. The potential for pollution prevention also exists in the
PCWP industry; however, there are no known and demonstrated pollution prevention techniques that can
be universally applied across the industry. The emissions from PCWP process units are associated with
the wood and/or resin processed. Thus, switching to alternative fuels (e.g., switching from wood fuel to
natural gas) would not significantly reduce emissions and would not be economical for many facilities
that use their wood waste as fuel. Facilities cannot readily switch wood types (e.g., from softwoods to
hardwoods) for several reasons: (1) equipment at each facility is often designed for a particular wood
type; (2) product characteristics would change; and (3) PCWP facilities are located near their wood
source. Over the past decades, the PCWP industry and its resin suppliers have responded to pressure to
reduce the HAP content of resins. It is expected that this trend will continue into the future (e.g., resins
with lower HAP content are likely to be developed). Resin reformulation is a slow, trial-and-error
process that must be completed by individual facilities and their resin suppliers so that product quality is
maintained. At this time, no information is available to determine the degree of emission reduction that
can be achieved through resin reformulation. The achievable emission reduction would be very facility-
specific, and may not be comparable to the emission reduction achievable with add-on control systems
because emissions from the wood would remain. At the present time, few (if any) facilities use pollution
prevention measures to achieve an emission reduction comparable to that of add-on incineration-based
control systems. Therefore, this analysis focuses on add-on control devices.
Available data on control device performance were reviewed to determine which add-on control
systems are best at reducing HAP emissions. Because total hydrocarbons (THC), formaldehyde, and
methanol are the most prevalent pollutants emitted from the PCWP industry and represent the majority of
the available data on control device performance, the control systems were analyzed based on their ability
to reduce emissions of these three pollutants. Although THC is not a HAP, control systems that are
effective in reducing THC emissions are generally effective in reducing HAP emissions.
The available control device performance data for the PCWP industry shows that only two types
of add-on air pollution control devices consistently and continuously reduce HAP emissions: incineration-
based controls (including regenerative thermal oxidizers [RTOs], regenerative catalytic oxidizers [RCOs],
and incineration of pollutants in onsite process combustion equipment [process incineration]) and
biofilters. For control systems that use onsite process combustion equipment (e.g., power boilers or fuel
cells) to reduce emissions, only those systems that route 100 percent of the process unit's exhaust to the
combustion equipment are included in the "incineration-based controls" category. Several of the process
incineration systems are fully integrated systems that combine heat/energy recovery with pollution
control. Systems that only incinerate a portion of the process unit exhaust stream (e.g., less than
75 percent) are referred to as "semi-incineration" and are not included in the incineration-based controls
category.
Those PCWP facilities that practice semi-incineration take a portion of the exhaust stream and
then route these emissions to a burner for use as combustion air. In those situations, the HAP emissions
in the slip stream are actually combusted. However, some facilities with direct-fired dryers (i.e., dryers
that receive hot exhaust air directly from combustion source) that practice semi-incineration may also use
the dryer exhaust gas slip stream (or fresh air) to cool the exhaust gas from the burners in "blend
chambers." 3 When the exhaust gas is routed to the blend chamber, the HAP in the exhaust gas are not
combusted in the dryer, and if the dryer emissions are uncontrolled, these HAP are ultimately emitted to
3-2
-------
the atmosphere. The amount of exhaust gas recycled either to the burner or to the blend chamber can vary
over time. Decisions about how much of the recycled exhaust stream are used as combustion air and
when and how much exhaust air is used in a blend chamber generally are made by the equipment
operators and are affected by process conditions such as the moisture content of the incoming wood
furnish (which affects the target dryer operating temperature) and the desired amount of water removal.4
Thus, semi-incineration is used to maintain the heat balance in the drying system (e.g., combustion unit
and dryer). There is a lack of detailed information on how the semi-incineration process works at each
facility, and thus, the actual HAP emission reductions that are achieved at PCWP facilities that practice
semi-incineration cannot be determined/verified. In addition, it may not be possible to retrofit semi-
incineration onto existing process units, and therefore, semi-incineration may not be an option for process
units that were not originally designed to incorporate semi-incineration. For the reasons stated above and
for the purpose of establishing MACT floors for the PCWP source category, semi-incineration is not
considered a verified control technique for reducing HAP emissions. However, as explained later in
Section IV.B of this memorandum, there are only two process unit groups (bagasse fiberboard mat dryers
and hardwood veneer dryers) where semi-incineration is the only available candidate for the MACT floor
technology.
The available control device efficiency data show that control devices installed for particulate
matter (PM) abatement had no effect on gaseous HAP or THC emissions.5 These control devices include
cyclones, multiclones (or multicyclones), baghouses (or fabric filters), and electrified filter beds (EFBs).
The performance data for wet electrostatic precipitators (WESPs) and wet scrubbers installed for PM
control also showed no effect on HAP and THC emissions. These wet systems may achieve short-term
reductions in THC or gaseous HAP emissions, however, the HAP and THC control efficiency data, which
range from slightly positive to negative values, indicate that the ability of these wet systems to absorb
water-soluble compounds (such as formaldehyde) diminishes as the recirculating scrubbing liquid
becomes saturated with these compounds.5 One wet scrubbing system, a combination water tray
tower/high energy venturi scrubber that uses treated water and is designed to minimize emissions of both
PM and odorous compounds from a hardboard press, did achieve notable HAP and THC emissions
reductions. This system reduces formaldehyde and methanol emissions by 65 percent and 50 percent,
respectively, and reduces THC emissions by 86 percent.5
The THC, methanol, and formaldehyde control device performance data for incineration-based
control and biofilters are presented in the MACT floor memo included in the public docket.6 The
information in this memo was extracted from a separate memorandum which provides information on the
available control device performance data for the various types of control devices applied to PCWP
process units. The performance data for the incineration-based controls and biofilters showed methanol
and formaldehyde emission reductions equal to or greater than 90 percent, except in some cases where the
pollutant loadings of the emission stream entering the control systems were very low. The performance
data for THC showed that incineration-based control systems could achieve THC emission reductions
equal to or greater than 90 percent. The THC emission reductions achieved with biofilters varied
somewhat, with THC reductions ranging from 73 percent to 90 percent. Although biofilters are effective
in reducing the HAP compounds emitted from process units in the PCWP industry, they can be less
effective in reducing some of the less water-soluble non-HAP compounds, such as pinenes, that can make
up a portion of the THC measurements.
The proposed MACT floor technology for the process units was either determined to be no
emission reduction or equivalent to the emission reduction achievable with incineration-based control
systems or biofilters. Although some process units are equipped with add-on controls that perform at a
level somewhere between zero (no control) and the performance level achievable with incineration-based
controls and biofilters, none of these control systems were identified as MACT floor control technologies
because they (1) do not reduce HAP emissions (e.g., bag houses) or (2) do not reduce HAP emissions on
3-3
-------
a consistent basis (e.g., wet electrostatic precipitators), or (3) achieve lower HAP emission reductions
than biofilters and incineration-based controls (e.g., semi-incineration). Therefore, the MACT floor
analysis focused on incineration-based controls and biofilters.
For the purpose of establishing the performance level of the MACT floor control systems, all
available data on incineration-based controls and biofilters were grouped together. This "group
approach" was used because some of the control systems treat HAP emissions from multiple types of
process units, such as primary tube dryers, reconstituted panel presses, and board coolers.
Determinations of the performance of the control system on emissions from each type of process
unit were not possible. Also, for some process unit groups, limited data were available for the control
systems applied to the process units in that group. By considering all of the performance data for
incineration-based controls and biofilters together, the amount of available data upon which the MACT
floor level of performance was based was maximized.
The available data for incineration-based controls and biofilters (provided in the MACT floor
memo)6 shows variability in performance from process unit to process unit and over time. In some cases,
it was not possible to directly compare the performance of different control systems because data were not
available for the same pollutant (i.e., not all test reports included data for THC, methanol, and
formaldehyde). Comparison of the performance of the different types of incineration-based control
systems with other incineration-based controls and with biofilters was also hampered by the fact that the
uncontrolled emissions being treated by the different control systems varied with respect to pollutant
loading (inlet concentration) and pollutant type. Because the control device efficiency is somewhat
dependent on the amount of HAP entering the device, the variability in the uncontrolled emissions from
process units both within and among the different process groups meant that the control device
efficiencies also varied. With a few exceptions, when the concentration of methanol, formaldehyde, or
THC in the uncontrolled emission stream was greater than 10 parts per million dry volume (ppmvd), the
associated HAP emission reductions ranged from 90 to 99 percent. In general, lower control efficiencies
were achieved when the inlet pollutant concentrations were below 10 ppmvd; however, in some cases, the
control efficiency exceeded 90 percent even at the lower inlet concentrations.
To account for the variability in the type and amount of HAP in the uncontrolled emissions from
the various process units and the effect of this variability on control system performance, it is
recommended that the MACT floor performance level be based on all three of the pollutants analyzed and
include maximum concentration levels in the outlet of the control systems as an alternative to emission
reductions. The MACT floor performance level is a 90 percent reduction in THC or methanol or
formaldehyde emissions. The maximum concentration level in the outlet of the MACT floor control
system is 20 ppmvd for THC, or 1 ppmvd for methanol, or 1 ppmvd for formaldehyde. The 20 ppmvd is
recommended as the alternative maximum concentration for THC because 20 ppmvd represents the
practical limit of control for THC. The 1 ppmvd is recommended as the maximum outlet concentration
for both methanol and formaldehyde because this concentration is achievable by the MACT floor control
systems and the method detection limits for these compounds using the National Council of the Paper
Industry for Air and Stream Improvement (NCASI) impinger/canister emission test method (NCASI
Method IM/CAN/WP-99.01) are less than 1 ppmvd.7
3.1.3 MACT Floor Options
The six recommended options for representing the MACT floor are shown in Exhibit 3-1. These
six options reflect the emission reductions and maximum outlet pollutant concentrations achievable at the
MACT floor for all process units with a MACT floor technology represented by incineration-based
controls or biofilters. As shown in Exhibit 3-1, it is recommended that a restriction be placed on the use
3-4
-------
of the outlet concentration options for methanol and formaldehyde. The proposed restriction would be
that the concentration of the pollutant (methanol or formaldehyde) entering the MACT control system
must be at least 10 ppmvd for the facility to use the outlet concentration option. The purpose for this
restriction is that some process units may have very low uncontrolled methanol or formaldehyde
emissions, while still emitting significant quantities of HAP, and facilities with these process units could
claim that they are achieving MACT floor levels of control without doing anything to reduce HAP
emissions. All of the MACT floor control systems evaluated can meet at least one of the six control
options for add-on control devices, based on the available data. Only a few of the MACT floor control
systems evaluated can meet all six options; in those cases, the control systems tend to be applied to
process units with both moderately high HAP emissions and moderately high THC emissions, which
would allow them to meet the outlet concentration-based options for methanol and formaldehyde as well
as the percent reduction options. Therefore, it is recommended that facilities be required to meet only one
of the six emission options in Exhibit 3-1.
Exhibit 3-1. MACT FLOOR CONTROL OPTIONS
Pollutant
Reduce by
OR achieve emissions <
methanol
90 percent
1 ppm3
...OR...
formaldehyde
90 percent
1 ppm3
...OR...
THCh
90 percent
20 ppm
"This option would only be applicable to units with uncontrolled emissions of that HAP that are > 10 ppm.
bMills will be allowed to adjust THC measurements to subtract methane.
3.1.4 Summary of MACTfor Existing and New Process Units
Exhibit 3-2 summarizes MACT for each type of process unit at new and existing PCWP
facilities. The MACT represents the level of control that would be required by the PCWP NESHAP. The
technologies listed below achieve that control level.
3-5
-------
Exhibit 3-2. SUMMARY OF MACT FOR PCWP PROCESS UNITS
AT NEW AND EXISTING SOURCES
Process unit
MACT for process units at existing
sources
MACT for process units at new
sources
Tube dryers; Rotary strand
dryers; Conveyor strand dryers;
Green particle rotary dryers;
Hardboard ovens; Softwood
veneer dryers; Pressurized
refiners
emission reduction achievable with
incineration-based control1
emission reduction achievable with
incineration-based control1
Reconstituted wood product
presses
emission reduction achievable with
incineration-based control1 or
biofilter
emission reduction achievable with
incineration-based control1 or
biofilter
Fiberboard mat dryers (wood);
Hardboard press preheat ovens
No emission reduction
emission reduction achievable with
incineration-based control1
Reconstituted wood product
board coolers
No emission reduction
emission reduction achievable with
incineration-based control1 or
biofilter
Rotary agricultural fiber dryers;
Dry particle rotary dryers;
Paddle-type particle dryers;
Hardboard humidifiers
Fiberboard mat dryers (bagasse);
Veneer kilns; Radio-frequency
veneer redryers; Hardwood
veneer dryers; Particleboard press
molds; Particleboard extruders;
Engineered wood products
presses; Agriboard presses;
Plywood presses; Stand-alone
digesters; Atmospheric refiners;
Blenders; Formers; Sanders
Saws; Fiber washers; Chippers;
Log vats; Lumber kilns
No emission reduction
No emission reduction
" Incineration-based control includes RTOs, RCOs, TCOs, TOs, and incineration of process exhaust in combustion
unit.
3.1.5 Beyond the MACT Floor Options and Related Technologies
Because the control devices that represent MACT levels of control are the same for all process
units that have a controlled MACT floor for both new and existing units, the only beyond the floor
options considered were for existing process unit groups that had MACT floors equal to "no emission
reduction." The annual total HAP emissions from the following equipment are very low compared to the
emissions from other process units used in the PCWP industry:
• agriboard dryers • particleboard press molds
• dry particle rotary dryers • particleboard extruders
• paddle-type particle dryers • engineered wood products presses
3-6
-------
hardboard humidifiers
bagasse fiber mat dryer
veneer kilns
agriboard presses
atmospheric refiners
lumber kilns
• RF veneer redryers
• hardwood veneer dryers
• resin storage tanks
• other miscellaneous equipment (formers,
sanders, saws, fiber washers, chippers, and log
vats)
No beyond-the-floor control options were considered for these equipment because emissions
from these process units would not be cost-effective to control. In addition, no beyond-the-floor analyses
of wastewater operations, wastewater tanks, and miscellaneous coating operations were conducted
because sufficient information is not available to make beyond-the-floor determinations and it is not
known (or expected) that emissions from these operations would justify control.
Based on a review of the HAP emissions data for process units with MACT floors of no emission
reduction, blenders and stand-alone digesters were selected for a beyond-the-MACT-floor analysis
because these equipment emit higher levels of HAP emissions relative to other process units. Beyond-
the-floor analyses were also conducted for process units with a MACT floor of no emission reduction for
existing units and a MACT floor represented by the emission reduction achievable with incineration-
based controls for new units. These process units include flberboard mat dryers, press preheat ovens, and
reconstituted wood products board coolers.
This analysis of beyond-the-floor options was based on the industry average exhaust flow for
each process unit, the typical number of each process unit per plant, the industry average amount of HAP
emitted from the process units, and assuming that an RTO would be used to control emissions from each
process unit. The average exhaust flow rates and typical number of process units per plant were
determined using the results from EPA's MACT survey. The average HAP emissions and emission
reductions were determined using the methodology described in the baseline emissions memo.9 The
annualized RTO cost was calculated based on flow rate using the methodology described later in this
chapter. The cost per ton values were calculated by dividing the total annualized cost (TAC) for the RTO
by the HAP reduction. This analysis assumes that facilities will not be able to route the emissions from
process units to an existing control device or to a new control device installed to meet the PCWP
standards (i.e., that a separate RTO must be purchased to handle the additional flow from the process
units). Exhibit 3-3 below presents the results of this analysis.
3-7
-------
Exhibit 3-3. Cost-Effectiveness Analysis Of Beyond-The-Floor Control Options
Process Unit
Average
flow, dscfm
Typical
no.
per plant
RTO
TAC
Average
HAP
emitted, tpy
Tons HAP
reduced, tpy
Cost
effectiveness
$/ton (1998
dollars)
Fiberboard mat dryer at FB
plant
49,389
1
$471,187
8
7.6
Fiberboard mat dryer at w/d
HB plant
19,491
1
$370,952
12
11
Fiberboard mat dryers
(average)
34,440
1
$421,070
10
9.3
$30,076
Press preheat oven - w/d HB
plant
21,812
1
$377,904
15
14
$26,520
Board cooler - PB
41,423
1
$442,096
5
4.8
Board cooler - MDF
79,483
1
$599,447
3
2.9
Board cooler (average)
60,453
1
$520,772
4
3.9
SI i\5^1
Stand-alone digester - FB
7,587
2
$358,359
14
13
Stand-alone digester - HB
7,587
2
$358,359
14
13
Stand-alone digester (average)
7,587
2
$358,359
14
13
$26,944
Blender - PB
13,590
2
$394,486
45
43
Blender - OSB
13,590
2
$394,486
11
10
Blender (average)
13,590
2
$394,486
28
27
$14,610
In all cases, the emission reductions that could be achieved from requiring controls for these
existing units did not appear to be justified by the cost. Many of the existing control devices at well-
controlled facilities would not have the additional capacity to treat the emissions from these process units,
and thus, these facilities would have to install new controls.
For more information, refer to the MACT memo for this proposed rule and the BID.8
3.1.6 Considerations of Possible Risk-Based Alternatives to Reduce Impacts to Sources
The Agency has made every effort in developing this proposal to minimize the cost to the
regulated community and allow maximum flexibility in compliance options consistent with our statutory
obligations. However, we recognize that the proposal may still require some facilities to take costly steps
to further control emissions even though their emissions may not result in exposures which could pose an
excess individual lifetime cancer risk greater than one in one million or which exceed thresholds
determined to provide an ample margin of safety for protecting public health and the environment from the
effects of hazardous air pollutants. We are, therefore, specifically soliciting comment on whether there are
further ways to structure the proposed rule to focus on the facilities which pose significant risks and avoid
the imposition of high costs on facilities that pose little risk to public health and the environment.
Industry representatives provided EPA with descriptions of three mechanisms that they believed
could be used to implement more cost-effective reductions in risk. The docket for today's proposed rule
contains "white papers" prepared by industry that outline their proposed approaches (see docket number
A-98-44, Item # II-D-525). The Agency is taking comment on these approaches. We believe that two of
3-8
-------
the three suggested approaches warrant further consideration. We believe they could be used to focus
regulatory controls on facilities with significant risks and avoid the imposition of high costs on facilities
that pose little risk to public health or the environment. One of the approaches, an applicability cutoff for
threshold pollutants, would be implemented under the authority of CAA section 112(d)(4); the other
approach, subcategorization and delisting, would be implemented under the authority of CAA sections
112(c)(1) and 112(c)(9). The EPA requests comment on the technical and legal viability of these
approaches, as well as any modifications to these approaches that commenters may wish to suggest. The
maximum achievable control technology, or MACT, program outlined in CAA section 112(d) is intended
to reduce emissions of HAP through the application of MACT to major sources of toxic air pollutants.
Section 112(c)(9) is intended to allow EPA to avoid setting MACT standards for categories or
subcategories of sources that pose little risk to public health and the environment. The EPA requests
comment on whether the proposals described here appropriately rely on these provisions of CAA section
112. While both approaches focus on assessing the inhalation exposures of HAP emitted by a source, EPA
specifically requests comment on the appropriateness and necessity of extending these approaches to
account for non-inhalation exposures of certain HAP which may deposit from the atmosphere after being
emitted into the air or to account for adverse environmental impacts. We are also interested in any
information or comment concerning technical limitations, environmental and cost impacts, compliance
assurance, legal authority, and implementation relevant to the approaches. We also request comment on
appropriate practicable and verifiable methods to ensure that sources' emissions remain below levels that
protect public health and the environment. We will evaluate all comments before determining whether
either of the two approaches will be included in the final rule.
3.1.6.1 Applicability Cutoffs for Threshold Pollutants Under Section 112(d)(4) of the CAA
The first approach is an "applicability cutoff for threshold pollutants that is based on EPA's
authority under CAA section 112(d)(4). A "threshold pollutant" is one for which there is a concentration
or dose below which adverse effects are not expected to occur over a lifetime of exposure. For such
pollutants, section 112(d)(4) allows EPA to consider the threshold level, with an ample margin of safety,
when establishing emissions standards. Specifically, section 112(d)(4) allows EPA to establish emission
standards that are not based upon the maximum achievable control technology (MACT) specified under
section 112(d)(2) for pollutants for which a health threshold has been established. Such standards may be
less stringent than MACT. Furthermore, EPA has interpreted 112(d)(4) to allow us to avoid further
regulation of categories of sources that emit only threshold pollutants, if those emissions result in ambient
levels that do not exceed the threshold, with an ample margin of safety.10
A different interpretation would allow us to exempt individual facilities within a source category
that meet the section 112(d)(4) requirements. There are three potential scenarios under this interpretation
of the section 112(d)(4) provision. One scenario would allow an exemption for individual facilities that
emit only threshold pollutants and can demonstrate that their emissions of threshold pollutants would not
result in air concentrations above the threshold levels, with an ample margin of safety, even if the category
is otherwise subject to MACT. A second scenario would allow the section 112(d)(4) provision to be
applied to both threshold and non-threshold pollutants, using the 1 in a million cancer risk level for
decisionmaking for non-threshold pollutants. A third scenario would allow a section 112(d)(4) exemption
at a facility that emits both threshold and non-threshold pollutants. For those emission points where only
threshold pollutants are emitted and where emissions of the threshold pollutants would not result in air
concentrations above the threshold levels, with an ample margin of safety, those emission points could be
exempt from the MACT standard. The MACT standard would still apply to non-threshold emissions from
1 See 63 FR 18754, 18765-66 (April 15, 1998) (Pulp and Paper Combustion Sources Proposed
NESHAP)
3-9
-------
other emission points at the source. For this third scenario, emission points that emit a combination of
threshold and non-threshold pollutants that are co-controlled by MACT would still be subject to the
MACT level of control. However, any threshold HAP eligible for exemption under section 112(d)(4) that
are controlled by control devices different from those controlling non-threshold HAP would be able to use
the exemption, and the facility would still be subject to the parts of the standard that control non-threshold
pollutants or that control both threshold and non-threshold pollutants.
Under the section 112(d) (4) approach, EPA would have to determine that emissions of each of
the threshold pollutants emitted by PCWP sources at the facility do not exceed the threshold levels, with
an ample margin of safety. The common approach for evaluating the potential hazard of a threshold air
pollutant is to calculate a "hazard quotient" by dividing the pollutant's inhalation exposure concentration
(often assumed to be equivalent to its estimated concentration in air at a location where people could be
exposed) by the pollutant's inhalation Reference Concentration (RfC). An RfC is defined as an estimate
(with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation exposure that, over a
lifetime, likely would not result in the occurrence of adverse health effects in humans, including sensitive
individuals. The EPA typically establishes an RfC by applying uncertainty factors to the critical toxic
effect derived from the lowest- or no-observed-adverse-effect level of a pollutant.11 A hazard quotient less
than one means that the exposure concentration of the pollutant is less than the RfC, and, therefore,
presumed to be without appreciable risk of adverse health effects. A hazard quotient greater than one
means that the exposure concentration of the pollutant is greater than the RfC. Further, EPA guidance for
assessing exposures to mixtures of threshold pollutants recommends calculating a "hazard index" by
summing the individual hazard quotients for those pollutants in the mixture that affect the same target
organ or system by the same mechanism.12 Hazard index (HI) values would be interpreted similarly to
hazard quotients; values below one would generally be considered to be without appreciable risk of
adverse health effects, and values above one would generally be cause for concern. For the determinations
discussed herein, EPA would generally plan to use RfC values contained in EPA's toxicology database,
the Integrated Risk Information System (IRIS). When a pollutant does not have an approved RfC in IRIS,
or when a pollutant is a carcinogen, EPA would have to determine whether a threshold exists based upon
the availability of specific data on the pollutant's mode or mechanism of action, potentially using a health
threshold value from an alternative source such as the Agency for Toxic Substances and Disease Registry
(ATSDR) or the California Environmental Protection Agency (CalEPA).
In the past, EPA routinely treated carcinogens as non-threshold pollutants. The EPA recognizes
that advances in risk assessment science and policy may affect the way EPA differentiates between
threshold and non-threshold HAP. The EPA's draft Guidelines for Carcinogen Risk Assessment13 suggest
that carcinogens be assigned non-linear dose-response relationships where data warrant. Moreover, it is
possible that dose-response curves for some pollutants may reach zero risk at a dose greater than zero,
creating a threshold for carcinogenic effects. It is possible that future evaluations of the carcinogens
emitted by this source category would determine that one or more of the carcinogens in the category is a
threshold carcinogen or is a carcinogen that exhibits a non-linear dose-response relationship but does not
have a threshold.
11 "Methods for Derivation of Inhalation Reference Concentrations and Applications of Inhalation
Dosimetry." EPA-600/8-90-066F, Office of Research and Development, USEPA, October 1994.
12 "Supplementary Guidance for Conducting Health Risk Assessment of Chemical Mixtures. Risk
Assessment Forum Technical Panel," EPA/630/R-00/002. USEPA, August 2000.
http://www.epa.gov/nceawwwl/pdfs/chem mix/chem mix 08 2001.pdf
13 "Draft Revised Guidelines for Carcinogen Risk Assessment." NCEA-F-0644. USEPA, Risk
Assessment Forum, July 1999. pp 3-9ff. http://www.epa.gov/ncea/raf/pdfs/cancer gls.pdf
3-10
-------
There are at least several options for establishing a hazard index limit for the section 112(d)(4)
analysis that reflect to varying degrees total public exposure. One option is to allow the hazard index
posed by all threshold HAP emitted by PCWP sources at the facility to be no greater than one. This
approach assumes that no additional threshold HAP exposures would be anticipated from other sources in
the vicinity or through other routes of exposure (i.e., through ingestion).
A second option is to adopt a "default percentage" approach, whereby the hazard index limit of the
HAP emitted by the facility is set at some percentage of one (e.g., 20% or 0.2). This approach recognizes
the fact that the facility in question is only one of many sources of threshold HAP to which people are
typically exposed every day. Because noncancer risk assessment is predicated on total exposure or dose,
and because risk assessments to focus only on an individual source, establishing a hazard index limit of 0.2
would account for an assumption that 20% of an individual's total exposure is from that individual source.
For the purposes of this discussion, we will call all sources of HAP, other than the facility in question,
"background" sources. If the facility is allowed to emit HAP such that its own impacts could result in HI
values of one, total exposures to threshold HAP in the vicinity of the facility could be substantially greater
than one due to background sources, and this would not be protective of public health, since only HI
values below one are considered "safe" (i.e., without appreciable risk of harmful effects). Thus, setting the
hazard index limit for the facility at some default percentage of one will provide a buffer which would help
to ensure that total exposures to threshold HAP near the facility (i.e., in combination with exposures due to
background sources) will generally not exceed one, and can generally be considered to be without
appreciable risk of adverse health effects. The EPA requests comment on using the "default percentage"
approach and on setting the default hazard index limit at 0.2. The EPA is also requesting comment on
whether an alternative HI limit, in some multiple of 1, would be a more appropriate applicability cutoff.
A third option is to use available data (from scientific literature or EPA studies, for example) to
determine background concentrations of HAP, possibly on a national or regional basis. These data would
be used to estimate the exposures to HAP from non-PCWP sources in the vicinity of an individual facility.
For example, the EPA's National-scale Air Toxics Assessment (NATA)14 and ATSDR's Toxicological
Profiles15 contain information about background concentrations of some HAP in the atmosphere and other
media. The combined exposures from PCWP sources and from other sources (as determined from the
literature or studies) would then not be allowed to exceed a hazard index limit of one. -
As an alternative to the third option, a fourth option is to allow facilities to estimate or measure
their own facility-specific background HAP concentrations for use in their analysis.
3.1.6.3 Subcategory Delisting Under Section 112(c)(9)(B) of the CAA
EPA is authorized to establish categories and subcategories of sources, as appropriate, pursuant to
CAA section 112(c)(1), in order to facilitate the development of MACT standards consistent with section
112 of the CAA. Further, section 112(c)(9)(B) allows EPA to delete a category (or subcategory) from the
list of major sources for which MACT standards are to be developed when the following can be
demonstrated: 1) in the case of carcinogenic pollutants, that "no source in the category . . . emits
[carcinogenic] air pollutants in quantities which may cause a lifetime risk of cancer greater than one in one
million to the individual in the population who is most exposed to emissions of such pollutants from the
source"; 2) in the case of pollutants that cause adverse noncancer health effects, that "emissions from no
source in the category or subcategory . . . exceed a level which is adequate to protect public health with an
14 See http://www.epa.gov/ttn/atw/nata
15 See http ://www. atsdr. cdc. gov/toxpro2.html
3-11
-------
ample margin of safety"; and 3) in the case of pollutants that cause adverse environmental effects, that "no
adverse environmental effect will result from emissions from any source."
Given these authorities and the suggestions from the white paper prepared by industry
representatives (see docket number A-98-44, Item # II-D-525), EPA is considering whether it would be
possible to establish a subcategory of facilities within the larger PCWP category that would meet the risk-
based criteria for delisting. Since each facility in such a subcategory would be a low-risk facility (i.e., if
each met these criteria), the subcategory could be delisted in accordance with section 112(c)(9), thereby
limiting the costs and impacts of the proposed MACT rule to only those facilities that do not qualify for
subcategorization and delisting. Facilities seeking to be included in the delisted subcategory would be
responsible for providing all data required to determine whether they are eligible for inclusion. Facilities
that could not demonstrate that they are eligible to be included in the low-risk subcategory would be
subject to MACT and possible future residual risk standards.
Establishing that a facility qualifies for the low-risk subcategory under section 112(c)(9) will
necessarily involve combining estimates of pollutant emissions with air dispersion modeling to predict
exposures. The EPA envisions that we would promote a tiered analytical approach for these
determinations. A tiered analysis involves making successive refinements in modeling methodologies and
input data to derive successively less conservative, more realistic estimates of pollutant concentrations in
air and estimates of risk.
As a first tier of analysis, EPA could develop a series of simple look-up tables based on the results
of air dispersion modeling conducted using conservative input assumptions. By specifying a limited
number of input parameters, such as stack height, distance to property line, and emission rate, a facility
could use these look-up tables to determine easily whether the emissions from their sources might cause a
hazard index limit to be exceeded.
A facility that does not pass this initial conservative screening analysis could implement
increasingly more site-specific but more resource-intensive tiers of analysis using EPA-approved modeling
procedures, in an attempt to demonstrate that their facility does not exceed the hazard index limit. The
EPA's guidance could provide the basis for conducting such a tiered analysis.16
Another approach would be to define a subcategory of facilities within the PCWP source category
based upon technological differences, such as differences in production rate, emission vent flow rates,
overall facility size, emissions characteristics, processes, or air pollution control device viability. If it
could then be determined that each source in this technologically-defined subcategory presents a low risk
to the surrounding community, the subcategory could then be delisted in accordance with 112(c)(9).
One concern that EPA has with respect to the section 112(c)(9) approach is the affect that it could
have on the MACT floors. If all of the well-controlled, low-risk facilities are subcategorized, that could
make the MACT floor less stringent for the remaining facilities. One approach that has been suggested to
mitigate this effect would be to establish the MACT floor now based on controls in place for the category
and to allow facilities to become part of the low-risk category in the future, after the MACT standard is
established. This would allow low risk facilities to use the 112(c)(9) exemption without affecting the
MACT floor calculation. EPA requests comment on this suggested approach.
16"A Tiered Modeling Approach for Assessing the Risks due to Sources of Hazardous Air
Pollutants." EPA-450/4-92-001. David E. Guinnup, Office of Air Quality Planning and Standards,
USEPA, March 1992.
3-12
-------
If this section 112(c)(9) approach were adopted, the rulemaking would likely indicate that the rule
does not apply to any source that demonstrates, based on a tiered approach that includes EPA-approved
modeling of the affected source's emissions, that it belongs in a subcategory which has been delisted under
section 112(c)(9).
3.2 Emissions and Emission Reductions
As mentioned in Chapter 1, the U.S. Environmental Protection Agency (EPA) is developing
national emission standards for hazardous air pollutants (NESHAP) for the plywood and composite wood
products source category. This part of the RIA presents emission reductions expected to occur from
compliance with the MACT floor alternative that is being proposed.
3.2.1 Some Results in Brief
The proposed plywood and composite wood products NESHAP will reduce HAP emissions by
about 11,000 tons in the third year after its issuance.9 The major HAP reduced, as mentioned in the rule
preamble, are acrolein, acetaldehyde, formaldehyde, phenol, propionaldehyde, and methanol. In addition,
nearly 27,000 tons of VOC (reported as total hydrocarbon) emissions will be reduced. Nearly 11,000
tons of CO emission reductions will occur, along with 13,000 tons of PM (coarse) emission reductions.
There will also be 5,000 tons of additional NOx emissions and 4,000 tons of S02 emissions added to the
atmosphere due to the additional incineration-based controls that may be necessary for affected facilities
to meet the MACT floor alternative.
3.2.2 Gen eral Approach
The methodology used to estimate the HAP emission reductions associated with this proposed
rule is summarized in this section. Before the emissions reductions could be estimated, the baseline
emissions level for each pollutant had to be determined. This was conducted by first estimating emissions
without considering current air pollution controls, and then calculating the emissions levels with current
controls applied. The first, uncontrolled emission estimates, are developed without consideration of air
pollution controls currently in use at wood products plants. Baseline estimates reflect the level of
pollution control that is presently used. The remainder of this section discusses the general methodology
used to estimate uncontrolled and baseline emissions.
Estimating uncontrolled and baseline emissions involves the following four steps:
(1) Identification of hazardous air pollutant (HAP) emission sources,
(2) Characterization of emission sources (e.g., assignment of throughput and other
characteristics),
(3) Application of emission factors, and
(4) Calculation of emissions.
3.2.2.1 Identifying Emission Sources
Emission sources were identified based on responses to the EPA's maximum achievable control
technology (MACT) surveys and available emissions test data. The EPA gathered plant-specific
information with three MACT surveys. The results of the three surveys are documented in separate
memoranda.10'11'12 Available emissions test data include data from nearly 100 test reports collected
through EPA's MACT survey, data from EPA's Compilation of Air Pollutant Emission Factors,
Volume I: Stationary Point and Area Sources (commonly referred to as AP-42), and extensive data from
3-13
-------
the industry-sponsored test program performed by the National Council of the Paper Industry for Air and
Stream Improvement (NCASI). Emissions were estimated for sources that were identified in the EPA's
MACT survey (e.g., dryers, presses, board coolers) and for additional miscellaneous sources (e.g.,
digesters, refiners, fiber washers) for which applicable emission test data were available. Process flow
diagrams submitted with the MACT surveys provided information about the presence (or absence) of
miscellaneous sources at plants. The EPA's MACT survey also provided information on the control
devices used for most unit operations.
Some plants have begun operation and other plants have added equipment and controls since
EPA conducted the MACT survey. Such changes have been accounted for in the nationwide emission
estimates. A separate memorandum summarizes the changes to plants that have occurred following the
EPA's MACT survey.13
3.2.2.2 Characterizing Emission Sources
After the emission sources were identified, each source was assigned a throughput and further
characterized. In most cases, plant-specific dryer and press throughput was provided in EPA's MACT
survey. If the dryer or press throughput was claimed confidential or was not provided, then a default
throughput was assigned. For dryers, the default throughput was the average throughput for the same
type dryers for the product manufactured. 101112 If available, plant production (or capacity if production
was unavailable) was used as the default throughput for presses; otherwise, presses were assigned the
average press throughput for the product manufactured.
Throughput for miscellaneous equipment was based on either dryer or press throughput,
depending on the units of measure for the applicable emission factor (e.g., pounds per oven dry ton
[lb/ODT] or pounds per thousand square feet [lb/MSF]). Collective throughputs for digesters, refiners,
fiber washers, blenders, and formers were generally approximated as the total dryer throughput (ODT/yr)
for the plant. Board cooler, sander, and saw throughput was generally approximated as the total press
throughput (MSF/yr) for the plant. By assigning either the total dryer or press throughput to
miscellaneous processes, no assumption about the number of miscellaneous operations (e.g., number of
refiners) at each plant was necessary.
Following assignment of throughput for each emission source, sources were further characterized
(as necessary for application of emission factors) based on wood species, resin type, or other
characteristics. Further characterization of emission sources is discussed in the baseline emissions memo
for this proposed rule. If characterization of an emission source was not possible due to claims of
confidentiality or missing information in the survey response, then default characterizations were applied
based on practices most commonly observed at other plants.
3.2.2.3 Applying Emission Factors
As emission sources were characterized, the available emission factors were reviewed for
applicability to each emission source. The emission factors used in developing the nationwide estimates
are documented in a separate memorandum.14 Emissions were generally estimated for total hydrocarbon
as carbon (THC as C , referred to as "THC" in this RIA) and the following HAP's:
acetaldehyde
acrolein
benzene
methyl isobutyl ketone (MIBK)
phenol
propionaldehyde
styrene
cumene
3-14
-------
formaldehyde toluene
methanol m,p-xylene
methylene chloride o-xylene
methyl ethyl ketone (MEK)
When used in this document, the terminology "complete set" refers to the set of emission factors
for the above list of HAP's, THC, and any other additional HAP's that may have been measured at levels
above a test method detection limit.
Total HAP was taken to be the sum of the individual HAP's. If all test data for a pollutant at a
source were below the test method detection limit (denoted as "BDL" in reference 5), then the emission
factor was treated as zero. For a few specific sources, HAP's other than the ones listed above were tested
for and detected. These HAP were included in the total HAP estimate for that source. Exhibit 3-4
illustrates how the total HAP emission factors were developed.
Exhibit 3-4. ILLUSTRATION OF TOTAL HAP CALCULATION
FOR AN EMISSION SOURCE
HAP
Emission Factor (from reference 5)
acetaldehyde
0.0012
acrolein
BDL3
benzene
BDL
cumene
BDL
formaldehyde
0.015
methanol
0.076
methylene chloride
BDL
MEK
BDL
MIBK
BDL
phenol
0.0047
propionaldehyde
BDL
styrene
BDL
toluene
BDL
m,p-xylene
BDL
o-xylene
BDL
Total HAP
0.0012 + 0.015 +0.076 + 0.0047 = 0.097
" BDL (below detection limit); all test runs for this pollutant and this source were below the test method detection
limit.
Test data were not available for all of the HAP's considered for some sources (i.e., a complete set
of emission factors was not available). In some cases, it was necessary to apply emission factors for one
source to a similar source for which factors were not available. In other situations, it was necessary to
3-15
-------
group emission factors so that emissions of all the likely pollutants could be estimated for a particular
emission source. Grouped emission factors were calculated from emission test averages using the
methodology described in the emission factor memo14. Specific application of the emission factors for
each unit operation is discussed in the baseline emissions memo for this proposed rule.
3.2.2.4 Calculating Emissions
Applicable emission factors were used to estimate uncontrolled emissions from each unit
operation as follows:
E = EF x T / 2000
where:
E = annual emissions (ton/yr)
EF = emission factor (lb/ODT or lb/MSF-specified basis)
T = process throughput (ODT/yr or MSF/yr-specified basis)
To estimate baseline emissions, the emission reduction achieved by air pollution control devices
(APCD's) in place on unit operations was taken into account. Control devices that achieve significant
reduction of HAP and THC include biofilters and incineration-based controls (e.g., regenerative thermal
oxidizers [RTO's], regenerative catalytic oxidizers [RCO's], thermal oxidizers [TO's], and thermal
catalytic oxidizers [TCO's]). Emission factors were available for several, but not all, of the unit
operations that are presently controlled with biofilters and incineration-based controls. If a complete set
of emission factors based on inlet and outlet test data for a single control device was available, the set of
emission factors was used to estimate baseline emissions. Otherwise, the achievable percent reduction in
emissions for the control device was used as follows:
E = EF x T / 2000 x (1-R)
where:
R = percent reduction achievable with the control device (see table below)
Control device
HAP reduction
THC reduction
Biofilter
95%
80%
RTO, RCO, TO, & TCO
95%
95%
If only a portion of an exhaust stream was controlled (e.g., as with semi-incineration where only a
portion of the exhaust is routed to a combustion unit), then controlled emissions were estimated for the
controlled portion of the exhaust and uncontrolled emissions were estimated for the remaining exhaust.
Because plant-specific capture efficiency information is not readily available, presses and board coolers
without a permanent total enclosure that are routed to an APCD were assumed to operate with 50 percent
capture efficiency.
Following estimation of uncontrolled and baseline emissions for each unit operation, annual
emissions for each facility were totaled. If there were facilities with no available information was
available (e.g., plants that claimed their entire MACT survey confidential or plants that never responded
to the survey), then the average facility-specific emissions for plants making the same product was used
3-16
-------
to approximate the emissions from plants with no available information. The emissions from all facilities
were summed to obtain nationwide emission estimates.
3.2.3 Nationwide HAP Emission Estimates
Exhibit 3-5 presents the total nationwide uncontrolled and baseline emission estimates for each
product type. Uncontrolled emissions are the emissions that occur before the application of a HAP
emission control device. Baseline emission estimates take into account the HAP emission controls
currently in place on HAP sources in the industry. Exhibits 3-6 and 3-7 present the nationwide
uncontrolled and baseline speciated HAP emission estimates for each product type.
To estimate the emissions and emission reductions associated with compliance with this proposed
rule, it is necessary to determine the number of major sources that may be subject to the rule. Major
sources are facilities with the potential to emit 10 or more tpy of any single HAP or 25 or more tpy of any
combination of HAP. The number of major sources was approximated using the uncontrolled emission
estimates (scaled up to reflect potential to emit) for each facility.
The emission estimates presented in this chapter are based on equipment throughput at plant
production levels. Plant capacity typically exceeds plant production. Thus, a facility's potential to emit
may be greater than the emissions estimated for each facility at plant production levels. To account for
this, an average ratio of plant production to plant capacity was determined based on the non-CBI
responses to EPA's MACT survey. On average, plywood (hardwood and softwood) and reconstituted
wood products plants were found to operate at around 75 percent of their plant capacity. Engineered
wood products plants were found, on average, to operate at 60 percent of their capacity. For purposes of
determining which facilities may be major sources based on potential to emit, the uncontrolled emission
estimates for each facility were scaled up by 25 or 40 percent before comparison to the 10- or 25-tpy
major source thresholds. Exhibit 3-8 presents the estimated number of major sources by product type.
Because of the uncertainty in the emission estimates and lack of knowledge about the specific
operations at facilities the numbers of major sources presented in Exhibit 3-8 are merely estimates. Major
source determinations depend on the types of operations at a facility and facility-specific factors. There
may be operations and HAP emission sources at wood products facilities that have not been accounted for
in the emission estimates (e.g., plants that manufacture furniture in addition to particleboard). Plants with
additional onsite operations may be major sources regardless of their plywood and composite wood
products operations. Then again, there may be plants that were determined to be major sources in this
analysis that are not major sources due to uncertainty in the emission estimates or potential to emit. The
purpose of the analyses discussed in this document is to estimate - on a nationwide scale - emissions
from plywood and composite wood products plants and the number of major sources. On a nationwide
scale, it is unlikely that the uncertainties in the emission estimates or number of major sources will have a
significant impact on the direction of the proposed plywood and composite wood products rulemaking.
The reduction in emissions of total HAP and THC is the difference between baseline emissions
and the emissions expected to remain following implementation of the MACT floor level of control
identified for the PCWP standards. Baseline emissions reflect the level of air pollution control that is
currently used at PCWP plants. The MACT floor control level reflects the level of control that will be
used following implementation of the PCWP standards. The following assumptions were used when
estimating emissions at the MACT floor control level: (1) plants will install RTO on all process units that
require controls to meet the MACT floor; (2) presses at conventional particleboard, MDF, OSB, and
hardboard plants will be fully enclosed by a PTE that captures and routes 100 percent of the emissions
from the press area to an RTO; and (3) WESP will be installed upstream of RTO for new RTO
installations on rotary strand dryers. The nationwide HAP and THC emission reduction was calculated
3-17
-------
by subtracting the emissions remaining at the MACT floor control level from the baseline emissions.
Exhibit 3-9 presents the nationwide HAP and THC emissions reduction.14
3.2.4 Nationwide Emission Estimates - Non-HAP Species
As mentioned earlier in this chapter, there are reductions in pollutants other than HAPs as a result
of compliance with this proposed rule. There are reductions of coarse particulate matter (PM10), volatile
organic compounds (VOC), and carbon monoxide (CO), and increases in nitrogen oxides (NOx), and
sulfur dioxide (S02). The reductions of PM10 are estimated at 13,000 tons, the reductions of VOC
(reported as total hydrocarbon) are estimated at 27,000 tons (see Exhibit 3-9), and the reductions of CO
are estimated at 11,000 tons. The increase of NOx emissions is estimated at 5,000 tons, and there are
potentially as many as 4,000 tons of additional S02 emissions. All emission estimates are estimated for
the fifth year after the issuance of the proposed rule. The methodology used to prepare these estimates is
contained in the BID for this proposed rule.15
3-18
-------
Exhibit 3-5. UNCONTROLLED AND BASELINE HAP EMISSIONS ESTIMATES
Product
No. of
plants3
Uncontrolled emissions, ton/yr1.
Baseline emissions, ton/yr"
Total HAP
THC as C
Total HAP
THC as C
MDF
24
4,000
8,200
2,400
4,800
Particleboardd
51
5,700
13,000
5,400
13,000
Hardboard
18
3,500
5,800
3,300
5,500
Fiberboard
7
78
400
78
400
OSB
37
7,100
19,000
3,500
5,400
Softwood plywood
105
4,000
24,000
3,700
20,000
Hardwood plywood
166
150
640
150
640
EWP
39
310
990
290
790
Nationwide total6
447
25,000
73,000
19,000
50,000
" Some plants make multiple products and are counted once for each product they make (e.g., a particleboard and softwood plywood plant).
b Uncontrolled emissions represent the emissions that occur before the application of HAP emission control devices.
c Baseline emissions reflect the application of HAP emission controls in the industry as of April 2000.
d Includes conventional and molded particleboard.
e Nationwide emission totals may not exactly match sum due to rounding.
3-19
-------
Exhibit 3-6. SPECIATED NATIONWIDE UNCONTROLLED HAP EMISSIONS BY PRODUCT TYPE
Product
Estimated HAP's emitted, ton/yr
Acetaldehyde
Acrolein
Formaldehyde
Methanol
Phenol
Propionaldehyde
Other HAPa
Totalf
MDF
48
1
1,700
2,200
93
1
27
4,000
Particleboardb
230
56
1,300
3,800
150
20
©
5,700
Hardboard
320
76
580
2,100
99
250
52d
3,500
Fiberboard
9
1
17
45
1
0
6
78
OSB
1,500
540
890
3,500
340
88
280e
7,100
Softwood
plywood
450
26
280
2,900
150
24
170
4,000
Hardwood
plywood
20
0
11
81
15
0
22
150
EWP
53
4
36
140
46
8
19e
310
Totalf
2,600
700
4,800
15,000
890
390
730
25,000
11 Other HAP's include benzene, cumene, methylene chloride, MEK, MIBK, styrene, toluene, m,p-xylene, and o-xylene.
b Includes conventional and molded particleboard.
c Includes HAPs listed in footnote "a" plus acetophenone, biphenyl, bis-(2-ethylhexyl phthalate), bromomethane, carbon disulfide, carbon tetrachloride,
chloroform, chloromethane, di-n-butyl phthalate, ethyl benzene, hydroquinone, n-Hexane, 1,1,1-trichloroethane, and 4-methyl-2-pentanone.
d Includes HAPs listed in footnote "a" plus chloroethane, chloromethane, ethyl benzene, m.p-cresol, and o-cresol.
e Includes HAPs listed in footnote "a" plus MDI.
f Totals may not exactly match sum due to rounding.
3-20
-------
Exhibit 3-7. SPECIATED NATIONWIDE BASELINE HAP EMISSIONS BY PRODUCT
Product
Estimated HAP's emitted, ton/yr
Acetaldehyde
Acrolein
Formaldehyde
Methanol
Phenol
Propionaldehyde
Other HAPa
Totalf
MDF
29
1
1,000
1,300
51
0
17
2,500
Particleboardb
200
50
1,200
3,700
140
18
150c
5,400
Hardboard
270
61
570
2,100
96
200
48d
3,300
Fiberboard
9
1
17
45
1
0
6
78
OSB
570
200
370
2,000
210
32
69e
3,500
Softwood
plywood
390
20
230
2,700
130
17
150
3,700
Hardwood
plywood
20
0
11
81
15
0
22
150
EWP
47
4
30
140
46
7
16e
290
Totalf
1,500
330
3,400
12,000
690
270
480
19,000
11 Other HAP's include benzene, cumene, methylene chloride, MEK, MIBK, styrene, toluene, m,p-xylene, and o-xylene.
b Includes conventional and molded particleboard.
c Includes HAPs listed in footnote "a" plus acetophenone, biphenyl, bis-(2-ethylhexyl phthalate), bromomethane, carbon disulfide, carbon tetrachloride,
chloroform, chloromethane, di-n-butyl phthalate, ethyl benzene, hydroquinone, n-Hexane, 1,1,1-trichloroethane, and 4-methyl-2-pentanone.
d Includes HAPs listed in footnote "a" plus chloroethane, chloromethane, ethyl benzene, m.p-cresol, and o-cresol.
e Includes HAPs listed in footnote "a" plus MDI.
f Totals may not exactly match sum due to rounding.
3-21
-------
Exhibit 3-8. ESTIMATED NUMBER OF MAJOR SOURCES BY PRODUCT
Product
No. of plants3
No. of major sourcesb
No. of potentially
non-major sourcesb
MDF
24
24
0
Particleboard0
51
42
9
Hardboard
18
18
0
Fiberboard
7
3
4
OSB
37
37
0
Softwood plywood
105
87
18
Hardwood plywood
166
0
166
EWP
39
12
27
Total
447
223
224
" Some plants make multiple products and are counted once for each product they make (e.g., a particleboard and softwood plywood plant).
b Major sources are defined as facilities with the potential to emit 10 or more tons per year (tpy) of any single HAP or 25 or more tpy of any combination of
HAP. Sources with HAP emissions estimated to be below the 10/25 thresholds in this analysis are labeled as potentially non-major sources. The emission
estimation methodology described in this document does not account for onsite operations (e.g., furniture manufacture) not included in the plywood and
composite wood products source category.
c Includes conventional and molded particleboard. Five of the potentially non-major particleboard sources manufacture molded particleboard.
3-22
-------
Exhibit 3-9. ESTIMATED NATIONWIDE REDUCTION IN TOTAL HAP AND THC
Product type
Total HAP (ton/yr)
THC (ton/yr)
Baseline3
MACT floor
Reduction
Baseline3
MACT floor
Reduction
Softwood plywood/veneer
3,700
3,043
657
19,631
9,709
9,922
Hardwood plywood/veneer
161
161
0b
640
640
0b
Medium density fiberboard
2,469
345
2,124
4,763
572
4,191
Oriented strandboard
3,513
753
2,760
5,362
1,755
3,607
Particleboardc
5,377
2,787
2,590
12,632
6,724
5,908
Hardboard
3,291
752
2,539
5,478
2,103
3,374
Fiberboard
78
78
0b
398
398
0b
Engineered wood products
298
230
68
793
617
176
TOTAL
18,933
8,196
10,737
49,706
22,529
27,178
" The baseline emissions presented in this table may differ slightly from the values presented in Exhibits 3-4 through 3-6 because of slight differences in
calculation procedures (i.e., use of a total HAP emission factor instead of summing speciated HAP emissions) and rounding.
b There is no impact because no plants are impacted by the PCWP standards at the MACT floor control level.
c Includes conventional and molded particleboard.
3-23
-------
3.3 Control Equipment and Costs
Traditional add-on pollution control devices are expected to be the types of control measures that
firms will choose in order to comply with the proposed rule. Add-on air pollution control devices that are
most likely to be used to comply with the plywood and composite wood products rule include
incineration-based controls. Among these types of controls are regenerative thermal oxidizers (RTO),
regenerative catalytic oxidizers (RCO), and process incineration. Biofilters is another add-on control that
may be applied. The control device most commonly used to control emissions from plywood and
composite wood products plants is the RTO. Therefore, it was assumed that most plants would install
RTO's to comply with the rule. A number of RCO's and biofilters are also presently used by plywood
and composite wood products plants, and plants may choose to install these technologies to comply with
the plywood and composite wood products rule. There may be cost advantages to using RCO's or
biofilters instead of RTO's for some plants. However, the cost analyses focus on use of RTO's for
simplicity (e.g., to minimize the number of cost algorithms developed and to avoid judgements regarding
which plants may choose a particular technology).
Several facilities with large capacity heat energy systems currently use process incineration as a
method of emissions control. Facilities may elect to use process incineration to comply with the rule.
However, the applicability of process incineration is limited to those plants that have or may later install
large onsite heat energy systems. The capital and operating costs of process incineration are expected to
be significantly lower than the costs of add-on controls.
The plywood and composite wood products rule contains emissions averaging provisions and
production-based emission limits. It may be possible for some plants to reduce their control costs by
complying with either the emissions averaging provisions or production-based emission limits. However,
because there is no way to predict which plants might use the emissions averaging provisions or
production-based emission limits, the cost estimates did not account for these options. Instead, the
control cost estimates were developed assuming that all facilities would install RTO(s) to meet the 90
percent HAP reduction emission limit in the plywood and composite wood products rule.
Oriented strandboard plants typically install wet electrostatic precipitators (WESP's) upstream of
rotary dryer RTO's to protect the RTO media from plugging. Thus, the capital and annual costs
associated with WESP's were modeled for rotary strand dryers.
Enclosures must be installed around presses in order to ensure capture of the press emissions that
are routed to a control device. Thus, the capital costs of permanent total enclosures (PTE's) were
included in the costing analyses. Annual costs associated with PTE's (if any) were assumed to be
minimal and were not included in the cost analyses.
This chapter presents the estimated nationwide capital and annualized costs for compliance with
the PCWP rule. Compliance costs include the costs of installing and operating air pollution control
equipment and the costs associated with demonstrating ongoing compliance (i.e., emissions testing,
monitoring, reporting, and recordkeeping costs). Section 3.3.1 discusses the estimated air pollution
control costs. Cost estimates associated with testing, monitoring, reporting, and recordkeeping are
discussed in the Paperwork Reduction Act submission for the proposed PCWP standards and are
summarized in Section 3.4.
3.3.1 Basis For Control Costs
As discussed above, add-on air pollution control devices most likely to be used to comply with
the PCWP rule include incineration-based controls (e.g., RTO, RCO, and process incineration) or
3-24
-------
biofilters. The control device most commonly used to control emissions from PCWP plants is the RTO.
Therefore, for costing purposes, it was assumed that most plants would install RTO's to comply with the
rule. A number of RCO's and biofilters are also presently used by PCWP plants. In addition, several
plants with large capacity heat energy systems currently use process incineration. However, the
applicability of process incineration is limited to those plants that have, or may later install, large onsite
heat energy systems. There may be cost advantages to using RCO's, biofilters, process incineration, other
add-on control devices, or pollution prevention measures instead of RTO's for some plants. Plants may
elect to use any of these technologies to comply with the rule, provided the technology limits HAP
emissions to the levels specified in the rule. However, the cost analyses described in this chapter focus on
use of RTO's due to their prevalence in the industry and to minimize the number of cost algorithms
developed and to avoid judgements regarding which plants may choose a particular technology.
Oriented strandboard plants typically install WESP's upstream of rotary dryer RTO's to protect
the RTO media from plugging. Thus, the capital and annualized costs associated with WESP's were
modeled for rotary strand dryers. Available information indicates that WESP's are not necessary for
protecting RTO's installed of other types of dryers (e.g., tube dryers) or on presses.16 Therefore, with the
noted exception of OSB dryers without WESP's, the existing particulate abatement equipment on process
units was assumed to be sufficient for protecting the RTO media.17
Enclosures must be installed around presses to ensure complete capture of the press emissions
before routing these emissions to a control device. Thus, the capital costs of permanent total enclosures
(PTE's) were included in the costing analyses. Annualized costs associated with PTE's were assumed to
be minimal and were not included in the cost analyses.
The following sections discuss the RTO, WESP, and PTE costs. Section 3.3.1.4 describes how
plant-by-plant control costs were estimated, and Section 3.3.1.5 summarizes the nationwide control costs.
3.3.1.1 RTO Costs
An RTO cost algorithm was developed based on: (1) information from an RTO vendor with
numerous RTO installations at PCWP plants, and (2) the costing methodology described in the EPA Air
Pollution Control Cost Manual.18'19 The RTO cost algorithm was used to determine RTO total capital
investment (TCI) and total annualized cost (TAC) based on the exhaust flow to be controlled and annual
operating hours. Development of the algorithm is discussed in Sections 3.4.1.1.1 and 3.4.1.1.2.
RTO Total Capital Investment.18,19
Equipment costs (including equipment, installation, and freight) were provided by the RTO
vendor for four sizes of RTO's. The 1997 equipment costs were not escalated because the Vatavuk Air
Pollution Control Cost Index (VAPCCI) for 1997 (107.9) was slightly greater than the preliminary
VAPCCI for RTO's in fourth quarter 1999 (107.8).20 According to the EPA Air Pollution Control Cost
Manual, instrumentation is typically 10 percent of equipment cost (RTO and auxiliary equipment); sales
tax is typically 3 percent of the equipment cost; and freight is typically 5 percent of the equipment cost.
Figure 3-1 presents the purchased equipment costs (PEC) supplied by the RTO vendor (minus freight),
and shows that the equipment costs vary linearly with gas flow rate. The regression equation presented in
Figure 3-1 was included in the RTO cost algorithm to calculate the equipment cost for the oxidizer and
auxiliary equipment for various gas flow rates. Instrumentation, sales tax, and freight were added to the
calculated equipment costs to obtain the total PEC.
3-25
-------
2,500,000 i
3
0.
500,000
0 -I , , , ,
0 50,000 100,000 150,000 200,000 250,000
Flow rate, dscfm
Figure 3-1. Variation in RTO purchased equipment cost with flow rate.
Direct installation costs for handling and erection, electrical, and piping were included in the
equipment cost provided by the RTO vendor. Start-up costs were also included in the equipment cost
provided by the RTO vendor. These costs are typically 22 percent of the PEC. Thus, these costs were
subtracted from the PEC before further calculations based on the PEC were performed. Direct installation
costs including foundation and support, insulation for ductwork, and painting were estimated according to
the procedures in the EPA Air Pollution Control Cost Manual. Because PTE's were costed separately, no
enclosure building was costed in the RTO algorithm. Site preparation costs and indirect installation costs
(e.g., engineering, field expense, contractor fees, performance tests, and contengencies) were estimated
according to the procedures in the EPA Air Pollution Control Cost Manual. The TCI was calculated by
summing the PEC, direct and indirect installation costs, and site preparation cost.
RTO Total Annualized Cost
Total annualized costs consist of operating and maintenance labor and material costs, utility costs,
and indirect operating costs (including capital recovery). Operating and maintenance labor and material
costs were estimated based on the RTO vendor information because the RTO vendor assumptions led to
higher costs than the EPA Air Pollution Control Cost Manual and were assumed to be more representative
of the PCWP industry. The operator labor rate supplied by the RTO vendor was $19.50 per hour.
The RTO electricity use and natural gas use was provided by the RTO vendor for the four RTO
sizes. Figures 3-2 and 3-3 present the relationships between flow rate and electricity and flow rate and
natural gas use, respectively. As shown in the figures, there is a linear relationship between RTO
electricity consumption and flow rate, and an exponential relationship between RTO fuel consumption
and flow rate.
3-26
-------
0 50000 100000 150000 200000 250000
How rate, dscfm
Figure 3-2. Relationship between RTO electricity consumption and flow rate.
Flow rate, dscfm
Figure 3-3. Relationship between RTO natural gas consumption and flow rate.
Electricity costs were estimated by the RTO vendor at $0,045 per kilowatt-hour (kWh). The
RTO vendor estimated natural gas costs at $3 per million British thermal units (MMBtu). Both of these
energy prices match closely with currently published nationwide average prices.21,22 Thus, the electricity
and natural gas prices supplied by the RTO vendor were used in the cost algorithm.
Indirect operating costs were estimated using the methodology described in the EPA Air
Pollution Control Cost Manual. The capital recovery cost was estimating assuming an RTO equipment
life of 15 years (based on the RTO vendor information) and a 7-percent interest rate. The TAC was
calculated by summing the direct and indirect annual operating costs.
Application of the RTO Cost Algorithm to Estimate Capital and Annualized Costs.
3-27
-------
The complete RTO cost algorithm, which predicts RTO capital and annualized costs as a function
of operating hours and flow rate, was run several times assuming 8,000 operating hours per year and
various flow rates. The 8,000-hr operating time was selected based on the results of the EPA's MACT
survey, which show industry average operating hours of slightly less than 8,000-hr/yr.1" Although several
plants operate process lines more than 8,000 hr/yr, their equipment and control devices may or may not be
operated for more than 8,000 hr/yr. Thus, 8,000 hr/yr was selected as the control device operating time
for purposes of costing.
The TCI and TAC values generated for each flow rate using the RTO cost algorithm are
presented in the BID for this proposed rule. A regression equation was developed based on the calculated
TCI and TAC for each flow rate. Figures 3-4 and 3-5 present the relationships between flow rate and
RTO capital costs and flow rate and annualized costs, respectively, and the associated regression
equations.
3.3.1.2 WESP Costs
A WESP cost model was developed based on: (1) information from a WESP vendor with many
WESP installations at wood products plants, and (2) the costing methodology described in the EPA Air
Pollution Control Cost Manual for electrostatic precipitators (ESP's).19,23 The cost model was used to
determine TCI and TAC for WESP's used to control particulate emissions from OSB rotary dryers. The
WESP vendor provided cost information for a WESP sized to treat 27,650 dry standard cubic feet per
minute (dscfm) of OSB rotary dryer exhaust. This flow rate matches closely with the flow rates for
uncontrolled OSB. Thus, the model TCI and TAC could be applied for each dryer to be controlled (i.e.,
the model need not calculate different costs for varying flow rates). The WESP cost model is presented in
the BID for the proposed rule. Development of the model is discussed below.
WESP Total Capital Investment.
The WESP and auxiliary equipment costs (which makeup the PEC) were provided by the WESP
vendor. These PEC include the cost of the WESP; pumps, piping, and tanks; ducting (including the
quench); fans; and a 1-gallon per minute (gpm) blowdown solids removal system. Instrumentation costs
were also provided by the WESP vendor. Sales tax and freight were added into the total PEC based on
the methodology described in the EPA Air Pollution Control Cost Manual.
Flow rate, dscfm
Figure 3-4. Variation in RTO total capital investment with flow.
3-28
-------
3,500,000
3,000,000
2,500,000
S 2,000,000
^ 1,500,000
1,000,000
500,000
0
Figure 3-5. Variation in RTO total annualized cost with flow.
The direct installation costs such as foundation and support, handling and erection, electrical,
piping, insulation for ductwork, and painting were included in the PEC provided by the WESP vendor. It
was assumed that no building would be necessary for the WESP and that there would be no additional site
preparation costs. Several indirect costs were also included in the equipment cost supplied by the WESP
vendor, including engineering, construction and field expense, start-up, and contingencies. Because
WESP's are already widely used at OSB plants, it was assumed that no model study would be necessary
for the WESP although the EPA Air Pollution Control Cost Manual mentions model-study costs for
ESP's.
The cost of a performance test was included in the WESP cost model. According to the EPA Air
Pollution Control Cost Manual, the performance test is typically 1 percent of the PEC. Thus, 1 percent of
the model PEC (minus the direct and indirect installation costs included in the PEC) was used as the cost
of the performance test. The direct and indirect costs were summed to arrive at the WESP TCI.
WESP Total Annualized Cost
The direct annualized costs include operating and maintenance labor and materials, utilities, and
waste disposal. The operating labor cost was based on 1,146 hr/yr (provided by the WESP vendor) at
$19.50/hr (the labor rate used in the RTO cost algorithm). The annual cost of operating materials,
including caustic and defoamer, was provided by the WESP vendor. The maintenance labor rate was
estimated as 110 percent of the operating labor rate. The maintenance hours per year were estimated
based on information supplied WESP vendor. The cost of maintenance materials (including replacement
of one pump seal per year, and one voltage controller every 4 years, and miscellaneous materials) was
supplied by the WESP vendor.
The electricity necessary to power the WESP components (approximately 2,076,000 kWh/yr for
all WESP components) was based on information provided by the WESP vendor. An electricity cost of
$0.045/kWh was used (the same as used in the RTO cost algorithm). A $0.20/gal cost for makeup water
50,000 100,000 150,000 200,000 250,000 300,000 350,000
Flow rate, dscfm
3-29
-------
was used based on the EPA Air Pollution Control Cost Manual. The WESP water recirculation rate,
makeup water addition rate, and blowdown generation rates were provided by the WESP vendor. The
EPA Air Pollution Control Cost Manual indicated that wastewater treatment costs may range from $1.30
to $2.15 /1,000 gallons. Methods of WESP wastewater treatment and disposal could include evaporation
from settling ponds, discharge to a municipal water treatment facility, or spray irrigation. The wastewater
treatment and disposal cost for the blowdown was assumed to be $2.15 per gallon. The wastewater
percent solids of 7.6 percent was based on the average from the MACT survey responses.11 It was
assumed that the solids would ultimately be disposed in a landfill (although they could be burned onsite
or used for soil amendment). The trucking cost for hauling sludge to landfill was estimated to be $0.20
yd3-mi.22 The landfill was assumed to be 20 miles away, and a $20/ton landfill tipping fee was used.22
The density of the solids was assumed to be 0.5 ton/yd3 for wet wood particulate (given that the density
of water is 0.84 ton/yd3 and the density of wood is from 30 to 50 percent of the density of water).
The indirect operating costs were estimated based on the methodology described in the EPA Air
Pollution Control Cost Manual. The capital recovery cost was estimated assuming a WESP equipment
life of 20 years (based on the EPA Air Pollution Control Cost Manual and WESP vendor information)
and a 7-percent interest rate. The TAC was calculated by summing the direct and indirect annual
operating costs.
3.3.1.3 Permanent Total Enclosure (PTE) Costs
The capital costs associated with installation of a PTE were based on available information in the
project files on the capital cost of PTE's for particleboard, MDF, and OSB presses.24 These costs
included the following elements:
installed cost of the PTE (including fan system)
ductwork
instrumentation and wiring
fire suppression (in some cases)
site supervision
start-up and testing
Based on the available cost information, the following algorithm was developed to estimate the PTE costs
for various exhaust flowrates:
TCIpte = 1.2031 x Qdscfm + 425,760
where:
TCIpte = the total capital cost of the permanent total enclosure, $
Qdscfm = design exhaust flow rate from PTE, dry standard cubic feet per minute
Available information on actual exhaust flow rates from PTE's installed around reconstituted wood
product presses was used to develop model flow rates for the various press applications.25 Information on
press vent flow rates from unenclosed presses was available, but not used, because unenclosed press flow
rates are altered when a PTE is installed around a press.26 The cost algorithm was then applied to the
model flow rates to estimate the capital costs of the model PTE's as shown in Exhibit 3-10. The costs
were rounded to the nearest $1,000. In the case of the particleboard press PTE, the cost was set at
$485,000 (instead of $481,000, which is the value derived from the cost algorithm) because the PTE
model flow rate was similar to those for the MDF and hardboard presses, and applying the same cost to
all three types of press PTE's simplified the costing analyses. Annualized costs were not developed for
3-30
-------
PTE's because the annualized cost of the fans is already accounted for in the estimated costs of the RTO's.
Exhibit 3-10. PRESS ENCLOSURE EXHAUST FLOW RATES AND CAPITAL COSTS
Equipment type
Flow rate, dscfrn
PTE capital cost
Particleboard press
45,524
$485,000
OSB press
97,509
$543,000
MDF or dry/dry hardboard press
49,413
$485,000
Wet/dry or wet/wet hardboard press
49,209
$485,000
3.3.2 Plant-by-Plant Costing Approach
The control costs associated with the PCWP standards were estimated for each plant and were
summed to arrive at a nationwide estimate of control costs. The PCWP standards apply only to major
sources of HAP emissions. Therefore, cost estimates were developed for only those plants that were
assumed to be major sources.15 The information used to estimate the plant-by-plant control costs is
described below.
3.3.2.1 Application of Control Costs to Process Units
The cost models discussed in Section 3.3.1 were applied to each plant that would likely need to
install air pollution controls in order to meet the PCWP standards. Plant-specific information on process
units (e.g., dryers, presses) and controls was taken from the MACT survey responses.101112 In addition,
information about the presence of PTE's on presses was taken from the MACT survey responses.10 If
information about press enclosures was not provided in the MACT survey responses, or was claimed
confidential, the press was assumed to be unenclosed if it was uncontrolled or enclosed if it was
controlled for purposes of costing.
Some plants have begun operation and other plants have added equipment or controls since EPA
conducted the MACT survey. Such changes were accounted for in the nationwide cost estimates. A
separate memorandum summarizes the changes to plants that have occurred following the EPA's MACT
survey.13
The process units and controls present at each plant were reviewed to determine which of the
assumed what control equipment (i.e., RTO, WESP, or PTE) the plant would need to install to meet the
PCWP standards based on the MACT floor control levels. The MACT floor control levels are based on
the information presented in the BID for this proposed rule.6 Exhibit 3-2 summarizes the process units for
which control equipment would be required to meet the MACT floor and the control equipment costed for
these process units. At each plant, the exhaust gas flow rates from the applicable uncontrolled process
units listed in Exhibit 3-11 were summed to yield a plant-wide uncontrolled exhaust gas flow rate.
Process units already equipped with controls to meet the MACT floor were not included in the plant-wide
uncontrolled gas flow rate estimates. The procedures for estimating the uncontrolled gas flow rates from
process units and the application of the cost algorithms is discussed in the following section.
3-31
-------
Exhibit 3-11. CONTROL EQUIPMENT COSTED FOR PROCESS UNITS WITH
CONTROLLED MACT FLOOR
Existing process units with
control requirements
Control
equipment costed
Notes
Tube dryers (primary and
secondary)
RTO
Tube dryers are located at particleboard, MDF, and
hardboard plants
Rotary strand dryers
WESP and RTO
Rotary strand dryers are located at OSB and LSL plants.
Assumed that the WESP is not needed for plants that already
have an RTO without a WESP. Assumed that plants that
currently operate an EFB or multiclone alone (i.e., with no
RTO) would install a WESP with the RTO.
Conveyor-type strand
dryers
RTO
Conveyor strand dryers are located at OSB and LSL plants.
Rotary green particle dryers
RTO
Rotary green particle dryers are located at particleboard,
MDF, or hardboard plants and process furnish with >30%
(dry basis) inlet moisture content at dryer inlet temperature
of >600°F
Hardboard ovens
RTO
Includes bake and tempering ovens
Softwood veneer dryers
RTO
Softwood veneer dryers are located at softwood plywood,
hardwood plywood, LVL, and PSL plants and dry > 50% (by
volume, annually) softwood veneer
Pressurized refiners
None
The exhaust from pressurized refiners typically passes
through a tube dryer and exits through the tube dryer control
device. Therefore, it was not necessary to cost separate
control equipment for pressurized refiners. Pressurized
refiners are located at MDF and hardboard plants.
Reconstituted wood
products presses
PTE and RTO
Reconstituted wood products presses are located at
hardboard, MDF, OSB, and particleboard plants
Exhaust Flow Rate to Be Controlled
If provided in the non-confidential MACT survey responses, process-unit specific exhaust flow
rate, temperature, and percent moisture were used to determine the dry standard flow rate for each process
unit. If sufficient information was not provided in the MACT survey response to determine dry standard
flow rates (or the information was claimed confidential), then default values were used for the flow rate.
The default values were based on the average value for other similar process units at plants that provided
enough non-confidential information to calculate the dry standard flow rate. Exhibit 3-12 summarizes the
default flow rates used in the costing analyses. The average flow rates from press enclosures are
described in Section 3.3.1 and were used for all presses.
3-32
-------
Exhibit 3-12. DEFAULT FLOW RATES
Process line
Equipment type
Flow rate (dscfm)
Particleboard
Rotary green particle dryer
35,731
Tube dryer
14,955
OSB
Rotary strand dryer
32,478
Conveyor-type strand dryer
37,810
MDF
Primary tube dryer (single-stage or first
stage of staged dryer)
79,173
Secondary tube dryer (second stage of staged
dryer)
18,195
Plywood
Softwood veneer dryer
12,062
Hardboard
Bake oven
4,742
Tempering oven
4,055
Primary tube dryer (single-stage or first
stage of staged dryer)
37,436
Secondary tube dryer (second stage of staged
dryer)
31,728
Several plants have multiple process units requiring controls. The flow rates for these process
units were summed and divided across control equipment as necessary. In most cases, the total dryer
flow was assumed to be routed to one or more RTO's dedicated to controlling dryer exhaust and the total
press flow was assumed to be routed to one or more RTO's dedicated to controlling press exhaust.
Because RTO fuel costs increase exponentially with gas flow rate, RTO sizes were assumed to remain
less than about 150,000 dscfm. (The largest RTO in mentioned in the MACT survey responses was
around 150,000 dscfm.)
In some cases, dryers and presses were assumed to be routed to the same RTO, provided that the
total dryer and press flow remained under 150,000 dscfm. For example, two RTO's (103,500 dscfm
each) would be costed for a MDF plant with 2 dryers (79,000 each) and 1 press (49,000) assuming that
the flow for both dryers and the press could be combined and split equally across the two RTO's. This
approach seems reasonable given that several MDF plants currently route dryer and press exhaust to the
same RTO.
Calculation of Nationwide Control Costs
The total plant-by-plant control cost was calculated by summing the control cost associated with
each RTO, WESP, and PTE costed for each plant. The number of control devices at each plant depended
on the number of process units and the exhaust flow to be controlled at the plant. In some cases, only one
control device was costed, while in other cases, multiple control devices were costed for a plant.
3-33
-------
Some plants claimed all relevant portions of their MACT survey responses confidential. In
addition, a MACT survey response was not available for a few plants likely to be impacted by the PCWP
standards. Without a non-confidential MACT survey response, information was not available to develop
plant-specific cost estimates. Therefore, the average cost for all other plants manufacturing the same
product was used to approximate the costs for plants for which there was no non-confidential plant-
specific information.
The nationwide capital and annualized control costs were determined by summing the total plant-
specific costs.
3.3.3 Summary of Nationwide Control Costs
Exhibit 3-13 summarizes the nationwide control costs for different product types. The
nationwide total capital cost for control equipment is estimated to be $473 million and the nationwide
total annual cost for control equipment is estimated as $136 million (1999 dollars). Exhibit 3-14 presents
the dollars of total annualized costs per ton of HAP and THC reduced.
3.4 Testing, Monitoring, Reporting, And Recordkeeping Costs
Compliance with the PCWP standards must be demonstrated through performance testing,
ongoing monitoring of process or control device operating parameters or emissions, periodic reporting to
the government agency that implements the PCWP rule, and recordkeeping. There are capital and
annualized costs associated with these testing, monitoring, reporting, and recordkeeping activities. These
costs, which are estimated and documented in the supporting statement for the Paperwork Reduction Act
submission, and are summarized in this section.27
3-34
-------
Exhibit 3-13. ESTIMATED NATIONWIDE CONTROL COSTS FOR THE PCWP INDUSTRY
Product type
No. of plants3
No. of plants
impacted'1
Process units impacted
Control equipment
Total capital
costs, $MM
Total annual
costs, $MM
Softwood plywood/veneer
105
66
softwood veneer dryers
RTO
$87.1
$28.4
Hardwood plywood/veneer
166
0
N/A
no control
$0.0
$0.0
Medium density fiberboard
24
18
dryers, presses
RTO for dryers and
PTE/RTO for presses
$71.3
$21.5
Oriented Strandboard
37
23
dryers, presses
WESP/RTO for dryers and
PTE/RTO for presses
$94.6
$25.5
Particleboard
(conventional and molded)
51
38
green rotary particle
dryers, presses
RTO for dryers and
PTE/RTO for presses
$125.2
$34.2
Particleboard (agriboard)
5
0
N/A
no control
$0.0
$0.0
Hardboard
18
18
tube dryers, presses,
ovens
RTO for dryers and
PTE/RTO for presses
$84.4
$23.5
Fiberboard
7
0
N/A
no control
$0.0
$0.0
Engineered wood products
41
3
softwood veneer dryers,
strand dryers
RTO for veneer dryers and
WESP/RTO for strand dryers
$10.3
$3.2
TOTAL:
454
166
$473
$136
" Some plants manufacture more than one product type. These plants are listed once for each product type manufactured.
b The number of plants impacted may be different from the number of plants nationwide for one of the following reasons: (1) some plants are not major sources;
(2) some plants already have all of the necessary control equipment; or (3) a few plants are major sources but do not operate any process units for which there
are control requirements (e.g., glu-lam plants).
3-35
-------
Exhibit 3-14. DOLLARS (IN TOTAL ANNUALIZED COSTS) PER TON OF
HAP AND THC REDUCED
Product type
HAP, $/ton
THC, $/ton
Softwood plywood/veneer
$43,000
$2,900
Hardwood plywood/veneer
NA
NA
Medium density fiberboard
$10,000
$5,100
Oriented Strandboard
$9,200
$7,100
Particleboard (all types)
$13,000
$5,800
Hardboard
$9,300
$7,000
Fiberboard
NA
NA
Engineered wood products
$47,000
$18,000
Overall industry
$13,000
$5,000
The annual costs associated with testing, monitoring, reporting, and recordkeeping activities
include reporting and recordkeeping labor; annualized capital for monitoring equipment, file cabinets, and
performance tests; and the operation and maintenance costs associated with monitoring equipment. The
capital costs include capital for monitoring equipment, file cabinets and performance tests. Performance
tests are considered to be capital costs because plants will typically hire a testing contractor to conduct the
performance tests.
The total nationwide capital cost associated with testing, monitoring, reporting, and
recordkeeping is estimated to be $5.8 million and the total nationwide annualized cost is estimated to be
$5.6 million (1999 dollars). These costs were developed based on the information presented in the
Paperwork Reduction Act submission for the first 3 years following the effective date of the PCWP rule.
The costs apply for the 223 PCWP plants that are expected to be major sources. There are 57 facilities
that incur monitoring, recordkeeping, and reporting costs but do not incur control costs from compliance
with this proposed rule.
3.5 REFERENCES
1. United States Congress. Clean Air Act, as amended October 1990. 42 U.S.C. 7401 et seq.
Washington, DC. U.S. Government Printing Office.
2. U. S. Environmental Protection Agency. National Emission Standards for Hazardous Air
Pollutants for Source Category: Organic Hazardous Air Pollutants from the Synthetic Organic
Chemical Manufacturing Industry and Other Processes Subject to the Negotiated Regulation for
Equipment Leaks; Determination of MACT "Floor." 59 FR 29196. Washington, DC. U.S.
Government Printing Office. June 6, 1994.
3. Memorandum from K. Hanks and B. Nicholson, MRI, to P. Lassiter, EPA/ESD. April 16, 1998
(Finalized 5/28/98). Trip report for February 12, 1998 site visit to Timber Products Company
particleboard and hardwood plywood plant in Medford, Oregon.
3-36
-------
4. R. Nicholson, MRI, to D. Word, NCASI. February 17, 2000 (Confirmed via Email on 9/7/00).
Contact report to (1) find out if oxygen measurement data were available from the NCASI MACT
emissions test program, and (2) to get some information on typical oxygen concentrations in
dryer and press exhaust at plywood and composite wood products plants.
5. Memorandum from R. Nicholson, MRI, to M. Kissell, EPA/ESD. May 26, 2000. Control Device
Efficiency Data for Add-on Control Devices at PCWP Plants.
6. Memorandum from K. Hanks and B. Nicholson, MRI, to M. Kissell, EPA/ESD. July 13, 2000.
Determination of MACT floors and MACT for the Plywood and Composite Wood Products
Industry.
7. National Council of the Paper Industry for Air and Stream Improvement (NCASI) Method
IM/CAN/WP-99.01, Impinger/Canister Source Sampling Method for Speciated HAPs at Wood
Products Facilities. 1999.
8. Reference 6.
9. Memorandum from K. Hanks and D. Bullock, MRI, to M. Kissell, EPA/ESD. June 9, 2000.
Baseline Emission Estimates for the Plywood and Composite Wood Products Industry.
10. Memorandum from D. Bullock, K. Hanks, and B. Nicholson, MRI to M. Kissell, EPA/ESD.
April 28, 2000. Summary of Responses to the 1998 EPA Information Collection Request
(MACT Survey) ~ General Survey.
11. K. Hanks, B. Threatt, and B. Nicholson, MRI to M. Kissell, EPA/ESD. May 19, 1999. Summary
of Responses to the 1998 EPA Information Collection Request (MACT Survey) ~ Hardwood
Plywood and Veneer.
12. K. Hanks, B. Threatt, and B. Nicholson, MRI to M. Kissell, EPA/ESD. January 20, 2000.
Summary of Responses to the 1998 EPA Information Collection Request (MACT Survey) ~
Engineered Wood Products.
13. Memorandum from K. Hanks, MRI, to Project Files. April 18, 2000. Changes in the population
of existing plywood and composite wood products plants and equipment following the
information collection request.
14. Memorandum from D. Bullock and K. Hanks, MRI, to M. Kissell, EPA/ESD. April 27, 2000.
Documentation of Emission Factor Development for the Plywood and Composite Wood Products
Manufacturing NESHAP.
15. Background Infomation Document for the Proposed Plywood and Composite Wood Products
NESHAP. U.S. EPA/OAQPS, September 2000.
16. Suchsland, O., and G. Woodson, FiberboardManufacturing Practices in the United States, John
Wiley & Sons, New York, 1991, pp. 151-153.
17. K. Corrigan, North Dakota Department of Health, Division of Environmental Engineering.
February 11, 1997. Performance Test Report Review: Primeboard, Inc.
3-37
-------
18. Facsimile from J. Seiwert, Smith Environmental Corporation, to L. Kesari, EPA/OECA.
October 31, 1997. Revised emissions abatement systems RTO pricing (Smith Proposal B07-95-
156-1, Trinity Consultants)
19. EPA Air Pollution Control Cost Manual (Sixth Edition), U.S. Environmental Protection Agency,
April 26, 2002. EPA -452/B-02-001. Found on the Internet at
www. epa. gov/ttn/catc/products ,html#cccinfo.
20. Vatavuk Air Pollution Control Cost Indexes (VAPCCI), Chemical Engineering, March 2000,
p. 150.
21. Energy Information Administration, Form EIA-826, "Monthly Electric Utility Sales and
Revenue Report with State Distributions."
22. Energy Information Administration, Natural Gas Monthly, February 2000, p. 60.
23. Letter and attachments from S. Jaasund, Geoenergy International Corporation, to B. Nicholson,
MRI. March 28, 2000. Geoenergy WESP Capital and Operating Costs.
24. Memorandum from B. Nicholson, MRI, to Plywood and Composite Wood Products Project File.
July 31, 2000. Cost of Permanent Total Enclosures. (Confidential Business Information)
25. Memorandum from B. Nicholson, MRI, to Plywood and Composite Wood Products Project File.
July 31, 2000. Exhaust Gas Flowrate Information for Enclosed Presses. (Confidential Business
Information)
26. Memorandum from D. Bullock and K. Hanks, MRI, to P. Lassiter, EPA/ESD. October 27, 1998.
Trip report for visit to Temple-Inland Forest Products plant in Diboll, Texas.
27. Paperwork Reduction Act Submission, Supporting Statement, Plywood and Composite Wood
Products, 2000.
3-38
-------
4
ECONOMIC IMPACT ANALYSIS
4.1 Results in Brief
This economic impact analysis presents the results of modeling the effects of the proposed
plywood and composite wood products NESHAP upon affected facilities, firms, markets, and industries.
The analysis shows that product prices increase for each of the four industry sectors included in the
economic model, from 0.9 percent for the softwood plywood industry to 2.5 percent for the particleboard
and medium density fiberboard industry. Output is expected to decrease from 0.1 percent for the
softwood plywood industry to 0.7 percent for the particleboard and medium density fiberboard industry.
Exports of these products are expected to decline by no more than 0.7 percent for these four industries,
while imports are expected to rise. Only one product line closure at an affected facility is expected, and
employment at affected firms is expected to decline by only 0.3 percent. In addition, an analysis of
effects of this proposal on the supply, distribution, or use of energy finds that these effects are not
significant.
4.2 Introduction
The U.S. EPA is proposing a rule that addresses the emissions of hazardous air pollutants (HAPs)
from facilities that produce plywood and composite wood products. As described in Chapter 3, the
proposed rule will result in some facilities in this industry incurring costs associated with controlling HAP
emissions. The addition of these control costs will directly affect the individual facilities because their
costs of production will increase. In addition, the addition of compliance costs may also have an effect on
the overall markets for plywood and wood composite products through market price changes. Depending
on market conditions, these changes could occur if facilities with compliance costs increase their prices,
reduce their output, or cease operations altogether.
This section presents estimates of the economic changes that are expected to occur as a result of
the proposed NESHAP rule for the plywood and composite wood industry U.S. EPA, 1999a). The goal
of the assessment is to develop estimates of the following impact measures.
Market price changes
Market quantity changes
International trade effects
Size and distribution of social costs
Chapter 2 presented a profile of the different sectors within the plywood and composite wood
industry, which provides market information necessary to design and implement the EIA for this industry.
The plywood and composite wood products manufacturing industry affected by the proposed NESHAP
rule includes five distinct market sectors.
Softwood Plywood and Veneer (SWPW)
Oriented Strandboard (OSB)
Other Wood Composites (OWC), including
Particleboard and Medium Density Fiberboard (PB/MDF)
Hardboard (HB)
Engineered Wood Products (EWP)
For all but the EWP market sector, the Agency applied partial equilibrium (P/E) modeling
techniques to estimate the economic impacts of the proposed NESHAP rule for plywood and wood
4-1
-------
composites. For reasons presented in Section 4.4 below, developing an estimate of the economic impacts
of the proposed rule on the EWP sector required a qualitative approach.
Section 4.3 presents the inputs for the P/E economic analysis, including producer
characterization, market characterization, and compliance costs of the regulation. Section 4.4 describes
the approach to estimating the economic impacts on the SWPW, OSB, PB/MDF, and HB industry sectors,
and Section 4.5 presents the results of the economic impact analysis. Section 4.6 presents the qualitative
analysis of the EWP market sector.
4.3 Economic Impact Analysis Inputs
The first step of an impact assessment is developing information used to characterize the baseline
conditions of the industry. Key information needed for EIA inputs is used to characterize the following.
Individual producers
Product markets
Compliance costs
4.3.1 Producer Characterization
The primary source of baseline data on individual producers used in the EIA is a database
developed using the data collected by 1998 ICR described in Chapter 2 (U.S. EPA, 1998). The
information in the facility database allowed the characterization of facilities according to several features
needed for the analysis, including major products (SWPW, OSB, etc), baseline production volumes, and
facility capacity. The baseline year for the analysis is 1997 as it corresponds to the year for which
plywood and composite wood producers provided this information.
For certain facilities, production and capacity information was not available, and was
supplemented by additional research. Those facilities without specific production and capacity data were
assigned an estimate based on the average production and capacity data for facilities within the same
market sector and size category (as determined by employment).
4.3.2 Plywood and Composite Wood Markets
The Plywood and Composite Wood Products industry is a broad category encompassing the four
distinct markets listed above: (1) SWPW, (2) OSB, (3) OWC, and (4) EWP. For reasons discussed in the
EIA, the OWC sector is decomposed into two markets: PB/MDF and HB. The Agency developed P/E
models representing each of these four markets.
Market level data used in this EIA are presented in Exhibit 4-1. The market prices and elasticities
for each of the four markets were obtained from various industry market reports and economic literature
as described in Chapter 2. Total market volumes used in the models for each product are the sum of the
production of all identified U.S. facilities (from the facility database) and imports from foreign producers.
Total U.S. production for each market is the total production of all facilities identified in the EPA's
facility database. The production was also separated in to two subsets - the total production of affected
facilities and the total production of unaffected facilities.17 The source of foreign trade data on exports
and imports of these products was presented in Chapter 2.
17As mentioned in Chapter I, some facilities in the "unaffected" category have monitoring, reporting, and
record keeping costs of $25,194 peryear.
4-2
-------
llxhihil 4-1: li;iseline ( liiir
;ieleri/iilion of PI\\\ood iiml C omposite Wood M;irkels:
1«)«)"»
Soil wood
Oriented
"Oilier ( om poshes"
Plvwood
Siriindhoiird
PIJ/MDI
IMS
Market price (1997$/cubic meter)
$235
$185
$169
$1,322
Price Elasticity of Demand
Construction
-0.1034
-0.1034
-0.1149
-0.1149
Manufacturing/Other
-0.2585
n/a
-0.2872
-0.2872
Price Elasticity of Supply
0.42
0.42
0.42
0.42
Market quantity
(thousand cubic meters)
17,568,254
9,595,121
11,646,227
1,768,930
Domestic production
17,568,162
9,590,456
11,644,523
1,768,545
Affected
11,680,778
4,691,645
9,670,639
1,768,545
Unaffected
5,887,384
4,898,811
1,973,884
n/a
Exports
1370
148
333
371
Imports
92
4,666
1,705
385
Sources: U.S. EPA facility database; Section 2; and Appendix B of the Economic Impact Analysis.
4.3.3 Facility Compliance Costs
As described in Chapter 3, the Agency developed compliance costs estimates for those facilities
that must control HAP emissions in accordance with the regulatory requirements of the proposed
NESHAP rule. The EIA uses these costs to develop a "with regulation" market equilibrium scenario used
to estimate changes in individual facility production, total market volumes, market price, and social costs.
Typically, the Agency adjusts the compliance cost estimates from nominal dollars to baseline dollars
using the plant cost indices to be consistent with the baseline industry characterization of the economic
model (U.S. EPA, 1999b). In this case, there was virtually no change in the plant cost indices between
the baseline year of 1997 and current period, so no adjustment was made to the costs used in this analysis.
4.4 Economic Impact Analysis Methodology
The following section presents a summary of the approach used to assess the economic impacts of
the proposed NESHAP rule. The EIA contains a more detailed description of the methodology used to
analyze the economic impact of this proposed regulation on each of the markets analyzed. The purpose of
the EIA is to model the responses of individual producers and the overall market to the imposition of
compliance costs. For this EIA, the agency used a market-based economic model that reflects the
production choices producers make in the face of changes in their individual production costs and
changes in overall market prices.
The economic model used in this analysis simulates the short run decisions of the producers in
response to operating cost and price changes within a given market. For each of the four markets, the
model used in this analysis assumes that the market is perfectly competitive, based on the conclusions
drawn from the information in the industry profile. In a competitive market, each individual facility takes
the price as determined by the market because they do not have the power to set the market price of the
product. The approach used for this EIA assumes that the competitive market for each of the four
products determines both prices and quantities (U.S. EPA, 1999b).
In the short run, a firm with an existing plant will decide, based on the market price, how much
output to produce with its capital stock (e.g., fixed investment in the plant and equipment) considered as
4-3
-------
constant. In this decision, the firm considers the costs of inputs that vary with output levels, such as
materials and labor (U.S. EPA, 1999b). When the market price is equal to the average variable
(operating) costs, the firm is only recouping the cost of its variable inputs. When the price is less than the
firm's variable costs, it will no longer produce output because it cannot recover all of its variable costs in
the short run. When the market price exceeds variable costs, the firm can also recover a part of its fixed
investment in the plant and equipment. In the long run, the firm must cover all of its fixed investment in
the plant and equipment. Under this more stringent condition, the market price must exceed its average
total costs, which include capital and variable input costs. In this analysis, which is short-run in
timeframe, the model assumes that all firms wish to maximize its profits and will produce output using
their existing plants as long as the market price even marginally exceeds their variable costs of producing
output.
The market model developed by the Agency for this EIA is based on a series of equations that
represent the market supply and demand functions for a given product. The demand functions use
baseline price, total domestic production, import, and export data, as well as estimates of the price
elasticity of demand for a given product as inputs. The market supply function is based on an estimated
supply function for each plywood or wood composite product at all production facilities. The Agency
developed a spreadsheet model for each of the four markets to represent the conceptual model described
below. The EIA provides a description of this process.
Figure 4-1 shows a generalized upward-sloping supply function that characterizes the production
function of each facility included in the analysis (affected and unaffected). In the EIA model, this
function represents the marginal cost curve for each supplier of the product within the market. The
minimum constraint on this function is zero, and the maximum constraint is each facility's capacity. If
the market price is above a given facility's average variable cost, the facility will produce output up to the
point where production equals the facility's production capacity. If a given facility's marginal production
cost is above the market price, it will produce zero output (EPA, 1999b).
lbs/year
Figure 4-1. Supply Curves for Affected Facilities
Source: U.S. EPA, 1999b.
4-4
-------
The model then aggregates the supply functions of the individual facilities within a market to
represent the market level supply curve. Next, the model equilibrates the market demand curve and the
market supply curve according to the baseline market price and quantity values. Figure 4-2 shows that
market prices and quantities are set in the baseline according to the intersection of the supply and demand
curves in the baseline (the period prior to the imposition of compliance costs on affected facilities). The
baseline scenario equilibrium market price and quantity (P, Q) are the points on the graph's axes points
where the downward-sloping market demand curve (DM) intersects with the upward-sloping market
supply curve (SM). In the baseline, at price P, the industry produces total output, Q, with affected
facilities producing the amount qa and unaffected facilities producing qu(EPA, 1999b).
Next the same facility-specific supply functions in the model are recalculated after taking into
account the increase in production costs associated with the imposition of annual compliance costs on
some producers. The costs are expresses in terms of dollars per unit of output in the baseline (in this case
per thousand cubic meter of product). Figure 4-1(b) show the effect of the compliance costs: the supply
curve for the affected producers shift upward Sa to Sa' . This raises the point at which market price must
cover variable production costs from pa to pa'. The supply curve Su for the unaffected facilities remain the
same. When the supply curves for the affected and unaffected facilities are aggregated to represent the
market level supply curve, the market supply curve also shifts upward from SM to SM. Using the original
market demand curve, DM, the new equilibrium price increases from P to P' and market output declines
from Q to Q'. This reduction in market output is the net result from reductions at affected facilities and
+ P
Affected Facilities Unaffected Facilities
SM /
/1 \
! DM
1
i
Q
Market
a) Baseline Equilibrium
Affected Facilities
P
= P
%
Unaffected Facilities
l<-|
J L
Q' Q
Market
b) With Regulation Equilibrium
Figure 4-2. Market Equilibrium Without and With Regulation
Source: U.S. EPA, 1999b.
4-5
-------
increases at unaffected facilities (EPA, 1999b). For more information on the economic impact analysis
methodology, refer to Appendix B of this report.
4.5 Economic Impact Analysis Results
The following section presents the results of the Agency's implementation of the EIA models
described in the previous section. The results are presented at three levels: market, industry, and societal.
The market level results present the impacts in terms of changes in price and quantity for each of the four
markets. The industry level impacts include changes in revenues, production, employment, and number
of operating production lines.
4.5.1 Market-Level Results
Exhibit 4-2 presents the expected impacts of the regulation at the market-level. These changes
include the new price and quantity for each product and changes in foreign trade. In each market, prices
increase and production quantities decrease due to the imposition of compliance costs on the affected
producers. The reduction in market quantities of each product reflect reductions in domestic production
(including production for exports) and increases in foreign imports. The reduction in domestic production
reflects production decreases by affected producers and increases by unaffected producers (EPA, 1999b).
For softwood plywood, the market price is expected to increase by 0.9 percent, while market
quantity declines by 0.1 percent, or 26,207 thousand cubic meters (or M cubic meters) per year. For
oriented strandboard, the market price is expected to increase by 1.3 percent, while market quantity
declines by 0.1 percent, or 12,945 M cubic meters per year. For particleboard and medium density
fiberboard, the market price is expected to increase by 2.5 percent, while market quantity declines by 0.7
percent, or 78,595 M cubic meters per year. Finally, for hardboard, the market price is expected to
increase by 1.0 percent, while market quantity will decline by 0.3 percent, or 4,727 M cubic meters per
year.
Generally, increases in market price result in changes in foreign trade of these products: exports
decrease and imports increase. Exhibit 4-2 shows that exports of softwood plywood from the United
States are expected to decline by 0.2 percent (or 2 M cubic meters per year); exports of oriented
strandboard are expected to decline by 0.1 percent (or 0.2 M cubic meters per year); exports of
particleboard and medium density fiberboard are expected to decline by 0.7 percent (or 2 M cubic meters
per year); and exports of hardboard are expected to decline by 0.2 percent (or 1 M cubic meter per year).
Imports of each of these products to the United States are expected to increase as follows: SWPW by 685
percent (or 630 M cubic meters per year), OSB by 28.82 percent (or 1,345 M cubic meters per year),
PB/MDF by 85 percent (or 1,451 M cubic meters per year), and HB by 100 percent (or 387 M cubic
meters per year).
Ilxhihil 4-2. M;irke(-I.e\el Iinpiicls of (lie Proposed M.SIIAP
^j,!, ('Minifies from ISiiseline
Indusln Sector ISiiseline Kefiiihilion AMsoMile I'ereenl
Softwood Plywood
Market price (1997$/cubic meter)
$235
$237.20
$2.20
0.9%
Market output (M cubic meters/yr)
17,568,254
17,542,048
-26,206
-0.1%
Domestic production
17,568,162
17,541,326
-26,837
-0.2%
Affected Facilities
11,680,778
11,629,258
-51,520
-0.4%
Unaffected Facilities
5,887,384
5,912,067
24,683
0.4%
4-6
-------
IMiibil 4-2. Markc(-I.e\c
1 Impacts of (lie Proposed MISHAP
Willi
Changes from ISasclinc
Indusln Seclor
ISaselinc
Keuulalion
Absolute
Pcrcenl
Exports
1,370
1,368
-2
-0.2%
Imports
92
722
630
685.1%
Oriented Strandboard
Market price (1997$/cubic meter)
$185
$187.43
$2.43
1.3%
Market output (M cubic meters/yr)
9,595,121
9,582,176
-12,945
-0.1%
Domestic production
9,590,456
9,576,165
-14,291
-0.1%
Affected Facilities
4,691,645
4,654,117
-37,528
-0.8%
Unaffected Facilities
4,898,811
4,922,048
23,237
0.5%
Exports
147.8
147.6
-0.2
-0.1%
Imports
4,666
6,011
1,345
28.8%
Particleboard and Medium Density Fiberboard
Market price (1997$/cubic meter)
$169
$173.29
$4.29
2.5%
Market output (M cubic meters/yr)
11,646,228
11,567,633
-78,595
-0.7%
Domestic production
11,644,523
11,564,477
-80,046
-0.7%
Affected Facilities
9,670,639
9,553,510
-117,129
-1.2%
Unaffected Facilities
1,973,884
2,010,967
37,083
1.9%
Exports
333
331
-2
-0.7%
Imports
1,705
3,156
1,451
85.1%
Hardboard
Market price (1997$/cubic meter)
$1,322
$1,335.17
$13.17
1.0%
Market output (M cubic meters/yr)
1,768,930
1,764,203
-4,727
-0.3%
Domestic production
1,768,545
1,763,431
-5,114
-0.3%
Affected Facilities
1,768,545
1,763,431
-5,114
-0.3%
Unaffected Facilities
n/a
n/a
0
n/a
Exports
371
370
-1
-0.2%
Imports
385
772
387
100.5%
4.5.2 Industry-Level Results
Industry impacts associated with the proposed NESHAP for the plywood and composite wood
industry are presented in Exhibit 4-3. Industry-level impacts include changes in revenue, production,
numbers of operating product-lines, and changes in employment. The estimates for changes in production
lines and employment are described in Section 4.5.3 below.
The EIA model estimates the change in prices and production levels after the imposition of the
compliance costs on the affected facilities. Exhibit 4-3 shows that overall revenues for each industry
increase slightly. Industry revenues increase because demand elasticities for these four markets mean that
the change in price (in percentage terms) is greater than the percentage reduction in output. Affected
facilities will experience an increase in revenues due to the increase in the market price, but this effect is
likely to be offset by the increase in production costs. Changes in profits, however, could not be estimated
due to lack of information on production costs at the facility level.
4-7
-------
r.xhihil 4-3. lnduMn-l.e\el I ill p;ic(s of (ho Proposed M.SIIAP
('halites from liaseline
liaseline
Willi Regulation
\bsoluk'
Pel Veil I
Softwood Plywood
Revenues ($ million/yr)
Production (M cubic meters/yr)
Operating product lines (#)
Employment*
$4,128.5
17,568162
108
36,877
$4,161
17,541,326
107
36,821
$32.4 0.78%
-26,836 -0.15%
-1 -0.93%
-56 -0.15%
Oriented Strandboard
Revenues ($ million/yr)
Production (M cubic meters/yr)
Operating product lines (#)
Employment*
$1,774.2
9,590,456
38
6,681
$1,794.9
9,576,164
38
6,671
$20.7 1.17%
-14,292 -0.15%
0 0.00%
-10 -0.15%
Particleboard & Medium Density Fiberboard
Revenues ($ million/yr) $1,967.9
Production (M cubic meters/yr) 11,644,523
Operating product lines (#) 83
Employment* 20,424
$2,004.0
11,564,477
83
20,284
$36.1 1.83%
-80,046 -0.69%
0 0.00%
-140 -0.69%
Hardboard
Revenues ($ million/yr)
Production (M cubic meters/yr)
Operating product lines (#)
Employment*
$2,338.0
1,768,545
18
6,271
$2,354.5
1,763,431
18
6,252
$16.5 0.71%
-5,114 -0.29%
0 0.00%
-18 -0.29%
*Baseline employment estimates are based total production and employment of all facilities in each market as
reported in the EPA facility database. Average production per employee was calculated using the sum of
facility-specific production and employment. Because facilities reported employment as a range, they are
assigned an employment estimate using the mid-point of the reported range. Post-regulation employment was
then estimated by dividing post-regulation production by the baseline production per employee.
4.5.3 Distribution of Impacts
The distribution of regulatory impacts is presented in Exhibit 4-4. This table presents the same
information as in Exhibit 4-3, but provides details on how the rule impacts affected and unaffected
facilities differently.
One important result from the EIA model is the projection of process line closures that could
occur following promulgation of the proposed NESHAP rule. The model's estimate of process line
closures may be sensitive to the accuracy of the baseline characterization of the facilities. Characteristics
such as baseline production levels, revenues (as a function of production and price), the underlying
supply function that represents production costs, and the compliance cost estimates are all factors that
affect the distribution of the rule's impacts.
4-8
-------
llxhihil 4-4: Distribution ol' lndustn-l.e\el 1 inp;icls of Proposed M.SIIAP:
Affected and I nalTecled Producers
Willi
Changes l"r
oni ISaseline
ISaseline
Regulation
Absolute
Percent
Softwood Plywood
Affected Process Lines
Revenues ($ million/yr)
$2,745.0
$2,758.5
$13.5
0.49%
Production (M cubic meters/yr)
11,680,778
11,629,258
-51,520
-0.44%
Operating process lines (#)
66
65
-1
-1.52%
Employment*
24,519
24,411
-108
-0.44%
Unaffected Process Lines
Revenues ($ million/yr)
$1,383.5
$1,402.4
$18.9
1.37%
Production
5,887,384
5,912,067
24,683
0.42%
Operating process lines (#)
42
42
0
0.00%
Employment*
12,358
12,410
52
0.42%
Oriented Strandboard
Affected Process Lines
Revenues ($ million/yr)
$868.0
$872.3
$4.4
0.50%
Production (M cubic meters/yr)
4,691,645
4,654,117
-37,528
-0.80%
Operating process lines (#)
20
20
0
0.00%
Employment*
3,268
3,242
-26
-0.80%
Unaffected Process Lines
Revenues ($ million/yr)
$906.3
$922.6
$16.3
1.80%
Production
4,898,811
4,922,048
23,237
0.47%
Operating process lines (#)
18
18
0
0.00%
Employment*
3,413
3,429
16
0.47%
Particleboard & Medium Density Fiberboard
Affected Process Lines
Revenues ($ million/yr)
$1,634.3
$1,655.6
$21.2
1.30%
Production (M cubic meters/yr)
9,670,639
9,553,510
-117,129
-1.21%
Operating process lines (#)
53
53
0
0.00%
Employment*
16,962
16,756
-205
-1.21%
Unaffected Process Lines
Revenues ($ million/yr)
$333.6
$348.5
$14.9
4.47%
Production
1,973,884
2,010,967
37,083
1.88%
Operating process lines (#)
30
30
0
0.00%
Employment*
3,462
3,527
65
1.88%
4-9
-------
llxhihil 4-4: Dislrihulion ol' lii(lus(r\-l.c\cl Impaels of Proposed M.SIIAP:
AITeeled and I nalTeeled Producers
Willi
Changes l"r
oni ISaseline
liaseline
Regulation
Absolute
Perce n (
Hardboard
Affected Process Lines
Revenues ($ million/yr)
$2,338.0
$2,354.5
$16.5
0.70%
Production (M cubic meters/yr)
1,768,545
1,763,431
-5,114
-0.29%
Operating process lines (#)
18
18
0
0.00%
Employment*
6,271
6,252
-18
-0.29%
Unaffected Process Lines
Revenues ($ million/yr)
n/a
n/a
n/a
n/a
Production
n/a
n/a
n/a
n/a
Operating process lines (#)
n/a
n/a
n/a
n/a
Employment*
n/a
n/a
n/a
n/a
Total
Affected Process Lines
Revenues ($ million/yr)
$7,585.3
$7,640.9
$55.6
0.73%
Production (M cubic meters/yr)
27,811,607
27,600,316
-211,291
-0.76%
Operating process lines (#)
157
156
-1
-0.64%
Employment*
51,020
50,662
-358
-0.70%
Unaffected Process Lines
Revenues ($ million/yr)
$2,623.4
$2,673.4
$50.1
1.91%
Production
12,760,079
12,845,082
85,003
0.67%
Operating process lines (#)
90
90
0
0.00%
Employment*
19,233
19,366
133
0.69%
All Process Lines (net)
Revenues ($ million/yr)
10,209
10,314
$105.6
1.03%
Production
40,571,686
40,445,398
-126,288
-0.31%
Operating process lines (#)
247
246
-1
-0.40%
Employment*
70,252
70,028
-225
-0.32%
n/a = not applicable
*Baseline employment estimates are based total production and employment of all facilities in each market as
reported in the EPA facility database. Average production per employee was calculated using the sum of
facility-specific production and employment. Because facilities reported employment as a range, they are
assigned an employment estimate using the mid-point of the reported range. Post-regulation employment was
then estimated by dividing post-regulation production by the baseline production per employee.
Exhibit 4-4 shows that one softwood plywood production line is expected to prematurely cease
operations. The model does not predict any closures in all of the other product markets. The affected
entity that closes following adoption of the regulation is a small SWPW producer that incurs higher
control costs per unit of production than other SWPW production lines. This facility, with one process
line, did not respond to the EPA's ICR survey. Therefore, it was necessary to estimate baseline
production and compliance costs based on very little facility-specific information. Total baseline
4-10
-------
production of SWPW by the facility that closes was roughly 9,500 M cubic meters per year, or less than
0.05 percent of the 17,568,162 M cubic meters of SWPW produced by all SWPW facilities during 1997.
As a result of the closure and reductions in production at other affected facilities, the EIA model
estimates a total net loss in employment of 225 employees (0.3 percent) attributable to the proposed
NESHAP across all four market sectors. Affected facilities experience employment loss of 358, which is
offset by employment gains at unaffected facilities of 133. Overall the four sectors together experience a
1 percent increase in revenues, and a 0.3 percent decrease in production.
4.5.4 Social Costs of the Proposed NESHAP
The social costs of a regulation are measured according to the impacts that it has on both
consumers and producers. The proposed NESHAP rule, because it is expected to result in changes in
both market price and market quantity, will impact the consumers and producers of softwood plywood
and composite wood products. Social costs, also called welfare impacts, are the measure of overall gains
and losses experienced by the two groups that may result from the imposition of costs associated with the
regulatory requirements.
The economic benefits producers experience when participating in a market is called producer
surplus. Producers experience impacts when their revenues change either because of a new market price,
increased production costs, or both. These impacts change the amount of producer surplus relative to the
baseline. Likewise, the economic benefits consumers experience is called consumer surplus. Consumer
surplus changes when consumers experience impacts when the amount of product they consume or the
product price changes.
The estimate of the social cost of the proposed NESHAP rule, presented in Exhibit 4-5, is the sum
of the change in producer and consumer surplus. The EIA model estimates the social cost of the proposed
NESHAP as $134.2 million annually (1999 dollars). These costs are distributed across both consumers
and producers of plywood and composite wood products according to projected changes in market price
and quantity associated with the proposed NESHAP.
Consumer surplus is reduced by $135.1 million annually due to the increase in prices and
reductions in consumption. Consumers of softwood plywood are worse off by $38.7 million annually;
consumers of oriented strandboard are worse off by $23.3 million annually; consumers of particleboard
and medium density fiberboard and of hardboard are worse off annually by $49.8 million and $23.3
million, respectively.
Producers (in aggregate) are slightly better off because of the imposition of the proposed
NESHAP, with an increase in producer surplus of just under $1 million annually. Essentially all of this is
change associated with domestic producers. Because the market prices for these products increase,
certain individual domestic producers gain at the expense of their competitors. The size of this gain at
any given facility depends on how much their production costs change (as a result of new compliance
costs) relative to the change in market price and quantity that increases revenues. The benefit to foreign
producers associated with higher market prices is quite small, under $100,000.
4-11
-------
I.xhihil 4-5: Dislrihiilion of Social Costs Associaled willi (lie Proposed M.SIIAP
( liange in Value (S million -1 *>*>*>
Stakeholder dollars)
Social Costs of Regulation ^ ^4 2
(Change in consumer surplus + Change in producer surplus)
Total Change in Consumer Surplus $-135.1
SWPW Consumers $-38.7
OSB Consumers $-23.3
PB/MDF Consumers $-49.8
HB Consumers $-23.3
Total Change in Producer Surplus $0.9
Softwood Plywood
Producer Surplus, total
Domestic producers
Affected Facilities
Unaffected Facilities
Foreign producers
Oriented Strandboard
Producer surplus, total
Domestic producers
Affected Facilities
Unaffected Facilities
Foreign producers
Particleboard & Medium Density Fiberboard
Producer surplus, total
Domestic producers
Affected Facilities
Unaffected Facilities
Foreign producers
Hardboard
Producer surplus, total $-0.8
Domestic producers $-0.8
Affected Facilities $-0.8
Unaffected Facilities n/a
Foreign producers $0.0
$8.3
$8.3
$-4.1
$12.4
$0.0
$-1.4
$-1.4
$-12.9
$11.5
$0.0
$-5.2
$-5.2
$-13.5
$8.3
$0.0
4.5.5 Energy Impact Analysis
Executive Order 13211, "Actions Concerning Regulations That Significantly Affect Energy
Supply, Distribution, or Use" (66 FR 28355, May 22, 2001), provides that agencies shall prepare and
submit to the Administrator of the Office of Information and Regulatory Affairs, Office of Management
4-12
-------
and Budget, a Statement of Energy Effects for certain actions identified as "significant energy actions."
Section 4(b) of Executive Order 13211 defines "significant energy actions" as "any action by an agency
(normally published in the Federal Register) that promulgates or is expected to lead to the promulgation
of a final rule or regulation, including notices of inquiry, advance notices of proposed rulemaking, and
notices of proposed rulemaking: (1) (i) that is a significant regulatory action under Executive Order 12866
or any successor order, and (ii) is likely to have a significant adverse effect on the supply, distribution, or
use of energy; or (2) that is designated by the Administrator of the Office of Information and Regulatory
Affairs as a significant energy action." The proposed rule is not a "significant energy action" because it is
not likely to have a significant adverse effect on the supply, distribution, or use of energy. The basis for
the determination is as follows.
As stated in Chapter 2, this proposed rule affects manufacturers in the softwood veneer and
plywood (NAICS 321212), reconstituted wood products (NAICS 321219), and engineered wood products
(NAICS 321213) industries. There is no crude oil, fuel, or coal production from these industries.
Hence, there is no direct effect on such energy production related to implementation of this proposal. In
fact, as mentioned in section IV.D. of this preamble, there will be an increase in energy consumption, and
hence an increase in energy production, resulting from installation of regenerative thermal oxidizers
(RTOs) and wet electrostatic precipitators (WESPs) likely needed for sources to meet the requirements of
the proposed rule. This increase in energy consumption is equal to 718 million killowatt-hours/year
(kWh/yr) for electricity and 45 million cubic meters/year (m3/yr) for natural gas. These increases are
equivalent to 0.012 percent of 1998 U.S electricity production and 0.000001 percent of 1998 U.S. natural
gas production.18 It should be noted, however, that the reduction in demand for product output from these
industries may lead to a negative indirect effect on such energy production, for the output reduction will
lead to less energy use by these industries and thus some reduction in overall energy production.
For fuel production, the result of this indirect effect from reduced product output is a reduction of
only about 1 barrel per day nationwide, or a 0.00001 percent reduction nationwide based on 1998 U.S.
fuel production data19. For coal production, the resulting indirect effect from reduced product output is a
reduction of only 2,000 tons per year nationwide, or only a 0.00001 percent reduction nationwide based
on 1998 U.S. coal production data. For electricity production, the resulting indirect effect from reduced
product output is a reduction of 42.8 million Kilowatt-hours per year (kWh-yr), or only a 0.00013 percent
reduction nationwide based on 1998 U.S. electricity production data. Given that the estimated price
increase for product output from any of the affected industries is no more than 2.5 percent, there should
be no price increase for any energy type by more than this amount. The cost of energy distribution should
not be affected by this proposal at all since the rule does not affect energy distribution facilities. Finally,
with changes in net exports being a minimal percentage of domestic output (0.01 percent) from the
affected industries, there will be only a negligible change in international trade, and hence in dependence
on foreign energy supplies. No other adverse outcomes are expected to occur with regards to energy
supplies. Thus, the net effect of this proposed rule on energy production is an increase in electricity
output of 0.012 percent compared to 1998 output data, and a negligible change in output of other energy
types. All of the results presented above account for the passthrough of costs to consumers, as well as the
cost impact to producers. These results also account for how energy use is related to product output for
18U.S. Department of Energy, Energy Information Administration. Annual Energy Review, End-Use
Energy Consumption for 1998. Located on the Internet at http://www.eia.doe.gov/emeu/aer/enduse.html.
Ibid.
4-13
-------
the affected industries.20 More detailed information on the estimated energy effects and the methodology
employed to estimated them are in the background memo21 that provides such details for the proposed
rule.
Therefore, we conclude that the proposed rule when implemented is not likely to have a
significant adverse effect on the supply, distribution, or use of energy.
4.6 Analysis of Economic Impacts on Engineered Wood Products Sector
4.6.1 Overview
The engineered wood products (EWP) sector is much different than the other four sectors
examined in this analysis. Due to both the nature of the products and their markets as mentioned in
Section 4.6.4 below, and the limited availability of data, a quantitative analysis was not performed.
EWP products are characterized by differentiated structural beam products, although some can
also be used as columns. EWPs are produced by firms using differentiated production systems. There are
a number of large firms with market power, allowing them to set prices. Although many of the products
in this market sector can be substituted for one another to an extent, each product's design usually
provides it with an advantage over others in the design of final structures. Product differentiation allows
each product to dominate particular market niches, but also makes many of the products complements as
well as substitutes. Even when substitution is possible, some large firms use their market power to
influence purchase decisions and persuade consumers to purchase their products. Consequently, these
products are not standard commodities as are products in the other plywood and composite wood sectors.
Therefore, commodity prices and other standard information that is characteristic of competitive markets
are not available. These factors limit the ability of EPA to quantify the impact of the proposed NESHAP
rule on this sector using the same quantitative method applied to the other sectors.
Three of the fifty-three EWP facilities in the U.S. will have compliance costs are a result of the
proposed NESHAP rule. This includes the country's two LSL plants, and one of the country's two PSL
plants. Weyerhaeuser, Inc. purchased these four plants when they acquired Trus Joist MacMillan.
(Weyerhaeuser-Trus Joist MacMillan will be referred to as W/TJM for the remainder of this report.)
W/TJM is currently the only manufacturer of LSL and PSL in the world.
4.6.2 Characteristics of EWP Products
EWPs include laminated veneer lumber (LVL), laminated strand lumber (LSL), parallel strand
lumber (PSL), wood I-joists (I-J), and glue laminated timber (GL). These are value added lumber
products designed to be used in applications that are not suitable for ordinary framing lumber. The design
and composition of each of these products differ, providing them with different strength, stiffness, cost,
and dimensional properties. Strength properties allow products to carry heavier loads without breaking.
Strength is particularly important for uses such as carrying beams in floors or bridge girders, which are
used to support structures. Stiffness prevents materials from shaking as objects are moved over them.
This is particularly important in floors and bridge decks where structure shakes or squeaks are
20 U.S. Department of Energy, Energy Information Administration. 1998 Manufacturing Energy
Consumption Survey. Located on the Internet at
http://www.eia.doe.gov/emeu/mecs/mecs98/datatables/contents.html.
21 U.S. Environmental Protection Agency. "Energy Impact Analysis of the Proposed Plywood and
Composite Wood Products NESHAP." July 30, 2001.
4-14
-------
uncomfortable or dangerous. Each product's strength and stiffness is partially a function of its
dimensional characteristics. For example, I-J are strong, stiff and typically less expensive than the other
products. However, their structural integrity does not hold beyond certain lengths.
Laminated-Strand Lumber
W/TJM introduced LSL in 1992. Its dimensional characteristics match nominal 2x4 and 4x4
framing lumber, but it provides greater and more uniform strength and stiffness properties.22 Unlike
many of the other EWPs, one of the primary uses of LSL is in vertical applications, as a stud or column.
PSL is also used in vertical applications, but its primary uses are in horizontal applications, as a beam.
The uses of LSL are summarized in Exhibit 4-6.
Lxhihil 4-(>: Priiium I son iinri Substitutes lor LSI.
Application
I sos
Substitutes
Columns or studs
Headers
Beams
Rim Board
Wall, window, and door framing
Framing lumber, PSL, solid sawn
lumber, steel
Garage door, other wide span doors and Framing lumber, GL, LVL, PSL, steel
windows
Light applications, low load bearing Solid sawn lumber, GL, LVL, PSL, 1-J
Floor systems, nailing surface for Plywood
sheathing, decks and siding
Source: http://www.trusjoist.com
Because W/TJM is the sole producer of LSL, very little information is available about their costs
or profits. Exhibit 4-7 below identifies the production information that is known. Note that neither plant
operates near full capacity. The two plants combined operate at less than 50 percent of their combined
capacity (EWP survey). This may be due to a decline in demand or an immature market.
IMiihil A--*:
( hiimclcrislics of LSL Phinls
Locution
Item
Deerwood, MN
Capacity
7,900,000 ft3/yr
Production
4,900,000 ft3/yr
Capacity
62%
Employees
100-249
Plant Age
1992 (estimate)
Chavies, KY
Capacity
14,900,000 ft3/yr
Production
5,400,000 ft3/yr
Capacity
36%
Employees
250-499
Plant Age
1992 (estimate)
Source: EPA facility survey, f998.
22 LSL bending strength = f700psi; stiffness (modulus of elasticity or MOE) = 1.3-1.5E.
4-15
-------
Parallel-Strand Lumber
PSL is a slightly older technology than LSL. It was introduced to the market in the mid-1980's
by W/TJM. PSL is a high performance beam and header product that can also be used as a column. It is
designed with exceptional strength, stiffness characteristics that is capable of spanning much longer
distances than most other EWPs.23 Only GL can match or beat PSL's ability to carry heavy loads over
long spans.24 Furthermore, PSL is a balanced beam, which means it has no top or bottom. This property
is particularly desirable to the residential framing industry. Balanced beams save on labor costs due to
lower skill required to install the product, providing PSL with an edge over GL for a number of years.
However, Anthony Forest Products recently developed a line of balanced GL beams with comparable
strength and stiffness properties to PSL, and a number of other GL manufacturers have since done the
same. This has lead to a competitive more environment. However, PSL still maintains a share of the
residential market. Exhibit 4-8 summarizes the uses of PSL.
Ilxhihil 4-X: Prim;ir\ I son ;iii(I Suhsiiiiiios lor PS I.
Application
I SOS
Suhsliliilcs
Columns or studs
Wall framing, street lights
Steel, solid sawn lumber, steel
Headers
Garage door, other wide span doors and
GL,LVL, steel
windows
Beams
Heavy applications, high load bearing
GL, steel, solid sawn lumber, LVL
Because W/TJM has been the only producer of this product, very little is known about their
production costs and profits. Exhibit 4-9 below identifies the production information that is known about
the two U.S. PSL production facilities.
Ilxhihil 4-*>: ( h;ir;icU'rislics of PS I. Pliinis
Locution
110111
Colbert, GA
(affected by rule)
Buckhannon, WV
(not affected by rule)
Capacity
Production
Capacity Factor
Employees
Plant Age
Capacity
Production
Capacity Factor
Employees
Plant Age
3,000,000ft7yr
2,770,000ft3/yr
92%
250-499
mid-1980s (estimate)
2,500,000ft3/yr
f,929,000ft3/yr
77%
250-499
mid-f 980s (estimate)
source: EPA facility survey, 1998.
Since PSL does compete directly with GL in some applications, there is slightly more information
available on it than on LSL. There is one published source of EWP prices that includes some retail PSL
delivered prices: Engineered Lumber Trends produced by The Irland Group in Winsor, ME. Exhibit 4-10
presents a comparison of retail prices for PSL and GL. However, it is not possible to estimate firm
23 PSL bending strength = 2900psi; stiffness = 2.0E.
24 GL bending strength = 2400-3000psi; stiffness = 1.8 - 2.fE
4-16
-------
revenue using these prices because they are retail prices and there is considerable markup from the
producer to retail level. The prices listed in the table below show that PSL is more expensive than GL.
r.\hihi( 4-10: Kciiiil Prices of (. I. ;ind PS I. ISoiims
Dclixcrcd lo l.os Angeles
lie;i in ¦ Width Dcplh Price
(inches) (inches) per linc;ir Tool
PSL
(strength = 2900psi 3-1/2 11-7/8 $8.16
stiffness = 2.0E)
GL
(strength = 2400psi 3-1/8 12 $5.90
stiffness = 1.8E)
GL **
(strength = 3 OOOpsi 3-1/8 12 $7.08
stiffness = 2. IE)
Source: Engineered Lumber Trends (December 1999)
*Beam bending strength measured in pounds per square inch and
stiffness the beams modulus of elasticity or E
**Price of the 3000-psi GL is estimated by marking up the 2400psi
GL by 20% based on conversations with GL manufacturers
W/TJM has stated that PSL could be manufactured using waste material from plywood and LVL.
Other industry members dispute this due to the high quality material required for the product. They feel
that there is actually considerable wasted material in the production process in order to acquire the high-
quality wood fiber. Consequently, there are two conflicting arguments why PSL is able to sell their
product at the higher price.
Consumers view it as a superior product despite the availability of substitutes, providing it with a
higher profit margin.
W/TJM has used strategic behavior to maintain the price it requires to provide its product on the
market.
4.6.3 Comparison to Other EWPProducts
Substitution among EWPs is a complicated dynamic. While the products can be viewed as
distinct commodities, they are integral components in the engineering design of a structure. Therefore, it
is possible for two EWP products to be both substitutes and complements. This is particularly true for
W/TJM's products because they sell both the individual products and entire framing systems that include
either LSL or PSL, or both.
LSL is more comparable to framing lumber and solid sawn lumber than most other engineered
wood products. As a header it competes with LVL, GL, and PSL. However, as a column or stud, it is
fairly unique. It is also a very new technology, which may partially explain why competing firms have
not developed comparable products to date. Furthermore, LSL is marketed as a complementary product
to W/TJM's other engineered wood products (I-J, LVL, PSL) in various structures they design.
As an individual product, PSL has more direct substitutes than LSL. In shorter spanning
applications (such as headers or floor joists) LVL, I-J, GL, or LSL can be substituted for each other
4-17
-------
depending on the application. In longer spanning applications, PSL competes with GL and steel. W/TJM
also designed PSL to complement their other products in pre-designed framing systems. If an architect or
engineer purchases the pre-designed framing system, PSL will be specified in the system. This eliminates
the possibility of substitutes in these applications. These characteristics are discussed in greater detail
below.
4.6.4 Market Ch aracterization
The EWP sector is characterized by imperfect competition among firms, particularly. In the case
of LSL and PSL, W/TJM is a monopoly producer because it operates all four facilities that produce
products affected by the rule. Although a number of other products can be substituted for both LSL and
PSL at the commodity level, W/TJM has a monopoly on these two products. In addition, they have some
ability to control material choice within the residential framing market because they produce Computer
Automated Design (CAD) programs used by architects, engineers, and lumber product distribution
companies to design structures. The individual using the program is actually purchasing a complete
architectural design, which competes with other architectural designs on the basis of total design cost.
Other EWP companies have not developed such programs to the same extent that W/TJM has.
By pre-specifying its products as material inputs in these design packages, W/TJM assures that
they are used in the structure. Moreover, the individual product prices are less important to W/TJM
because the price the buyer using the program is concerned with is the price of the whole structure.
Further, because W/TJM makes all the products used in the design, they can sell any one of the
components at a loss provided the loss is made up by profits on the whole structure. In addition, W/TJM
also sells pre-designed floor, roof, and wall systems that incorporate its EWP products.
Therefore, W/TJM is a price setter in both these markets. The high price of PSL allows producers
of comparable GL a substantial markup over its costs. However, GL manufacturers offered the opinion
that PSL could not raise its price without suffering a loss in demand . Those same producers believe that
PSL is already selling at a loss to promote the sale of its other products, although this information cannot
be confirmed. Another complicating factor is the purchase of TJM by Weyerhaeuser, Inc. Even if TJM
had been selling PSL at a loss, Weyerhaeuser may not be willing to do this, particularly if costs increase
due to the proposed NESHAP rule.
4.6.5 Appropriateness of Partial Equilibrium Economic Analysis for the EWP Market
It is not appropriate to use a partial equilibrium analysis to quantify the impact of the rule on the
EWP sector. The EWP facilities that will have compliance costs due to the rule are not part of a
competitive market. W/TJM has a monopoly on both PSL and LSL because they are the only company in
the world producing both these products. Although the products compete to an extent with other EWPs,
W/TJM is able to use its market power to influence consumer purchasing decisions and prices. Also,
W/TJM may be able to absorb cost increases on one product with profits from another.
4.6.6 Analysis
The estimated total annual compliance costs for the three impacted EWP facilities are $3.2
million, which is only 2.4 percent of the overall cost of the rule of $142 million per year. Moreover, the
costs are incurred by one of the largest wood products companies in the world. The costs incurred by
W/TJM are less than 1 percent the corporation's total annual revenues, which now include revenues from
the facilities purchased from TJM. Although the specific impact on W/TJM's profitability cannot be
determined, it is likely that W/TJM can afford the costs at the corporate level.
4-18
-------
The Agency cannot be certain of the impact that these cost increases will have on the markets for
LSL and PSL. W/TJM could decide use their price setting power to increase the price of PSL and/or
LSL. This could lead to a decrease in consumer surplus. However, it is important to note the
distributional impact of this potential price increase. EWPs are often used in high end residential and
commercial construction. They typically provide luxury features, such as squeak proof floors, large open
rooms with high ceilings and no support columns.
Alternatively, W/TJM may choose not to raise the prices of PSL and LSL. For example, PSL is
already more expensive than GL, a substitute product. A rise in PSL prices could lead to a loss in market
share. Furthermore, production data showed that the PSL market is functioning at about 50 percent of its
full capacity. These factors indicate that a PSL price increase may not be favorable to W/TJM. W/TJM
could choose to absorb the cost increase with profits from its other products in the short run. In the long
run they could abandon production of these products. In order to meet the demand for their pre-designed
structural systems, they could either produce or buy a substitute product to use in the structural systems.
If W/TJM chooses to shut down these plants, there may be a number of unfavorable short run
impacts. First, the closures could mean the loss of 600-1250 jobs, spread over three communities. EPA
expects that the reduction in employment at these facilities would be offset by an increase in employment
at a competing firm, but the community impacts may remain.
The impact of the proposed NESHAP rule on both consumers and the individual facilities and
firms will depend upon corporate strategy. Given the acquisition of Trus Joist MacMillan by
Weyerhaeuser, corporate decisions and long term strategy are more influential than what could be
represented in a model.
4.6.7 Conclusions
Although the specific impacts to the EWPs sector cannot be determined, it is unlikely that
substantial economic losses will result. The cost burden of this sector is minimal in comparison to the
other sectors. Furthermore, the affected facilities are all owned by W/TJM, which has sufficient resources
to handle the compliance costs. Even if W/TJM passes this cost on to consumers, the price increase is
likely to be minimal as substitute products are available.
4.7 References
Abt Associates Inc. 1999. Profile of the Plywood and Wood Composite Industries. Prepared for Larry
Sorrels, Innovative Strategies and Economic Group, Air Quality Strategies and Standards
Division, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC. December
Abt Associates Inc. 2000. Economic Analysis Methodology for the Plywood and Wood Composite
Industries. Prepared for Larry Sorrels, Innovative Strategies and Economic Group, Air Quality
Strategies and Standards Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC. March.
U.S. Environmental Protection Agency. 1998. Industry Specific Information Collection Request (ICR) for
the Development of Plywood and ParticleboardMaximum Achievable Control Technology
(MACT) Standards. Air Quality Strategies and Standards Division, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
4-19
-------
U.S. Environmental Protection Agency. 1999a. OAQPSEconomic Analysis Resource Document. Air
Quality Strategies and Standards Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1999b. Economic Impact Analysis for the Polymers & Resins III
NESHAP. Air Quality Strategies and Standards Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
4-20
-------
5 SMALL BUSINESS IMPACTS
Under the requirements of the Regulatory Flexibility Act (RFA) of 1980, Federal regulatory
agencies must give special consideration to small entities that are affected by regulatory actions. In 1996,
amendments to the RFA under the Small Business Regulatory Enforcement Fairness Act (SBREFA)
added certain requirements associated with analyses and procedures associated with determining whether
a regulatory action will have a significant impact on a substantial number of small entities (U.S. EPA,
1999a, 1999b).
5.1 Results in Brief
The screening analysis of small business impacts presented in this chapter indicates that 17 of the
83 businesses affected by this proposed rule are small. Of these 17 small firms, ten have annual
compliance costs of 1 percent or greater of their sales. Of these 10 firms, three have annual compliance
costs of 3 percent or greater of their sales. Of the 32 facilities owned by these 17 small firms, only one is
predicted to close in order to avoid incurring costs associated with compliance with the proposed rule.
This analysis supports a certification of no significant impact on a substantial number of small entities
(SISNOSE) for this rule because, while a few small firms may experience significant impacts, there will
not be a substantial number incurring such a burden.
5.2 Introduction
The proposed NESHAP rule for the plywood and composite wood industries will affect the
owners of the facilities that will incur compliance costs to control their HAP emissions. The owners,
either firms or individuals, are the entities that will bear the financial impacts associated with these
additional operating costs. The proposed rule has the potential to impact all firms owning affected
facilities, both large and small. This section presents the results of EPA's small business impact analysis
of the impact the proposed NESHAP for the Plywood and Composite Wood Industries on small
businesses. The analysis was performed in two stages: a screening-level analysis, and an examination of
impacts on small businesses developed using the models described in Chapter 4.
The screening analysis provides EPA with a preliminary estimate of the magnitude of impacts the
proposed NESHAP may have on the ultimate domestic parent companies that own facilities EPA expects
to be impacted by the standard. The analysis focused on small firms because they may have more
difficulty complying with a new regulation or affording the costs associated with meeting the new
standard. This section first describes the data sources used in the screening analysis, the methodology we
applied to develop estimates of impacts, the results of the analysis, and conclusions about the results. The
results of the impact assessment specific to small businesses that own affected facilities follows. Detailed
documentation of the screening analysis presented in this section is contained in the economic impact
analysis (EIA).
5.3 Screening Analysis Data Sources
The screening analysis was based on the following information:
Industry Specific Information Request (ICR) for the Development of Plywood and Particleboard
MACT Standards (U.S. EPA, 1998).
Profile of the Plywood and Composite Wood Products Industries (Abt Associates, January 2000).
Estimated Nationwide Costs of Control: Plywood and Composite Wood Products (MRI, April 14,
2000).
5-1
-------
Dun & Bradstreet, Ward's Business Directory, and Internet research on facility and firm
employment and sales.
5.4 Screening Analysis Methodology
Cost Analysis
After summing annual compliance costs for each affected facility associated with a given parent
firm, EPA developed ratios of firm-level compliance costs to firm-level sales. The preparation of cost to
sales ratios is a typical part of screening analyses such as this one. The analysis incorporated firm level
compliance costs from the most recent facility compliance cost estimates for all affected facilities. The
firm sales data for the majority of the firms with affected facilities was obtained through a search of the
Dun & Bradstreet (D&B) company database. The information obtained from D&B was supplemented
with data from firm web sites and several other business databases available through the Internet (e.g.,
Zapdata, Hoover's Online, Thomas' Register, and Lycos Companies Online).
Profitability Analysis
For the second step in the screening analysis, EPA reviewed the profitability of only those firms
with affected facilities with a cost to sales ratio (C/S) greater than 1 percent. Specifically, EPA examined
a measure of profitability called the net profit margin, also known as the return on sales ratio (R/S),
calculated by dividing net profit after taxes by annual net sales. Profitability data is not publicly available
for those affected firms with cost /sales ratios greater than 1 percent because they are all privately held
firms. EPA estimated profitability according to industry-wide average profitability measures available
from public sources. Exhibit 5-1 presents the profitability measure of R/S by product category.
Ilxhihil 5-1: Nol Profit M;iriiins In Product l \|K'
h'uducl ( ak'ums
Softwood, Plywood, and Veneer
Oriental Strandboard
Other Wood Composites
|w~ kcliii'iioii Sales kalio
1.7
3.5
3.5
Engineered Wood Products* 5.0
Multiple Processes** 2.6
Notes:
Source: Dun & Bradstreet (1999). Indicator values are based on median
values of the industrial sample.
*Includes 1998 data for Structural Wood Members, the only data
reported for this sector.
"Firms with multiple product lines were assigned an R/S ratio based on
the average of the R/S ratios for firms with Softwood Plywood and
Other Wood Composite product lines.
5.5 Screening Analysis Assumptions
Because there were certain gaps in the data, EPA had to make assumptions regarding some
affected firms' employment and sales data as follows.
For those affected facilities for which no parent company data were available, EPA assumed that
the facility represents the ultimate parent. EPA then assumed that the employees or the sales
associated with the facility were the same for the parent firm.
5-2
-------
For 8 affected firms for which Dun & Bradstreet, Ward's, or Internet sales data was not available
at the firm or facility level, EPA applied a sales estimate based on the average sales for firms in
the same product and employee size category.
• For one affected firm, EPA obtained employment information for all three of its identified
facilities and sales information for two of the three. EPA extrapolated the sales data on a dollar
per employee basis from the two facilities with complete data to the facility with only
employment data and added those sales to estimate the firm's total sales.
5.6 Screening Analysis Results
Cost Analysis
Based on the results of the C/S ratio test, EPA developed the following summary information.
The total number and percent of affected firms with C/S ratios greater than 3 percent.
The total number and percent of affected small firms with C/S ratios greater than 3 percent.
The total number and percent of affected firms with C/S ratios greater than 1 percent.
The total number and percent of affected small firms with C/S ratios greater than 1 percent.
The median and mean C/S ratios for the following groups of firms:
• All firms.
• Small firms.
Firms owning affected softwood plywood and veneer facilities.
Firms owning affected oriented strand board facilities.
Firms owning affected other wood composite products facilities.25
Firms owning affected engineered wood product facilities.
Firms owning affected facilities that make multiple products.
The screening analysis showed that of the 52 firms that own facilities incurring capital and
monitoring, recordkeeping , and reporting (MRR costs), 17 of them (33 percent) are small firms according
to the U.S. Small Business Administration's "Small Business Size Standards Matched to NAICS Codes"
(U.S. SBA, 2000). Small firms with affected facilities had a median C/S of 1.22 percent. The remaining
35 firms (67 percent) are large, with a median C/S ratio of 0.33 percent. Overall, the weighted median
C/S ratio for all firms is 0.62 percent. Exhibit 5-2 summarizes this information, along with the mean,
maximum and minimum C/S by firm size category.
25Medium density fiberboard, hardboard, conventional particle board and molded particleboard.
5-3
-------
llxhihil 5-2:
AIToclod Firms In Si/o
Nil ill hoi' of
Pcrconl ol'
I'irin
AIToclod
l oliil AIToolod
Moduli C/S
Mosul
Miixiimi in
Minimum
Si/o1
l"irms:
linns
Kill i«i
C/S Kiilio
C/S Kiilio
C/S Kiilio
Small
17
33.0%
1.2%
2.3%
8.3%
0.53%
Large
35
67.0%
0.3%
0.6%
5.1%
0.01%
Total/
52
100.0%
0.6%
1.2%
8.3%
0.01%
Weighted
Average
Notes:
'For those firms for which firm size information was not available, EPA assumed a typical firm size
within the product type category.
2Based on affected facilities only. Includes the firm that has capital costs but no annual costs.
When screened by process type,26 EPA found that the affected softwood plywood and veneer
firms (40 percent of all firms), other composite wood firms (25 percent of all firms) and firms with
multiple processes (29 percent of all firms) make up the majority of affected firms. The firms owning
facilities that produce softwood plywood have the highest median C/S ratio, 0.82 percent, followed by the
owners of other composite wood facilities, with a mean C/S ratio of 0.41 percent. See Exhibit 5-3 for a
summary of the data by process type, including mean, maximum and minimum C/S ratios.
r.\hil)il 5-3: Alloc lot I I'irms l>> Process T>po
NiiihIkt ill'
Al'l'iikil
IVrivnl ill' Tiihil
Mi'dhin
C/S
M;i\iiiiiiin
Minimum
Process Tjpo"
l"irmv'
AITi-ik-d Firms
C/S K;iiin
k;ilin
C/S K;iiin
( IS Uiiiin
Softwood Plywood/Veneer
21
40.4%
0.8%
1.1%
8.3%
0.01%
Oriented Strand Board
3
5.8%
0.2%
0.7%
1.9%
0.01%
Other Wood Composites2
13
25.0%
0.4%
2.1%
8.2%
0.03%
Engineered Wood Products
0
0.0%
n/a
n/a
n/a
n/a
Multiple Processes
15
28.8%
0.4%
0.6%
2.0%
0.01%
Total/Weighted Average3
52
100.0%
0.6%
1.2%
8.3%
0.01%
Notes:
'Firms categorized according to the process types associated with the affected facilities. Firms owing facilities
with more than one process type were assigned to the "Multiple Processes" category,
includes Medium Density Fiberboard, Hardboard, and Particleboard (conventional and molded).
3Based on affected facilities only (facilities incurring capital and MRR costs). Includes one firm that has
capital costs but no annual costs.
26Firms categorized according to the process types associated with the affected facilities. Firms owing
facilities with more than one process type were assigned to the "Multiple Processes" category.
5-4
-------
Of the four firms with C/S ratios of 3 percent or greater, three are small. Ten small firms have
C/S ratios of one percent or greater out of 16 in this category. One large firm has a C/S ratio of 3 percent
or greater and six of them have C/S ratios of one percent or greater. The other wood composite category
has the most firms with C/S ratios of 3 percent or greater (three out of the four). The softwood plywood
has seven firms followed by other wood composites with five firms with C/S ratios of one percent or
greater (out of 16). The C/S screening results are presented in Exhibits 4 and 5, below. These tables also
compare the number of affected firms (i.e., firms with facilities incurring only MRR costs as well as those
incurring capital and MRR costs) to the estimated total number of firms nationally.
I-Aliihil 5-4: AITcclcri l-'i nils with C/S Kniios of 3 IViron( or (>iv;ilcr
Number of l inns
Number of
l-'i rms ;is ;i IViron (
l-'i rms ;is ;i IVrceul
CsiU'Kon
N;i(ioimi(k-
AIToclod l-'irms
of Niiiioiiiil linns
of AITccU'd l inns
Firm Size
Small
38
3
7.9%
17.6%
Large
42
1
2.4%
2.9%
Undetermined
3
n/a
n/a
n/a
Total/Weighted Average
83
Process Type
4
4.8%
7.7%
Softwood Plywood/Veneer
30
1
3.3%
??
Oriented Strand Board
2
0
0.0%
??
Other Wood Composites
19
3
15.8%
??
Engineered Wood Products
11
n/a
n/a
n/a
Multiple Processes
21
0
0.0%
??
Total/Weighted Average
83
4
4.8%
7.7%
Notes:
See notes to Exhibits 5-2 and 5-3 above.
* Estimate.
5-5
-------
Exhibit 5-5:
AITcclcd 1
Firms with C/S Ratios of 1 Pi
Tccnl or (ircalci
Firms :is ;i
Finns ;is ;i lYm-nl of
NiiihIkt
\
Category
N alio mm idc- Firms
.Vilioiuil Finns
Category
Firm Size
Small
38
lO
26.3%
58.8%
Large
42
6
14.3%
17.1%
Undetermined
3
n/a
n/a
n/a
Total/Weighted Average
83
16
19.3%
30.8%
Process Type
Softwood Plywood/Veneer
30
7
23.3%
??
Oriented Strand Board
2
1
50.0%
??
Other Wood Composites
19
5
26.3%
??
Engineered Wood Products
11
0
0.0%
n/a
Multiple Processes
21
3
14.3%
??
Total/Weighted Average
83
16
19.3%
??
Notes:
See notes to Exhibits 5-2 and 5-3 above.
* Estimate. In a few cases, a firm's process type changed when all facilities were taken into account.
Profitability Analysis
Based on the results of the profitability analysis performed as described above, EPA developed
the following information.
The total number and percent of affected firms27 whose C/S ratio: 1) exceeds their profitability
ratio by 50 percent or more; 2) is between zero and 50 percent; or 3) is less than or equal to their
profitability ratio.
The total number and percent of affected small firms whose C/S ratio: 1) exceeds their
profitability ratio by 50 percent or more; 2) is between zero and 50 percent; or 3) is less than or
equal to their profitability ratio.
When EPA compared the C/S ratio to the R/S ratio for those firms with C/S ratios greater than one
percent, EPA found that in 14 of the 16 cases, the C/S ratio exceeded the R/S by over 50 percent. Two
firms' C/S ratio exceeded their R/S ratio by between zero and 50 percent. Exhibit 5-6 presents these
results in tabular form.
27 That is, firms with C/S ratios greater than one percent.
5-6
-------
llxhihil 5-(»: C/S lo R/S comp
irisnn lor linns with C/S of one
percent or «
vsilcr
C/S e\c<
i ¦ k
C/S is less
¦
lh;in hi- i'(|ii;il In K/S
;iihI 50 pereenl
rirm Si/.i-
NiiihIkt
hI' l"imis
lYri'i'iil ill' I'iiUiI
I'irnis with Ciisis
Number
hI' firms
lYrienl ill' Tiihil
firms with ( usis
Number
! nl'I'inns
lYlTl'lll llfTlllill
I'irnis w iili Costs
Small
10
??
1
??
i °
n/a
Large
4
??
1
??
! °
n/a
Total/
Weighted
Average
14
??
2
??
i o
n/a
Notes:
See notes to Exhibits 5-2 and 5-3 above.
EPA focused its review of the results for the 16 firms with C/S ratios greater than one percent.
For 4 of the 10 small firms in this category, no sales data were available. EPA developed sales estimates
for these firms according to the average sales for firms with the same number of employees in the same
product category. It is possible that in reality, these firms have parent-level sales that differ from those
assumed in the current analysis. However, based on extensive research into domestic parent-level
employment and sales data, EPA expects that this information is not publicly available for these firms.28
For the remaining 6 small firms with C/S ratios greater than one percent, EPA assumed that the domestic
parent firm sales information obtained from D&B, Wards, or the other Internet sources are reliable for the
purpose of this analysis.
EPA assumed that the firm size and sales data are reliable for 5 of the 6 large firms with C/S
ratios greater than 1 percent. One firm, Sierra Pine (a California Limited Partnership), should be regarded
as a special case. Sierra Pine recently purchased three affected facilities from Weyerhaeuser, greatly
increasing the total costs associated with Sierra Pine. However, Sierra Pine's sales data is from a query of
D&B data performed in late 1999. As a result, the sales associated with recent acquisition of the three
plants are not reflected in Sierra Pine's sales data. If, for instance, Sierra Pine's total compliance costs
were prorated to exclude the costs attributed to the plants previously owned by Weyerhaeuser, Sierra
Pine's C/S would change from 4.9 percent to 1.7 percent.
EPA also tested the sales estimates applied to large firms. The one large firm for which EPA
assumed sales had a C/S ratio well below one percent. In this case, the firm's actual sales would have to
be approximately half of the assumed sales in order for the firm's C/S ratio to exceed one percent. EPA
believes that while the assumed sales are potentially higher than the actual sales of this firm, it is less
likely that the firm has actual sales that would result in a C/S ratio over one percent.
28Research indicates that one of the small firms with particularly high impacts, Dominance Industries (d.b.a.
Pan Pacific Products) is a single location private corporation owned Philip Ling. However, Dr. Ling is also the
owner of a group of companies around the world and is the chairman of Malaysian-based Pan Pacific Asia Berhad,
an investment holding company that also provides management services. Pan Pacific Asia's 1998 sales were $88
million (U.S.). The group manufactures and distributes timber logs and timber moldings. Other activities of the
company are stockbroking and investment holding. While this information is not necessarily applicable to the
ultimate domestic parent of Dominance Industries' affected facility, is indicates that the facility's owner has access
to financial capital beyond what the sales from the facility generate.
5-7
-------
EPA used industry-wide measures of profitability because such information is not available for
privately held firms. Because all the affected firms with C/S ratios above one percent are privately held,
firm-specific information regarding R/S is not publicly available. If firm specific profitability measures
were available, it is likely that a comparison of firms' C/S ratio to their R/S ratio would produce
significantly different results. It should be noted that while overall industry return on sales may be low, it
is not necessarily the case for any given firm.
All ten small firms with C/S ratios of one percent or greater had a C/S ratio that exceeded the
industry profitability measure of return on sales. For the three small firms whose C/S ratios were three
percent or greater, the comparison of compliance costs and profitability measures showed that their C/S
ratios exceeded the industry R/S by over 100 percent. This divergence between the C/S ratios for these
firms and the industry R/S is an indicator that these three firms may experience high impacts as a result of
incurring the costs of compliance associated with the proposed rule. However, the results of the
economic impact analysis show that the affected facilities owned by these firms will continue to operate
after controls have been applied to comply with the proposed rule. Therefore, the impact on these firms,
while relatively high, are not enough to lead them to cease operations at these facilities.
5.7 Screening Analysis Conclusions
The Regulatory Flexibility Act (RFA) generally requires an agency to prepare a regulatory
flexibility analysis of any rule subject to notice and comment rulemaking requirements under the
Administrative Procedure Act or any other statute unless the agency certifies that the rule will not have a
significant economic impact on a substantial number of small entities. Small entities include small
businesses, small organizations, and small governmental jurisdictions.
For purposes of assessing the impacts of today's proposed rule on small entities, small entity is
defined as: (1) a small business according to Small Business Administration size standards by 5-digit
NAICS code of the owning entity (in this case, 500 employees); (2) a small governmental jurisdiction that
is a government of a city, county, town, school district or special district with a population of less than
50,000; and (3) a small organization that is any not-for-profit enterprise which is independently owned
and operated and is not dominant in its field.
After considering the economic impact of today's proposed rule on small entities, we certify that
this action will not have a significant impact on a substantial number of small entities. In accordance with
the RFA, we conducted an assessment of the proposed standard on small businesses in the industries
affected by the rule. Based on SB A size definitions for the affected industries and reported sales and
employment data, the Agency identified 17 of the 52 companies, or 32 percent, owning affected facilities
as small businesses. These facilities incur capital and MRR costs associated with the proposed rule.
There 31 other firms that only incur MRR costs; and all of these firms are small. Although small
businesses represent 32 percent of the affected companies within the source category, they are expected to
incur only 6 percent of the total industry compliance costs of $142 million. There are only three small
firms with compliance costs equal to or greater than 3 percent of their sales. In addition, there are seven
small firms with cost-to-sales ratios between 1 and 3 percent.
We performed an economic impact analysis to estimate the changes in product price and
production quantities for the firms affected by this proposed rule. The analysis shows that of the 32
facilities owned by affected small firms, only one would be expected to shut down rather than incur the
cost of compliance with the proposed rule. Although any facility closure is cause for concern, it should
be noted that the baseline economic condition of the facilities predicted to close affects the closure
estimate provided by the economic model. Facilities which are already experiencing adverse economic
5-8
-------
conditions for reasons unconnected to this rule are more vulnerable to the impact of any new costs than
those that are not.
This analysis indicates that the proposed rule should not generate a significant impact on a
substantial number of small entities for the coatings manufacturing source category for the following
reasons. First, of the 10 small firms that have compliance costs greater than 1 percent of sales, only 3
have compliance costs of greater than 3 percent of sales. Second, the results of the economic impact
analysis show that only 1 facility owned by a small firm out of the 32 facilities owned by affected small
firms may close due to the implementation of this rule. The facility that may close rather than incur the
cost of compliance appear to have low profitability levels currently. It also should be noted that the
estimate of compliance costs for this facility is likely to be an overestimate due to the lack of facility-
specific data available to assign a precise control cost in this case. In sum, this analysis supports today's
certification under the RFA because, while a few small firms may experience significant impacts, there
will not be a substantial number incurring such a burden.
Although this proposed rule will not have a significant economic impact on a substantial number
of small entities, we minimized the impact of this rule on small entities in several ways. First, we
considered subcategorization based on production and throughput level to determine whether smaller
process units would have a different MACT floor than larger process units. Our data show that
subcategorization based on size would not result in a less stringent level of control for the smaller process
units. Second, in light of cost considerations, we chose to set the emission limitation at the MACT floor
control level and not at a control level more stringent than the MACT floor control level. Thus, the
control level specified in the proposed PCWP rule is the least stringent allowed by the CAA. Third, the
proposed rule contains multiple compliance options to provide facilities with the flexibility to comply in
the least costly manner while maintaining a workable and enforceable rule. The compliance options
include emissions averaging and production-based emission limits which allow inherently low-emitting
process units to comply without installing add-on control devices and facilities to use innovative
technology and pollution prevention methods. Fourth, the proposed rule includes multiple test method
options for measuring methanol, formaldehyde, and total HAP. We continue to be interested in the
potential impacts of the proposed rule on small entities and welcome comments on issues related to such
impacts.
5.8 Economic Impact Analysis Results for Small Businesses
Exhibit 5-7 provides a summary of the economic impact on small businesses associated with the
estimated market adjustments due to compliance with the proposed NESHAP. As shown, the Agency's
economic analysis indicates that the 17 small businesses that own 18 affected process lines will be
affected as follows:
5-9
-------
Exhibit 5-7: Economic Impacts on Small Businesses Associated with
Projected Market Adjustments*
Changes
from Baseline
Baseline
With Regulation
Absolute
Percent
Revenues($
thousands/yr)**
394,393
387,229
7,164
-1.82
Production (million
m3/yr)
1,791,408
1,737,969
53,439
-2.98
Compliance Costs ($
thousands/yr)
0
9,194
9,194
n/a
Operating Process
Lines
18
17
1
-5.56
Employment loss
(FTEs)
3,621
3,513
108
-2.98
Notes:
* Does not include small businesses that own facilities with
MRR costs only.
** Estimated using production and price data in economic
impact model
FTEs = full-time equivalents
The one process line closure predicted by the economic impact model is owned by a small
business. This results in a 5.6 percent decrease in the number of process lines owned by small business.
Overall, the small businesses' revenues decrease by just under 2 percent, their production decreases by
just under 3 percent, and their total employment decreases by just under 3 percent, or 108 FTEs. The
estimate of employment loss assumes that the production per employee at the affected facilities owned by
small firms is the same as the industry average.
5.9 References
Abt Associates Inc. 2000a. Small Business Screening Analysis for the Plywood and Wood Composite
Industries: Revised Draft. Prepared for Larry Sorrels, Innovative Strategies and Economic Group,
Air Quality Strategies and Standards Division, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, NC. June.
Abt Associates Inc. 2000b. Corporate employment and sales data collected through Internet business
information resources, including Dun & Bradstreet, Hoover's On-line, Lycos Companies On-line,
Thomas' Register, Ward's Business Directory, and Zapdata.com.
Abt Associates Inc. 1999. Profile of the Plywood and Wood Composite Industries. Prepared for Larry
Sorrels, Innovative Strategies and Economic Group, Air Quality Strategies and Standards
Division, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC. December.
5-10
-------
Midwest Research Institute. 1999. Memorandum from Becky Nicholson and Melissa Icenhour to Mary
Tom Kissell and Larry Sorrels, U.S. EPA: "Preliminary facility-specific cost estimates for
implementation of the plywood and composite wood products NESHAP." October 20.
Midwest Research Institute. 2000. "Estimated Nationwide Costs of Control: Plywood and Wood
Composite Products." April 14.
U.S. Environmental Protection Agency. 1998. Industry Specific Information Collection Request (ICR) for
the Development of Plywood and ParticleboardMaximum Achievable Control Technology
(MACT) Standards. Air Quality Strategies and Standards Division, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1999a. OAQPS Economic Analysis Resource Document. Air
Quality Strategies and Standards Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC.
U.S. Small Business Administration. 2000. "Small Business Size Standards Matched to NAICS Codes."
5-11
-------
Appendix A Size Determination for Individual Firms
Determination of whether a firm is large or small is based on each facility's ultimate domestic
parent, the parent firm's SIC code, the parent firm's employment or revenue, and the SBA standard
threshold for that SIC code. EPA asked facilities to provide their ultimate domestic parent firm name and
employment in the ICR. However, many facilities did not provide this information, or provided
inaccurate information for ultimate parent, and rather provided a subsidiary of another domestic firm. For
this reason, a DUNS facility search was conducted based on the 416 unique facility names specified in the
ICR and the additional 15 facilities identified after the survey date. Ultimate parent name, primary SIC
code, employment, and sales for 302 of the facilities in the ICR database were retrieved from DUNS,
resulting in identification of 92 unique ultimate domestic parent firms. Ultimate parent employment
reported in the DUNS database was compared to the SBA standard threshold for that SIC code to
determine whether the ultimate parent was a small or large firm. The remaining 129 facilities were
classified based on information from DUNS and the ICR or through other sources as described below:
(1) Facilities with no identifiable ultimate parent.
Twenty-nine facilities were listed in the DUNS database without information on an ultimate
parent. These facilities were assumed not have another ultimate parent and the facility SIC code in
DUNS was used as the ultimate parent's SIC as well. For these facilities, the primary SIC code for the
facility was used to determine the appropriate SBA standard. Information provided by facilities on their
ultimate parent's employment was then used to determine the size relative to the SBA standard.
(2) Facilities with no DUNS listing.
For those impacted facilities for which no parent company data were available, we assumed that
the facility represents the ultimate parent. We then assumed that the employees or the revenues
associated with the facility were the same for the parent firm.
For 8 impacted firms for which Dun & Bradstreet, Ward's, or internet revenue data was not
available at the firm or facility level, we applied a revenue estimate based on the average revenue
for firms in the same product and employee size category.
For one impacted firm, we obtained employment information for all three of its idenitified
facilities and revenue information for two of the three. We extrapolated the revenue data on a
dollar per employee basis from the two facilities with complete data to the facility with only
employment data and added those revenues to estimate the firm's total revenues.
Exhibit A-l below lists all firms that own facilities projected to have compliance cost impacts and
relevant information used to make size determinations. The source column identifies which of the three
sources (DUNS, ICR, or other) was used to make the size determination.
A-l
-------
Exhibit A-l: Size Determination for Individual Impacted Firms
Ultimate Domestic Parent Firm
Facilities**
Standard
Size
Source
Bassett Furniture Inds Inc
8
500
Large
DUNS
Bessemer Plywood Corporation
1
500
Small
OTHER
Boise Cascade Corporation
12
100
Large
DUNS
Champion International Corp
3
750
Large
DUNS
Clyde Lang
1
500
Small
SURVEY
Coastal Lumber Company
1
500
Large
DUNS
Collins Products Lie
3
500
Large
DUNS
Emerald Forest Products Inc
3
500
Small
DUNS
-ibe resin Industries Inc
1
500
Small
DUNS
-reres Lumber Company, Incorporated
1
500
Small
DUNS
Furniture Brands International
7
500
Large
DUNS
Georgia-pacific Corporation
43
500
Large
DUNS
Hambro Forest Products Inc
1
500
Small
DUNS
Hollingsworth & Vose Company
1
750
Large
DUNS
Hood Industries Inc
2
500
Large
DUNS
Hunt Forest Products Inc
2
500
Large
DUNS
Litco International Inc.
1
500
Small
DUNS
International Paper Company
17
750
Large
DUNS
] M Huber Corporation
4
750
Large
DUNS
Icld-wcn Inc
4
500
Large
SURVEY
loint Noranda Forest, Inc, & The Mead
Corp
1
500
Large
SURVEY
Simpson Investment Co.
1
750
Large
OTHER
EClukwan Inc
1
$17*
Large
DUNS
Langdale Company Inc
2
500
Large
DUNS
Louisiana-pacific Corporation
32
500
Large
DUNS
Vlacmillan Bloedel Limited
1
500
Large
DUNS
Vlartco Partnership
2
500
Large
DUNS
Vlasco Corporation
1
500
Large
DUNS
Vlurphy Company
1
500
Small
DUNS
Sforanda Inc
2
500
Large
DUNS
Plum Creek Timber Company L p
4
100
Large
DUNS
Potlatch Corporation
6
750
Large
DUNS
R L C Industries co
6
500
Large
DUNS
Ron Fallard
1
500
Small
SURVEY
Scotch Plywood Company of Alabama
1
500
Small
SURVEY
Sierra Pine a Cal Ltd Partnr
7
500
Large
DUNS
Smurfit-stone
2
750
Large
DUNS
Sticky Pitch, Incorporated
1
500
Small
SURVEY
A-2
-------
Exhibit A-l: Size Determination for Individual Impacted Firms
Ultimate Domestic Parent Firm
Facilities**
Standard
Size
Source
Stimson Lumber Company
4
500
Large
DUNS
Superior Lumber co
1
500
Small
DUNS
Temple-inland Inc
10
500
Large
DUNS
Timber Pdts Co Ltd Partnershi
6
500
Large
DUNS
Us Forest Industries, Incorporated
2
500
Large
SURVEY
Webb Furniture Enterprises, Inc
1
500
Large
SURVEY
Weyerhaeuser Company
19
500
Large
DUNS
Willamette Industries Inc
23
500
Large
DUNS
Ponderosa Products Inc
1
500
Small
DUNS
Dominance Industries
1
500
Small
OTHER
3ds Lumber Company
1
500
Small
OTHER
Canfibre
1
500
Small
OTHER
Quality Veneer & Lumber
1
500
Large
OTHER
Fourply, Inc.
2
500
Small
OTHER
Westbrook Wood Products
1
500
Small
OTHER
* Size was determined by firm revenues which were $71.5 million based on DUNS information.
** Includes affected and unaffected facilities owned by firm.
Sources: Dun & Bradstreet DUNS Database (1999), U.S. Environmental Protection Agency (1998), MRI (1999),
Abt Associates (2000).
A-3
-------
Appendix B EIA Methodology
B.1 Introduction
This appendix provides a description of the methodology EPA used to perform the economic
impact assessment (EIA) of the proposed NESHAP that will apply to the Plywood and Composite Wood
(PWC) Manufacturing Sector. Specifically, the Agency applied this methodology to assessing the impacts
of the proposed NESHAP to the manufacturers of softwood plywood (SWPW), oriented strandboard
(OSB), particleboard (PB), medium density fiberboard (MDF), and hardboard (HB). See Section 4.4 for
a discussion of potential impacts on the manufacturers of engineered wood products (EWP).
This appendix presents the general background for the EIA, describes the market segments and
components of demand associated with PWC manufacturing, and summarizes the assumptions and data
inputs used in the analysis. The document then describes the approach to the EIA, including the
development of the supply and demand components of the partial equilibrium model, and the
computational tool used to implement the model and perform the analysis. The final section discusses the
results produced by the analysis and the uses of those results.
B.2 Background
In the process of proposing NESHAPs, the EPA performs EIA to estimate the impacts the
proposed rule will have on market, facilities, and total social welfare. The framework for performing this
EIA is predicated on modeling the market-based behavioral response of facilities to compliance costs
associated with proposed NESHAP. Specifically, this framework is based on a partial equilibrium
analysis that uses mathematical representations of supply and demand functions. The framework
estimates the short run facility-level, industry-level, and market-level impacts that the imposition of
compliance costs associated with the NESHAP are expected to have.
At the market level, the analysis employs a partial equilibrium analysis to estimate changes in
market prices, domestic production and consumption, and foreign trade. The partial equilibrium
economic analysis examines each impacted market alone, ignoring the possible interactions that the
market impacts being studied may have on other markets. Because facility-specific information is
available for baseline production levels and compliance costs estimated to be incurred as a result of the
proposed NESHAP, it is possible to provide estimates of facility-specific changes in production and
revenues. Facility-level compliance costs vary based on characteristics of the producing facilities.
Using this methodology, the EIA also estimates the net social burden of the proposed NESHAP,
taking into account market adjustments to the rule. The social burden includes:
the producers' compliance costs,
the loss consumers incur should plywood and wood composite prices rise,
the loss both consumers and affected producers incur when plywood and wood composite
production decreases, and
increases in profits for unaffected plywood and wood composite producers who enjoy higher
market prices and/or production without any additional costs.
B-l
-------
This social burden or cost is weighed by policymakers against the social gains associated with the
regulation, including the value of reduced adverse impacts on human health and the environment, when
evaluating regulatory options.
Social costs vary based on characteristics of markets in which these facilities participate. At the
market level, the imposition of compliance costs on some producers adds to the cost of producing
plywood or composite wood products, and thus results in an upward shift of the market supply curve. In
a perfectly competitive market, the supply curve measures the marginal cost of supplying plywood or
composite wood products to the market as a function of the quantity supplied. The area under the supply
curve from 0 to any number, Q, measures the total variable cost of supplying Q of plywood or composite
wood products. An upward shift in the supply curve means that it costs more to supply a given quantity
of plywood or composite wood products. This is what would be expected if some producers have to
switch to lower HAP technologies to meet the requirements of the proposed NESHAP.
The demand for plywood or composite wood products measures the quantity of plywood or wood
composites consumers will purchase as a function of the price of the product, assuming all else is held
constant. The approach used for this EIA assumes that the demand for plywood or composite wood
products at any given price is unchanged by the regulation, unless the quality of the product changes
substantially due to required changes in product inputs or production processes. Figure 1 below
graphically presents the theoretical impact of the regulation on the market for plywood or composite
wood products. Note that the shift in the supply curve results in an increase in price and a decrease in the
quantity of plywood or composite wood products sold. This figure is discussed in detail later in the
document.
P
S post-
compliance
Pi
Po
Qi Q»
Q
Figure B-l: Market Supply and Demand
B-2
-------
B.3 Market Segments and Demand Components
The Plywood and Composite Wood Products Industry contains two main industry sectors
expected to experience impacts associated with the proposed Plywood and Composite Wood Products
NESHAP: Softwood Veneer and Plywood (SIC 2436) and Reconstituted Wood Products (SIC 2493).
The Reconstituted Wood Products industry sector is further disaggregated into two sectors: oriented
strand board (OSB) and other wood composites. Other wood composites include particleboard (PB),
medium density fiberboard (MDF), and hardboard (HB). Each of the three industry sectors has a
corresponding market segment. The domestic facilities that produce output in a given sector are the
domestic suppliers in the corresponding market segment. EPA analyzed each of these three market
segments separately.
Within the third market segment, "other composite wood products," a further division was made.
As shown below (see Exhibit B-3), the baseline prices of particleboard and medium density fiberboard are
very different from the baseline price of hardboard. However, information on the other inputs to the
analysis was available only at the more aggregate level of "other composite wood products." In
particular, many data sources present these products together because the facilities have the same
industrial classification (SIC or NAICS), making market level distinctions among the products difficult.
Because of the substantially different baseline prices, however, PB/MDF and HB were analyzed as two
distinct markets that share values for all inputs except baseline price.
In three of the four market segments (softwood plywood, PB/MDF, and HB), there are two major
components of domestic demand: housing construction and manufacturing. There is only one component
of domestic demand for OSB: housing construction. The EIA methodology for these markets takes into
account separate domestic demand components that make up total demand. Exhibit B-l lists the market
segments and domestic demand components, along with the percent of output from producers going to
each source of demand, to the extent that the data is currently available.
l.xhihil I?-1. DoiihiihI ( omponeiUs ol' llio Plywood ;iinl Wood ( omposiu* Industries
IVrconl ol' Domoslic Dcniiind
' ( onslriiclion M;imiI'licluriiiLt
Softwood Plywood and Veneer (SWPW) 64% 36%
Oriented Strandboard (OSB) 100% 0%
Other Composite Wood Products: 10% 90%
PB/MDF
Other Composite Wood Products: HB29 10% 90%
Source: Abt Associates, 1999.
290ur research into the domestic demand components of HB did not produce specific data regarding the
share of HB products going to construction and manufacturing. For the purpose of this analysis, we will assume that
the demand components for HB are the same as for PB and MDF.
B-3
-------
B.4 Data Inputs
The economic impact analyses required specific data inputs pertaining to supply and demand
within each market segment. Exhibit B-2 presents the data inputs needed to analyze the economic
impacts resulting from the imposition of compliance costs on the facilities in a given market segment
(e.g., softwood plywood). The data elements required are essentially the same for all market segments.
Exhibit B-2: Data Inputs for Analyzing Impacts on a Given Market Segment
Market Component Data Element
Price Baseline market equilibrium price of the affected product
Supply Side
Supply from affected
domestic facilities
Supply from unaffected
domestic facilities
Supply from the foreign
sector
Demand
Separate domestic
demand components
(e.g., manufacturing
sector and housing
construction)
Foreign demand Baseline quantity demanded by the foreign sector*
Own-price elasticity of demand
Cross-price elasticities of demand (if available)
*Needed to estimate the multiplicative constant in the demand equation.
B.5 Assumptions
Based on EPA's understanding of the plywood and composite wood industry, review of available
data, and application of the analytical approach described in the next section, EPA made the following
assumptions.
The baseline year for the analysis is 1997.
All market segments are perfectly competitive.
Each affected product is sold in a single national market for homogenous output at one price.
Affected domestic products, unaffected domestic products, and foreign products in the same
market segment are perfect substitutes for each other.
All factors that affect demand, other than price(s), are constant in the short-run.
30We assume that the EPA's ICR collected production quantities from all domestic facilities, both affected
and unaffected.
Baseline quantity supplied by each affected facility30
Compliance costs for each affected facility
Price elasticities of supply for domestic facilities (if available)
Baseline quantity supplied by each unaffected facility
Price elasticities of supply for domestic facilities (if available)
Baseline quantity supplied by the foreign sector
Price elasticity of supply for the foreign sector (if available)
Baseline quantity demanded by each demand component*
Own-price elasticity of demand
Cross-price elasticities of demand (if available)
B-4
-------
Industry supply elasticity for foreign suppliers is the same as industry supply elasticity for
domestic suppliers.
The price elasticities for domestic and foreign demand are the same.
The production costs of foreign suppliers are unaffected by emission controls.
The production costs of unaffected domestic suppliers and of substitutes are unaffected by
emission controls.
B.6 Analytical Approach
This section describes EPA's approach to the EIA for the plywood and composite wood
economic impact analysis. As the previous section described, this analysis disaggregated plywood and
composite wood products into the following three market segments.
Softwood plywood and veneer (SWPW).
Oriented strandboard (OSB).
Other composite wood products (OWC):
Particleboard and medium density fiberboard (PB/MDF), and
Hardboard (HB).
A separate market equilibrium analysis was performed for each market segment, incorporating
separate equations for domestic supply, foreign supply, domestic demand, and foreign demand.31 For
each market, these equations were combined into a single equation that sets the quantity supplied equal to
the quantity demanded in the post-regulation scenario, determining the new market equilibrium price at
which that equality holds.
Although specific parameter estimates may differ from one market to another, the methodology
used to model supply and demand and to estimate the new equilibrium prices and quantities in these
markets is basically the same. These markets are therefore discussed together, with market-specific
distinctions made as appropriate.
B.7 Notation
To facilitate the discussion below of supply and demand functions in the baseline and the post-
regulation scenarios, a brief explanation of the notation used will be helpful. In general, Q denotes
quantity and P denotes price. Price and quantities in the baseline are subscripted with a "0," and price and
quantities in the post-regulation scenario are subscripted with a "1." Domestic and foreign supply and
demand in a market are denoted by the following superscripts:
DS denotes domestic supply;
FS denotes foreign supply (i.e., imports);
DD denotes domestic demand;
FD denotes foreign demand (e.g., exports);
TS denotes total supply; and
TD denotes total demand.
31 Because there are separate components of domestic demand, there is one demand equation for each
component. See Exhibit B-1.
B-5
-------
The quantity supplied by the jth domestic facility in a market is denoted by subscripting the QDS with a
"jThe quantity demanded by the kth demand component in a market is denoted by subscripting the QDD
or the QFD with a "k." It is necessary to distinguish between the multiple domestic demand components
associated with several of the markets analyzed here.
The following are some examples of the notation used below, based on the above notational rules:
Q jcf 's t'lc quantity supplied by the jth domestic facility in the baseline;
Q'f is the quantity supplied by the jth domestic facility in the post-regulation scenario.
Qkf's t'10 quantity demanded by the kth domestic demand component (e.g., domestic
manufacturing) in the post-regulation scenario;
P0 is the baseline market equilibrium price; and
P, is the post-regulation market equilibrium price.
B.8 The Supply Side of the Market
This analysis modeled two broad sources of supply — domestic supply and foreign supply —
that satisfy demand from two broad sources — domestic demand and foreign demand — in each market.
Because facility-specific information on quantities produced and compliance costs is available, this
analysis modeled facility-specific supply functions. Although many domestic facilities will have
compliance costs, some will not. In addition, foreign facilities will have no compliance costs. For some
products, the foreign component of supply is substantial. For example, in 1995, 17 percent of
consumption of reconstituted wood products in the United States came from imports (foreign supply).
Because some domestic facilities and all foreign facilities will have no compliance costs, and because,
among those domestic facilities that will have compliance costs, the cost per unit of product produced is
likely to vary from one facility to another, the analysis will capture differential impacts across facilities.
While some facilities (those with compliance costs) are likely to supply reduced quantities of products to
their markets, other facilities with zero compliance costs are likely to supply greater quantities.
B.8.1 Functional Form of Supply Curves
There is little empirical support for any particular functional form of the supply functions in these
markets. Although facility-specific production quantities and compliance cost estimates are available,
facility-specific estimates of average production cost are not. It is therefore not possible to estimate step
supply functions in any of these markets. Nor are there facility-specific estimates of price elasticity of
supply. In the absence of any empirical basis to support one functional form over another, the analysis
assumed that the average domestic facility supply curve is the supply function corresponding to a
generalized Leontief profit function. This supply function has the following form:
where Q'^ is the average market equilibrium quantity supplied, P is the market equilibrium price, and P
and y are parameters (P< 0; y>0). This is a well recognized and reasonable form for a supply function. It
(1)
B-6
-------
can be shown that the price elasticity of supply for this supply function, g , is a function of the price, the
quantity and the parameter p, and that the industry supply function has the same elasticity of supply at the
equilibrium price. Given this, the parameter p can be solved for as a function of this elasticity, and the
baseline price and quantity. In particular,
|3 = - 4s *Q°Sg *P1/2 (2)
Therefore, given information about the baseline market equilibrium price and the baseline
equilibrium average quantity per facility, and a reasonable estimate of the industry elasticity of supply at
the baseline price, it is possible to estimate p. The assumption about the industry average elasticity
affects not only the value of p but all aspects of the economic analysis (as explained below). The model
has the capability to produce results through sensitivity analyses around these assumptions that would
allow the Agency to examine how the analytic outcomes (including change in equilibrium price and
quantities, closures, and social welfare impacts) change as the input assumptions about this elasticity
change. This is straightforward in a spreadsheet analysis. However, no sensitivity analyses were
performed for this EIA.
B. 8.2 Domestic Supply
The total domestic quantity supplied in a given market (e.g., the total domestic supply of
softwood plywood used for all final goods) is the sum of the domestic facility-specific supply quantities.
Under the assumption that the supply functions of individual domestic facilities have the same functional
form as the average domestic supply function (equation (1)) and the same value for the parameter p (as
calculated in equation (2)), facility-specific supply functions were derived by solving for the facility-
specific value of y that is consistent with the facility's baseline quantity supplied:
Yi = Qf -(j)*P-"2 P)
where Q'>s is the quantity supplied by the jth domestic facility and P is the market equilibrium price.
Using the baseline equilibrium price, P0, the jth domestic facility's quantity supplied in the baseline,
QjoDS, and a value for p, we can derive a value for yr32 The jth domestic facility's supply function in the
baseline is therefore
= Tj + I2J *p°"2 <4)
The jth facility's shutdown price (the price at which it supplies zero units) can be shown to be
B2
-pshutdown _ r
iO ~ a 2 • W
jo
4y j
32 Using this approach, the elasticity of supply, which is a function of both price and quantity supplied, will
vary from one facility to another. The shutdown price — the price at which the facility will supply zero units — will
also vary from one facility to another.
B-7
-------
The total quantity supplied by domestic facilities in the baseline is the sum of the facility-specific
quantities:
Qf = IQj? • ®
j
In the post-regulation scenario, the supply function of the jth facility, with per unit compliance cost Cj > 0,
becomes
(7)
where P, is the new (post-regulation) market equilibrium price. The jth facility's shutdown price is now
P2
T) shutdown >
J1 " 4y 2 CJ ' (8)
The jth facility's post-regulation supply decision can therefore be written as
Qjf = <
/ ,
y, + -] *(pi - c,r1/2 if pi > p,futdown
v 2 J /9\
q p ^ pshutdown ' v '
The total quantity supplied by all domestic facilities in the post-regulation period is
QiDS = E Q? • <"»
Baseline quantities are available for all domestic facilities from the EPA's 1998 Information
Collection Request (ICR) issued to plywood and wood composite producers (EPA, 1998). EPA also
developed facility-specific compliance costs. EPA calculated both total domestic baseline quantity
supplied and facility-specific per unit output compliance costs from this information. SWPW, OSB, and
PB/MDF baseline prices were based on the industry-reported free on board (f.o.b.) prices. These are the
product prices prior to shipping costs and distributor mark-ups.33 Because f.o.b prices were not available
for HB, the agency used wholesale price quotes that were publicly available. Exhibit B-3 presents the
prices in 1996 dollars and the extrapolated baseline prices in 1997 dollars.
"Except for hardboard, which includes delivery costs.
B-8
-------
r.xliihil 15-3. Prices in !•)% ;ind l$;iselino Prices in I')')"7
Product
Prices in I')')(>
(S/cuhic mc(cr)
liiisclinc Prices in I'J')"7
(S/cuhic mc(cr)
Softwood Plywood
Oriented Strandboard
Other Composite Wood
Products
231
184
235
185
Particleboard/MDF
165
169
Hardboard
n/a
1,322
Source: SWPW, OSB, and PB/MDF prices: Abt Associates, 1999. HB prices: Get-a-Quote.net. 4' x 8'
sheets, used as underlayment, 1/8" tempered .
Note: SWPW, OSB, and PB/MDF prices are based on 1996 prices adjusted by average annual price
change from 1989 to 1996. HB prices are in 2000 dollars.
As noted above and as shown in Exhibit B-3, the baseline prices of PB/MDF and hardboard, within the
broader category "other composite wood products," are very different. Each of these two product groups
(PB/MDF and HB) was therefore analyzed as a separate market. However, because of lack of
information specific to each (PB/MDF and HB) on the other inputs to the analysis ~ i.e., elasticities of
supply and demand, and percentage that goes to manufacturing versus construction ~ estimates for the
overall category "other wood composites" for these inputs were applied to each of the specific products.
Although there are no available estimates of the price elasticities of supply of softwood plywood,
OSB, or other composite wood products, there are some estimates of elasticities of supply of different
softwood lumbers, which served as a proxy for the elasticities of supply of related products such as those
considered in this analysis.
Lewandrowski et al. (1994) provide several estimates of supply elasticities. While their estimates
differ by an order of magnitude depending on the underlying assumptions they make, they note that most
published lumber supply elasticities are less than 1.00. Because they model inventories of lumber, there
are differences between production and supply: the products may either be sold or stored. They note
further that, "if inventory adjustments tend to balance out over time, production should about equal
supply over a year," and that their lumber production elasticities (0.358 for Southern Pine, 0.443 for
Douglas Fir, and 0.446 for Western Pine) may therefore be closer to supply elasticities. This analysis
used the (unweighted) average of these elasticities, 0.42, as the price elasticity of domestic industry
supply for all market segments.
As noted above, while the supply elasticities of individual domestic facilities depend (indirectly)
on this industry supply elasticity, they also depend on facility-specific quantities supplied. Domestic
facilities' supply elasticities therefore vary from one facility to another. As noted above, the models are
designed to that, if desired, the Agency can perform a sensitivity analysis to investigate the impact of
alternative estimates of the industry average supply elasticity on the results of the analysis.
B-9
-------
B.8.3 Foreign Supply
Supply from the foreign sector (imports) can be modeled similarly, and this supply was added to
the domestic supply. Unlike domestic facilities, foreign facilities will not experience any increases in
production costs due to the proposed NESHAP. This analysis therefore combined and treated them as
one "facility" with supply function:
QFS = yFS + (|)*p-1/2. (11)
The industry profile presents information on the baseline quantities imported (Abt Associates,
1999). Because there is no information on the elasticities of foreign supply of softwood plywood, OSB,
or other wood composites, the analysis assumed that these elasticities of supply are the same as the
corresponding elasticities of supply for the corresponding domestic industries. The analysis further
assumed that prices in the foreign sector are the same as those in the corresponding domestic markets.
B. 8.4 Total Market Supply
The total quantity of a product supplied at the new market equilibrium price, P,. is the sum of the
quantities supplied by domestic facilities and the quantity supplied by the foreign sector:34
Qr=Q?s+Qfs (12)
Substituting from equations (9) and (10) for the domestic supply (Q[ )S ) and from equation (11) for the
foreign supply ( QFS ), the total quantity supplied to the market post-regulation is:
QP = Qfs + QP = Z Qjf + QP (13)
j
34 The supply and demand curves typically used by economists when showing supply and demand in a
market are actually inverse supply and demand functions because quantity is on the x-axis and price is on the y-axis.
Rather than showing the quantity supplied (demanded) at a given price, the inverse supply (demand) function shows
the price at which a given quantity is supplied (demanded). The (quantity, price) point at which the market inverse
supply curve and the market inverse demand curve cross is the market equilibrium point. The (inverse) market
supply curve is the horizontal sum of the individual supply curves. If at price p the jth supplier will supply qj units to
the market, then the total quantity supplied at price p is ^ . Similarly, if at price p the kth demand component
j
demands q, units, then the total quantity demanded at price p is ^ Qk . The market is in equilibrium when
k
X Qj = X Qk •
j k
B-10
-------
Qf = lQ™ + rFS +
(14)
where, as shown in equation (9),
\Pl-c])-y2ifPl>P]
) shutdown
A
0 ifPl
-------
where
the quantity demanded by the kth domestic demand component;
the market equilibrium price. (Note that for each market there is a single market price
which is paid by all demand components in that market.);
a demand-component-specific constant calibrated to the baseline price and quantity;
the own-price elasticity of demand by the kth demand component, and
P
rk=e"k ¦
The multiplicative constant, yk, is calibrated to the baseline price and quantity demanded by the kth
domestic demand component (using equation (16)).
B.9.2 Substitution
In some cases, there may be substitution possibilities between the products being considered in
this analysis — most notably, softwood plywood and OSB may both be used as structural panels in
construction. If the necessary information were available — in particular the cross-price elasticities of
demand — it would be possible to modify the above demand relationship to include the effect of changes
in the price of the substitute product. Cross-price elasticities, however, are very difficult to estimate and
estimates of the cross-price elasticities that would be relevant to this analysis are not obtainable.
The omission of short-run substitution possibilities from the model, however, is not likely to have
any significant impact on the results. In the construction industry, inputs such as plywood and
reconstituted wood products are used in combination with other specific inputs and may best be modeled
as part of a composite (or aggregate) input to the final good, a residence for example. To change from
plywood to OSB in housing construction, for example, would require a change not only of plywood to
OSB but also a change of those inputs that must be used in combination with plywood to those inputs that
must be used in combination with OSB — i.e., a change from one composite input to another. This
would also be true of non-wood substitutes for plywood and composite wood products.
This model of production as using composite inputs is described by Chambers (1988) and was
applied to lumber used in residential construction by Adams et al. (1992). In such a model, the demand
for a product (e.g., lumber) is a function of the "price" of the composite input that includes the product
and the price of the composite input that does not include the product (but instead includes a substitute).
A change in the relative prices of two inputs (e.g., lumber and a substitute for lumber, or plywood and
OSB) which are components of two composite inputs which may be substituted for each other will induce
substitution only to the extent that this change induces a change in the relative prices of the two
composite inputs.37
37 The view of these products as being components of composite inputs in the construction industry was
corroborated in a personal communication with Prof. J. Buongiorno, a leading economist in the area of forest sector
economics. Professor Buongiorno noted that it is the cost of the product in place that matters in construction — it is
not just a matter of substituting one material input for another, that such substitution requires other substitutions that
also have cost implications. In effect, what is required is a substitution of one composite input for another. It is
therefore the relative "prices" of the composite inputs that matters, not the relative prices of the two material inputs
(e.g., plywood and OSB) in these composite inputs that determines the extent of substitution.
B-12
-------
Similarly, plywood and wood composite input substitution in furniture manufacturing is likely to
be minimal. Structural properties and design characteristics of certain wood products used in
manufacturing processes are specific to the necessary product. Furthermore, these processes are designed
to take advantage of certain wood products and substitution in the short-term would require high capital
investments. The American Plywood Association's (APA) case studies note how changes in the use of
wood products in manufacturing significantly alters costs. Bassett Furniture states "plywood provides
labor efficiencies because the [manufacturing] equipment eliminates many manual operations" and Rowe
furniture claims "converting to a plywood frame enabled Rowe to streamline the manufacturing process"
(APA, 1999). This also supports the omission of substitution effects in this analysis.
Another factor that supports the reasonableness of omitting substitution possibilities from the
model is the presumption that the price changes induced by the regulation are likely to be small. The
estimated annualized costs of the rule, $120 million, are small relative to the roughly $11 billion value
segments from the industries. Given that the price of any product being considered is only a portion of
the "price" of the composite input of which it is a part, the changes in the relative prices of the composite
inputs to the construction industry are likely to be small as well.
B. 9.3 Own-Price Demand Elasticities
As shown in Exhibit B-l, there are two domestic demand components for softwood plywood and
two for other composite wood products — the construction industry and the manufacturing industry.
Although own-price elasticities for softwood plywood, OSB, and other composite wood products are not
available for these specific demand components, short-term own-price elasticities of total demand have
been estimated by Buongiorno for plywood, particleboard, and fiberboard (Buongiorno, 1996).38 These
total demand elasticities are -0.16, -0.27, and -0.10, respectively Exhibit B-4).
Spelter estimated short-term (one year) elasticities of demand for softwood plywood and
structural particleboard over several-year time spans (1970-1981 for softwood plywood and 1977-1981
for structural particleboard), showing a general downward trend over time, reflecting the effect of
maturing markets (Spelter, 1984). The most recent year for which short-term elasticities of demand are
available from Spelter is 1981. Spelter estimated the price elasticities of demand for softwood plywood
and structural particleboard in 1981 to be -0.10 and -0.56, respectively. These estimated elasticities of
demand from Spelter are supportive of those in Buongiorno. The estimated demand elasticities for
plywood in the two papers are quite similar (-0.16 in Buongiorno, versus -0.10 in Spelter).
Although Spelter's estimated demand elasticity for structural particleboard is somewhat higher
than Buongiorno's estimated demand elasticity for particleboard (-0.56 versus -0.27), the Spelter estimate
corresponds to 1981 whereas the Buongiorno estimate is more recent. If the general downward trend in
short-term price elasticities of demand observed by Spelter continued from 1981 onward, then the more
recent Buongiorno estimate of -0.27 would be consistent with the earlier Spelter estimate of -0.56.
Because the Buongiorno estimates are more recent, and because they are supported by the estimates from
Spelter (1984), this analysis uses the estimates from Buongiorno (1996), given in Exhibit B-4. In the
absence of an estimated elasticity of demand specifically for OSB, the analysis assumes that it is the same
38Fiberboard, a type of reconstituted wood product, is not impacted by the proposed rule.
B-13
-------
as for softwood plywood used in construction.39 This assumption is reasonable considering the following
characteristics.
Softwood plywood and OSB have approximately equal shares of the structural panel market.
Factors of demand for both products are similar.
Both products have reached similar levels of life-cycle maturity.
I'l\hihil l$-4. Own-Price l-'l;is(iciliosol'iToiiih Dcniiind I sod in Aiiiihsis
Product
()«n-Pricc r.l;is(ici(> ol'(To(;ih Dcniiind
Softwood Plywood*
-0.16
Oriented Strandboard**
-0.1034
Other Composite Wood Products*
-0.27
*Based on elasticities of demand estimated in Buongiorno (1996) for plywood (-0.16)
and particleboard (-0.27).
** Assumed to be the same elasticity of demand as for softwood plywood used in
construction, see Exhibit B-5.
For the purposes of this analysis, EPA applied the estimated elasticity of particleboard to the three
products in the other composite wood products category for the following reasons.
Research did not produce estimates of demand elasticities for medium density fiberboard or for
hardboard.
Particleboard constitutes the largest share of production of the other wood composites category.
There are roughly twice the number particleboard plants as medium density fiberboard and
hardboard plants combined.
The elasticities of demand given in Exhibit B-4 refer to total demand. Estimating the elasticities
of demand specific to the construction industry and the manufacturing industry requires either (1)
assuming that they are both equal to the elasticity of total demand or (2) deriving elasticities of demand
for each of these domestic demand components in such a way that the weighted average of these
elasticities (weighted by the demand shares in the market) equals the elasticity of total demand. As
explained below, EPA selected the second approach.
The estimated elasticities of demand for the end products, the production of which requires these
products as inputs, differ substantially. The demand for durable manufactured goods and for wood
furniture have been shown to be substantially more elastic than the demand for residential construction.
The own-price elasticities of demand for durable manufactured goods was estimated to be about -1.4 and
wood furniture was estimated to be about -3.4, whereas the (short-term) elasticity of demand for
39 It should be noted that there is undoubtedly substantial statistical uncertainty surrounding all of the
published elasticity of demand estimates. Although specific point estimates are necessary for the analysis, the
demand elasticities should not be taken to be more precise than they actually are. While the Buongiorno estimate for
plywood (-0.16) was statistically significant, the estimate for particleboard (-0.27) was not. Particularly for OSB and
other wood composites, only a more modest statement (e.g., that these demand elasticities are relatively small,
probably somewhere in the range of 0 to -0.4) is truly supported by the literature. However, given that an actual
point estimate is necessary for the analysis, the point estimate in Buongiorno (1996) is a reasonable estimate.
B-14
-------
residential building construction is estimated to be about -0.58 (U.S. EPA, 1995). The demand for these
products for use in the construction industry is therefore likely to be substantially less elastic than the
demand for use in manufacturing (e.g., of furniture). Therefore, EPA used the second approach — that is,
deriving elasticities of demand for each of these domestic demand components in such a way that the
weighted average of these elasticities (weighted by the demand shares in the market) equals the elasticity
of total demand.
To do this, EPA assumed a ratio (based on reasonable judgment) between the elasticity of
demand for a given product for use in construction versus for use in manufacture, and to calculate the
demand-component-specific elasticities, given that ratio, that are consistent with the domestic demand
component proportions in Exhibit B-l and the elasticity of total demand in Exhibit B-4. For example,
suppose the domestic demand for softwood plywood to be used in manufacturing is twice as elastic as the
domestic demand for softwood plywood to be used in construction. Given the proportions of domestic
demand for softwood plywood by each industry (64% in construction versus 36% in manufacturing) and
the elasticity of total demand (-0.16), the elasticity of demand for softwood plywood to be used in
construction would be calculated as
-0.16
s , = =-0.118 . (17)
construction (Q 36*2+0 64)
The elasticity of demand for softwood plywood to be used in manufacturing would be twice that,
or -0.235. For this purpose of this analysis, EPA assumed that the ratio of the elasticity of demand for
wood products for manufacturing to the elasticity of demand for construction is 2.5. EPA arrived at this
assumption by examining the ratio of elasticities of the finished products for which these goods are
inputs: residential construction (-0.58) and durable manufacturing (-1.40), which is roughly 2.5 (EPA,
1995). Exhibit B-5 presents the base case demand-component-specific elasticities of demand.
r.\hihil l?-5. IH;islicilk's ol' Dciimnd lor Specific l)oiii;ind ( omponciUs
Product
l'l;islkil\ til'Ti>l;d
l)i-m;ind
Assumi-d K;ilii> id'
Khisliiilii's ol' l)i-m;ind:
M:iniiii rin^ In
(iiiisiriiiliiiii
Implied r.hisikilk's id' l)i-m;iiid
(iinslrui'liiiii M;iiiiir;K'luriii^
Softwood Plywood
-0.16
2.5
-0.1034 -0.260
OSB
-0.1034
n/a
-0.1034 n/a
Other Wood
Composites
-0.27
2.5
-0.115 -0.287
(PB/MDF & HB)
Source: Exhibit B-4, Equation 17, Abt Associates Inc. assumptions.
The remaining parameters in the demand-component-specific domestic demand functions (the
yk's from equation (16)) were calibrated to the baseline prices and quantities, given in Exhibit B-6. To
calibrate yk to the baseline price and quantity, the analytical model solved equation (16) for yk and
calculated the value of yk that results if the price is the baseline price and the quantity demanded is the
quantity demanded in the baseline, given a value for the elasticity. Solving equation (16) for yk yields:
B-15
-------
QkDD
Using the 1997 baseline price and quantity of softwood plywood (SWPW) demanded for
construction (k=l) shown in Exhibit B-6, for example, and the price elasticity of demand for SWPW to be
used in construction (shown in Exhibit B-5), the value ofy , swpw is calculated as
Q,T 11,498
y = ^10 = = 21 542 451
r i,swpw 235"0115 .
With this value ofy , SWPW in equation (16), and given the elasticity of demand for SWPW for
construction and the baseline price of SWPW, equation (16) predicts the observed baseline quantity of
SWPW.
Under some circumstances, a reasonable alternative to calibrating yk to the baseline price and
quantity might be to estimate yk based on several years' worth of price/quantity data — that is, to find the
value of yk that results in the best fit of the demand function to the price/quantity data over several years,
given the elasticity of demand. In this approach, it would be preferable to find the best fit value of ak in
equation (15), which is the linear form of equation (16), using linear regression, and then to calculate
Yk = eat .
The regression approach has the advantage that it is based on more than a single year and is
therefore likely to be more robust. However, it will produce a demand curve that will not go through the
baseline equilibrium price/quantity because it will not fit any particular year exactly. Because the
changes in equilibrium prices and quantities in these markets is likely to be small, reasonable estimation
of the new market equilibrium requires that the baseline market equilibrium that we observe is reflected in
the analysis — that is, that the baseline supply and demand curves intersect at the empirical baseline
price/quantity point (i.e., that both curves go through the empirical baseline price/quantity). We will
therefore use the calibration approach.
B-16
-------
IMiihil I5-6. IW
ISiisolino Prices ;iml Qu;iu(i(ii's
li;isi'liiH' Oii:i 111 i 1 \ Sunnlii'd in:
Product
li;isi-liiu- Priii'
(S per iul)ii- iiK'li i )
( iinslrui'liiiii liuluslrx M;i 1111 11 rin^ liuliiMrx
(MlV) (M I'l1)
Softwood Plywood
(3/8 in basis)
235
11,244
6,324
OSB
(3/8 in basis)
185
9,595
0
Other Wood Composites
Particleboard/MDF
(3/4 in basis)
169
1,164
10,481
Hardboard
(1/8 in basis)
1,322
177
1,592
Source: Abt Associates, 1999. Exhibit B-3. Quantities split using data from Exhibit B-l.
The total domestic demand for a product (e.g., softwood plywood) at a given price is the sum of the
domestic demand components — ^ Q™ . In the case of softwood plywood and other wood composites,
k
there are two domestic demand components — i.e., k=l, 2:
QDD = Q™ + Q?D = r?DP«+r?DP'* ,
where
Qi° = domestic demand for the product to be used in construction (k=l);
Q®d = domestic demand for the product to be used in manufacturing (k=2);
P = the price of the product;
o, = the own-price elasticity of demand for the product for use in construction;
S2 = the own-price elasticity of demand for the product for use in manufacturing; and
y["' and y'-!'' = multiplicative constants specific to the domestic demands for the product from
construction and manufacturing, respectively.
In the case of OSB, there is only one domestic demand component — the construction industry. The
s~\DD _ s~\DD _ DD -psosB, 1
domestic demand equation for OSB would be Szosb ~ Sdosb,\~ / osB,\rosB . This is just a special
case, however, of equation (18) above in which YosB 2's zero-
B-17
-------
B.9.4 Foreign Demand
Foreign demand for (exports of) each of these products is small relative to domestic demand. For
example, on a value basis, only 6 percent of softwood plywood and reconstituted wood products are
exported. Moreover, information that would allow any distinction to be made between the elasticities of
demand for these products in foreign markets as opposed to domestic markets is not available. Therefore,
although a foreign demand (export) component was included in the demand side of each market model, it
was necessary to assume that the elasticities of demand are the same in the foreign market as in the
domestic market. In particular, the foreign demand analysis was based on the following assumptions.
The price of a product is the same in foreign markets for the product as it is in the domestic
market for the product.
The price elasticity of foreign demand for the product to be used in a given industry (construction
or manufacturing) is the same as the price elasticity of domestic demand for the product to be
used in that industry.
The proportions of the product in foreign markets that go to the construction versus
manufacturing industries are the same as in the United States.
Given these assumptions, the distinction between domestic and foreign demand is a formality. If the
domestic demand for the product for use by the kth industry (k=l, 2) is given by
QkD = 7kDPSk (19)
and the foreign demand for the product for use by the same industry is given by
QFkD=rlDPsk, (20)
then the total demand for the product for use in the kth industry (both domestically and in foreign
countries) is
QlD=7uDPSk + rlDPSk=Yk PSk, (21)
where = y®D + y'k!). As with the supply parameters, yk, the demand parameters are calibrated to the
baseline quantity demanded and the baseline price.
The total demand for the product (for all uses) in the post-regulation scenario , then, is
q™=Hq™ • (22)
k
B.10 Market Equilibrium
In equilibrium, the total quantity supplied equals the total quantity demanded. To solve for the
new market equilibrium price of the product (e.g., softwood plywood), then, the analysis set the total
B-18
-------
quantity of the product supplied at the new market equilibrium price equal to the total quantity demanded
by all demand components:
er=aro,
(23)
or, substituting in from equations (13) and (22),
£&?
3
k
Substituting in from equations (14) and (21), this can be written:
£0,?+/*+
j
(24)
where the jth domestic facility's quantity supplied post-regulation, (j'f, depends on whether the new
market equilibrium price, P,. is greater than or less than the facility's new shutdown price, as shown
above in equation (9):
This is an equation in one unknown — P,. the new market equilibrium price. The value of P, that
satisfies the new market equilibrium (equation (24)) was solved for iteratively.
B.11 Facility Closures
In the short run, a facility will close if its average variable cost exceeds the market price. Because
facility-specific information on average variable costs in the baseline is not available, however, it is not
possible to base facility closure predictions on empirical data and estimated compliance costs. To the
extent that the assumed facility-specific supply functions are reasonable depictions, however, they can be
used to allow predictions of facility closures in the post-regulation scenario. The facility supply function
shows the quantity that the facility will supply at any given price. The point at which the supply function
intersects the price axis (the y-axis) is therefore the price at which the facility will supply zero. In a short-
run analysis, the price at which the facility will close is called the shutdown price. For those facilities
with positive compliance costs, the shutdown price in the post-regulation scenario is higher than it was in
the baseline, as shown in equation (8), reproduced below:
0 ifPl
-------
The facility's closure decision, originally given in equation (9), is reproduced above as part of equation
(24). Computationally, for any value of P,. the jth facility's quantity produced can be calculated using
equation (7) above:
with the added constraint that if the predicted Qjf < 0 then set Q*f = 0 . Calculation of facility closures
is thus part of the iterative process of calculating the post-regulation market equilibrium price and
quantity. The number of facility closures is simply the number of facilities that produce zero in the post-
regulation scenario.
B.12 Social Cost of the Regulation
Social costs vary based on the characteristics of the markets in which facilities participate. In a
perfectly competitive market, an individual facility's supply curve measures the marginal cost of
production at different quantities of production. Because marginal costs will increase with the imposition
of compliance costs, the supply curves for those facilities with compliance costs will shift up as a result of
the regulation. Because the market level supply curve is just the horizontal sum of individual facility
supply curves, the market supply curve will shift up as well. The (derived) demand for any of these
products measures the total quantity of the product demanded (summed over all demand components) at a
given price, assuming all else (e.g., technology and output of products going to final demand) held
constant. The demands are unchanged by the regulation. Figure 2 below graphically represents the
impact of the regulation on the market for a particular product, which, for the purpose of illustration, can
be softwood plywood.
Figure B-2: Impact of Regulation on Social Welfare
P+
S post- , Q
compliance /
/ baseline
1
Pi a
Po h
g
D
f
Q1 Q»
Q
B-20
-------
The imposition of compliance costs on softwood plywood producers results in changes in both
consumer and producer surplus in the softwood plywood market:
~~~ Change in Consumer Surplus: Consumer surplus measures the difference between what
consumers are willing to pay and what they actually pay for a product. Graphically, this is the
area under the market demand curve and above the market price line. In the baseline (before
imposition of the regulation), this is area hci in Figure 2. When the supply curve shifts up and
the price of softwood plywood increases, consumers lose some surplus. After the imposition of
the regulation, consumer surplus is reduced to area abi. The change in consumer surplus is
therefore area abi- area hci, or a loss of area abch in Figure 2. Note that because softwood
plywood does not go to final demand (i.e., it is used to produce other products that are then
purchased by consumers), the "consumers" of softwood plywood are the producers of the final
goods. The concept of consumer surplus, however, is the same, whether the demand is final
demand or derived demand.
~~~ Change in Producer Surplus: Producer surplus measures the difference between what it costs
suppliers to produce Q units of softwood plywood and the price they receive for those Q units —
their economic profit. Graphically, this is the area above the market supply curve and below the
market price line. In the baseline, before the imposition of the regulation, producer surplus is
represented by the area hcf. When compliance costs are imposed, some producer surplus is
gained if the price received increases, and some producer surplus is lost because the cost of
producing softwood plywood also increases. After the imposition of the regulation producer
surplus is represented by area abg. The change in producer surplus is a mixture of a loss (area
gecf and a gain (area abeh ) to producers. Note that area abeh is actually a transfer from
consumers of softwood plywood to producers of softwood plywood — that is, area abeh is a loss
to consumers of softwood plywood but a gain to producers.
The change in social welfare connected with the ith market (e.g., the market for softwood
plywood) is the sum of the changes in consumer surplus in the ith market (ACSj) and producer surplus in
the ith market (APS,). Using the graphic areas in Figure 2, the total change in social welfare connected
with the ith market resulting from the regulation would be:
ACSj + APSj = ( abi - hci) + ( abeh - gecf)
= - abch + ( abeh - gecf).
Decomposing area abch into its two component parts — abeh and bee,
ACSj + APSj = - (abeh + bee) + ( abeh - gecf)
= - bee - gecf = - gbcf
The net impact in each market is a social cost — that is, the change in social welfare (producer
plus consumer surplus) is negative. The social cost associated with the changes in the softwood plywood
market can be calculated by calculating area gbcf for that market — that is, given the specific market
supply and demand curves. The total social cost of the regulation is the sum of the social costs associated
with each of the affected markets — for softwood plywood, OSB, and other wood composites.
B-21
-------
B.12.1 Calculation of Change in Social Welfare
The change in social welfare is the sum of the change in producer surplus and the change in
consumer surplus. The derivation of the changes in producer and consumer surplus are presented below.
B.12.1.1 Change in Producer Surplus
Figure B-3: Facility-Specific Supply Curve
Inverse Supply Function Supply Function
Q0(P) = 2 Qjo(P)
Quantity Price
Producer surplus in the baseline
Let A0 = producer surplus in the baseline.
The jth facility's supply curve in the baseline is given by:
Qjo = max{0,x7- + (/?/2)JP~12 }
The market supply curve is the sum of the facility-specific supply curves40:
&=L 2,0=1 max{0,/, + (j3/2)P-m}
J j
40 The foreign sector is treated as a "foreign facility".
B-22
-------
Producer surplus in the baseline = area A0 =
Po Po
Iq„(P)cIP= JI maxj0,^+(/;/2)p-'2 }clP
Po Po j
Po
=Z \\r,Hm)P~v*VP
j ps
rj o
where /'0V is the price at which zero is supplied to the market in the baseline (the "market shutdown
price"), and PJ0 is the jth facility's shutdown price in the baseline. Note that the jth facility's supply
curve is flat (i.e., Qj = 0) up to Pf0 and then Qj>0:
Figure B-4: Constrained Supply Curve
J Pjo
p p
= Z;/./ \dP+(J3l2)2 \p-ll2dp
^ Pjo J Pjo
A=I r., [ n ¦- Pfl 1+ PL I Pf
j j
B-23
-------
Producer surplus in the post-regulation scenario
Let Aj = producer surplus in the post regulation scenario.
The jth facility's supply curve in the post-regulation scenario is given by:
Qn = max{0/; + (P!2)(P- c] )1/2} .
Where Cj is the jth facility's per unit output compliance cost. The market supply curve is:
a=Ze„=I max{0,r, HpiTiP-c, r"2}
j i
Producer surplus in the post-regulation scenario = area A1 (analogous to A0 in the baseline):
Pi Pi
A,= jQ,(P)dP= J£ max{0,rJHff2KP-cjyh2 }dP
P" PI j
Pi
=1 J[/,+(/;/2)(p-c.)-i2]ip
j pf,
where J\' is the price at which zero is supplied to the market in the post-regulation scenario (the "market
shutdown price"), and I'-\ is the jth facility's shutdown price in the post-regulation scenario.
Now doing a change of variable, let x = P - Cj > P = x + c, and dP = dx. Then
Pi-Cj
A=I \\_y}+(PI2)xl/2Ykc
1 Pji-Cj
Pl ~cj Pl ~cj
=Ydrj \dx+(J3!2)X \x~l/2dx
J Pjl~°j J Ph~Cj
pi-cj
\x mdx
/ j Pj\-Cj
4=L r, [ ^ - -P/, ]+ PL [(^1 - c,) "2 - < p'a - c, y1,2 ]
j j
B-24
-------
The change in producer surplus, APS = Aj - A0.
B. 12.1.2 Change in Consumer Surplus
Figure B-5: Market Level Demand Curve
Inverse Demand Function
Demand Function
Qi* Qo
Quantity
•fQo*
S3
C5
Qi!
QTD(P)
/
Price
Demand by the kth demand component:
Qk=rkPSk ¦
The total demand at price P is
QTD=YJrlP"=rtP'^rIP'7
k
Note: Foreign and domestic demand components are combined in this analysis, as explained in
"Economic Analysis Methodology for the Plywood and Wood Composites Industries." There are
therefore two demand components : construction and manufacturing.
The change in consumer surplus is just the area under the demand curve from P, to P0:
Po
a cs=\[rlPSl+r2pS2W
B-25
-------
= yl\pSldP+y2\PSldP
p, p,
A CS= , Yl „[i30gl+1-i31gl+1]+ 72
(ffi + 1)
02 + 1)
B.13 Computational Tools
The analysis calculated the post-regulation equilibria for the four separate market sectors using an
Excel spreadsheet for each. Given parameter values for the relevant supply and demand equations in a
given market, and facility-specific per unit compliance costs, EPA programmed the following.
The relevant market segment demand equations (incorporating both domestic and foreign
demand, as discussed above),
The foreign supply equation, and
Facility-specific domestic supply equations.
EPA used an iterative procedure in which, on each iteration, a new equilibrium market price (P,)
is input to the equations, and the resulting demand and supply are compared. The value of P, on the
subsequent iteration will be adjusted according to whether market demand exceeds market supply (in
which case, P, would be increased) or market supply exceeds market demand (in which case, P, would be
decreased) until market supply equals market demand — i.e., equilibrium is achieved.
B.14 Sensitivity Analyses
All aspects of the economic analysis depend on the assumptions made about the supply and
demand functions. The closure analysis and the estimation of social cost may be particularly sensitive to
the values of the parameters chosen for the analysis. Because of this, EPA may choose to perform a
sensitivity analysis to show how the analysis outcomes — equilibrium price and quantity, number of
closures, and social cost of the regulation — vary as the input parameter values vary. This is
straightforward to carry out in a spreadsheet analysis in which parameter values in designated cells can be
easily changed and the computations rerun.
B.15 EIA Results
As described throughout this appendix, the economic analysis produced estimates of how the
imposition of compliance costs associated with the regulatory proposal will change the market
equilibrium price and quantity. The methodology described here provides estimates of how production
will shift across impacted producers, non-impacted producers, and foreign suppliers. This approach also
provides estimates of net social costs associated with the proposed rule.
There are some additional insights into the rule's impacts that can be gained from examination of
the EIA's results. EPA examined the overall EIA result of the change in market quantity as an indicator
of how much less production capacity will be required to meet overall demand. Initially, it is of interest,
B-26
-------
to the extent possible, to examine the facility-level impacts of the rule associated with this decline in
demand.
EPA assumed that the facilities with the highest baseline production costs will experience the
greatest impacts associated with the rule. The impacts will depend on the facility's response to increased
costs, and whether the increase in costs places the facility's overall production costs above market prices.
Responses will vary widely from facility to facility: they could close, reduce production by closing a
process line, or cut costs elsewhere, among others. The analytical approach described in the earlier
section on domestic supply allows EPA to estimate facility closures in the post-regulation scenario in the
absence of actual facility-specific information on average variable production costs.
EPA iteratively calculated the post-regulation equilibrium market price. As part of the iterative
procedure EPA determined, for each facility, it's new shutdown price in the post-regulation scenario and
therefore whether or not it closes. If the quantity it would supply at the post-regulation market price is
<0, then the facility will close. In other words, the facility will close if the post-regulation market
equilibrium price is less than or equal to the facility's post-regulation shutdown price.
B.16 References
Abt Associates Inc. 1999. Profile of the Plywood and Wood Composite Industries. Prepared for Larry
Sorrels, Innovative Strategies and Economic Group, Air Quality Strategies and Standards
Division, Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, NC.
Adams, Darius M., R. Boyd, and J Angle. 1992. "Evaluating the Stability of Softwood Lumber Demand
by End-Use Sector: A Stochastic Parameter Approach." Forest Science, Vol. 38, No. 4, pp 825-
841.
American Plywood Association. 1999. Upholstered Furniture Case Studies, http://www.apawood.org.
Buongiorno, Joseph. 1996. "Forest Sector Modeling: A Synthesis of Econometrics, Mathematical
Programming, and System Dynamics Methods." International Journal of Forestry, 12, pp. 329-
343.
Buongiorno, Joseph. 1999. Personal communication with Ellen Post, Abt Associates Inc. November 23,
1999.
Chambers, Robert G. 1988. "Applied Production Analysis: A dual approach." Cambridge University
Press.
Get-a-Quote.net. 2000 National Construction Estimator, Panelized Wood Wall Panels.
http://www.get-a-quote.net, page 395.
Lewandrowski, Jan K., M. K. Wohlgenant, T. J. Grennes. 1994. "Finished Product Inventories and
Prices Expectations in the Softwood Lumber Industry." American Journal of Agricultural
Economics, 76, pp. 83-93.
B-27
-------
Spelter, Henry. 1984. Price Elasticities for Softwood Plywood and Structural Particleboard in the
United States. Forest Products Laboratory, U.S. Department of Agriculture Forest Service,
Madison, WI.
U.S. Environmental Protection Agency. 1995. Appendix B: Complete List of Own-price Demand and
Supply Elasticities. Air Quality Strategies and Standards Division, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1998. Industry Specific Information Collection Request (ICR) for
the Development of Plywood and Particleboard Maximum Achievable Control Technology
(MACT) Standards. Air Quality Strategies and Standards Division, Office of Air Quality
Planning and Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1999a. OAQPS Economic Analysis Resource Document. Air
Quality Strategies and Standards Division, Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1999b. Economic Impact Analysis for the Polymers & Resins III
NESHAP. Air Quality Strategies and Standards Division, Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC.
U.S. Environmental Protection Agency. 1999c. 1999 Revised RFA/SBREFA Guidance for EPA
Rulemakers.
B-28
------- |