United States Office of Air Quality EPA-450/3-89-006
Environmental Protection Planning and Standards August 1B89
Agency Research Triangle Park, NC 27711
Air
vvEPA Economic Impact
of Air Pollutant
Emission Standards
for New Municipal
Waste Com bustors
This document is printed on recycled paper.
-------
ERRATA SHEET
October 12, 1989
Economic Impact of Air Pollutant Emission S^ndards
for New Municipal Waste Combustors
Page Number Error Correction
1-3 676 (PV of Social Capital Costs 679
for Reg. Alt. IV)
6-9 41.4 (Present Value of 47.4
Capital Costs for Reg. Alt. I)
6-14 676 (PV of Social Capital Costs 679
for Reg. Alt. IV)
9-2 676 (PV of Social Capital Costs 679
for Reg. Alt. IV)
9-2 1,190 (Present Value of Social 1,910
Costs for Reg. Alt. IIA)
-------
EPA-450/3-89-006
EPA Contract Number RT| Project Number
68D80073 233U-4300-12-09-FR
Economic Impact of Air Pollutant
Emission Standards for New
Municipal Waste Combustors
Final Report
August 1989
Prepared for
John Robson
Office of Air Quality Planning and Standards
U S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared by
Glenn E. Morris
Brenda L. Jellicorse
Katherine B. Heller
P Timothy Neely
Tayler H. Bingham
Center for Economics Research
Research Triangle Institute
Research Triangle Park, NC 27709
-------
This report has been reviewed by the Emission Standards Division, Office of Air
Quality Planning and Standards, Office of Air and Radiation, U.S. Environmental
Protection Agency (EPA), and approved for publication. It is issued by EPA to
report technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and grantees,
and non-profit organizations—as supplies permit—from the Library Services
Office (MD-35), U.S. Environmental Protection Agency, Research Triangle Park
NC 27711, or may be obtained, for a fee, from the National Technical
Information Service, 5285 Port Royal Road, Springfield, VA 22161.
This report was furnished to EPA by the Center for Economics Research,
Research Triangle Institute, Research Triangle Park, NC 27709, in fulfillment of
assignments under EPA Contracts 68-02-4321 and 68D80073. The contents are
reproduced herein as received from the Contractor. The opinions, findings, and
conclusions expressed are those of the authors and not necessarily those of EPA
Mention of company or product names does not constitute endorsement by EPA.
Publication No. EPA-450/3-89-006
-------
ACKNOWLEDGMENTS
The authors would like to acknowledge the guidance of our EPA Project Officer, John
Robson, who, along with other EPA personnel, provided us with important insights throughout
the preparation of this report. We also appreciate the dedicated editorial and clerical support of
Maria Bachteal, Craig Hollingsworth, Andrew Jessup, and Judy King.
111
-------
CONTENTS
Chapter
Page
1 Introduction and Summary 1-1
1.1 Costs of Regulation 1"2
1.2 Emission Reductions I"5
1.3 Distribution of Economic Impacts 1-6
1.3.1 MWC Plants and Technologies 1-7
1.3.2 Households !'8
1.3.3 Government Units 1-1°
2 Demand Conditions 2-l
2.1 Generators 2-1
2.2 Generator Behavior 2-3
2.2.1 Household Demand 2-3
2.2.2 Firm Demand 2-7
2.3 Waste Disposal Services Demand 2-8
3 Supply Conditions 3-1
3.1 Production Processes 3-2
3.1.1 Combustion 3-3
3.1.2 Landfilling 3-8
3.1.3 Collection and Transportation 3-13
3.1.4 Recycling 3-14
3.2 Production Costs 3-15
3.2.1 Combustion 3-18
3.2.2 Landfilling 3-20
3.2.3 Collection and Transportation 3-20
-------
CONTENTS (continued)
Chapter Pa8e
4 Municipal Solid Waste Management 4-1
4.1 Public Involvement in the MSW System 4-1
4.1.1 Local Government 4-1
4.1.2 State Government 4-9
4.1.3 Federal Government 4-11
4.2 MSW Decision Making 4-12
4.2.1 Private Decision Making 4-12
4.2.2 Government Decision Making : 4-13
4.2.3 Cost Minimization 4-14
5 Analytical Approach to Estimation of Cost and Emission Impacts 5-1
5.1 Baseline Projections 5-1
5.1.1 Initial Conditions 5-1
5.1.2 Projections 5-4
5.1.3 Baseline Combustioa 5-5
5.2 Scenarios 5-6
5.3 Regulatory Alternatives 5-9
5.3.1 Baseline Emissions 5-11
5.3.2 Regulatory Alternative 1 5-11
5.3.3 Regulatory Alternative IIA 5-11
5.3.4 Regulatory Alternative IIB 5-12
5.3.5 Regulatory Alternative III 5-12
5.3.6 Regulatory Alternative IV 5-12
5.4 Cost and Emission Reduction Estimation 5-14
5.4.1 Scenario I: No Substitution 5-14
5.4.2 Scenario II: MWC Substitution 5-16
5.4.3 Scenario III: MWC/Landfill Substitution 5-17
-------
CONTENTS (continued)
Chapter Pa§e
6 Cost and Price Impacts 6-1
6.1 Model Plants and the Cost of Regulatory Alternatives 6-1
6.2 National Enterprise Costs of Each Regulatory Alternative 6-9
6.3 Price Impacts 6-10
6.4 Social Costs 6-12
7 Emission Reductions and Cost-Effectiveness 7-1
7.1 Emission Reductions and Energy Impacts 7-1
7.2 Cost-Effectiveness 7-3
8 Economic Impact on Sectors of the Economy 8-1
8.1 Regulatory Flexibility Analysis 8-1
8.2 Private Business Impacts 8-2
8.2.1 Private Owner Profile 8-3
8.2.2 Private Supplier Profile 8-7
8.3 Impacts on Households and Government Entities 8-7
8.3.1 Household Impacts 8-12
8.3.2 Governmental Impacts 8-16
9 Sensitivity Analysis 9-1
References R'l
APPENDIX
A Estimation of the Real Discount Rate for Private Firms and
Public Entities A-l
vn
-------
TABLES
Number Page
1-1 NSPSs Economic Impact Scenarios 1-2
1-2 NSPSs National Cost Impacts (1987 $) 1-3
1-3 NSPSs National Baseline Emissions and Emissions Reductions (Mg per Yr.) 1-6
1 -4 Enterprise Costs of Control for Publicly Owned NSPSs Model Plants (1987$):
Scenario I, Regulatory Alternative IV 1-7
1-5 Enterprise Cost of Control for Publicly Owned NSPSs Model Plants Under
Scenario I (1987 $) 1-8
2-1 Materials in the Municipal Waste Stream, 1986 2-2
3-1 Estimated Flows of Municipal Solid Waste to Municipal Waste Combustion
Plants, 1980 Through 1986 3-4
3-2 Characteristics of NSPSs Model Plants 3-9
3-3 Production Characteristics of NSPSs Model Plants 3-10
3-4 Production-Cost Relationships of NSPSs Model Plants ($1987) 3-19
3-5 Production-Cost Relationships of Landfills 3-21
3-6 Costs of Hazardous Waste Landfilling 3-22
3-7 Costs of Collecting and Transporting Municipal Solid Waste 3-22
4-1 Types of Solid Waste Collection (Percent of Generators Served) 4-2
4-2 Methods of Financing Solid Waste Collection by Collection Agency,
1964 (Number of Cities) 4-3
4-3 Methods of Financing Solid Waste Collection by City Size, 1964 (Percent) 4-4
4-4 Ownership of Municipal Waste Combustion Plants, by Size 4-5
4-5 Ownership of Landfills, by Size 4-6
4-6 Total Operating Subsidies as a Share of Total Revenues (Municipal Waste
Combustion Plants Only) 4-7
viu
-------
TABLES (continued)
Number Page
4-7 Grants as a Share of Total Capital Costs (Municipal Waste Combustion,
Plants Only) 4-7
4-8 State Solid Waste Laws Enacted in 1988 4-10
5-1 Baseline Plant Capacity and Waste Flow Estimates for MWC Plants Subject
toNSPSs 5-6
5-2 Maximum Emissions by Regulatory Alternative 5-10
5-3 Air Pollution Controls by Regulatory Alternative 5-13
5-4 Control Options by NSPSs Model Plant for Each Regulatory Alternative 5-14
5-5 Scaling Factors Used to Obtain Scenario I National Cost Estimates 5-17
5-6 Estimated Waste Flow Shares 5-18
5-7 Estimated Waste Flows Subject to Disposal Choice by Technology and
Regulatory Alternative (106 Mg/Yr) 5-19
5-8 NSPSs Model Plant National Waste Flows and Scaling Factors 5-21
6-1 NSPSs Enterprise Costs of Control for Publicly Owned Model Plants (1987$)a 6-2
6-2 NSPSs Enterprise Costs of Control for Publicly Owned Model Plants:
Scenario I, Regulatory Alternative IV 6-7
6-3 NSPSs Enterprise Costs for Publicly Owned Model Plants3 (1987$):
Scenario I Cost per Mg of Municipal Solid Waste and Percentage Changes
in Cost over the Baseline for Each Regulatory Alternative 6-8
6-4 NSPSs National Cost Impacts: Enterprise Costs for Publicly Owned
Model Plants (1987 $) 6-9
6-5 Percentage Price Increases Based on Full Pass Through of Estimated
NSPSs Enterprise Costs of Control per Mg of Municipal Solid Waste 6-12
6-6 NSPSs National Cost Impacts (1987 $) 6-14
7-1 NSPSs National Baseline Emissions and Emissions Reductions (Mg per Yr.) 7-2
7-2 NSPSs National Energy Impacts 7-3
ix
-------
TABLES (continued)
Number Page
8-1 Private Owner Profile 8-4
8-2 NSPSs Enterprise Costs of Control for Privately Owned Model Plants Under
Regulatory Alternative IV: Ordered by Design Capacity 8-6
8-3 MWC Systems Private Supplier Profile: Project Managers 8-8
8-4 MWC Systems Private Supplier Profile: Technology Supplier 8-9
8-5 Estimated Ownership of NSPSs Plants 8-11
9-1 NSPSs National Cost Impacts: Social Costs Using a Two-Step Discounting
Procedure (1987 $) 9-2
9-2 NSPSs National Cost Impacts: Social Costs Using a 10 Percent Discount
Rate (1987$) 9-3
9-3 NSPSs National Cost Impacts: Social Costs Using a 3 Percent Discount
Rate (1987$) 9-4
9-4 Scenario I Scaling Factors Calculated Using a Higher Capacity Utilization 9-5
9-5 NSPSs National Cost Impacts: Social Costs Using a Higher Capacity
Utilization (1987 $) 9-6
9-6 NSPSs National Baseline Emissions and Emissions Reductions (Mg per Year):
Higher Capacity Utilization 9-7
-------
FIGURES
Number Pa§e
1-1 Distribution of Household Impacts Under NSPSs by Number of Service Areas
and Regulatory Alternative: Index 1 1-9
2-1 Sources of Municipal Solid Waste, 1986 2-2
2-2 Effect of Income Changes on Household Demand for MSW Collection and
Disposal Services 2-4
2-3 Effect of Collection and Disposal Price Changes on the Household Demand
for MSW Collection and Disposal Services 2-6
3-1 Solid Waste Management Options 3-1
3-2a Operating Decisions: Choices Involved with at Least One Fixed Input 3-3
3-2b Investment Decisions: Choices Involved with No Fixed Inputs 3-3
3-3 Percentages of Total Capacity of Mass Burn, RDF, FBC, and Modular
Technologies that Will Be Affected by the NSPSs 3-5
5-1 Solid Waste Flow Projections, 1986 Through 1996 5-2
5-2 Comparison of MWC Capacity Projections, 1986 to 1996 5-7
5-3 Municipal Waste Combustion Response under Scenario I: No Substitution 5-15
8-1 Distribution of Household Impacts Under NSPSs by Number of Service Areas
and Regulatory Alternative: Index 1 8-14
8-2 Distribution of Household Impacts Under NSPSs by Number of Service Areas and
Regulatory Alternative: Index 2 8-15
8-3 Distribution of Household Impacts Under NSPSs by Service Area Population and
Regulatory Alternative: Index 1 8-17
8-4 Distribution of Household Impacts Under NSPSs by Service Area Population and
Regulatory Alternative: Index 2 8-18
8-5 Distribution of Government Impacts Under NSPSs: Preliminary Screening
Results 8-20
XI
-------
FIGURES (continued)
Number Page
8-6 Distribution of Government Impacts Under NSPSs by Service Area Population and
Regulatory Alternativeb: Index 1 8-22
8-7 Distribution of Government Impacts Under NSPSs by Service Area Population and
Regulatory Alternative13: Index 2 8-23
8-8 Distribution of Government Impacts Under NSPSs by Service Area Population and
Regulatory Alternative15: Index 3 8-24
xn
-------
CONVERSIONS AND DEFINITIONS
This report uses metric units, as well as acronyms and terms that may not be familiar to
all readers. Following is a short guide to conversions and definitions for a selection of the units,
acronyms, and terms.
CONVERSIONS
To Approximate
Mg
(megagram)
As
Ton
(2,000 Ib)
Multiply by
1.1025
Examples from Text
45 Mg « 50 tons
225 Mg - 250 tons
g/dscm
(grams/dry standard
cubic meter)
TJ
(terajoule)
TJ
(terajoule)
km
(kilometer)
gr/dscf 0.44
(grains/dry standard
cubic foot)"
106Btu 948
(million British
Thermal Units)
MWh 278
(megawatt
hours)
mile 0.62
0.02 g/dscm
0.18 g/dscm
8.54 TJ
34.2 TJ
4.32 TJ
13 TJ
10km
25km
50km
100km
0.01 gr/dscf
0.08 gr/dscf
8,i00106Btu
32,400 106Btu
1,200 MWh
3,600 MWh
5 miles
15 miles
30 miles
60 miles
("Celsius)
OTHER MEASURES
(Fahrenheit)
[F = (9/5) C + 32]
150°C « 300°F
175°C - 350°F
230°C - 450°F
hectare 1,000 square meters (m^)
Nanogram-one billionth of a gram
Normal cubic meter (A normal cubic meter is at 0°C, while a standard
cubic meter is at 20°C; both at 1 atmosphere of pressure.)
103; 106 Thousands; Millions
POLLUTANTS
CDD/CDF
CO
HC1
PM
SO2
Polychlorinated dibenzo-p-dioxins and dibenzofurans
Carbon monoxide
Hydrogen chloride
Paniculate matter
Sulfur dioxide
<; ']
-------
GENERAL ACRONYMS
APCD Air pollution control device
FBC Fluidized bed combustion
MSW Municipal solid waste
MWC Municipal waste combustor
RDF Refuse-derived fuel
ECONOMIC TERMS
National The sum of the regulatory costs incurred by each MWC, discounted and
enterprise annualized at market interest rates
cost
National The sum of the regulatory costs incurred by each MWC, discounted and
social cost annualized at interest rates reflecting society's opportunity costs for capital
and consumption
Net present The estimated present value (PV) of the offsetting revenue required to
value (NPV) cover the full cost of the NSPSs.
Net present The sum of PV capital costs and PV of operating costs net of PV of
cost (NPC) salvage recovery.
1987$ Constant (real) dollars at their fourth quarter 1987 value
Tipping fee The charge for incinerating or landfilling MSW, usually $/Mg, imposed by
MWCs or landfill operators on MSW collectors. Tipping fees, where they
are charged, do not reflect the cost of collecting and transporting MSW to
the disposal site and often fail to reflect the full cost of incineration or
landfilling.
Unit cost The full cost of incinerating MSW, in $/Mg, after subtracting credits for
electricity and steam
WACC Weighted average cost of capital (See Appendix A)
REGULATORY AND LEGISLATIVE TERMS
Baseline
Guidelines
Conditions that would exist were there to be no new Clean Air Act
§11 l(b) and (d) regulation of MWCs
Clean Air Act §11 l(d) emission standards for existing sources
Model Plant A hypothetical MWC representative of a class of MWCs; used to analyze
impacts of regulatory alternatives
NSPSs Clean Air Act §11 l(b) new source performance standards
RCRA Resource Conservation and Recovery Act
Regulatory Sets of performance standards and related requirements for controlling
Alternatives emissions; used by EPA to help select the stringency of regulations. (See
Tables 5-2 and 5-3.)
RFA
Regulatory Flexibility Act; also regulatory flexibility analysis, a study of
the impact of regulations on small entities (businesses, governments, and
organizations)
xiv
-------
§ 111 (b) Clean Air Act section governing emission standards for new sources
(NSPSs)
§11 l(d) Clean Air Act section governing emission standards for existing sources
Subtitle C RCRA subtitle governing hazardous waste landfills
Subtitle D RCRA subtitle governing sanitary landfills
xv
-------
ASSUMPTIONS AND CONVENTIONS
Myriad assumptions, analytical conventions, and underlying calculations form the basis
for projecting the economic impacts of EPA regulations. This page summarizes the principal
assumptions, conventions, and calculated values used in this report. Chapter 9 describes how
projected impacts would be different if some of these assumptions, conventions, and values are
changed.
- Effective date for the § 111 (b) NSPS: January 1, 1990
• Affected MWCs: All MWCs placed under construction on or after the effective date
• Date for which impacts are evaluated: January 1, 1995 (This analysis covers MWCs to be
placed under construction 1990 through 1994.)
• Lifetimes of physical facilities:
- MWCs: 30 years after incurring initial compliance costs
- APCDs: 15 years
• % utilization of daily capacity (There are some exceptions. These percents remain constant
over time.):
- Mass burn: 85%
- RDFandFBC: 83%
- Modular: 82%
• Monetary units: Constant (real) 1987 dollars, usually for the 4th quarter
• Capital costs for each MWC and APCD:
- Incurred only at the outset of operation of the MWC or APCD
- Amortized over the lifetime of the MWC or APCD when included in annualized costs
• Annual operating costs and revenues for each MWC or APCD:
- Invariant over the lifetime of the MWC or APCD
- Proportional to MWC capacity utilization (for analysis purposes when alternative
capacity utilization rates are introduced)
• Market discount rates for computing accounting costs:
- 8% real WACC for private MWCs
- 4% real municipal revenue bond rate of interest for public MWCs
• Social discount rates for computing social costs:
- 10% for capital costs
- 3% for operating costs
xvi
-------
CHAPTER 1
INTRODUCTION AND SUMMARY
The U.S. Environmental Protection Agency (EPA) plans to propose New Source
Performance Standards (NSPSs) for air emissions from new municipal waste combustors
(MWCs) in late 1989. * Affected plants include all MWC plants that are placed under
construction after regulations are proposed in the Federal Register! These regulations will
affect the number of plants built and the combustion technology selected. These regulations will
also significantly affect the cost of owning and operating these new plants.
This report uses three economic scenarios to examine the economic impact of the five
regulatory alternatives under most active consideration by EPA.3 We order the scenarios to
reflect increasing levels of cost-reducing waste management choices as shown in Table 1-1.
With the exception of Regulatory Alternative III the regulatory alternatives are ordered to reflect
both increasing stringency of air emission limits and broader industry coverage of more stringent
limits. As shown in Table 1-1, we provide quantitative estimates of the economic impact of each
regulatory alternative under Scenarios I and III. We expect that quantitative results for these two
scenarios in most instances bracket the results for Scenario II, had Scenario II impacts been
computed.
The economic impacts reported here are based on a wide variety of estimates and
assumptions. The following major assumptions frame the analysis:
• In the absence of the NSPSs, emissions from new MWC plants will just meet current
federal limits.
• All plants that burn municipal solid waste (MSW), however small the fraction of MSW
in the fuel stream, are subject to the NSPSs.
• Affected facilities are those projected for construction over the period 1990-1994.
1 New Source Performance Standards (NSPSs) covering MWC plants are sometimes referred to as 11 l(b) Standards
because provisions of Section 11 l(b) of the Clean Air Act require EPA to propose regulations establishing
federal standards of performance for new stationary sources that contribute significantly to air pollution that
may endanger public health or welfare.
2 Concurrent with the NSPSs, EPA plans to propose a parallel regulation that provides emission guidelines for
existing MWCs under the authority granted by Section 111 (d) of the Clean Air Act Another economic impact
analysis (EPA, 1989a) addresses the effects of these 11 l(d) Guidelines.
3This analysis does not incorporate the impacts of materials separation requirements and nitrogen oxide emission
reduction requirements currently being considered under the NSPSs because of their late inclusion in the
regulatory structure.
1-'
-------
TABLE 1-1. NSPSs ECONOMIC IMPACT SCENARIOS
Regulatory
Alternatives
I
HA
Extent and Stringency
of Coverage
Small Plant Large Plant
(<225 Mg/day) (>225 Mg/day)
GCPs3 GCPs
Moderate PMb Best PM
GCPs GCPs
Moderate PM Good Acid Gasc
Best PM
Economic Impact Scenarios
Scenario I:
Baseline Levels of
MWC Activity
#
#
Scenario 11:
Cost-Reducing Choices
of MWC Technology
*-
+/-
Scenario 111:
Cost-Reducing Choices
of Waste Disposal
#
#
HB
m
IV
GCPs
Good Acid Gas
Best PM
GCPs
Moderate PM
GCPs
Good Acid Gas
Best PM
GCPs
Good Acid Gas
Best PM
GCPs
Best Acid Gasd
Best PM
GCPs
Best Acid Gas
Best PM
#
#
#
- Good combustion practices (GCPs) include proper design and operation of the combustor. Exhaust gas temperature control is also
included in all alternatives with GCPs.
b Paniculate matter (PM) control levels are shown in Table 5-2.
c Good Acid Gas control reduces emissions through the use of dry sorbent injection.
d Best Acid Gas control reduces emissions through the use of spray dryers and fabric filters.
Key: "#" = Quantitative analysis
"+/-" = Qualitative analysis
-------
The first two assumptions tend to boost both the cost and emission reductions attributed
to the NSPSs. Under the third assumption, we confine the analysis to plants built in the initial
five years of the NSPSs. While the analysis covers the entire 30-year life of these plants, the
third assumption still limits cost and emission reduction -stimates attributed to the NSPSs to
those plants covered during the initial period of regulation.
1.1 COSTS OF REGULATION
Table 1-2 lists the estimated national social costs of these regulatory alternatives under
two of the economic scenarios. Under Scenario I we project that roughly 67 plants representing
capacity of 15 million Mg per year will be affected. Under Scenario III, the projected number of
MWC plants ranges from 64 for Regulatory Alternative I to 49 for Regulatory Alternative IV.
The level of MWC capacity also varies with each regulatory alternative—from 14.5 million Mg
per year for Regulatory Alternative I to 9.65 million Mg per year for Regulatory Alternative IV.
TABLE 1-2. NSPSs NATIONAL COST IMPACTS (1987 $)
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
Scenario HI
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative III
Regulatory Alternative IV
PVof
Social
Capital
Costs
($106)a
37.8
227
268
638
676
36.9
150
185
398
430
Annualized
Social Costsb
($10«/yr)
6.41
97.2
115
150
168
6.26
67.6
82.0
93.8
107
Annualized
Social Costs
per Mg
MSWb>d
($/Mg)
0.46
6.99
7.70
10.80
11.20
0.46
6.86
7.69
10.60
11.10
Annualized
Enterprise
Costs per
Mg MSWc'd
($/Mg)
0.37
6.45
7.10
9.23
9.69
0.37
6.35
7.11
9.09
9.65
a Present value of capital costs are based on 2 consecutive, 15-year life cycles for APCD
equipment over the 30-year plant life. These assumptions make no difference in the
annualized cost impacts (Robson, 1989).
b Annualized social costs are the sum of capital costs, annualized at 10 percent, and annual
operating costs.
c Annualized public enterprise costs are the sum of capital costs, annualized at 4 percent, and
annual operating costs.
d Computed by dividing total annualized cost by the estimated amount of MSW processed per
year.
-------
Estimated annualized social costs of the NSPSs (exclusive of the cost of building or
controlling emissions from substitute landfills) increase substantially with the scope and
stringency of regulatory alternatives. The large magnitudes result from both the substantial cost
of installing and operating additional control equipment on any individual plant and from our
baseline projection and scenario projections of the number of plants affected.
As shown in Table 1-2, moving from Scenario I to Scenario III under Regulatory
Alternative I results in a 2.3 percent reduction in total annualized social cost and a 2.4 percent
reduction in social capital cost. This reduction results from our factoring into the analysis the
projected replacement of some of the baseline MWC plants with new landfills because of the
additional cost of the MWC plants under the regulations. As the regulatory alternatives become
more stringent the percent reduction in costs from Scenario I to Scenario III increases. Moving
from Scenario I to Scenario III under Regulatory Alternative IV results in a 36.3 percent
reduction in total annualized social cost and a 36.4 percent reduction in social capital costs due to
the increasing number of MWC plants that are replaced with new landfill capacity. Note again,
however, that the cost of control for Scenario III do not reflect the cost of building or controlling
emissions at the landfills that replace some of the affected MWCs.
The capital cost reported in Table 1-2 represents the estimated purchase and installation
cost of capital equipment consistent with the NSPSs. About two-thirds of this capital cost
represents the "first cost" of the air pollution control devices (APCDs); the other third represents
the present value of the capital and installation cost when replacing the original APCDs after 15
years. These expenditures can be amortized over a 30-year operating life and have therefore
been included in the annualized cost data. The capital costs of APCDs, however, also represent a
substantial initial expense. Because APCDs represent from 5.1 to 22.3 percent of plant capital
cost, they possibly increase the financial risk associated with building a MWC plant. For
comparison, in 1986 the government enterprise expenditure for fixed capital for air pollution
control (primarily for control of municipally owned power plants) was $330 million (1987
dollars) and the capital cost for all solid waste collection and disposal by local government was
$1,060 million (1987 dollars) (Farber and Rutledge, 1988).
Table 1-2 also presents the average cost per Mg of waste (annualized cost divided by the
amount of MSW processed by affected plants). Such measures are often referred to as costs per
unit of waste disposed, or "unit costs." The unit social cost is based on the social costs of the
regulation: the cost seen from a social, opportunity cost perspective. The unit enterprise cost is
1-4
-------
based on ,he enterprise cos, of .ne regulation: ,he cos. as seen in *e accounts of*. .f-ed
entities in .his case primarily municipaH.ies or pub.ic authorities. Because of .he d.fference ,„
the basis for measurement, the uni, social cos.s are about 8 ,o 24 percent greater than urn,
erprise costs. The uni. costs increase substantial with the rectory aHernat.ves, .nd.canng
L nlh higher average costs associated with the broader scope and more suingen, con,ro,s of
.he higher regulatory al.ernatives.
To help pu, these uni, costs in rough perspective, .he average price for disposing of a Mg
of Was,e a, a MWC ,ha, charged a ".ipping fee" (a fee paid by the .ash hauler for the priv.lege of
dumping or .ipping .ash a, .he MWC plan,, in ,988 was $4,7
-------
TABLE 1-3. NSPSs NATIONAL BASELINE EMISSIONS AND EMISSIONS
REDUCTIONS (Mg per yr.)
Scenario and
Regulatory
Alternative
CDD/CDF
CO
PM
SO2
HCI
Pb
Ashb
Scenario I
Baseline Emissions
Emissions Reductions
Regulatory Alternative I
Regulatory Alternative HA
Regulatory Alternative IIB
Regulatory Alternative III
Regulatory Alternative IV
Scenario III
Baseline Emissions*
0.0152 5,470 7,540 42,000 49,300 127 3,700,000
0 0 5,220 0
0.0107 0 5,220 18,300
0.0115 0 5,960 19,300
0.0139 0 5,220 35,400
0.0146 0 5,960 36,400
0 88.5
36,700 108
39,500 124
44,400 108
47,200 124
-154,000
-383,000
-401,000
-314,000
-332,000
Emissions Reductions
Regulatory
Regulatory
Regulatory
Regulatory
Regulatory
Alternative
Alternative
Alternative
Alternative
Alternative
I
HA
HB
m
IV
0
0.00781
0.00841
0.00916
0.00974
0
0
0
0
0
5,120
3,440
4,040
2,980
3,550
0
13,700
14,500
22,400
23,100
0
25,700
27,900
27,700
29,900
86.3
77.9
90.3
68.6
81.6
-138,000
-308,000
-322,000
-232,000
-246,000
a Scenario III baseline emissions vary with each regulatory alternative because the level of total
waste flows handled by MWC varies with each regulatory alternative under Scenario HI.
b Includes bottom ash and fly ash with some residual quench water. Negative values reflect
increases in ash emissions relative to the baseline.
KEY: polychlorinated dibenzo-p-dioxins and dibenzofurans (CDD/CDF), carbon monoxide (CO),
paniculate matter (PM), sulfur dioxide (SO2), hydrogen chloride (HCI), and lead (Pb).
1.3 DISTRIBUTION OF ECONOMIC IMPACTS
Because of the localized nature of solid waste management markets and institutions, it is
difficult to generalize about the economic sectors that would be affected by the NSPSs and about
the magnitudes of those impacts. This analysis, therefore, examines in some detail the economic
impacts on three economic sectors: affected MWC plants, households served by MWC plants,
and government units that own and operate MWC plants.
1-6
-------
1.3.1 MWC Plants and Technologies
The regulatory alternatives considered by EPA distinguish between large and small
plants, varying the emission requirements and associated controls for the different size plants
within the same regulatory alternative. At the same time, differences in design capacity,
technology, etc. contribute to variations between MWC plants in the cost of controlling at the
same level. These differences in cost can be seen in the data of Table 1-4 for Scenario I,
Regulatory Alternative IV. This table presents estimated control costs for 12 model plants used
in the analysis. It shows that the absolute magnitude of capital and operating costs, as well as
their relative magnitudes, vary considerably by the size and type of plant.
TABLE 1-4. ENTERPRISE COSTS OF CONTROL FOR PUBLICLY OWNED NSPSs
MODEL PLANTS (1987$)«: SCENARIO I, REGULATORY
ALTERNATIVE IV
Model
Plant
#
1
2
3
4
5
6
7
8
9
10
11
12
Model Plant
Description1*
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
Model
Plant
Capacity
(Mg/day)
180
730
2,040
450
950
1,810
1,810
220
45
90
820
820
PVof
Capital
Costc
($103)
1,710
13,800
28,500
13,700
17,800
30,600
30,600
1,960
2,180
1,960
13,300
13,300
Total
Annualized
Costd
($103/yr)
639
2,250
4,910
2,100
2,940
5,400
3,240
573
448
460
2,110
2,110
Cost
of Control
per Mg*
($/Mg)
16.90
10.00
7.78
15.00
9.97
9.86
11.90
8.83
47.40
17.00
8.54
8.54
a Control costs are costs over the baseline model plant costs of Chapter 3. These costs are
incurred to meet the emission requirements of the NSPSs.
b Definition of terms used to describe model plants contained in Table 3-2.
c Present value of capital costs are based on 2 consecutive, 15-year life cycles for APCD
equipment over the 30-year plant life.
d Total annualized costs based on a 30-year facility life, 15-year APCD equipment life, and a real
discount rate of 4 percent.
e Computed by dividing total annualized cost by the estimated amount of MSW processed per
year at the model plant.
1-7
-------
The unit enterprise cost data of the table show these costs after adjustment for the size of
the MWC. They range from $7.78, for a large mass burn waterwall plant, to over $47 per Mg for
a small modular plant. In general, for plants with non-zero control costs, the smaller the MWC
plant, the greater the economic impact, especially under Regulatory Alternatives IIB and IV in
which smaller plants must meet more stringent emission requirements than under the other
regulatory alternatives. Since modular plants are generally the most cost-competitive small
plants, modular technology will be most affected by the NSPSs.
Table 1-5 provides a broader picture of plant impacts across regulatory alternatives.
These data show that a wide range of unit enterprise costs apply to model plants. While
recognizing that the tipping fee is as much an administrative convention as it is a measure of
cost, data values at the high end of the unit enterprise cost range are from 35 to 110 percent of the
average tipping fee reported in 1988. Disregarding zero values, the low end values range from 9
to 21 percent of the average 1988 tipping fee.
TABLE 1-5. ENTERPRISE COST OF CONTROL PER Mg FOR PUBLICLY OWNED
NSPSs MODEL PLANTS UNDER SCENARIO I (1987 $)a
MWC Plant Capacity
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
a Cost per Mg based on cash flow
Small—
<225 Mg/day
($/Mg)
0.00-12.70
0.00-12.70
8.83-47.40
0.00-12.70
8.83-47.40
analyses of publicly owned
Large —
>225 Mg/day
($/Mg)
0.00-0.75
3.88-9.38
3.88-9.38
7.78-15.00
7.78-15.00
plants described in Chapter 3.
1.3.2 Households
By matching Census of Governments and Census of Population data for particular MWC
sites with model plant cost data, we obtained estimates of the economic impact of the regulatory
alternatives on households in the MWC service areas that would likely have new plants. In most
cases, the cost per household is estimated to be under $30 per year and only in a few cases is it
estimated to be as high as $90 per year. The distribution of these costs by regulatory alternative
is shown in Figure 1-1.
1-8
-------
Number of 25 n
Service
Areas b-c
20 -
15 -
10 -
$0-$10
$10-$20
$20-$30
$30-$40
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative MB
| | Regulatory Alternative III
BHj Regulatory Alternative IV
$80-$90
$220
Average Cost per
Household per Year a
Costs refer to control costs only; no baseline costs are included.
'Service areas w'rth less than 2,500 total population were not included in the sample because census data for these service areas
were not available.
' Service areas with implicit capacity utilization less than 40 percent for modular pla-tb ,es !han 60 percent for other -echnoSoaies, or
greater than 400 percent for ail technologies were not included in the sample. See text in Chapter 8 for discussion.
Household impacts were defined as "severe" if average cost exceeds $220 per household per year.
Figure 1-1. Distr:. lion of Household Impacts Under NSPSs by
Number of Servr.* Areas and Regulatory Alternative: Index 1
-------
While there is a great deal of variation in a household's solid waste disposal collection
and disposal budget, it probably ranges from $100 to $200 per year. Many of the costs estimated
in this analysis would represent a significant (greater than ten percent) increase in that budget if
passed on to households in their entirety. Even so, because the household budget for solid waste
collection and disposal is so low, these costs do not exceed the threshold criteria for "severe"
impacts recently applied to another regulation affecting solid waste disposal (EPA, 1988b).
1.3.3 Government Units
The NSPSs increase the cost of MWC plants by different amounts depending on the
technology, size, and emission controls of a given plant. Public entities that plan to build plants
for which the shift is largest bear a larger share of the regulatory cost than those for which the
shift is smaller. A government unit's economic impact, then, depends on the particular MWC
plant they build, or are served by, respectively, in conjunction with the regulatory alternative
ultimately selected as the basis for the NSPSs. As in the case of households, this government
unit impact analysis matches communities with cost impacts for particular MWC plants.
Because of data limitations, we made assumptions that, on the whole, amplified the impacts
while at the same time limited the number of matches that could be made. We applied several
criteria for measuring the severity of impacts on government units. For all but one of these
criteria, the impacts were not found to be severe. For the remaining criterion, 9 of the 17
communities examined showed severe impacts. We therefore made follow-up phone calls to
these communities and found that the assumptions we used to estimate the population served by
the MWC did indeed amplify the impacts. In most cases, the financial base in our analysis was
underestimated. For example, the combustor plant located in Dayton, Ohio, serves Montgomery
County as well as other nearby counties and municipalities which may dispose of waste at the
combustor plant for a fee. When we corrected the data to better reflect actual site conditions, the
impact measures were no longer severe for these communities.
1-10
-------
CHAPTER 2
DEMAND CONDITIONS
The demand for municipal waste combustion is derived from the demand for services that
collect and dispose of the large volume and variety of wastes we produce each year. Most, but
not all, of the material burned at municipal waste combustors (MWCs) is classified as municipal
solid waste (MSW). About 130 million Mg of MSW was generated in 1986 (Franklin
Associates, Ltd., 1988). This represents an annual average of one-half Mg per capita based on
the middle-series projection of total population in Statistical Abstract of the U.S. (U.S.
Department of Commerce, 1987).
MSW consists of all the major materials used in the modern industrial state. Table 2-1
presents the estimated quantities and shares of these materials. Paper and paperboard products
comprise over 35 percent of the total. Glass, metals and plastics are each about one-fifth to one-
quarter of the paper and paperboard amount. Yard waste (e.g., grass clippings, tree trimmings,
and leaves) represent the second largest portion of MSW—about 20 percent.
2.1 GENERATORS
Generators of MSW demand services that collect and dispose of MSW. These generators
provide most of the demand, often a "derived demand," for MWC services. As shown in
Figure 2-1, the demand for MSW collection and disposal services can be classified into four
broad source categories:
• Residential: Waste from single- and multiple-family homes.
• Commercial: Waste from retail stores, shopping centers, office buildings, restaurants,
hotels, airports, wholesalers, auto garages, and other commercial establishments.
• Industrial: Waste such as corrugated boxes and other packaging, cafeteria waste, and
paper towels from factories or other industrial buildings. This term does not include
waste from industrial processes, whether hazardous or nonhazardous.
• Other: Waste from public works such as street sweepings and tree and brush trimmings
and institutional waste from schools and colleges, hospitals, prisons, and similar public
or quasi-public buildings. Infectious and hazardous waste from these types of facilities
are managed separately from MSW.
Households are the primary direct source of MSW, followed by the commercial sector.
On average, each U.S. household directly generated 0.79 Mg of solid waste in 1986. The
commercial, industrial, and other sectors each directly generate smaller portions of MSW than
households (see Figure 2-1). In particular, the industrial sector manages most of its own solid
2-1
-------
TABLE 2-1. MATERIALS IN THE MUNICIPAL WASTE STREAM, 1986
Materials
106 Mg
TOTAL
128.1
Percent
Paper and Paperboard
Glass
Metals
Plastics
Rubber and Leather
Textiles
Wood
Food Wastes
Yard Wastes
Miscellaneous Wastes
45.6
10.7
11.5
9.4
3.5
2.5
5.3
11.4
25.8
2.5
35.6
8.4
8.9
7.3
2.8
2.0
4.1
8.9
20.1
1.8
100.0
Source: Franklin Associates Ltd., 1988. Characterization of Municipal Solid Waste in the
United States, 1960 to 2000. Final report prepared for U.S. Environmental Protection Agency.
Other (13.4%)
Industrial (4.2%)
Commercial (27.9%)
Residential (54.5%)
Figure 2-1. Sources of Municipal Solid Waste, 1986
Source: U.S. Environmental Protection Agency. 1988b. National Survey of Solid Waste
(Municipal) Landfill Facilities. Final Report. Prepared by Westat, Inc. EPA/68-01-7359.
Table 7-3.
2-2
-------
residuals, whether MSW or industrial process wastes, by either recycling, reuse, or self disposal.
Thus, direct generation of MSW by industry accounts for only a small share of the MSW flows,
although some industrial process wastes do end up being collected and disposed of along with
MSW.
As shown by Bingham et al. (1976), through derived demand relationships, households
and other components of "final demand" (business spending on plants and equipment,
government spending, and exports) indirectly affect the amounts and composition of residuals
discarded to the environment, including solid wastes directly generated by other sectors. For
example, when food items are shipped to the supermarket in cardboard boxes, the boxes are
unpacked and items shelved at the store. When the shipping containers are discarded to the
MSW system, the household has indirectly contributed to the amount and composition of MSW.
Although the waste is attributed to the commercial sector, the store directly generated MSW as a
results of the household's demand for food.
Little empirical evidence is available about the factors that affect waste generation rates.
However, without substantial changes in market conditions or policies that promote more
recycling and the use of less residual-intensive production, packaging, and consumption
methods, increases in economic activity and in the population indicate that MSW will increase in
the future. Franklin Associates (1988) estimates that MSW will increase at an annual rate of
approximately 1.5 percent over the 1984-2000 period. This growth rate is slightly more than the
population growth rate, indicating an increase in expected per-capita waste generation. A recent
Frost and Sullivan report (Coal and Synfuels Technology, July 25, 1988) estimates that future
MSW generation will be proportional to population growth.
2.2 GENERATOR BEHAVIOR
The responsiveness of the quantity of MSW generated by each generator is important
because regulatory actions may change the conditions under which households and firms make
MSW generation and collection choices. Little empirical information is available regarding these
choices. However, some conjectures are advanced below. In each case a demand relationship is
hypothesized and used to organize the subsequent discussions.
2.2.1 Household Demand
Final consumption purchases (e.g., food items) and household production activities (e.g.,
yard care) result in the generation of MSW. Since these wastes do not provide the household
with utility they are an economic "bad" whose collection and removal is a service of value to the
2-3
-------
household. A household may be viewed as having a demand for solid waste collection and
disposal services, Qc, just as it has a demand for food and other consumer goods:
where
Qc = /l(Y,Pc,S,C) (2.1)
Y = household income,
Pc= price of waste collection and disposal services,
S = service conditions (e.g., frequency of collection and site of collection, degree of
waste separation required, materials accepted), and
C = cost of self-management (e.g., recycling, incinerating, burying, littering).
Household income changes affect the household's demand for MSW collection and
disposal services. Increases in the household's income increase consumption spending; however,
because of savings, the relationship is not one-for-one. These consumption increases include
increases for commodities that generate solid wastes. Solid waste collection and disposal is
likely a normal good—as income increases, all other arguments in the demand function held
constant, the demand for solid waste collection and disposal services increases (see Figure 2-2).
Wertz (1976) has argued that the income elasticity of demand for collection and disposal
services, (3Qc/Qc)/(dY/Y), is likely to be positive, but small. Goddard (1975), while noting
$/Q,
D' (Y = Y)
Q,
Figure 2-2. Effect of Income Changes on Household Demand for MSW
Collection and Disposal Services
2-4
-------
serious data and methodology problems in a study of demand for waste collection in Chicago,
reports an income elasticity of demand estimate of 0.4.
In most communities today, MSW collection and disposal services are financed by
general tax revenues. If increased costs for these sendees result in increased tax rates, disposable
household income will be reduced. Given a positive income elasticity of demand for waste
collection and disposal services, this would, in turn, reduce MSW generation. Because both the
income elasticity and the cost of MSW collection and disposal as a share of all taxes are small,
however, this effect is unlikely to be significant.
The relationship between quantity demanded and price is an inverse one—increases in the
price for MSW collection reduce the quantity demanded of these services. This inverse
relationship has been empirically demonstrated for a large variety of commodities; MSW
collection and disposal services should not be an exception to these findings. However, it is
difficult to demonstrate this relationship for MSW collection and disposal services and estimate
the numerical relationship because of
• the variety of MSW collection service arrangements,
• the absence of MSW collection pricing on a per-unit-of-service basis, and
• the lack of adequate micro data on household waste generation rates.
As noted above, in most communities today there is no price mechanism through which
changes in the cost of MSW collection and disposal services provide incentives to households to
adjust their use of collection and disposal services. When households are not charged, the price
of collection and disposal services is zero and the quantity demanded is Qj in Figure 2-3. In
some communities households are charged a flat fee per week or month for a specified service
(e.g., solid waste collected from four containers twice weekly). At best, this provides a weak link
between the fee (or price of service) and the amount of MSW generated since the fee does not
vary with the amount of waste generated by any given household.
In a few communities, such as Seattle, households are charged on a per-container basis.
In such instances the linkage between price and the quantity of waste generated is strengthened.
Increasing in the price per container certainly encourages households to find ways to reduce the
number of containers used and likely has an effect on the amount of waste generated. As the
containers become small relative to the amount of waste generated, the household demand begins
to resemble that depicted in Figure 2-3. For a price change from P£ to P-?, the quantity of MSW
generated declines from Q? to Q?.
2-5
-------
Q
Q:
Q
Figure 2-3. Effect of Collection and Disposal Price Changes on the
Household Demand for MSW Collection and Disposal Services
In communities where price provides an incentive to adjust the amount of MSW
generated, the household price elasticity of demand for MSW collection and disposal services
(3Qc/Qc)/(dPc/Pc) is negative but the magnitude is likely to be small. Goddard (1975) reports on
a 1972 cross-section study of California communities that charged their citizens different flat
rates for MSW collection. While again noting data and methodological difficulties in the study,
Goddard reports that the researchers estimated statistically significant coefficients that are akin to
price elasticities for two forms of the demand equation. The point estimates of these values were
-0.7 and -0.5 with 95 percent confidence limits of -0.5 to -1.0 and -0.3 to -0.8, respectively.
Part of the household's costs of MSW collection and disposal services is the household's
implicit cost of storing the waste before it is collected, sorting materials as required by the
collector, and moving the wastes to the place of collection (e.g., front yard, back yard). Wertz
(1976) cites evidence that the frequency of service elasticity (3Qc/Qc)/(3S/S) is likely to be
positive and high: as collection frequency increases, or collection site convenience is improved,
collection demand increases.
Increasing the inconvenience of disposing of certain wastes (e.g., aluminum containers)
by requiring that they be sorted is likely to reduce generation of those wastes. Households may
substitute products that produce waste that needs to be sorted with ones that do not; the net effect
2-6
-------
on total waste generation is uncertain. We have not found any empirically based estimates of the
effect of different sorting requirements or opportunities on total waste generation.
Households may self-manage the solid wastes they do generate through recycling,
burying, incinerating, compacting, discarding to the sewer system or to others' collection
systems, or by littering. Further, they may transport their wastes directly to the disposal site.
These activities have costs to the household, either explicit or implicit, and thus by themselves
reduce the household's welfare. However, these activities also offset the costs of MSW
collection. Increases in the costs for any self-management option are expected to lead to greater
use of other such options or to greater use of the MSW collection and disposal services provided
by other parlies.
2.2.2 Firm Demand
The firm's derived demand, Qc, for MSW collection and disposal services is
Qc = /2(Pc,Px,S,C) (2.2)
where
Px = price of the firm's output and, other terms are as defined above.
The price elasticity of demand, (3Qc/Qc) / (dPc/Pc). or simply ec, for collection and
disposal services can be shown to equal
ec = vc (hx + sc)
where
vc = ratio of collection and disposal cost to all costs of production,
hx = demand elasticity for the output x,
sc - Allen elasticity of substitution between waste collection and disposal services and
all other inputs in the production of output x.1
Finns' payments for municipal waste collection and disposal Cervices, vc, likely comprise
a small share of production costs. Product demand elastic ir.es, hx, vary from commodity to
commodity but typically range from -0.5 to -5.0. The elasticity of substitution between waste
1 The Allen elasticity of substitution is the ratio of the output-constant cross elasticity of demand for the factors of
production (waste colleclion and other inputs to production) and the vc of waste collection.
2-7
-------
collection and all other inputs is positive, but there are no estimates of this value. However, even
if firms have substantial opportunities to make solid waste reducing process changes, the
associated elasticity of substitution is weighted by vc, a value much less than one. Thus, in many
cases, we expect the product vcsc to be fairly low, This implies an inelastic demand by firms for
MSW collection and disposal services; that is, a price nasticity of demand less than 1 in absolute
value.
In summary, because of current MSW collection and disposal service financing methods
and demand relationships it seems unlikely that NSPSs affecting the costs of MWC will
significantly influence the overall demand for municipal waste collection and disposal services,
the composition of waste, or the expected increases in MSW generation over time. Even in cases
where increased costs are passed on to waste generators in the form of higher prices for waste
collection and disposal services, price inelasticity will moderate the impact of pi ice changes on
the quantities of waste generated. Most of the important effects are expected in the demand for
disposal services per se, particularly the demand for MWC disposal services, and in the supply
side of the market for these disposal services.
2.3 WASTE DISPOSAL SERVICES DEMAND
It is sometimes helpful to think of waste collection as distinct from waste disposal. Given
such a distinction, waste collectors have a demand for waste disposal services in addition to their
demand for labor, equipment, and other inputs used in the production of the service they provide.
The change in this derived demand in response to changes in the cost of disposal can be analyzed
through use of the general elasticity of demand expression introduced above. This expression,
adapted to examine the determinants of the elasticity of demand for waste disposal, e
-------
share of disposal to now be in the neighborhood of 50 percent (Morris, 1987). Consequently, the
share of collection costs represented by disposal services is probably large. The elasticity of
demand for collection services, as argued above, is likely to be inelastic (small) and, depending
on institutional conditions, virtually zero in some cases. Also, since every unit of collected
wastes must be disposed of, it is difficult to credit the notion of a large elasticity of substitution
between disposal services and other inputs to production of waste collection and disposal. Thus,
while the v^ coefficient in the elasticity expression may be large, the two terms inside the
parentheses are small, suggesting a moderate to small elasticity of demand for waste disposal
services. Anything short of a major increase in the price of disposal services is unlikely to result
in much of a change in the demand for those services.
Not all waste collected is disposed of in the same fashion. While households and firms
demand collection and disposal services, the organizations that collect the waste may be able to
choose among a variety of locations and technologies for disposing of MSW. These collectors
may be part of an integrated system of MSW management services, private collectors, or
generators that self-collect. They may be subject to a very different set of legal, institutional, and
market conditions. Even so, we can further decompose the analysis of waste disposal services
demand and further adapt the general derived demand relationship introduced above to examine
the determinants of the elasticity of demand for MWC disposal services, emwc, in particular.
= vmwc (hd + smwc) (2.5)
where
Vmwc = the ratio of MWC disposal services to the total cost of disposal (the cost share),
hd = demand elasticity for disposal services,
Smwc = Allen elasticity of substitution between MWC disposal services and all other
inputs to waste disposal.
For this equation for demand elasticity, the cost share will likely vary with the local
MSW management area in question and, as argued above, the elasticity of demand for waste
disposal services is likely to be moderate to low. The Allen elasticity of substitution, smwc, may
well be quite high if landfills, recycling centers, etc., provide essentially the same disposal
services at a similar price. The total effect of these influences on the demand for MWC disposal
services is uncertain, but conditions may well exist where demand is elastic and the increased
cost imposed by the NSPSs could change the mix 01" ', i SW disposal technologies, if not total
waste generation or disposal.
2-9
-------
CHAPTER 3
SUPPLY CONDITIONS
Solid waste management today often consists of a set of interrelated activities for the
collection and transport, processing, and disposal of solid wastes. The material flows associated
with most of these activities are illustrated in Figure 3-1. In light of both the preceding demand
analysis and the regulation in question, this discussion of supply conditions focuses on the supply
of municipal waste combustion and landfilling disposal services. Also examined, but to a lesser
extent, are waste recycling and transportation services.1
Independent
Dispos
Centralized
Recycling
Marketable
Materials
Disposing
Combustion
Processing
Nonhazardous
Ash
1
Hazardous
Ash
Sanitary
Landfilling
Hazardous
Waste
Landfilling
Figure 3-1. Solid Waste Management Options
1 Not included in Figure 3-1 are strategies relating to the design, manufacturing, packaging, and use of products so
as to reduce the quantity and toxicity of solid waste, especially MSW. These "source reduction" activities are
technically related to the substitution options that households and firms would consider if there were an effective
price mechanism that provided an incentive for reductions in solid waste generation. While source reduction is
part of EPA's national strategy for solid waste (EPA, 1989d), it probably will not be pursued under authority of
the Clean Air Act. Consequently, we don't address source reduction in this report. To the extent that source
reduction may alter the cost-effectiveness of the air pollution control devices (APCDs) considered in this report,
this exclusion is a shortcoming.
3-1
-------
Landfilling is the predominant method of solid waste disposal in the U.S. today. In 1986
approximately 80 percent of all MSW was directly landfilled, 10 percent recycled, and 10
percent combusted. The percentage of recycled or combusted discards has been growing since
1960, however, and these technologies are expected to have an even greater impact on MSW
disposal in the future (Franklin Associates, Ltd., 1988).
3.1 PRODUCTION PROCESSES
Understanding the production processes involved in solid waste management is the first
step toward understanding its economics. First we examine the production functions. For each
process involved in the management of MSW, production functions describing input and output
relationships can be written in the form
Qi,...,Qm = /(X1,...^cn), (3.1)
where Qi,...,Qm represent outputs l...m produced and Xi,...,Xn represent inputs l...n consumed
during the production period. Recognizing multiple outputs is particularly important in this
analysis because many combustors both dispose of waste (into the air with a 20 to 30 percent
solid waste residual) and produce energy.
In a simple setting this relationship reduces to one output and two inputs:
(3.2)
These production relationships may be examined from two perspectives — the operating
decision and the investment decision. For simplification the analysis assumes that existing firms
concern themselves with the operating decision in the short run, whereas new firms are faced
with the investment decision along with the subsequent operating decision. Existing firms decide
on the operating rate within the constraint of the fixed resource (Figure 3-2a). Owners of the
new firm make decisions regarding scale and resource substitution without the constraint of a
fixed resource (Figure 3-2b). This distinction, too, is important in this analysis since combustor
managers will chose between the best means of operating the existing plant (an operating
decision) and the best new combustor investment (an investment decision).
Figure 3-2a illustrates the choices involved in operating decisions when one of the inputs,
\2, is fixed. The firm must choose the optimum operating rate, Qn, and input rate, Xj. The
specific relationships between inputs and outputs is an empirical issue. Figure 3-2a illustrates a
situation where output increases rapidly as additional Xj is added over some range of input.
3-2
-------
Q
Q=/(Xl Mother)
Figure 3-2a.
Operating Decisions: Choices Involved
with at Least One Fixed Input
Q = /(X1,X2)
Figure 3-2b.
Investment Decisions: Choices
Involved with No Fixed Inputs
However, the ability of additional Xj to generate additional output deteriorates as more X j is
used. Finally, output reaches a maximum level. Beyond this point, additional Xj decreases
output.
Figure 3-2b illustrates the choices involved when both inputs are variable, as is the case
with the investment decision. The output rate Q shown is constant. The curve shows the
alternative combinations of inputs Xl and X2 that produce that rate of output.
We now use the concept of the production function to examine the major components of
MSW management associated with air emission NSPSs for MWCs: combustion, landfilling,
collection and transportation, and recycling.
3.1.1 Combustion
Municipal waste combustion (MWC) is the process of reducing the volume of MSW
through incineration. Because MWC reduces waste volume by as much as 70 to 90 percent, this
method of waste management has the potential to significantly reduce the need for landfills.
Industry Conditions
Combustion was once the principal way of disposing of MSW, especially in the
metropolitan areas of the U.S. These plants were dirty and smelly, however, and virtually all
were closed in the two decades following World War II. A renewed interest in the technology,
coincident with reductions in available, convenient landfill capacity and the search for alternative
energy sources, occurred in the mid-1970s. In the early and mid-1980s, new MWC capacity was
3-3
-------
added rapidly, as shown in Table 3-1. The amount of waste combusted increased by a factor of
three or four from 1980 to 1986.
TABLE 3-1. ESTIMATED FLOWS OF MUNICIPAL SOLID WASTE TO MUNICIPAL
WASTE COMBUSTION PLANTS, 1980 THROUGH 1986
Franklin
Year Associates Radianb
MWC Waste Flows8 MWC Waste Flows
(106 Mg/yr) (106 Mg/yr)
1980
1981
1982
1983
1984
1985
1986
2.45
2.09
3.17
4.54
5.90
6.89
8.71
2.47
3.45
4.44
6.02
7.60
9.60
10.20
a Franklin Associates estimates MWC energy recovery waste flows only (Franklin Assoc., 1988,
p. 18).
b In a profile of existing facilities, Radian (1988a) reports estimates for non-heat recovery
capacity, heat recovery capacity, and capacity for plants that co-fire MSW with other
materials (e.g. wood, tires, sewage sludge). The values reported above include heat recovery
capacity for plants processing at least but not less than 50 percent MSW. Average capacity
utilization values reported in the 1988-89 Resource Recovery Yearbook (Gould, 1988) were
applied to MWC capacity estimates to calculate waste flows.
Radian Corporation's report, Municipal Waste Combustion Industry Profile-Facilities
Subject to Section lll(d) Guidelines (Radian, 1988a), lists MWC plants that were operating, .
under construction, or planned in the mid-1980s. The report identifies 281 plants that might be
affected by Guidelines: 161 plants in actual operation at the report's printing date and another
120 that were projected to begin construction before 1990. While quite a number of new plants
have begun operating in the past few years, construction plans for many of the projected plants
have been deferred or cancelled due to local opposition and/or revisions in community waste
management plans. Consequently, the estimated number of plants subject to the Guidelines has
been revised downward to reflect these changing market conditions. We estimate that
approximately 200 plants (39 projected plus the 161 currently operating plants) will be affected
by the Guidelines.
The baseline conditions for MWC plants subject to the NSPSs were derived in part from
information compiled by Radian and presented in the NSPSs cost report (EPA, 1989c). Plants
3-4
-------
identified in this report are those expected to begin construction within the 5-year period
following the initial publication of the proposed NSPSs in the Federal Register, planned for late
1989. For the purpose of this analysis, MWC plants subject to NSPSs are called "NSPSs plants."
The Radian NSPSs cost report projected 138 such NSPSs plants.
For the NSPSs cost report Radian estimates the design capacities and technologies of
these NSPSs plants based on a list of plants it compiled from information on plants in the early
planning stages. To make these estimates, Radian sometimes used information about plants in
advanced planning or early construction stages to assign capacity and technology distribution to
plants in early planning stages. Radian projects that approximately 41.3 million Mg per year of
capacity will be distributed among 138 plants beginning construction between November 1989
and the end of 1994. This capacity was distributed across MWC technologies as follows: 64
percent mass burn, 27 percent RDF, 7 percent FBC, and 3 percent modular (see Figure 3-3). In
constructing the baseline projections for this report, we used Radian's estimates of the
distribution of facilities by technology but modified the projections to cover roughly 67 NSPSs
plants with a total capacity of 19.3 million Mg per year. This process is described in detail in
Chapter 5.
FBC (7%)
Modular (3%)
RDF (27%)
Mass Burn (64%)
Figure 3-3. Percentages of total capacity of mass burn, RDF, FBC, and modular
technologies that will be affected by the NSPSs (Radian, 1988a).
3-5
-------
In 1987 the market for MWC construction services sent mixed signals to suppliers of
those services. According to a Kidder Peabody report (McCoy, 1988), more capacity was
cancelled than was ordered in 1987, with a resulting total scheduled decline in capacity of 10
percent. Public opposition to siting and construction of new plants, combined with the
uncertainty regarding proposed legislation, are thought to be the major obstacles facing the
vendors of MWC systems. These vendors include project developers, manufacturers, and
engineering construction firms. According to the report, 28 companies participate in the
combustion industry with no one firm being dominant. Of the 28 firms, Ogden Martin is the
1987 industry leader in terms of capacity, claiming a 20 percent market share. Wheelabrator is in
second place with 18 percent, ahead of American Ref-Fuel with 8 percent, Combustion
Engineering with 7 percent, and Westinghouse Electric with 5 percent Generally, the capital
services vendors are not the MWC plant owners. Plant ownership has typically been the domain
of state or local governments who find it important to retain control over municipal waste
disposal services.
Technologies
MWC plants range widely in design capacity from less than 25 to more than 2,000 Mg
per day of MSW throughput. As the name suggests, mass burn combustion requires no
processing aside from the removal of oversized items and some mixing to produce a more
homogeneous fuel (EPA, 1987). Refuse is moved through the plant through the use of grates
and/or rams. A traveling grate may carry the MSW through the combustor without agitation, or a
rocking (reciprocating) grate may be used to agitate the waste as it moves through the combustor.
The rotary design uses a different process to achieve agitation through rotation of the waste.
Agitation allows more waste surface to be exposed, increasing efficiency (Robinson, 1986).
Because mass burn plants are built on site, variation in design and capacity are
characteristic of this type of MWC. Two typical mass burn design technologies are waterwall
and refractory designs. Virtually all waterwall furnaces incorporate energy recovery, but the
same is not true for refractory furnaces. The refractory design is an older, less efficient
technology and, for this reason, most new mass burn plants are expected to have waterwall
boilers.
Modular combustors, like mass burn combustors, require minimal processing of waste.
Modular plants consist of one or more prefabricated combustor units and range in capacity from
approximately 25 to 500 Mg of MSW throughput per day using either grates or rams to move
waste through the combustor. Modular combustors are constructed as "starved air " or "excess
air" designs. Both types use similar design components but differ in the amount of oxygen
3-6
-------
present in the combustion chamber. Excess air combustors incinerate waste with no limits on
the amount of oxygen present. Starved air combustors control the amount of oxygen to achieve
pyrolysis of MSW.
The modular combustor has primary and secondary combustion chambers. Partial
combustion of MSW in the primary chamber is followed by more complete combustion in the
secondary chamber assisted by an auxiliary burner and additional air (Robinson, 1986).
The third major category of MWC uses sorted and processed municipal waste referred to
as refuse-derived fuel (RDF). The sorting and separating of waste materials is typically
accomplished by a system of shredders, magnets, screens, air classifiers, and conveyers, which
produce a fuel (waste) that yields a higher heat value, lower ash volume, and more complete
combustion than nonprocessed waste. Processing may vary from shredding of refuse to fine
separation of waste to produce a fuel suitable for cofiring with a fossil fuel.
The fourth and final category of MWC technology dealt with in this report is Fluidized
Bed Combustion (FBC). Several refuse entry points are necessary with FBC to provide even fuel
distribution. Combustion chambers use either a waterwall or refractory design for temperature
control and energy recovery. Fluidized bed technology allows firing a wide variety of fuels with
relative insensitivity to the amount and type of ash in the fuel. By making the waste behave as a
liquid or gas, FBC combustion units burn MSW more efficiently than mass burn or RDF units.
At present, however, fluidized bed technology is relatively new and still undergoing
development.
MWC has two principal products, MSW volume reduction and energy generation, along
with the residual products of ash and emissions to the ambient air. The production function for
MWC can be expressed by the equation
QMWC'Qe,Qa'Qenv = /2(capital services, operating services,QMSW) (3.3)
where QMWC *s the volume reduction in MSW, Qg is the quantity of energy produced, Qj, is the
quantity of ash residue generated, and Qenv is a measure of the environmental impacts, including
air emissions, resulting from incineration. The inputs are capital services (e.g., combustor unit,
land, building, air pollution control devices), operating services (e.g., labor services, maintenance
services, fuel for cofiring, utility services), and QMSW (raw MSW for fuel).
Representative plants, called "model plants," are used extensively in this economic
impact analysis to represent NSPSs plants nationwide. Radian specifies the technical features
3-7
-------
of 12 representative model plants in the NSPSs cost report (EPA, 1989c) and assigns the plants
identified in the cost report to 1 of these 12 technologies. Table 3-2 provides a list of the
model plants and their characteristics including a description of the energy recovery
capabilities.
Table 3-3 shows the relationship between inputs and outputs for the model plants used in
this analysis. MSW input corresponds to QMSW m Equation (3.3) and represents design capacity
for each model plant. Waste reduction and energy recovery are the principal products of MWC,
referred to as output. Estimates for MSW reduction were calculated by subtracting the projected
annual volume of residua! ash from the projected MSW input per year. Energy recovery figures
are based on energy revenue projections divided by a S/IO6 Btu energy value factor derived by
Radian (1988c). 106 Btu's were then converted to TJ using a conversion factor. The volume of
ash and total emission estimates for CDD/CDF, CO, PM, SC>2, HC1, and Pb are included as
residua! products of MWC.
3.1.2 LandfilJing
Two types of landfills are used to manage MSW—sanitary landfills and hazardous waste
landfills. Sanitary landfills receive only nonhazardous waste (primarily from household and
commercial sources) with the exception of small quantity generator hazardous waste. Sanitary
landfilling is defined as a method of waste disposal through a process that includes (Robinson,
1986):
(1) spreading the collected waste into thin layers in the landfill,
(2) compacting the waste into the smallest practical volume, and
(3) covering the waste with soil on a daily basis.
The potential environmental impacts of landfilling (e.g., possible groundwater
contamination, air emissions, odor, traffic, dust, and danger of explosion) are becoming widely
known. Likewise, health and safety regulations surrounding landfill design, siting, and operation
are also becoming increasingly stringent, making landfilling more expensive (Glebs, 1988). As a
result of these factors, along with increasing land scarcity, many landfills have closed and
communities are facing increasing difficulty developing new landfill sites.
Since 1980 the number of landfills opening each year has continually declined (Temple,
Barker, & Sloan, Inc., et al., 1987). Approximately three-quarters of all municipal solid waste
landfills currently in operation are expected to close within the next 15 years. Municipalities,
especially those in New Jersey, New York, Connecticut, Florida, and California, have been
forced to look toward other MSW management options such as recycling and combustion that
reduce the quantity of waste to be landfilled.
3-8
-------
TABLE 3-2. CHARACTERISTICS OF NSPSs MODEL PLANTS
Model
Plant*
1
2
3
4
5
6
7
8
9
10
11
12
Abbreviated
Term
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/CB
FBC/BB
Definition
of Term
Mass Burn/Waterwall (small)
Mass Burn/Waterwall (mid-size)
Mass Burn/Waterwall (large)
Mass Burn/Refractory Wall
Mass Burn/Rotary Combustor
Refuse Derived Fuel
Refuse Derived Fuel/Co-fired
Modular/Excess Air
Modular/Starved Air (small)
Modular/Starved Air (mid-size)
Fluidized Bed Combustion (Circulating Bed)
Fluidized Bed Combustion (Bubbling Bed)
Model Plant
Capacity
(Mg/day)
180
730
2,040
450
950
1,810
1,810
220
45
90
820
820
Model Plant
Size
Category3
S
L
L
S
L
L
L
S
S
S
L
L
Energy
Recovery
steam
electric
electric
electric
electric
electric
electric
electric
none
electric
electric
electric
a Model Plants with design capacity less than or equal to 225 Mg/day are classified as small and plants with capacity greater than 225
Mg/day are classified as large. Specified control technologies arc assigned to model plants under various regulatory alternatives
according to this size classification.
-------
TABLE 3-3. PRODUCTION CHARACTERISTICS OF NSPSs MODEL PLANTS3
MSW Input
Model Combustor Annual
Plant Type Hours
# Operation1* (Mg/day)c
Mass Burn
1 MB/WW (small)
4 MB/REF
2 MB/WW (mid-size)
5 MB/RC
3 MB/WW (large)
Refuse-Derived Fuel
1 RDF/CF
6 RDF
Modular
9 MOD/SA (small)
10 MOD/SA (mid-size)
8 MOD/EA
FluidizedBed Combustion
11 FBC/BB
12 FBC/CB
7,420
7,420
7,420
7,420
3,620d
7,250
7J60
7,160
7,270
7,270
180
450
730
950
2,040
1,810
1,810
45
90
220
820
820
(Mg/yr)>>
37,800
140,000
224,000
294,000
631,000
274,000
548,000
9,450
27,000
64,900
247,000
247,000
Outputs
MSW
Reduction6
(Mg/yr)
27,700
95,500
153,000
200,000
430,000
196,000
467,000
6,610
18,900
44,200
229,000
229,000
Energy
Recoveryf
(TJ/yr)
0.404
1.18
2.41
3.25
6.75
7.83
7.06
0
0.290
0.695
3.19
3.19
Residuals
Selected Air
AshS Emissions'1
(Mg/yr) (Mg/yr)
12,100
44,700
71,600
94,000
201,000
77,900
80,900
2,840
8,120
20,700
18,700
18,700
255
980
1,450
1,970
4,060
2,860
5,060
66
183
457
1,590
1,630
a Values are rounded to three significant digits. Differences across columns due to rounding. Model plants represent average characteristics for projected MWC
facilities.
b Calculated using average capacity utilization reported in the 1988-89 Resource Recovery Yearbook (Gould, 1988). Allowance is made for increased downtime
for model plants 1 and 9; and allowance is also made for model plant 7 which co-fires 50 percent wood.
c Based on a 24-hour operating day.
d Reflects special conditions resulting from increased estimated downtime for smaller plants and co-firing other materials with MSW (e.g. wood chips, tires,
sludge, etc.).
e Calculated by subtracting ash residual from MSW input (Mg/yr).
f Based on energy revenue credits and $/106 Btu reported in a memorandum from Radian (1988c).
8 Ash values include some residual quench water.
n Represents sum of 6 types of emissions: polychlorinated dibenzo-p-dioxins and dibenzofurans (CDD/CDF), carbon monoxide (CO), paniculate matter (PM), sulfur
, rhlnriHp mm and lead (Pb).
t,,,H,
-------
Landfill sites may be constructed according to various design technologies (Robinson,
1986). The trench method involves excavating soil at a slight angle to facilitate drainage.
Facilities using this type of landfill technology must consider soil depth and groundwater
conditions. MSW is then spread in the trench, compacted, and covered with material taken from
the spoil of excavation with the excess material used for berms or area landfills. The trench
method:
• makes cover material readily available,
• exposes a minimum-size working face,
• gives optimum drainage during filling operations, and
• is easily adapted to wide variation in size of operation.
In the area method of landfilling, waste is spread on the ground with no prior excavation.
After the waste is compacted, it is covered with imported soil. This method is generally used in
locations with land depressions or gently sloping land. The area method:
• accommodates very large operations Oarge working face), and
• is acceptable where no below ground excavation is feasible.
Combinations of the trench and area methods allow for the greatest flexibility in adapting
site construction to the particular needs of a community. The progressive slope or ramp method
is one variation in which a small amount of soil is excavated directly adjacent to the working face
and spread over one day's waste. The depression is then filled with a portion of the next day's
waste, which is covered with soil from another adjacent excavation. Using this method
eliminates the need to import cover material and allows a portion of the discarded waste to be
deposited below the original surface.
The state-of-the-art landfill may include clay or synthetic liners, leachate collection and
monitoring, gas collection and monitoring, surface water controls, and groundwater monitoring.
However, the majority of landfills have little environmental protection equipment in place.
Ninety-five percent have no leachate collection, and 85 percent do not have a liner. Only 25
percent monitor groundwater, and only 12 percent practice surface water monitoring. Landfill
gas is monitored even less, with 5 percent of the sites incorporating methane monitoring (Glebs
1988).
Subtitle D regulations under the Resource Conservation and Recovery Act of 1976
(RCRA) proposed in the August 30, 1988, Federal Register may significantly increase the cost
of developing landfill sites. The proposed rule would require existing landfills to incorporate
closure and post-closure care including groundwater, surface water, and gas monitoring and to
3-11
-------
provide final cover integrity. Additional regulations for new landfills require that sites be
developed with a liner as well as other detection and monitoring equipment necessary to ensure
the integrity of groundwater and surface water within concentration limits set by the EPA.
As the cost of landfilling rises, the industry may be able to offset some of the costs
through energy recovery. Methane gas is formed as solid waste decomposes. The concentration
and quality of methane vary according to the extent of MSW decomposition, quality of the MSW
being landfilled, climatic conditions, and parameters of the landfill. Typically, the gas collection
system consists of vertical wells and horizontal collection headers distributed over the surface of
the landfill site (Jansen, 1986). After the gas is collected it is either upgraded for delivery to
utility companies, used as boiler fuel, or converted into electricity by an internal combustion
engine (Robinson, 1986). Energy recovery through gas collection is generally limited to large,
established landfills since methane formation may take a period of several years and require a
relatively sizable area of land for mining operations.
The production function for landfilling MSW can be expressed by
= /3(capital services, land, operating services,QMSW) (3.4)
where QL is the quantity of MSW landfilled per year, Qg is the quantity of energy produced per
year from the combustion of methane and other combustible gases generated from decaying
waste, and Qenv is a measure of the environmental impacts per year resulting from landfilling.
The inputs are capital services (e.g., bulldozers, scales, buildings, air pollution control devices),
land, operating services (e.g., labor services, maintenance services, utility services), and QMSW
(municipal sob'd waste).
The amount of land needed for sanitary landfills depends on the depth of the site, the
degree of waste compaction, and the desired closing height of the landfill. However, a rule of
thumb is that one hectare is required annually for every 12,000 people, or, based on average
waste generation rates, for every 6,200 Mg of MSW generated.
According to the Solid Waste Landfill Survey conducted by the EPA (1988b), 80 percent
of sanitary landfills are owned by local governments. An additional 5 percent are owned by state
and federal governments with the remaining 15 percent privately owned. A correlation between
type of ownership and size of the landfill has been observed, with publicly owned landfills more
likely to be small and privately owned landfills large. Half of the landfills in operation (the
smallest ones) receive less than 2 percent of the waste while 2.6 percent of landfills (the largest
ones) receive 40 percent of the waste (RTI, 1988b).
3-12
-------
Hazardous waste landfilling is the placement of hazardous waste in or on land, often in
cells that are subsequently covered with clay, asphalt, or concrete. Hazardous waste landfills are
made leak resistant by the use of combination clay and synthetic liners. Leachate detection and
groundwater monitoring systems alert operators to the existence of a leak (Environmental Law
Institute, 1983). The treatment, storage, and disposal of hazardous wastes are regulated under
RCRA Subtitle C, requiring operators of these landfills to keep a careful account of the types of
wastes disposed and ensure the protection of groundwater and surface water within established
environmental performance standards.
Because MWC fly ash has a high heavy metals content, there have been proposals to treat
it as a hazardous waste. The Science Advisory Board, a body of independent scientists, is
currently reviewing an overall approach to the handling, transportation, and disposal of ash from
municipal incineration. Rep. Thomas Luken (D-Ohio) introduced an incinerator ash bill (H.R.
2162) on May 1, 1989, which would mandate EPA regulation of solid waste incineration and
disposal of ash. Under the provisions of this bill, incinerator ash that fails to meet minimum
technical standards would be restricted to a RCRA Subtitle D monofill or codisposal facility
(Hazardous Waste News, May 8,1989).
The production function for hazardous waste landfilling of fly ash can be expressed by
QLA'Qenv = /4(caPital services, land, operating services,QA) (3.5)
where QLA is the quantity of ash landfilled per year, Qenv is a measure of the environmental
impacts per year resulting from landfilling, and the inputs are capital services (e.g., bulldozers,
scales, building, air pollution control devices, leachate containment systems), land, operating
services (e.g., labor services, maintenance services, utility services), and QA (ash).
3.1.3 Collection and Transportation
Collection and transportation of MSW are common components of every MSW
management system. MSW is collected from generators and transported to the treatment plant
or directly to the landfill. Where recycling or combustion are performed, the residue is
subsequently transported to the ultimate disposal site. Transfer stations are widely accepted in
large metropolitan areas as a means of reducing transportation costs, especially when landfill
sites are remote. A transfer station is a waste holding place located between waste collection
points and disposal plants. When a transfer station is used, transportation is also needed from
3-13
-------
the transfer station to the primary management plant. Other advantages of transfer stations
include (Robinson, 1986):
• better haul roads for collection vehicles,
• greater traffic control,
• fewer trucks on the sanitary landfill haul route, and
• improved landfill operating efficiency.
The production function for transporting MSW can be expressed by
QMSW = /i(capital services, operating services) (3.6)
where QMSW i§ me quantity per kilometer of MSW transported per year and the inputs are capital
services (e.g., trucks, buildings, land) and operating services (e.g., labor services, repair services,
fuel, management services).
3.1.4 Recycling
Materials may be recovered from MSW and recycled using one or both of two methods—
generator sorting (sometimes called curbside recycling in a residential context) and centralized
recycling. In the U.S., most materials are first recovered through generator sorting, in which
generators manually separate materials for reuse or recycling from waste for disposal. For
example, in some communities households are required to separate paper, glass, and metals from
their waste for curbside collection. Centralized recycling plants separate salable materials from
mixed waste after collection, often as part of transfer station operation.
Centralized recycling is often practiced in conjunction with combustion of RDF.
Operators of MWC plants that burn RDF have an interest in promoting the recycling of bottles
and cans to reduce the amount of these materials appearing in the fuel (waste). The sorting and
separating required for plants using RDF couples naturally with recycling programs.
Combustion Engineering agreed in 1987 to build a combination materials recovery plant and
combustion plant for a community in New York (Salimando, 1988). Increased quality of waste,
reduced need for landfill services, and reduced amount of ferrous metal in ash residue are some
of the benefits associated with arrangements of this type.
Recycling increases landfill life since processed materials tend to require less capacity
than nonprocessed waste. Enhancing the efficiency of existing landfill sites reduces the need for
new site development. However, institutional problems associated with source separation
include reinvestment problems and labor problems. Single compartment collection vehicles are
not efficient for curbside collection of separated materials. In addition, the retraining of a labor
3-14
-------
force that has traditionally viewed discarded materials as "waste" further compounds the
difficulties associated with recycling. As a result, recycling programs are often not regarded as
viable options for decision makers who are risk-averse.
3.2 PRODUCTION COSTS
The production function of Equation 3.2 describes the relationship between inputs and the
maximum output rate. Correspondingly, at a given output rate, Q, the minimum input
requirements can be identified as
Xi ......... ,Xm = g(Q). (3.7)
The cost of production at Q is the sum of the amount of X employed times its price. For each
output, Q, the minimum cost way of producing that output is defined by the cost function
,Xm;Q). (3.8)
This cost-minimizing set of inputs for a given level of Q can be estimated for each of the
waste management processes identified in Section 3.1. The following sections of this chapter
describe production cost estimates developed for three of the four waste management activities
discussed above — combustion, landfilling, and collection and transportation. Costs for recycling
are not developed because this analysis treats changes in the waste stream associated with
recycling or source reduction as exogenous. That is, reductions in the volume or composition of
waste to be transported and disposed by landfills and combustors due to generator sorting and
recycling can be varied but are taken as given in the analysis.
Given input prices, the C variable can be expressed as three cash flows: capital costs,
operating costs, and closure costs or salvage value. In general, the net present cost of producing
a given output of collection and transportation, landfill, or combustion service can be expressed
by the cash flow expression
( Net "i ( A } ( B W c "i
Present = Present value of + Present value of - Present Value of
i^ Cost J ^ Capital Costs J ^Operating Costs J (.Salvage RecoveryJ
In the first portion of the equation, (A), capital costs may be a stream of lumpy
investments or a one-time investment. Operating costs (B) for each period are calculated net of
energy recovery revenue, and salvage value (C) of plant and equipment, if any, is also accounted
for. Actual calculations involve some differences for private and public ownership due to
different cost-of-capital figures and tax obligations for those forms of ownership.
3-15
-------
The net present cost equations for the two types of ownership are as follows:
Private Financial Cash Flow
NPCp =
t=i
C
[S(l+rp)-T]
B
t=i
(3.9)
Public Financial Cash Flow (municipalities)
NPCm =
t=l
+
u
i: C R l+r -t"
t=l
c
- [Sd+rm)-1]
(3.10)
where
Kt_j = Capital cost at time t-1 (the beginning of period t), including land, equipment,
and structures.
t = Time period with initial construction at t = 0.
Ct = Operating costs, including labor, materials and supplies, interest on debt,
management and administration, working capital, property taxes, and
insurance. These occur at the end of each period t.
= Salvage value of equipment and land net of decommissioning costs.
= Operating life of the plant with T designating the final point in time for
purposes of the analysis.
= Private and public real rates of discount (rp = .08; rm = .04). See Appendix A
for a discussion of estimation of these parameters.
= Expected real inflation rate (r, = .04).
= Depreciation accrued in the t"1 period. It is calculated as a straight line over th
life of the plant.
xe = Effective tax rate equal to xs + (1 - xs)xf, where xs and Xf are the state and
federal average tax rates (xs = .07; Xf = .35).
Rt = Credits from associated sale of electricity and steam.
S
T
rp'rm
D
3-16
-------
Time periods in the analysis are denoted by t. Capital expenditures are considered
incurred at the beginning of a period. As a consequence, the capital expenditures of the first
period are denoted K^. Cash flows, operating expenditures, and revenues are considered to
occur at the end of a period and are subscripted accordingly. For example, the first period costs
are Cj and the fifth-period revenues are R5.
There are T periods in the analysis, the final point in time designated by T. Salvage value
revenues (S) are shown as occurring at the end of operation, and are thus discounted by the factor
av~T-
Since these expressions are for net present value of "costs," the cost components have
positive signs and the revenue components—sales, tax savings, and salvage value income—have
negative signs.
All flows are expressed in real terms or adjusted to be in real terms (e.g., the depreciation
tax saving). The following assumptions were used in formulating and implementing the
equations:
1. The economic life and the engineering life of the plant are both T.
2. Depreciation for tax purposes is "straight line."
3. rp and rm are symmetric with respect to borrowing or investing. There are no
differences in risk or transaction costs between these opportunities.
4. Costs in excess of revenues can be charged against revenues from other activities in a
period to obtain the tax deductibility of current period costs.
5. Marginal tax rates equal average tax rates.
6. Property taxes are included in operating costs.
A major interest in calculating cash flows is to find the increase in tipping fees that would
be needed to offset the after-tax cost. For private firms, the net present value (NPVp) of
offsetting revenue is shown in Equation (3.11).
NPC
1~* V^T)
= Txf (3.H)
The denominator (l-x«) allows for taxes that the private firm must pay on net revenues. The
"cost," in this context, is the estimated present value of offsetting revenue required by the plant's
owners if the full cost of the regulation to the plant is passed through to users in the form of
tipping fee increases. This adjustment is not needed for calculating NPVm because public
entities don't pay taxes on net revenues [NPVm = NPQnJ.
3-17
-------
Using these NPV equations to derive present value costs allows for comparison among
various management alternatives facing a single entity. In some situations the equations may not
be appropriate for comparing costs between two entities if those entities are different from each
other in financial structure. For example, two private firms, each with unique reinvestment
options and tax liabilities, require two different equations, each tailored to its firm's financial
situation. Comparisons between public and private firm cost of control made with these
equations are presented in Chapter 6, Table 6-2. Because public and private firms are inherently
different in financial and, perhaps, cost structure, the comparisons must be interpreted with care.
In Chapter 6 the NPV values are used to calculate annualized equivalents for the cost of
control for different model plants as well as for different plant ownership. The latter are then
divided by annual throughputs to obtain annualized $/Mg figures that are convenient for
comparing regulatory alternatives. Again, the same caution applies when comparing $/Mg costs.
The annualized $/Mg cost for a private firm or public entity is an approximation of the amount it
would have to charge in additional tipping fees in order to pass all costs along to MSW collectors
or generators, but there can be quite a bit of variation depending on the circumstances of
individual firms and public entities.
3.2.1 Combustion
Analysis of MWC costs using the cash flow model presented above provides a framework
for comparing costs as among plants and estimating cost impacts of regulatory alternatives for a
given plant. For the purposes of depreciation, the life of new plants is assumed to be 30 years.
Control equipment is assumed to have two consecutive 15-year life cycles. Salvage value, net of
decommissioning costs, is assumed to be zero in our calculations.
Table 3-4 provides estimated costs for NSPSs model plants. As noted above, these model
plants are used in this report as representative of certain types of MWC plants based on MWC
technology and capacity. The NSPSs model plants and their costs are described in detail in
Radian's cost report (EPA, I989c). The dollar measures used throughout the report for model
plants are December 1987 dollars.
3-18
-------
TABLE 3-4. PRODUCTION-COST RELATIONSHIPS OF NSPSs MODEL PLANTS
($1987)3
Costs
Annualized Capital Plus
Operating Costs (Net of
Energy Recovery Revenue)*1
Model
Plant
# MSW Input
(Mg/yr)c
Capital Operating
Costs Costs
C$103)
(SioV;
) ($/Mg)
Revenue
from Energy
Recovery
Privately
Owned
Facilities6
(SlfrVyr) ($/Mg) (SlOty
r) ($/Mg)
Publicly
Owned
Facilities'
($103/yr
) ($/Mg)
Mass Burn
1
4
2
5
3
37,800d
140,000
224,000
294,000
631,000
19,700
40,400
53,200
73,300
117,000
2,680
6,750
7,610
10,200
16,200
71.00
48.10
33.90
34.60
25.70
1,040
3,020
6,160
8,310
17.300
27.40
21.50
27.50
28.20
27.40
4,260
9,360
8,580
11,700
14,300
113.00
61.90
35.50
36.80
21.00
2,850
6,170
4,640
6,260
5,930
75.40
44.00
20.70
21.30
9.40
Refuse-Derived Fuel
1
6
Modular
9
10
8
FBC
11
12
274,000
584,000
9,450d
27,000
64,900
247,000
247,000
152,000
143,000
1,270
5,880
14,500
73,900
73,900
15,900
15,300
439
1,050
2,490
8,720
8,720
29.00
27.90
46.50
38.60
38.40
35.20
35.20
24,200
21,800
0
582
1,400
9,860
9,860
44.20
39.90
0
21.50
21.50
39.90
39.90
10,700
11,600
605
1,310
3,140
8,460
8,460
17.75
19.11
64.00
43.40
43.30
31.10
31.10
4,850
1.961
512
845
1,980
3,250
3,250
17.70
3.58
54.20
31.20
30.40
13.20
13.20
a Cost and input numbers are rounded to three significant digits. Model plants represent average characteristics for
new MWC facilities. Differences across columns due to rounding.
b Based on a 24-hour operating day
c Calculated using average capacity utilization reported in the 1988-89 Resource Recovery Yearbook (Gould, 1988).
Allowance is made for increased downtime for model plants 1 and 9; and allowance is also made for model plant
7 which co-fires 50 percent wood.
d Differences in annualized operating costs for privately and publicly owned facilities are due to differences in the
cash flow models for these firms as well as differences in cash flow model parameters, especially the discount
rate. See text for a discussion of these differences.
e Private discount rate = 8 percent
* Public discount rate = 4 percent
Operating costs for NSPSs plants in Table 3-4 include ongoing costs related to operation
and maintenance of the combustor unit, ash disposal, and auxiliary fuel use. Not included in the
baseline cost figures are any additional costs associated with the proposed NSPSs. Annualized
$/Mg unit costs of NSPSs plants decrease as MSW input increases, indicating economies of
scale, over the ranges for each technology used in this analysis.
3-19
-------
3.2.2 Landfilling
Landfill capital costs have two components—capital equipment and land. By combining
these two costs and treating land resale value as a component of salvage value, the total cost of a
MSW landfill can be represented by the net present cost equation used for combustors. As with
MWC plants, landfill costs are expressed in both private and public financial forms (see Table
3-5).
EPA has implemented revisions to Subtitle D criteria for MSW sanitary landfills. These
regulations impose standards that will increase landfill costs because they require landfills to
provide closure and post-closure care including groundwater, surface water, and gas monitoring
systems and final cover integrity. Table 3-5 also presents these Subtitle D control costs for
sanitary landfills. Unit costs decrease with increased throughput, indicating the existence of
economies of scale in MSW landfilling.
The data presented in Tables 3-5, 3-6, and 3-7 have not been estimated using the cash
flow model and parameters discussed previously and applied to combustor model plants. To
provide the reader with a general idea of the costs of these MSW management activities, related
but not strictly comparable data are presented in these tables.
Hazardous waste landfills are more expensive than sanitary landfills. This is due
*
primarily to the systems designed to prevent and detect groundwater contamination. Closure of
hazardous waste landfills is also highly regulated and much more expensive than closure of
sanitary landfills. The capital, land, and operating costs for hazardous waste landfilling are
presented in Table 3-6. The unit costs decrease with increased throughput, indicating the
presence of economies of scale.
3.2.3 Collection and Transportation
The cost of transporting MSW increases with the distance the waste is hauled and with
the traffic congestion along the haul route (Robinson, 1986). Transportation cost estimates are
presented in Table 3-7 for alternative distances and traffic congestion situations. Costs increase
with distance at a constant rate. Average costs decrease (i.e., cost/Mg/km) due to the presence of
fixed costs.
3-20
-------
TABLE 3-5. PRODUCTION-COST RELATIONSHIPS OF LANDFILLS a»b
Baseline Annualized
MSW
(Mg/day)
10
25
70
160
340
680
1,360
Input
(Mg/yr)
2,360
5,900
17,700
41,300
88,400
177,000
354,000
Capital
Costs
($103)
1,700
3,020
6,200
8,050
11,700
17,000
34,500
Operating
Costs
($103/yr)
44,400
35,800
91,400
95,500
87,600
156,000
229,000
Cost per Mg MSW
Public
($/Mg)
71.90
43.80
31.00
16.70
10.80
7.97
7.83
Private
($/Mg)
92.20
58.20
40.90
22.20
14.50
10.70
10.60
Subtitle D Annualized
Cost per Mg MSW
Public
($/Mg)
18.50
12.90
7.89
5.98
4.33
2.85
2.82
Private
($/Mg)
23.80
17.20
10.40
7.96
5.84
3.83
3.82
Total Annualized
Cost per Mg MSW
Public
($/Mg)
90.40
56.70
38.80
22.70
15.10
10.80
10.70
Private
($/Mg)
116.00
75.40
51.30
30.15
20.30
14.50
14.40
a Cost and input numbers are rounded to three significant digits. Details may not add to totals due to rounding.
b Differences in annualized operating costs for privately and publicly owned facilities are due to differences in the discount rate.
Public discount rate: 4%; Private discount rate: 8%.
Source: Temple, Barker, and Sloan, Inc., ICF, Inc., and Pope-Reid Associates. 1987. Draft Regulatory Impact Analysis of Proposed
Revisions to Subtitle D Criteria for Municipal Solid Waste Landfills. Prepared for the U.S. Environmental Protection Agency,
Office of Solid Waste.
-------
TABLE 3-6. COSTS OF HAZARDOUS WASTE LANDFILLING8
Capital
Land
Operating
TOTAL
Small (450 Mg/yr)
($/Mg)
186.80
21.94
383.20
591.94
Large (55,000 Mg/yr)
($/Mg)
8.48
1.08
45.65
55.21
a Values converted to $1987 dollars using the GNP implicit price deflator.
Source: Research Triangle Institute. 1986. A Profile of the Market for Hazardous Waste
Management Services, pp.D-67 to D-68. Draft Report prepared for the U.S. Environmental
Protection Agency.
TABLE 3-7. COSTS OF COLLECTING AND TRANSPORTING MUNICIPAL SOLID
WASTE*
Traffic Congestion (%rural/%suburban/%urban)
10 kilometers
25 kilometers
50 kilometers
100 kilometers
100/0/0
($/Mg)
2.95
4.90
7.84
13.71
50/25/25
($/Mg)
3.11
5.26
8.48
14.92
25/50/25
($/Mg)
3.15
5.35
8.62
15.19
25/25/50
($/Mg)
3.22
5.50
8.92
15.75
a Costs refer to one-way trips. Values converted to $1987 using GNP implicit price deflator.
Source: Robinson, William D., ed. 1986. The Solid Waste Handbook: A Practical Guide.
Wiley-Interscience.
3-22
-------
CHAPTER 4
MUNICIPAL SOLID WASTE MANAGEMENT
The discussion of demand and supply conditions in Chapters 2 and 3 focuses on the
historic activity levels and technical relationships associated with municipal waste combustion
and, more generally, solid waste management. As that discussion shows, analysis of the
municipal solid waste (MSW) management system is complicated by the many process options
available. In this chapter we examine another complicating feature of the municipal waste
management system: the role that public entities play as both shapers of, and participants in, that
system. In particular, we examine their role in two interrelated exchanges: the exchange
between waste generator and waste collector and that between waste collector and waste
disposer. We then conclude with a discussion of solid waste management decision making by
public and private entities.
4.1 PUBLIC INVOLVEMENT IN THE MSW SYSTEM
Public entities—local, state, and federal—play a large role in regulating and operating
MSW management systems. Their influence, however, is not unlimited. Material, engineering,
geographic, cost, and other technical and economic conditions certainly apply to both public and
private entities.
In addition, all MSW management systems ultimately involve private decision makers
and associated markets. Households and private firms generate MSW, collect and transport
MSW, build and operate MSW disposal systems, provide financing, and provide markets for
recycled material. In some settings these private activities compete with public operations; in
others, they provide factors of production and demand for outputs from public operations.
Whatever the case, these technical and market relationships are important factors in conditioning
the influence of public entities on MSW management generally and the economic impact of
changes in the cost of municipal waste combustion in particular.
Having noted this, however, we now examine in more detail the nature and extent of
public involvement in MSW management and the way this involvement shapes the economic
impact of NSPSs on municipal waste combustors (MWCs).
4.1.1 Local Government
Local communities, especially in more urbanized areas, often take the lead in organizing
MSW management and, in many cases, providing collection and disposal services. A wide
4-1
-------
variety of reasons explain this involvement: concern for the public health threat of uncollected or
improperly disposed MSW, natural economies of scale in organizing and performing MSW
collection and disposal, and a concern for the negative externalities sometimes associated with
private collection and disposal (e.g., litter, noise, smells, traffic) that, while not necessarily
unhealthy, may diminish public welfare.
How extensive is the local government role? Stevens (1978) identifies four market
structures for MSW collection:
• Public monopoly—public agency collects all MSW.
• Private monopoly—private firm(s) collect(s) all MSW in a specific area under a
franchise agreement and is (are) reimbursed by a public entity.
• Competitive—public agency and private firm(s) both collect MSW.
• Self service—generators haul their MSW to disposal sites.
Savas and Niemczewski (1976) estimate that over 80 percent of residential refuse is
collected under the first three market structures. Goddard (1975) estimated the share of public
and private collection from households, commercial, and industrial customers. These data,
presented in Table 4-1, show that public collection is most common for household refuse,
accounting for about 50 percent of collection service. Furthermore, some significant fraction of
private service was probably provided by contractors selected by public entities. In that case, the
public entity played a role in selecting the private collection firm, specifying the terms and
conditions of collection, and paying the private collector for the service.
TABLE 4-1. TYPES OF SOLID WASTE COLLECTION (PERCENT OF
GENERATORS SERVED)
Collection Agency
Generator
Household
Commercial
Industrial
Public
50
25
13
Private
32
(50)
62
(91)
57
(94)
Self
12
13
30
Sources: U.S. DHEW. 1968. The National Solid Wastes Survey: An Interim Report.
Cincinnati, Ohio. U.S. Public Health Service. (Cited in Goddard, 1975.) Estimates in
parentheses are from the National Solid Wastes Management Association. 1972. The
Private Sector in Solid Waste Management. Washington, DC. (Cited in Goddard, 1975.)
4-2
-------
Local public policy with respect to financing waste collection is important to determining
the economic impact of the NSPSs. First, the price for waste collection paid by generators may
affect their waste generation—higher prices elicit lower generation and vice versa. Second,
where waste generators don't directly pay for collection costs, then others, usually taxpayers, do.
Goddard (1975) presents two tables on financing solid waste collection, which are reproduced
here as Tables 4-2 and 4-3. We observe that half of the municipalities use only general taxes to
pay for collection and that larger cities are more likely to pay the cost of collection out of tax
revenues. Since these data were gathered, however, there has been a trend toward greater
emphasis on user fees as a source of local government revenue. This trend includes MSW
collection. Even so, user fees assigned to individual MSW generators often do not vary with the
amount of MSW produced. Generators are usually charged a flat fee per container or up to a
given number of containers. This type of pricing structure diminishes the impact of any changes
in the price of collection on the amount of waste generated. Furthermore, to the extent that
control costs for compliance with NSPSs are financed out of general revenues rather than
through charges, called tipping fees, paid by those unloading at the combustor, the additional
costs will not even have a chance to influence waste generation in those jurisdictions with a
system of user fees for collection.
TABLE 4-2. METHODS OF FINANCING SOLID WASTE COLLECTION BY
COLLECTION AGENCY, 1964" (NUMBER OF CITIES)
^^~ ' ' " _''•••^•J, _ .. mill ij^ii
Collection Agency
Method of Finance
General tax^
Service charge
Tax and service charge
Other
M
209
149
69
2
M,C
25
24
10
1
M,P
81
37
29
0
M,C,P
20
4
4
1
C
68
74
8
1
C,P
26
11
4
0
Percent
50
35
14
1
Key: M=Municipal
C=Contract
P-Private
a Sample size is 957.
b Such as property, income, or sales tax.
Source: Goddard, Haynes C. 1975. Managing Solid Waste. New York: Praeger Publishers.
4-3
-------
TABLE 4-3. METHODS OF FINANCING SOLID WASTE COLLECTION BY CITY SIZE, 1964 (PERCENT)
Population
Method
of
Finance
General Tax
Service Charge
Tax and Service Charge
Other
Number of Cities
5,000
to
9,000
47.2
39.0
13.4
0.6
180
10,000
to
24,999
46.0
38.0
16.6
0.0
307
25,000
to
49,999
51.5
32.7
14.2
1.6
190
50,000
to
99,999
58.0
28.0
12.9
1.1
93
100,000
to
999,999
59.6
27.0
13.5
0.0
74
1,000,000
and
Over
66.6
0.0
33.4
0.0
6
TOTAL
50.1
34.9
14.4
0.6
850
Source: Goddard, Haynes C. 197'5. Managing Solid Waste. New York: Praeger Publishers, p. 41.
-------
Waste disposal facilities, especially landfills, are more likely to be owned or operated by
government entities than are collection services.1 The EPA survey of 149 existing MWC
systems (1988a) requested information on owners and operators. The 106 responses received
and processed to date are tabulated in Table 4-4. The 25 respondents that provided distinct
ownership information reported that 78 percent were publicly owned and 64 percent were
publicly owned and operated.
TABLE 4-4. OWNERSHIP OF MUNICIPAL WASTE COMBUSTION PLANTS, BY
SIZE
Waste Received (103 Mg per year)
Ownership <10 10-30
Public 7 12
Private 5 5
Unknown3
TOTAL
30-50 50-100 100-250 >250 ALL
8 8 14 10 59
140 1 16
31
106
Ownership data not currently available.
Source: Research Triangle Institute. 1988. Memo to EPA on responses to Section 114 survey
letters.
A similar table on public and private ownership and operation of landfills was complied
from an EPA survey of nearly 20 percent of the operating landfill units (EPA, 1988b). Table 4-5
shows that over 70 percent of the landfills are publicly owned. In general, local entities feel a
strong responsibility to ensure that MSW generated and collected in their jurisdiction has a
proper place to go. They often believe that owning or operating the disposal facility provides
them with the necessary control. For instance, while several private firms recently submitted
bids to operate the Wake County, NC, landfills, the City of Raleigh substantially underbid its
competition to win the contract. The difference in the bids was not entirely a reflection of higher
operating productivity on the part of the City; it appears to have been all or in large part a
reflection of the City of Raleigh's desire to ensure that its residents have an accessible and
appropriately run disposal site for MSW (Tucker, 1988).
!In waste disposal facilities in particular it is not uncommon for the operator of the facility to be different from the
owner.
4-5
-------
TABLE 4-5. OWNERSHIP OF LANDFILLS, BY SIZE
Waste Received (103 Mg per year)
Ownership
Public
Private
Unknown
TOTAL
<0.9
1,596
198
15
1,809
0.9-9
1,548
178
8
1,734
9-45
1,098
147
28
1,273
45-90
362
112
34
508
90-180
225
24
11
260
>180
222
143
85
450
TOTAL
5,051
802
181
6,034
Source: U.S. Environmental Protection Agency. 1988. National Survey of Solid Waste
(Municipal) Landfill Facilities. Final report prepared by Westat, Inc., under contract 68-
An interesting finding of both surveys is that a sizable fraction of combustors and
landfills are owned or operated by a regional entity. For example, the combustor in Duluth, MN,
is owned and operated by a special service district created by the state legislature. This district
has responsibility for wastewater and MSW treatment of some half dozen municipalities in the
Duluth area. There is a trend—due to economies of scale in certain MSW management
operations and state government policy and legislation—toward forming sanitation authorities
that encompass multiple local government jurisdictions.
Financing of disposal systems owned by a public entity, like financing of MSW
collection, can be based on user fees, general tax revenues, or some combination of the two.
About 25 percent of landfills charge a price or tipping fee for MSW disposal services (Pettit,
1988). In 1988, the average tipping fee at landfills that have this charge was $29.70 per Mg; the
charge at resource recovery (combustion) facilities was $43.96 per Mg (Pettit, 1989). Since
collectors would be indifferent between a landfill or a resource recovery facility, the service
provided must be different at the two types of facilities. The most likely explanation for the fee
difference is that it reflects regional differences. When fees are charged they do not usually
cover all the costs of disposal; general tax revenues make up the difference between revenues and
costs. When fees are not charged, access to disposal facilities is typically restricted to service
area collection crews and, sometimes, residents.
State and federal grants have also been used to finance some of the cost of publicly owned
disposal facilities, particularly when experimental disposal systems are involved. Tables 4-6 and
4-6
-------
4-7 show tabulations of responses related to plant financing as reported to EPA in its survey of
existing facilities. These tables present the reported extent to which MWC owners' revenues
come from sources other than the sale of waste management services, energy, or recycled
materials. Such sources include subsidies and grants from state and federal government agencies.
TABLE 4-6. TOTAL OPERATING SUBSIDIES AS A SHARE OF TOTAL REVENUES
(MUNICIPAL WASTE COMBUSTION PLANTS ONLY)
Percentage
of Total Revenues
0
1-20
20-40
40-60
60-80
80-100
Total
Frequency
47
2
4
3
5
2
63
Relative Frequency
.746
.032
.063
.048
.079
.032
1.000
Source: Research Triangle Institute. 1988. Memo to EPA on responses to Section 114 survey
letters.
TABLE 4-7. GRANTS AS A SHARE OF TOTAL CAPITAL COSTS (MUNICIPAL
WASTE COMBUSTION PLANTS ONLY)
Percentage
of Capital Costs
0
1-10
10-20
20-30
30-40
40-100
Total
Frequency
39
1
3
5
7
2
57
Relative Frequency
.683
.018
.053
.088
.123
.035
1.000
Source: Research Triangle Institute. 1988. Memo to EPA on responses to Section 114 survey
letters.
4-7
-------
Table 4-6 shows total operating subsidies as a percentage of total facility revenues. This
is a crude measure of the extent to which facility operations receive support from government,
either at the state or federal level. Only 63 of the 106 facilities in the current Section 114 survey
database answered these questions. Of these 63 respondents, 75 percent reported receiving zero
subsidy. Based on this rough measure, it appears that the owners of MWC facilities do not, in
general, rely heavily on continuing direct payments from the state or federal government to cover
budget shortfalls. Two facilities, however, reported their subsidy revenue was more than 90
percent of their total revenues.
Table 4-7 shows grants as a percentage of total capital costs. This percentage is a crude
measure of the state and federal government support received by facility owners to cover their
capital costs. Again, only a fraction of the facilities responding to the survey answered this
question. Of the 57 facilities responding, 39 (68 percent) stated that they receive no grants.
Again, based on this crude measure, it appears that facility owners do not rely heavily on support
from the state and federal government to meet their capital costs, although a sizable minority
(15.8 percent) received over 30 percent of their capitalization in the form of state and federal
grants.
When a public entity collects MSW for disposal with either a private facility or a facility
owned by another public entity it usually negotiates a long-term contract for disposal with the
entity providing disposal. Private collection firms franchised by public entities also frequently
have such long-term contracts. These MSW disposal contracts usually have provisions for
passing on costs that arise due to circumstances outside the control of the disposal system
operator. Requirements related to air emission control NSPSs most likely would fall under such
provisions. As such, where tipping fees are used, the costs of regulation for an existing facility
would be passed on to the collectors that hold long-term rights to dispose of MSW at that
disposal site. Similar terms may also occur in contracts between private waste collectors and
public entities.
Local public entities also participate in MSW management in their capacity as regulators
of land use and guardians of public health and welfare. Siting and operation of private MSW
disposal facilities are usually subject to local government review and approval. Successfully
addressing local citizens' concerns has become very difficult because of awareness of both
hypothesized and actual effects disposal facilities have on local property values, health, and the
environment. Local fear and strident opposition to landfills and combustors are fairly
commonplace (Wall Street Journal, Sept. 4, 1987). One, albeit remote, economic impact of the
4-8
-------
NSPSs may be reduced costs of siting and building new MWC plants because local opposition
will wane as new MWC plants emit fewer residuals into the atmosphere.
4.1.2 State Government
In recent years, states have taken a more active role in shaping the MSW management
practices in their jurisdiction. While the nature and level of state initiatives vary tremendously,
many states have become active in providing a framework for organizing and planning local
MSW management (Kovacs, 1988). States most prominent in this area tend to be those
confronting serious MSW management problems due to dense populations (because of the large
amount of waste such populations generate and the limited space for disposal sites) and the
vulnerability of their natural environment (e.g., states with limited water resources and/or high
water tables).
Goddard (1975) notes that California and Connecticut were among the leaders in this
regard, setting up a Solid Waste Management Board in California and the Connecticut Resource
Recovery Authority in Connecticut in the early 1970s.
More recent examples include New York State's Department of Environmental
Conservation, which issued solid waste rules covering liner requirements for landfills; disposal of
combustor fly ash; and emission limits on particulates, dioxin, and nitrogen oxide for waste
incinerators (Solid Waste Report, Sept. 12,1988). In June 1988, Florida passed a solid waste law
that, among other things, required each county to initiate a recycling program, set 1994 as the
target date by which 30 percent of the State's waste would be recycled, established qualifications
to be met by operators of waste management facilities, and required owners and operators of
landfills to set fees to ensure the proper closure of landfills (Solid Waste Report, July 11, 1988).
Many other states have either passed legislation that is similar in intent to these examples
or are seriously considering such legislation—for example, Massachusetts (Cowen, 1987).
Table 4-8 summarizes state solid waste laws enacted in 1988. Of particular interest are the
regulations regarding general solid waste management, mandatory source separation, and waste-
to-energy facility requirements. Ohio's H592, signed June 24,1988, effectively doubles solid
waste disposal permitting fees and calls for a comprehensive state solid waste management plan
(Solid Waste Report, October 17, 1988). In effect, states have become very active in establishing
the terms and conditions under which local governments must operate as they seek to structure
their local MSW management systems.
4-9
-------
TABLE 4-8. STATE SOLID WASTE LAWS ENACTED IN 1988
ISSUES
General solid waste
management
Purchasing preferences
for recyclables
RECYCLING: Waste reduction
program requirements
RECYCLING: Planning and
goals
RECYCLING: Study
RECYCLING: Mandatory
source separation
NONDEGRADABLES: Bans/
restrictions on use
NONDEGRADABLES: Taxes
NONDEGRADABLES:
Incentives to recycle
Labeling of products made
with recyclable/degradable
material
Waste-to-energy facility
requirements
AZ
•
CA
•
•
•
•
CT
•
FL
•
•
•
•
•
•
•
•
HI
•
IL
•
•
•
IA
'
•
LA
•
MD
•
•
MA
•
•
ME
•
•
•
Ml
•
•
MN
•
•
•
NH
•
NJ
i
NY
•
•
•
OH
•
OK
•
•
•
•
PA
•
•
•
•
•
Rl
•
•
TN
•
VA
•
•
WA
•
•
Wl
•
•
Source: Solid Waste Report, October 17,1988
-------
4.1.3 Federal Government
The Federal Government, by virtue of both legislation and regulation, influences solid
waste management in a variety of ways. Sections of the Resource Conservation and Recovery
Act of 1976 (RCRA), the National Energy Conservation Policy Act of 1978, and the
Comprehensive Environmental Response, Compensation, and Liability Act of 1980 all address
issues of solid waste and MSW management. Under RCRA and subsequent amendments, for
example, regulations requiring stricter control of waste at sanitary landfills, referred to as
Subtitle D regulations, have been proposed and procurement procedures aimed as fostering
recycling initiated.
Two Federal actions that have been especially important to MWC activity have been the
Public Utility Regulatory Policy Act of 1978 (PURPA) and the Tax Reform Act of 1986. Under
PURPA regulations, independent small power generators have better opportunities, under more
favorable financial terms, to provide electricity to electric utilities. This, in combination with
higher electricity prices generally, has spurred the development of MWCs that co-produce steam
and electric energy. The Tax Reform Act of 1986, on the other hand, has reduced the tax and
financing advantages available to private owners of new MWCs (Hilgendorff, 1989). While
Hilgendorff estimates that between 50 and 100 upcoming MWC projects are "grandfathered"
under the Act, "most analysts agree that the eventual effect... will definitely compel municipal
ownership."
Recently, EPA has led a federal effort specifically aimed at reshaping the way in which
solid waste is managed in the U.S. In its reports, The Solid Waste Dilemma: An Agenda for
Action Final Report (1989d) and Background Document (1989e), EPA calls for "integrated waste
management" in keeping with the following preferred hierarchy of processes:
• source reduction (including reuse of products),
• recycling of materials (including composting),
• waste combustion (with energy recovery), and
• landfilling.
The final report and supporting documents go on to discuss obstacles to establishing the
integrated waste management system envisioned and options for overcoming those obstacles
through information, demonstration, and incentive programs.
4-11
-------
4.2 MSW DECISION MAKING
The design and operation of a MSW management system requires decision makers to
make a number of choices. These choices are constrained by technical and economic factors and
will have implications for generators and managers of solid waste as well as suppliers of inputs
of MSW management services (e.g., labor services, MWC suppliers, owners of land) and area
residents.
For example, governments must decide which institutional arrangement to use for waste
collection and disposal. Depending on the institutional arrangements, public and private decision
makers must choose the amount and type of solid waste to generate (the demand side) and
disposal processes to provide—combustion, recycling; and landfilling (the supply side). For each
process they must determine the appropriate location, scale, and design lifetime. Imposition of
the NSPSs will likely affect these choices. To help analyze this effect, we now briefly discuss
the objective(s) that guide the response of private and public decision makers.
4.2.1 Private Decision Making
In conventional economic analysis, households are utility maximizers and firms are profit
maximizers. They are bound together by markets—market supply and demand balance
competing interests given finite resources, limited technical knowledge, and institutional
conditions, including the structure of the markets themselves (e.g., perfect competition,
monopoly). This analysis assumes that the private components of most local solid waste
management systems will follow the conventional economic paradigm. As described in
Chapter 2, households and firms that generate waste respond to increases in the price of waste
collection by reducing their amount of waste generated. The proportionate reduction may not be
large, but if the price increase is large enough, the effect will be noticeable. By the same token,
however, if cost increases in the MWC portion of the management system are not fully reflected
in prices or tied to the amount of waste generated, there will be little or no change in waste
generation attributable to the NSPSs beyond that associated with "income effects" due to changes
in taxes.
Similarly, firms that collect, transport, and/or dispose of waste are assumed to respond to
changes in the cost of production by adjusting their input mix to keep costs down and by passing
as much of the remaining cost increase on to their customers (as price increases) as market
conditions and contractual arrangements allow. In the case of a new MWC plant, the prospective
owner has a broader set of available options for reducing cost, including changing the disposal
4-12
-------
technology selected and not building a plant at all. In the end, however, local market conditions
must be seen by the prospective owner as supporting a disposal price that covers the minimum
cost method of providing disposal services. The extent that MWC owners can raise prices and
the degree to which this is accompanied by a change in the amount of waste combusted depends
on the demand for combustion services in the local market. As noted in Chapter 2, this demand
depends on the share of combustion services used, the availability of substitutes, and the demand
of disposal services generally. While institutional conditions associated with local, state, or
federal policies will affect these determinants of demand, in this analysis private parties accept
such institutional conditions as given and make production and investment decisions so as to
maximize profit or, equivalently in a competitive market, minimize the cost.
4.2.2 Government Decision Making
Government decision making is of particular concern in this analysis given the large role
that government plays at every level, but especially the local level, in shaping MSW management
systems. Theoretical and applied literature does not provide much positive guidance on the
behavior of governments (e.g., Rubinfeld, 1987). On the other hand, normative literature on
MSW management decision making, much of it aimed at decision making by public officials,
often assumes that cost minimization, sometimes referred to a "project economics," is the
implicit basis on which decisions regarding MSW management arc reached. Consequently, this
literature addresses methods and means by which the decision maker can make the correct, cost-
minimizing choice (Robinson, 1986). Examples also include authors who aren't necessarily
enthusiastic about conventional waste disposal options. Kirshner and Stern (1985) and the
Institute for Local Self-Reliance (Morris, 1987) both couch their arguments against extensive
MSW incineration and for recycling in cost-minimizing terms.
RTI contacted eight communities that had recently made decisions to build MSW
combustors to learn their basis for selecting a particular disposal option and technology (Berry et
al., 1988). While this was not a scientific survey, the individuals contacted in the public works or
a similar department described cost as the over-riding consideration in the community's decision.
The decision was subject to the conditions that they have some means of meeting their
community's MSW disposal requirements and that this means is compatible with environmental
and other considerations. For example, most said either that MWC was more economical than a
landfill in their circumstances or that they couldn't get a permit to build a new landfill in their
area because of environmental constraints. The exception to this was an individual who noted
that the least costly MWC system was selected even though a landfill might have been a cheaper
4-13
-------
way to dispose of MSW per se because the community also had to provide steam to a public
facility and, given the wider scope of the decision, a MWC system was most economical.
While this evidence is not absolute proof of the proposition, it seems sufficient to justify
that local or municipal MSW system decision making represents cost minimization, subject to
the constraint that all MSW must be collected and properly disposed. It differs from private
decision making in that the minimum cost is assessed based on centralized enterprise of costs of
various combinations of MSW management alternatives, some of which may not be feasible for a
private firm due to institutional or financial constraints.
4.2.3 Cost Minimization
In this analysis cost-minimizing decision making is used as the basis for the enterprise
cost and emission reduction estimates of Scenario I. In other words, the control options
identified in the engineering analysis that supports this economic analysis were developed to
meet the alternative emission standards at minimum cost, and we assume that private and public
decision makers will select these control options for their new MWC plants after promulgation of
the NSPSs. In Scenario n, the decision makers consider a broader range of choices, reviewing
plant technology as well as control technology with the intent of minimizing the total system
cost. Making quantitative impact estimates for Scenario II is complicated, however, because of
both the different institutional and market conditions that characterize local waste management
systems. Rather than make quantitative estimates for Scenario II in this analysis, we take
advantage of a complimentary analysis of an even broader array of waste disposal choices
(Mathtech, 1989) and use it to estimate economic impacts under Scenario in. In Scenario in,
decision makers make investment decisions for solid waste disposal from among three MWC
technologies and a landfill of appropriate size and revise their choices based on the new costs of
control associated with the NSPSs for each of the MWC technologies.
4-14
-------
CHAPTER 5
ANALYTICAL APPROACH TO ESTIMATION OF COST AND EMISSION IMPACTS
To estimate the impact of MWC New Source Performance Standards (NSPSs), estimated
levels of new activity in relevant markets must be compared before and after the regulation. First
we project baseline or "without NSPSs" conditions. Then we compare these conditions with five
"with NSPSs" regulatory alternatives under three economic scenarios. As their names suggest,
the regulatory alternatives represent the different regulatory designs under active consideration
by EPA, and the economic scenarios represent varying degrees of market response to the NSPSs.
5.1 BASELINE PROJECTIONS
The baseline level of MWC activity depends on how much solid waste is generated and
what other waste handling methods are available and in use, including landfills, materials
recovery, self disposal, and littering. For this reason, we make baseline projections not only for
solid waste combustion but also for solid waste generation and the amount of solid waste
managed by other major methods.
5.1.1 Initial Conditions
We selected 1986 as the initial year (i) for baseline projections because it is the most
recent year for which a wide variety of quantitative data on waste management activities are
available. We used these data to build up a profile of 1986 solid waste flows (Qwi) associated
with MWCs and the major waste management options that are alternatives to combustion. This
procedure is sometimes complicated because of differences in the definition and scope of the
various source documents.
From the results of the 1986 landfill survey (EPA, 1988b) Westat estimated that a total of
189.4 million Mg of waste was landfilled in 1986. This estimate includes the estimated amount
of municipal incinerator ash landfilled in 1986: 1.3 million Mg. We subtracted this value from
the total because other estimates do not include incinerator ash. Thus, the Westat estimate of
waste landfilled in 1986, net of municipal incinerator ash (Qli), equals 188.1 Mg. This value
makes up the first component of 1986 solid waste flows illustrated in Figure 5-1.
A recent Franklin Associates report (1988) provides useful information on 1986
municipal solid waste (MSW) flows. These flows are based on estimates of consumption activity
and include some waste components that are not landfilled or combusted at a municipal
incinerator (such as self-disposed materials, litter, and recycled materials) and exclude some
5-1
-------
10b Mg/yr
17.2
188.1
1986
1991
1996
H!
HI
Combustion
Materials Recovery j/^/
Landfilling
Franklin
Associates
F-ctimzifac
C_OU///Cl(CsO
Figure 5-1. Solid Waste Flow Projections, 1986 through 1996
5-2
-------
solid wastes that are landfilled and combusted (such as car bodies, sewerage sludge, mausuuu
and commercial waste, construction wastes, and foreign import packing materials). Franklin
Associates estimates that 143.0 million Mg of MSW (excluding municipal incinerator ash) was
generated in 1986. They estimate the disposition of this MSW as follows: 15.3 million Mg of
material recovery (recycling), 1 19.0 million Mg landfilled (net discards), and 8.7 million Mg
combusted (by energy recovery facilities).
The Franklin Associates (1988) materials recovery estimate (Qmi) is added to the waste
flows represented in Figure 5-1. The Franklin Associates MSW flows to landfills and
combustors are shown as components of the landfill and combustion flows estimated in
Figure 5-1. These components are somewhat smaller than the corresponding landfill and
combustion flows used to specify initial conditions in baseline. This is because landfills and
combustors receive more solid waste than included in the definition of MSW used by Franklin
Associates. Still, as Figure 5-1 shows, most waste flowing to these treatment technologies
appears to originate as MSW.
The Radian census of existing municipal waste combustors (EPA, 1989b) is the primary
source for our estimate of baseline solid waste flows to combustors in 1986. This census
includes all combustors that burn MSW. We estimated waste flows to combustors by applying
average capacity utilization factors to each identified combustor. These factors varied with the
combustor technology (mass burn, modular, or RDF) and heat recovery capability as reported by
the 1988-89 Resource Recovery Yearbook (Gould, 1988) and estimated in Radian's retrofit report
(EPA, 1989b). We also adjusted combustor solid waste flows to account for the fraction of
facilities that co-fired solid waste with another fuel. As a result of these calculations we estimate
that combustors handled 14.4 million Mg of waste flow (Qci) in 1986. This is higher than the
Franklin Associates estimate of MSW flows to combustors in 1986 partly because Franklin
Associates only counts "energy recovery" facilities among its combustors.
Summing components as shown in Equation 5.1, total non-hazardous solid waste flow
handled by landfills, recycling, and combustion in 1986 is estimated to be 218 million Mg.
Q . = Q,. + O . + Q . (5.1)
^ ^ ^ V '
This total, along with the component major waste handling methods, is the point of
departure for baseline projections. As can be seen by Equation (5.1), our analysis does not
specifically deal with littering and self-disposal even though our total waste generation estimates
do include these components. These waste flows are even more difficult to estimate accurately
5-3
-------
and do not comprise a significant segment of the total waste stream. By the same token, we
don't attempt to estimate or project changes in materials recycling of industrial or commercial
waste that would otherwise be landfilled or ash from combustion, net of any ash sent to a
dedicated ash landfill, that is landfilled as MSW.
EPA's goal for the nation is to incinerate 20 percent and to recycle 25 percent of the
MSW stream by 1992 (EPA, 1989d). The projection of 44.3 and 19.7 million Mg in 1996 for
combustion and recycling, respectively, amounts to 17.5 and 7.8 percent of the MSW stream.
The projection used here is based on a slightly different definition of the composition of MSW
and is only for analytical purposes; it should not be interpreted as an alternative goal. To the
extent the nation recycles more than 7.8 percent of MSW in 1996, there will be a reduced
demand for MWC and landfill services, and the nationwide costs and emission reductions
associated with the NSPSs will be smaller. On the other hand, to the extent the nation burns
more than 17.5 percent of MSW in 1996, the demand for MWC services will be larger and the
nationwide costs and emission reductions associated with the NSPSs will be larger. The costs
and emissions associated with landfills are not analyzed here.
5.1.2 Projections
The baseline projection used in this analysis draws heavily on the assumptions and
projections made by Franklin Associates (1988). In summary, we project future total waste
flows and associated flows to landfills and materials recycling. Projected landfill and materials
recycling flows are subtracted from the projected total to provide a projection of the flow of
waste to combustors in 1996. This procedure is characterized in Equation (5.2), where the p
subscript represents baseline projections.
Qcp = Qwp - (Qmp + Qip) (5.2)
We select 1996 because of the five-year period of analysis adopted for this report. All
plants that begin construction within five years after publication of the NSPSs in the Federal
Register (anticipated for late 1989) are expected to be operating in 1996. Plants under
construction before the date the NSPSs are proposed will be subject to a separate regulation
(MWC Guidelines) to be published at the same time as the NSPSs.
Franklin Associates assumes that "waste landfilled will be relatively constant." Our
baseline projection actually fixes landfilled waste at 188.1 Mg per year. Materials recycling of
MSW is projected to grow at the annual rates estimated by Franklin Associates, reflecting "no
5-4
-------
dramatic changes in current practices." This amounts to 19.7 million Mg. in 1996. The growth
in total waste flows is based on the annual growth rates in MSW gross discards projected by
Franklin Associates In making its projection, Franklin Associates relied on documentation of
"historical production (or consumption) of materials and products that enter the municipal waste
stream."
The total solid waste flows projected for the baseline in this analysis (using Franklin
Associates' projected growth rate for gross discards) is 252.1 million Mg. Equation (5.2) gives a
baseline projection of waste combusted in 1996 of 44.3 million Mg. These projections are shown
in Figure 5-1 along with the comparable projections made by Franklin Associates for MSW
flows to landfills and combustors. Figure 5-1 also shows projections of total waste flows and its
components for 1991, based on the procedures described above and that apply to the MWC
Guidelines. These data are denoted by the subscript g in Equation (5.3) below.
5.1.3 Baseline Combustion
As shown in Equation (5.3), the difference in the projection of waste flows combusted in
1996 and those in 1991 is the waste flow to combustors, Qcn, that is affected by the NSPSs.
Qcn = Qcp - Qcg (5.3)
The baseline projection of waste flow to affected combustors is estimated to be 15.0
million Mg/year.
Using baseline projections of waste flows affected by the.NSPSs, we constructed
estimates of affected MWC capacity by MWC technology and NSPSs model plant category.
First we adjusted the waste flows implicit in the model NSPSs plant description and Radian's
cost report (EPA, 1989c) to reflect the capacity utilization estimates described above and plants
that co-fire with materials other than MSW. Then we allocated projected baseline waste flows,
Qcn, to each of the 12 model NSPSs plant categories in proportion to the waste flow shares
implicit in Radian's report. Table 5-1 displays the allocation of these flows to combustor
technologies and model NSPSs plant categories. Table 5-1 also presents corresponding
information on the total MSW capacity for each NSPSs model plant category.
5-5
-------
TABLE 5-1. BASELINE PLANT CAPACITY AND WASTE FLOW ESTIMATES FOR
MWC PLANTS SUBJECT TO NSPSs
Model Plant
Number
1
2
3
4
5
6
7
8
9
10
11
12
Model Plant
Description
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
Capacity3
(106 Mg/yr)
1.11
1.93
6.33
0.54
1.12
3.57
2.19
0.27
0.03
0.24
0.61
1.35
19.29
Waste FIowb
(106 Mg/yr)
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
Capacity estimates based on Radian's model plants description and cost report (EPA, 1989c).
bWaste flow estimates calculated based on the annual operating hours reported in Table 3-3. For
9 of the model plants these hours reflect the average capacity utilization, by plant type,
reported in the 1988-89 Resource Recovery Yearbook (Gould, 1988). Allowance is made for
increased downtime for model plants 1 and 9; and allowance is also made for model plant 7
which co-fires 50 percent wood.
Figure 5-2 compares the baseline projections of future combustion developed here with
projections recently made by other organizations. Most alternative projections are made in terms
of capacity, so that is the basis for the comparison adopted in Figure 5-2. The different
projections are adjusted for this comparison to reflect a common scope—that of the Radian
combustor profile. Figure 5-2 shows that projections of combustor capacity for 1996 range from
Franklin Associates' estimate of 26.3 million Mg to Radian's estimate of 104.7 million Mg. In
comparison to the alternative projections of MWC capacity made by other organizations
(excluding Radian), the "Baseline" projection used for this report exceeds all but the Radian
estimates and is 50 to 100 percent higher than the two lower projections.
5.2 SCENARIOS
Analyzing the economic impacts of MWC NSPSs under each regulatory alternative is
especially challenging because of the MSW industry's complicated organizational structure and
decision-making policies. Even assuming that cost minimization is a primary decision criterion,
it is difficult to predict how public and private decision makers—uncertain about the future,
constrained by a wide variety of institutional considerations, and examining a multiplicity of
interrelated waste management options—will respond to tighter air emission NSPSs on
5-6
-------
110.00 -i-
100.00 - -
90.00 - -
CAPACITY
MILLION 60.00 - -
Mg/yr
10.00
RADIAN
BASELINE
FROST &
SULLIVAN
FRANKLIN
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
YEAR
Figure 5-2. Comparison of MWC Capacity Projections, 1986 to 1996
5-7
-------
combustors. Will they adopt the air pollution control devices (APCDs) designated in the
engineering analysis for each model plant? Or, confronted by the prospect of potentially
expensive modifications to their new plant, will they review their entire waste management
program and devise an alternative strategy in light of the NSPSs?
Because of the uncertainty surrounding response to NSPSs, three economic impact
scenarios are considered.
Scenario I No substitution; employ APCD modification as specified
in the model plant analysis.
Scenario H MWC substitution; different MWC technology
replacement for baseline MWC where feasible and cost-
minimizing.
Scenario m Greater substitution; landfill or different MWC
technology allowed to replace baseline MWC.
These scenarios, applied to each of the regulatory alternatives, are analyzed using the model
plants introduced in Chapter 3. The scenarios differ in the extent to which decision makers
substitute away from the projected level and mix of baseline projections of new combustion
activity. Scenario I assumes that there is no such substitution: all plants projected in the baseline
will be built as planned. These plants, as represented by their respective model plants, are
assumed to be built as planned except that they incorporate the additional APCDs identified in
the cost report (EPA, 1989c) for each regulatory alternative. Of course, these changes also
impose additional cost on the plant and reduce plant emissions.
Scenario II assumes that decision makers consider substituting a different MWC plant for
a plant affected by the NSPSs. In this scenario substitution is based on cost and capacity
comparisons among the NSPSs model plants. In Scenario II, then, baseline levels of MSW
combusted remains the same, but the mix of MWC plants changes. This change reduces the cost
of the regulatory alternatives relative to those of Scenario I. In this report, we don't provide
quantitative estimates of these costs and any associated emission reductions of each regulatory
alternative for Scenario H; we examine in a qualitative way the impact of NSPSs when decision
makers include the option of building a different type of MWC in place of the one originally
planned. To a certain extent, however, this substitution is incorporated in Scenario in, for which
quantitative estimates of impact are provided.
5-8
-------
Under Scenario III we examine the impact of NSPSs when decision makers include
increased landfilling as well as other MWCs, as alternatives to the original NSPSs plant. Unlike
Scenarios I and n, then, Scenario III admits the possibility that the NSPSs will reduce the amount
of MSW combusted relative to the baseline. To estimate the extent of such substitution under
each regulatory alternative, an econometric analysis of historic disposal choice was undertaken
(Bentley and Spitz, 1989). RTI used the sensitivity of disposal choice to cost increases as
estimated by the econometric model to project the extent to which baseline estimates of the level
and mix of combustion would change in response to the cost increases associated with the
regulatory alternatives. The revised forecast of affected plants is then used as the basis for
projecting national cost impacts using the model plant method described above for Scenario I.
5.3 REGULATORY ALTERNATIVES
Six conditions are relevant to estimating the economic impacts of the NSPSs: the
baseline and five regulatory alternatives. The baseline conditions that establish the number and
type of affected NSPSs plants and the amount of MSW combusted in these plants are described
in the previous sections of this chapter. The baseline conditions described here refer to control
technologies and emission levels for MWC plants in the absence of the NSPSs. The costs of
these baseline control technologies have already been presented for model plants in Chapter 3.
The five regulatory alternatives considered in this analysis specify the emission levels the
affected plants would have to comply with.
Table 5-2 outlines the five regulatory alternatives for the NSPSs. The NSPSs would
impose varying emissions limits on MWCs. Air emissions affected include polychlorinated
dibenzo-p-dioxins and dibenzofurans (CDD/CDF), carbon monoxide (CO), paniculate matter
(PM), hydrogen chloride (HC1), sulfur dioxide (SO2), and lead (Pb). In certain circumstances the
residue remaining after combustion is changed due to the APCD changes.
Radian has estimated baseline emission levels in the NSPSs model plant description and
cost report (EPA, 1989c). These baseline emissions are the emissions produced if there are no
additional controls or changes in operating conditions at the affected facilities. The analysis of
regulatory impacts requires these data on the emission levels associated with the baseline and the
regulatory alternatives for affected facilities.
5-9
-------
TABLE 5-2. MAXIMUM EMISSIONS BY REGULATORY ALTERNATIVE
Plant Capacity (Mg per day)
Regulatory Control
Alternative Parameters
I CDD/CDF
PM
Temperature3
HA CDD/CDF
PM
Temperature2
IIB CDD/CDF
PM
Temperature3
HI CDD/CDF
PM
Temperature3
IV CDD/CDF
PM
Temperature3
Small
<225
300 ng/Nm3
0.1 8 g/dscm
230 °C
300 ng/Nm3
0.1 8 g/dscm
230 °C
75 ng/Nm3
0.02 g/dscm
175 °C
300 ng/Nm3
0.1 8 g/dscm
230 °C
75 ng/Nm3
0.02 g/dscm
175 °C
Large
>225
300 ng/Nm3
0.02 g/dscm
230 °C
75 ng/Nm3
0.02 g/dscm
175 °C
75 ng/Nm3
0.02 g/dscm
175 °C
5 ng/Nm3
0.02 g/dscm
150°C
5 ng/Nm3
0.02 g/dscm
150°C
Source: Radian. 1989. Background Paper, Municipal Waste Combustors. Prepared for the
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
Under each of the five regulatory alternatives, each model plant is associated with one of
three engineering control options examined by Radian (EPA, 1989c). The three control options
consist of various combinations of four types of control technologies. The technologies followed
by the emissions they are designed to control are listed below:
• Good combustion practice (GCP)—CDD/CDF and Pb
• Flue gas temperature reduction—CDD/CDF
• Paniculate matter (PM) control—PM
• Acid gas control (dry sorbent injection/ESP or spray dryers/fabric filters)—HC1 SO?
and CDD/CDF ' 2)
PM control is further defined as moderate, good, or best depending on the level of
emission reduction required to meet the relevant regulatory alternative. Acid gas control is also
5-10
-------
described as good or best in reference to the level of emissions allowed under each regulatory
alternative. Good acid gas control is achieved through the use of dry sorbent injection.
Reduction of emissions to the level required for best acid gas control is achieved through spray
dryers and fabric filters.
The levels of control reflected by the regulatory alternatives do not follow a straight
forward progression of least to most stringent when moving from Regulatory Alternative I
through Regulatory Alternative IV. Compared to HA, Regulatory Alternative IIB imposes no
further control over large plants (greater than 225 Mg/day design capacity) but tightens controls
over small plants (less than or equal to 225 Mg/day design capacity). Moving from IIA to III, on
the other hand, imposes no further control for small plants while bringing greater control over
large plants. The two paths of least to most stringent are characterized below:
HA IV
5.3.1 Baseline Emissions
Baseline combustion practice and emission rates vary from plant to plant. In the baseline,
all model plants meet the federal standards, which limit PM emissions to a maximum of 0.18
g/dscm for MWC plants with the exception of plants with design capacity of 45 Mg/day or less.
As a result, all model plants in the analysis limit PM emissions to 0.18 g/dscm or less in the
baseline except model plant 9 which has design capacity of 45 Mg/day.
5.3.2 Regulatory Alternative I
Regulatory Alternative I is the least stringent of all the regulatory alternatives. Under this
alternative, all plants must achieve GCPs and reduce flue gas temperatures to 230 °C. CDD/CDF
emissions would be reduced to 300 ng/Nm3 with these controls in place. Small plants are
required to practice moderate PM control, limiting PM emissions to 0.18 g/dscm, and large plants
are required to achieve best PM control, limiting PM emissions to 0.023 g/dscm. No acid gas
control is required for plants of any size under this alternative.
5.3.3 Regulatory Alternative IIA
Regulatory Alternative IIA is the next most stringent alternative and also requires GCPs
for all plants. Controls for small plants are no different under Regulatory Alternative I than HA.
5-11
-------
Moderate PM control, 230 °C flue gas temperatures, and no acid gas control are required of smal
plants. However, large plants are required to further reduce flue gas temperatures to 175°C
achieve good acid gas control through the use of add-on control equipment, and practice best PM
control. This has the affect of reducing CDD/CCF to 75 ng/Nm3 and PM to 0.023 g/dscm. HC1
and S02 emissions are reduced 80 and 40 percent over the baseline, respectively.
5.3.4 Regulatory Alternative IIB
One progression from least to most stringent regulation is characterized in the move from
IIA to IIB. Regulatory Alternative HB has the same controls for large plants and more stringent
controls for small plants when compared to Regulatory Alternative IIA. As with Regulatory
Alternative IIA, GCPs, 175 °C flue gas temperatures, best PM control, and good acid gas control
are requmxi of large plants. Small plants, however, must also reduce flue gas temperatures to
175 °C, achieve good acid gas control, and practice best PM control under Regulatory
Alternative IIB. These controls reduce CDD/CDF to 75 ng/Nm3 and PM to 0.023 g/dscm An
80 percent reduction in HC1 and a 40 percent reduction in SO2 over baseline emission levels
would result from these control measures.
5.3.5 Regulatory Alternative IH
Another progression from least to most stringent occurs when moving from IIA to m.
Regulatory Alternative IH imposes the same controls on small plants as IIA while enforcing
tighter controls on large plants. As before, all plants are required to achieve GCPs but small
plants are only required to practice moderate PM and maintain 230 °C flue gas temperatures. No
acid gas control is required for small plants under this alternative. Large plants must meet the
most stringent standards imposed in the progression thus far including GCPs, 150 °C flue gas
temperatures, best PM control, and best acid gas control. Emissions for large plants are reduced
to 5 ng/Nm3 for CDD/CDF and 0.023 g/dscm for PM. HC1 and SO2 emissions are reduced 97
and 90 percent over the baseline, respectively.
5.3.6 Regulatory Alternative IV
Regulatory Alternative IV is the most costly and the most stringent. Both small and large
plants are required to practice GCPs and achieve best PM control. Good acid gas control and
flue gas temperature reduction to 175 °C are required of small plants under this alternative.
Large plants must achieve best acid gas control and flue gas temperature reduction to 150 °C.
Emissions for large plants are identical to those under Regulatory Alternative III, and emissions
for small plants are identical to those under Regulatory Alternative HB for the same size
category.
5-12
-------
The technologies employed to control emissions are listed in Table 5-3. Table 5-4
identifies the control option described in the Radian cost report (EPA, 1989c) used for each plant
capacity range as applied to each regulation. For example, large plants can meet their emission
limits under Regulatory Alternative I using Control Option 1. Therefore, data on costs and
emissions for large plants under Regulatory Alternative I are for Control Option 1.
TABLE 5-3. AIR POLLUTION CONTROLS BY REGULATORY ALTERNATIVE
Regulatory
Alternative
Baseline
I
IIA
Plant Capacity (Mg
Small
<225
GCPsa
Moderate or no PM^
GCPsc
Moderate PMd
GCPs
Moderate PM
per day)
Large
>225
GCPsa
Moderate or good PM
GCPs
Best PM
GCPs
Good Acid Gase
Best PM
GCPs GCPs
Good Acid Gas Good Acid Gas
Best PM Best PM
HI GCPs GCPs
Moderate PM Best Acid Gasf
Best PM
IV GCPs GCPs
Good Acid Gas Best Acid Gas
Best PM Best PM
aModel-plant specific.
bAll 11 l(b) model plants meet Industrial Boiler Standards in the baseline requiring all plants with
design capacity over 45 Mg/yr to limit PM emissions the level required under moderate PM
control. All small plants with the exception of model plant #9 (45 Mg/day capacity) meet
this limit for PM emissions in the baseline.
cGood combustion practices (GCPs) include proper design and operation of the combustor.
Exhaust gas temperature control is also included in all alternatives with GCPs.
d"Particulate matter (PM)" control levels are shown in Table 5-2.
eGood Acid Gas control reduces emissions through the use of dry sorbent injection.
fBest Acid Gas control reduces emissions through the use of spray dryers and fabric filters.
Source: Radian. 1989. Background Paper, Municipal Waste Combustors. Prepared for the
U.S. Environmental Protection Agency, Office of Air Quality Planning and Standards.
5-13
-------
TABLE 5-4 CONTROL OPTIONS BY NSPSs MODEL PLANT FOR EACH
REGULATORY ALTERNATIVE
Model Model
Plant Plant
# Type
1
2
3
4
5
6
7
8
9
10
11
12
MB/WB (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
I
Baseline
#1
#1
#1
#1
#1
#1
Baseline
#lAa
Baseline
#1
#1
Regulatory Alternative
HA IIB IH
Baseline
#2
#2
#2
#2
#2
#2
Baseline
#1A
Baseline
#2
#2
#2
#2
#2
#2
#2
#2
#2
#2
#2
#2
#2
#2
Baseline
#3
#3
#3
#3
#3
#3
Baseline
#1A
Baseline
#3
#3
IV
#2
#3
#3
#3
#3
#3
#3
#2
#2
#2
#3
#3
Control option #1A appears in parentheses in EPA (1989c).
Sources: U.S. Environmental Protection Agency. 1989c. Municipal Waste Combustors—
Background Information to Proposed Standards: lll(b) Model Plant Description and
Costs of Control. EPA-45013-89-275.
Radian Corporation. 1988b. Memorandum to U.S. Environmental Protection Agency
August 30.
5.4 COST AND EMISSION REDUCTION ESTIMATION
5.4.1 Scenario I:
No Substitution
In this scenario two assumptions are made: (1) model plants are representative of the
many segments of the MWC industry affected by the NSPSs and (2) decision makers in these
industry segments will, on average, respond to each regulatory alternative by modifying their
MWC plant in the manner described by the model plant cost report (EPA, 1989c). Figure 5-3
depicts the second assumption of Scenario I. Any model NSPSs plant would incinerate Qmwc
Mg of MSW per year at a baseline cost of CBmwc per Mg MSW. Upon promulgation of the
NSPSs, the plant owners would undertake the equipment and operating changes necessary to
meet the NSPSs as estimated in the cost report (EPA, 1989c). The associated capital and
operating costs incurred by the operator become the basis for estimation of the economic impact
of the NSPSs under Scenario I.
For example, if the expenditures necessary to meet Regulatory Alternative I at this model
plant increase the per-Mg MSW cost of operation to C^wc, the annualized cost of the regulation
(from the producer's point of view) is the cross-hatched area of Figure 5-3. If Regulatory
5-14
-------
Cost IIA
($/Mg) Cmwc
^i
mwc
mwc
Qmwc Annual Throughput
Existing Model Plant (Mg MSW/yr)
Figure 5-3. Municipal Waste Combustion Response under Scenario I:
No Substitution
5-15
-------
Alternative IIA requires more stringent control, the cost per Mg of MSW may rise to CIIAmwc,
and the incremental cost of Regulatory Alternative IIA relative to Regulatory Alternative I is the
shaded area of the diagram. The total cost of this regulation relative to the baseline is the sum of
the cross-hatched and shaded areas of Figure 5-3; it is this latter cost that we report as this plant's
contribution to the annualized cost of Regulatory Alternative IIA.
In Scenario I then, the assumption is that the MWC decision makers adopt the controls
(and experience the costs) specified in the cost report (EPA, 1989c). In economic terms this is
equivalent to specifying the absence of any more attractive substitution possibilities for meeting
the emission requirements of the NSPSs. We also assume a perfectly inelastic demand for the
services each model plant provides. This could be due either to inelastic demand per se or an
institutional arrangement that does not allow any increase in costs to be passed along directly to
waste collectors or, ultimately, to waste generators. Both are extreme assumptions when applied
to all affected plants and probably result in some overestimation of the costs of regulatory
alternatives under Scenario I. These same assumptions also allow us to use the model plant
emission reductions estimated in the cost report to estimate incremental and total emission
reductions for each regulatory alternative.
Radian provided basic capital cost, operating cost, and emissions data for the 12 model
plants under baseline and controlled conditions. Using these data, together with baseline
estimates of the amount of MWC capacity in each model plant category, we scaled model plant
costs and emission reductions to cost and emission reduction estimates for each model plant
category. The scale factors represented the number of model plants necessary to match the
amount of MSW estimated to be handled by existing plants in that model plant category.
Summing over the 12 model plant categories results in a national measure of cost or emission
reductions. We also used the scale factors to estimate the number of plants in each model plant
category as well as the total number of plants that will be subject to the NSPSs. Based on the
scaling factors are shown in Table 5-5, approximately 67 MWC plants will be subject to the
NSPSs.
5.4.2 Scenario II: MWC Substitution
When published, the NSPSs will be specified as emission limits, without reference to a
specific abatement technology. This permits owners of prospective NSPSs plants to consider
alternatives to the emission control approach selected in the engineering analysis, especially if
these alternatives reduce the cost of disposing of MSW. In Scenario H the model plant cost data
5-16
-------
TABLE 5-5. SCALING FACTORS USED TO OBTAIN SCENARIO I
NATIONAL COST ESTIMATES8
Model
Plant
#
1
2
3
4
5
6
1
8
9
10
11
12
Model
Plant
Type*
MB/WB (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
Model Plant
Capacity
(Mg/day)
180
730
2,040
450
950
1,810
1,810
220
45
90
820
820
Scenario I
Scaling
Factors'5'0
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
a Scenario III scaling factors differ with each regulatory alternative.
b These scaling factors are based on the annual operating hours reported in Table 3-3.
For 9 of the model plants these hours reflect the average capacity utilization, by plant
type, reported in the 1988-89 Resource Recovery Yearbook (Gould, 1988).
Allowance is made for increased downtime for model plants 1 and 9; and allowance is
also made for model plant 7 which co-fires 50 percent wood.
c These scaling factors are used to estimate the number of MWC plants subject to NSPSs
under Scenario I.
developed for the economic impact analysis could be used to to examine whether, for each model
plant, constructing an alternative MWC technology that meets the emission limits would be less
costly overall.
Such an approach was used for the MWC Guidelines economic impact analysis (RTI,
1989). We do not make quantitative impact estimates based on such a method in this impact
analysis for two reasons. First, new technologies such as RDF and FBC systems are much more
integral to the NSPSs baseline, and considerable uncertainty still surrounds their cost and
performance. Second, Scenario Ell impact estimates are to be quantified, and these estimates
embody the MWC technology substitution identified with Scenario II.
5.4.3 Scenario HI: MWC/Landfill Substitution
The point of departure for the Scenario HI analysis is a report by Bentley and Spitz
(1989) on solid waste disposal choice. The report used recent data (1980 to 1986) on actual
5-17
-------
choices to build either new landfills or three types of new MWC plants (mass burn, modular, and
RDF) to estimate a relationship between these choices and features of both the community and
technology. Bentley and Spitz used engineering cost data from the Radian retrofit study (EPA,
1989b), plant characteristics data from the Radian cost report (EPA, 1989c), and financial and
operating parameters developed by RTI for this analysis to estimate the cost of new MWC
technologies as a function of plant capacity. They obtained data on landfill characteristics from
an EPA landfill survey (EPA, 1988b) and landfill costs from EPA's Office of Solid Waste and its
draft regulatory impact analysis of RCRA Subtitle D regulations on municipal waste landfills
(Temple, Barker, and Sloan, Inc. et al., 1987). These data were combined with site information
on population density, educational attainment of residents, and manufacturing employment and
used to make econometric estimates of the likelihood of choosing one or the other solid waste
disposal technology given the cost, plant, and site characteristics.
Bentley and Spitz provided the results of their analysis to RTI in form of a spreadsheet
that uses the estimated choice equations to predict solid waste disposal technology with changing
costs of landfills and MWC technologies. RTI modified this spreadsheet by inserting estimates
of the average increase in landfilling costs due to Subtitle D regulations as a function of landfill
solid waste flows. These cost estimates are based on data contained in the draft economic impact
analysis for Subtitle D landfill regulations, adjusted for the financial parameters used in this
analysis for public ownership. The waste flow shares estimated for each of the disposal
technologies both before and after the inclusion of Subtitle D costs for landfills are shown in
Table 5-6. These shares for the landfill, mass burn, modular, and RDF disposal options are
denoted by s\, Sb', sm', and sr', respectively. RTI then added the estimates of the cost of control
for each regulatory alternative to the cost the MWC technologies and obtained revised estimates
of the share of waste flows processed by the four disposal options. These revised share estimates
are denoted si, Sb, sm, and sr. These changes in shares are used in Scenario in to estimate the
marginal change in solid waste disposal choice due to the NSPSs.
TABLE 5-6. ESTIMATED WASTE FLOW SHARES
Before Inclusion
Technology of Subtitle D Costs
Mass Burn
Modular
Refuse-Derived Fuel
Landfill
TOTAL
13.79%
5.94%
8.02%
72.25%
100%
After Inclusion
of Subtitle D Costs
15.24%
6.49%
8.94%
69.34%
100%
Source: Bentley, Jerome T. and William Spitz. 1989. A Model of
the MSW Choice Decision. Prepared for the U.S. Environmental
Protection Agency. Princeton, NJ: Mathtech Incorporated.
5-18
-------
In order to use this share information, however, one needs estimates of the baseline level
of affected waste flows for both MWCs affected by NSPSs and landfills that would filled in the
same time frame. We obtained the MWC waste flows from the baseline allocation of flows to
model plants displayed in Table 5-1 and derived from information on planned MWC plants. We
obtained estimates of landfill waste flows subject to disposal choice from Westat's analysis of
the remaining capacity and waste flow rates obtained from the landfill survey data (EPA, 1988b).
If a landfill was scheduled to be filled in the period of analysis (1992-1996), we assumed that
disposal of that amount of waste was subject to a new waste disposal choice. These estimated
waste flows are shown in Table 5-7. We denote these initial values of waste flows subject to a
disposal choice by Qin for landfills and Qbn, Qmn, and Q™ for the respective combustor
technologies.i When summed, these baseline waste flows subject to disposal choice equal Qn
(72.5 million Mg).
TABLE 5-7. ESTIMATED WASTE FLOWS SUBJECT TO DISPOSAL CHOICE BY
TECHNOLOGY AND REGULATORY ALTERNATIVE (10* Mg/yr)
Technology Baseline
Regulatory Alternative
I
HA
Before Adjustment for Constant Total Waste Flow
Mass Burn 9.03 8.96 665
Modular 0.43 0.39 046
RDF 3.86 3.77 220
Landfill 59.14 59.93 65.43
TOTAL
72.45
73.05
After Adjustment for Constant Total Waste Flow
Mass Burn 9.03 8.89
Modular 0.43 0 38
RDF 3.86 3.74
Landfill 59.14 59.44
TOTAL
72.45
72.45
74.74
72.45
HB
6.61
0.46
2.20
65.52
74.78
72.45
in
5.87
0.49
1.84
66.91
75.10
72.45
rv
5.83
0.48
1.84
66.99
75.15
6.45
0.44
2.13
63.43
6.40
0.44
2.13
63.48
5.67
0.47
1.77
64.55
5.62
0.47
1.77
64.59
72.45
1
5-19
-------
Using ratios of the waste flow share estimates derived from Bentley and Spitz's choice
equations, we then made an initial adjustment in the components of waste flows subject to
disposal choice. Equation (5.4) shows this adjustment and the sum of initially adjusted values,
On'-
Qln (si'Vsi1) + Qbn (Sb'/Sb) + Qmn (sm7smf) + Qm (sr"/sr') = On' (5.4)
As this equation shows, application of these shares adjusts waste flows going to new
landfills or MWC plants in proportion to the change in share estimated by Bentley and Spitz's
choice analysis. If, for example, the costs of a regulatory alternative increase the choice of
modular combustors so that the modular share of waste flows in the historic data set increased,
the modular combustor's share of waste flows subject to disposal choice in the period of the
NSPSs analysis is estimated to increase proportionately.
One difficulty with this particular method is that Qn' (the initially adjusted value) does not
necessarily equal Qn (the baseline flows subject to disposal choice). Since we assume that all the
solid waste that was subject to a new disposal choice has to be disposed of in some way, we
make a final adjustment by multiplying the waste flows estimated by using the share ratios s"/s'
of Equation (5.4) by the ratio Qn/Qn'- These revised waste flows are the estimated waste flows
for each technology used in Scenario HI. We estimated a new set of waste flows for each
regulatory alternative based on the new waste flow share estimates derived when the estimated
costs of control for that regulatory alternative are introduced into the choice equations. These are
denoted Ojj, Qw, Qmj, and CM for Regulatory Alternative I; QmA, QbllA, QmllA, and OJ-HA for
Regulatory Alternative IIA; etc. These estimated waste flows for each technology and each
regulatory alternative are also shown on Table 5-7.
In the final steps of Scenario IE, RTI took the estimates of waste flows for each MWC
technology and allocated them to model plants using that same MWC technology. We made this
allocation proportional to the Scenario I baseline waste flows for that technology going the
particular model plant. For example, Scenario in mass burn waste flows were allocated to mode
plants that used mass burn technology and the waste flows estimated for a given model plant
were based on that model plant's share of mass burn waste flows in the Scenario I baseline.
These flows were then used to estimate the number of model plants under each regulatory
alternative of Scenario in (see Table 5-8). We then applied the cost and emission reduction
estimates for the model plants to obtain the national cost and emission reduction estimates for
Scenario III.
5-20
-------
TABLE 5-8. NSPSs MODEL PLANT NATIONAL WASTE FLOWS AND SCALING
FACTORS a,b
Scenario I
Model Plant
Model Plant Model Plant Waste Flow Scaling
Number Description (Mg/day) Factor
Scenario III
Total
Waste Flows Scaling
(106 Mg/yr) Factor
Total
Waste Flows
(106 Mg/yr)
Regulatory Alternative I
1
2
3
4
5
6
7
8
9
10
11
12
Regulatory
1
2
3
4
5
6
7
8
9
10
11
12
Regulatory
1
2
3
4
5
6
7
8
9
10
11
12
MB/WW (small)
MB/WW (mid- size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
Alternative II A
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
Alternative I IB
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
104
615
1,730
384
807
1,500
750
178
25.9
74.1
678
678
104
615
1,730
384
807
1,500
750
178
25.9
74.1
678
678
104
615
1,730
384
807
1,500
750
178
25.9
74.1-
678
678
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
16.54
7.16
8.36
3.18
3.18
5.22
3.21
3.01
1.62
6.39
1.85
4.07
63.81
11.99
5.20
6.06
2.31
2.31
2.98
1.83
3.49
1.87
7.41
2.15
4.72
52.32
11.91
5.16
6.02
2.29
2.29
2.98
1.83
3.47
1.86
7.37
2.14
4.70
52.03
0.62
1.61
5.27
0.45
0.94
2.86
0.88
0.20
0.02
0.17
0.46
1.01
14.48
0.45
1.17
3.82
0.32
0.68
1.63
0.50
0.23
0.02
0.20
0.53
1.17
10.72
0.45
1 16
A • A. \J
3.80
0.32
0.68
1.63
0.50
0.23
0.02
0.20
0.53
1.16
10.67
CONTINUED
5-21
-------
TABLE 5-8. NSPSs MODEL PLANT NATIONAL WASTE FLOWS AND SCALING
FACTORS a,b
Scenario I
Model Plant
Model Plant Model Plant Waste Flow
Number Description (Mg/day)
Scaling
Factor
Scenario III
Total Total
Waste Flows Scaling Waste Flows
(106 Mg/yr) Factor (106 Me/yr)
Regulatory Alternative HI
1
2
3
4
5
6
7
8
9
10
11
12
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
TOTAL
104
615
1,730
384
807
1,500
750
178
25.9
74.1
678
678
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
10.54
4.57
5.33
2.03
2.03
2.48
1.52
3.68
1.98
7.82
2.26
4.98
49.22
0 40
V/«"W
1 02
A ,\J±*
3.36
0.28
0.60
1 36
A • *J\J
0.42
0.24
002
\j * \j+*
0.21
0.56
1.23
9.70
Regulatory Alternative FV
1 MB/WW (small) 104
2 MB/WW (mid-size) 615
3 MB/WW (large) 1730
4 MB/REF 384
5 MB/RC 807
6 RDF 1500
7 RDF/CF 750
8 MOD/EA 178
9 MOD/SA (small) 25.9
10 MOD/SA (mid-size) 74.1
11 FBC/BB 678
12 FBC/CB 678
TOTAL
16.81
7.28
8.49
3.24
3.24
5.39
3.31
3.35
1.80
7.13
2.06
4.54
66.64
0.64
1.63
5.36
0.45
0.95
2.95
0.91
0.22
0.02
0.19
0.51
1.12
14.95
10.46
4.53
5.29
2.01
2.01
2.47
1.52
3.66
1.97
7.78
2.25
4.96
48.93
0.40
1.02
3.34
0.28
0.59
1.35
0.42
0.24
0.02
0.21
0.56
1.23
9.65
a These scaling factors are based on the annual operating hours reported in Table 3-3 For 9 of
the model plants these hours reflect the average capacity utilization, by plant type, reported
in the 1988-89 Resource Recovery Yearbook (Gould, 1988). Allowance is made for
increased downtime for model plants 1 and 9; and allowance is also made for model plant 7
which co-fires 50 percent wood.
b These scaling factors are used to estimate the number of MWC plants subject to NSPSs.
5-22
-------
CHAPTER 6
COST AND PRICE IMPACTS
This chapter presents the estimated costs of each regulatory alternative and scenario for
each of the model plants and for the nation as a whole. These costs are based on engineering
control cost estimates for 12 model plants that represent those plants affected by the NSPSs
(EPA, 1989c). The national cost estimates are computed from both an enterprise and a social
perspective.
6.1 MODEL PLANTS AND THE COST OF REGULATORY ALTERNATIVES
Table 6-1 lists the estimated additional capital and annual operating costs required by the
12 model plants to meet the emissions limits of each regulatory alternative. Also presented are
the associated present value of costs, annualized costs, and costs per Mg of MSW. The cost data
presented here are used directly to compute the cost of regulatory alternatives under Scenarios I
and HI. Costs are calculated based on two 15-year operating cycles for control equipment for all
scenarios. This treatment assumes that the plant as a whole has an economic life of 30 years but
that the APCD equipment has an economic life of 15 years. The operating costs are annual
values based on the capacity utilization specified for the model plant. The operating costs
include a credit for revenues when there is energy recovery during combustion. The present
value of costs, the annualized costs, and the costs per Mg of MSW in Table 6-1 are based on
public revenue bond financing of control expenditures by a public entity. The basis for the
associated discounting and annualization procedures is discussed in Chapter 3. Costs in the table
are zero when the model plant is small enough to qualify for exemption from emission limits or
was originally designed to meet these limits.
In looking at these costs, note again that, while values reported on these tables are
internally consistent, they are not necessarily comparable to costs reported in other studies. The
basis for computing costs per Mg of MSW, for example, varies greatly in the literature.
Differences include different base year dollars, nominal vs. real dollar flows, conventions for
treatment and timing of cost and revenue categories, and scope of the analysis. However, since
costs used in this report have been computed consistently, decision makers can use them to
represent the relative economic attractiveness of the model plants when making choices among
plants and technologies. The only qualification, an important one, is the assumption that the cost
and credit data for the model plants used in this impact analysis are representative of similar
MWC plants.
6-1
-------
TABLE 6-1. NSPSs ENTERPRISE COSTS OF CONTROL FOR PUBLICLY OWNED
MODEL PLANTS (1987 $)»
Model Model Plant
Plant Description
#
Regulatory Alternative I
1
2
3
4
5
6
7
8
9
10
11
12
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
PV of Annual
Capital Operating
Costb CosF
($103) ($103/yr)
(1)
0
747
2,100
964
840
1,930
1,930
0
827
0
0
0
(2)
0
39.0
114
50.0
35.2
109
49.8
0
72
0
0
0
PV of Total
Control
Costd
($103)
(3)
0
1,420
4,070
1,830
1,450
3,810
2,790
0
2,070
0
0
0
Total Annualized
Annualized Cost per
Cost6 Mg MSWf
($l
-------
TABLE 6-1. NSPSs ENTERPRISE COSTS OF CONTROL FOR PUBLICLY OWNED
MODEL PLANTS (1987 $)a (CONTINUED)
Model
Plant
#
Model Plant
Description
Regulatory Alternative III
1
2
3
4
5
6
7
8
9
10
11
12
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size
FBC/BB
FBC/CB
PVof
Capital
Costb
Annual
Operating
Costc
($103) ($103/yr)
(1)
0
8,870
18,300
8,820
11,500
19,700
19,700
0
532
0
8,560
8,560
(2)
0
1,450
3,260
1,310
1,900
3,630
1,480
0
72
0
1,340
1,340
PVof Total
Control
Costd
($l
-------
When choices arise, decision makers will be guided by the cost of choices as they see
them. These are termed private or business enterprise costs even if a public entity is doing the
accounting. As discussed in Chapter 3, this viewpoint will differ slightly for private and public
entities' decision makers given their different financial environments—particularly their tax
liabilities and opportunity costs of capital. Using the same cash flow model and financial
parameters as in Chapter 3, the present value and per-Mg cost are computed for public and
private plants. To illustrate how decision makers with different perspectives evaluate the control
costs, Table 6-2 provides results for public and private financial environments for Regulatory
Alternative IV.
The data of Table 6-2 show how an average private firm would view the two measures of
the cost of control, after allowance for tax effects, of passing the full cost along to the customer,
for the different model plants. Similarly, the data show how an average public entity would view
the cost of control for different model plants given the public entities' cost of capital and lack of
tax obligation. These measures provide a good basis for a single entity to compare different
investment choices having the same revenue effects or meeting the same regulatory requirement.
For example, the NPV measure can be the basis for a private firm's choice between building a
new MWC (with an associated NPV or cost per Mg) or modifying a current MWC at the cost
shown in Table 6-2.
The data of Table 6-2 also allow one to compare the cost of control between privately
owned and publicly owned MWCs. Such comparisons, however, should be made with care
because these entities have very different financial conditions. The NPV of total control costs
are lower for private ownership for all model plants but the relative difference between public
and private ownership varies from one model plant to another. These variations arise from a
combination of differences in the cost of capital and tax obligations for privately owned and
publicly owned MWCs and differences from one model plant to another in the share of capital
and operating costs for APCD equipment.
The annualized costs per Mg of MSW, however, show that the average publicly owned
MWC, with its financial conditions, would be able to meet the regulations at a lower cost
(tipping fee increase), than a privately owned MWC in any model plant category. This measure
is equivalent to asking, "What must the public entity or private firm receive per unit of waste
disposed if it is to cover the costs experienced when investing in the pollution control
equipment?" For the range of model plant capital and operating costs of control considered here,
the cost of control per Mg of MSW ranges from 12 to 43 percent higher for private ownership
6-4
-------
TABLE 6-2. NSPSs ENTERPRISE COSTS OF CONTROL FOR PUBLICLY AND PRIVATELY OWNED MODEL
PLANTS: SCENARIO I, REGULATORY ALTERNATIVE IV*
Public
Model Model Plant
Plant Description
Capital
Cost
($103)
(1)
Operating
Cost
($103/yr)
(2)
PV of Total
Control
Costc»e
($103/yr)
(3)
Annualized
Cost per
Mg/MSW«.«
($/Mg)
(4)
Private
PV of Total
Control
Costd'«
($103/yr)
(5)
Annualized
Cost per
Mg/MSW<>>
($/Mg)
(6)
ON
1 MBAVW (small)
2 MB/WW (mid-size)
3 MB/WW (large)
4 MB/REF
5 MB/RC
6 RDF
7 RDF/CF
8 MOD/EA
9 MOD/SA (small)
10 MOD/SA (mid-size)
11 FBC/BB
12 FBC/CB
1100
8,870
18,300
8,820
11,500
19,700
19,700
1,260
1,400
1,260
8,560
8,560
540
1,450
3,260
1,310
1,900
3,630
1,480
460
322
347
1,340
1,340
11,000
38,900
84,900
36,400
50,800
93,400
56,100
9,910
7,750
7,960
36,500
36,500
16.90
10.00
7.78
15.00
9.97
9.86
11.90
8.83
47.40
17.00
8.54
8.54
8,100
32,600
70,300
30,900
42,500
76,900
52,600
7,490
6,190
6,220
30,800
30,800
19.00
12.90
9.89
19.60
12.80
12.50
17.10
10.20
58.20
20.40
11.10
11.10
,.,.for Privately and publicly owned facilities are due to differences in the cash flow models for
differences in cash flow model parameters, especially the discount rate. See Chapter 3 for a
°Peratin g
of thse differences
bCapital costs presented here are for one APCD equipment cycle only
consecmive' 15"year APCD equipment cycles- 3°-year plant life- and lhe cash flow modd fOT
eThe NPV of total control costs and the annualized costs are estimates of the revenue required, in present value and annualized terms
respectively, by the public and private entities if tipping fees are to cover all of the cost. .wnuaiizea terms,
-------
than for public ownership (the average difference is 25 percent). As in the case of NPV values,
variations in the percentage differences in per-Mg MSW cost are due in part to the relative mix
of capital and operating cost—as the capital cost increases relative to operating cost, the per-Mg
cost differences shift in favor of public ownership.
These data suggest that, in situations where public and private entities compete to provide
disposal of MSW by combustion, imposing the NSPSs would substantially favor public
ownership. As noted above, this result must be interpreted with caution for the following
reasons. First, it doesn't reflect any differences in the productivity that may exist for public and
private owners of MWC plants. For example, private owners may be more efficient in installing
and operating controls by virtue of multiple project experience. Second, the measured
differences are also the product of the average financial parameters used to represent public and
private ownership. The differences in per-Mg cost of MSW between public and private
ownership are not so large that they are outside the range in variation associated with financial
parameters of public entities and private firms. For example, Hilgendorff (1989) notes the
backlog of MWC projects that qualify for the more favorable pre-1986 Tax Reform Act tax
treatment under "grandfather" provisions of the Act. Even so, these cost-per-Mg data appear to
corroborate the contention that, in the future, publicly owned MWC facilities will continue to be
the norm.1 The costs of meeting the NSPSs probably amplify the public ownership advantage in
most cases.
Because of this finding and the already prominent role of public ownership of current
MWC plants, cost impacts of the NSPSs are presented in the next two sections of this report
using the cash flow model for public financing of control equipment. A discussion of social cost
impacts follows later in this chapter.
Table 6-3 presents estimates of baseline costs per Mg MSW estimated for the model
plants (based on capital and operating costs of the model plant prior to the NSPSs) and the per-
Mg MSW cost for each regulatory alternative wherein the per-Mg MSW cost of APCD capital
equipment and operation are added to the baseline costs per Mg of MSW. These data provide
measures of the size of the costs of regulatory alternatives for each model plant relative to both
the baseline cost and other regulatory alternatives. These same relative costs apply to Scenario
III, only for fewer plants.
1Hilgendorff (1989) notes the existence of a large number of planned plants that qualify as private participants for
advantageous tax treatment under grandfather provisions of the 1986 Reform Act For this group of plants,
private firm accounting cost would be lower than portrayed in Table 6-2.
6-6
-------
TABLE 6-3. NSPSs ENTERPRISE COSTS FOR PUBLICLY OWNED MODEL PLANTS3 (1987$): SCENARIO I COST
PER Mg OF MUNICIPAL SOLID WASTE AND PERCENTAGE CHANGES IN COST OVER THE
BASELINE FOR EACH REGULATORY ALTERNATIVE
o\
Regulatory Alt. I
Model
Plant
#
1
2
3
4
5
6
7
8
9
10
11
12
Model
Plant
Description
MB/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
Baseline
Cost per
MgMSWb
($/Mg)
75.40
20.70
9.40
44.00
21.30
3.58
17.70
30.40
54.20
31.20
13.20
13.20
Cost
per
MgMSWc
<$/Mg)
75.40
21.10
9.77
44.70
21.50
3.98
18.30
30.40
66.90
31.20
13.20
13.20
Percentage
Change
over
Baseline
0
1.77
3.97
1.72
1.34
11.2
3.33
0
23.4
0
0
0
Regulatory Alt. HA
Cost
per
MgMSWc
($/Mg)
75.40
27.70
15.10
53.40
28.00
10.80
25.00
30.40
66.90
31.20
17.00
19.50
Percentage
Change
over
Baseline
0
33.7
60.6
21.3
31.8
201.0
40.9
0
23.4
0
29.5
47.9
Regulatory Alt. IIB
Cost
per
Mg MSWC
($/Mg)
92.30
27.70
15.10
53.40
28.00
10.80
25.00
39.30
102.00
48.30
17.00
19.50
Percentage
Change
over
Baseline
22.4
33.7
60.6
21.3
31.8
201.0
40.9
29.0
87.4
54.5
29.5
48.0
CONTINUED
-------
00
TABLE 6-3. NSPS ENTERPRISE COSTS FOR PUBLICLY OWNED MODEL PLANTS8 (1987$): SCENARIO I COST
PER Mg OF MUNICIPAL SOLID WASTE AND PERCENTAGE CHANGES IN COST OVER THE
BASELINE FOR EACH REGULATORY ALTERNATIVE (CONTINUED)
Regulatory Alt. Ill
Model
Plant
#
1
2
3
4
5
6
7
8
9
10
11
12
Model
Plant
Description
M8/WW (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
Baseline
Cost per
Mg MSWb
($/Mg)
75.40
20.70
9.40
44.00
21.30
3.58
17.70
30.40
54.20
31.20
13.20
13.20
Cost
per
Mg MSWC
(VMg)
75.40
30.70
17.20
59.00
31.20
13.40
29.60
30.40
66.90
31.20
21.70
21.70
Percentage
Change
over
Baseline
0
48.5
82.8
34.1
46.9
275.0
66.9
0
23.4
0
64.9
64.9
Regulatory Alt. IV
Cost
per
Mg MSWC
($/Mg)
92.30
30.70
17.20
59.00
31.20
13.40
29.60
39.30
102.00
48.30
21.70
21.70
Percentage
Change
over
Baseline
22.4
48.5
82.8
34.1
46.9
275.0
66.9
29.0
87.4
54.5
64.9
64.9
a Costs based on annualization of control costs over 30 years with a real discount rate of 4 percent
b Baseline costs are computed from capital costs annualized over 30 years, with a real discount rate of 4 percent, plus operating costs.
c Costs for regulatory alternatives are baseline costs plus the cost of control relative to the baseline.
-------
6.2 NATIONAL ENTERPRISE COSTS OF EACH REGULATORY ALTERNATIVE
Table 6-4 lists the national enterprise cost impacts estimated for each regulatory
alternative under Scenarios I and III based on public ownership of all the plants and the
aggregation method described in Chapter 5. The capital cost is the initial equipment and
installation cost for APCDs plus the discounted cost of these expenditures for replacement in the
15th year of operation. Estimated annualized cost and costs per Mg of MSW use the cash flow
model described in Chapter 3 and the accounting and amortization procedures of Section 6.1.
TABLE 6-4. NSPSs NATIONAL COST IMPACTS: ENTERPRISE COSTS FOR
PUBLICLY OWNED MODEL PLANTS (1987 $)
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative I1A
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
Scenario III
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative III
Regulatory Alternative IV
Present
Value of
Capital
Costs
($106)
41.4
285
336
800
852
46.2
188
232
499
540
Annualized
Costs3
($106/yr)
5.15
89.6
106
128
145
5.02
62.5
75.8
80.5
93.0
Annuaiized
Costs
per Mg MSWb
($/Mg)
0.37
6.45
7.10
9.23
9.69
0.37
6.35
7.11
9.09
9.65
a Costs based on annualization of control costs over 30 years with a real discount rate of 4 percent.
b Computed by dividing total annualized cost by the estimated amount of MSW processed per
year. r
The capital cost estimates show that these regulations will require a substantial initial
financial commitment on the part of affected MWCs and that the regulatory alternatives differ
substantially in the financial commitment required. If roughly 70 plants will be affected, the
6-9
-------
average capital cost under Scenario I ranges from $0.44 million per plant for Regulatory
Alternative I to $7.8 million per plant for Regulatory Alternative IV. The capital costs for
Scenario III are lower than for Scenario I because fewer MWC are projected to be built due to
the NSPSs. The capital cost reduction under the different regulatory alternatives ranges from 2.6
percent under Regulatory Alternative I to 37 percent under Regulatory Alternative IV. We note
again, however, that for Scenario in we haven't included in these cost data either the total or
incremental capital cost of new landfills that would be needed to manage the solid waste that,
without the NSPSs, would have been managed by new MWCs.
Annualized costs cover operating costs and amortized capital for a 30-year period. For
Scenario I they exceed $100 million per year for all Regulatory Alternatives more stringent than
IIA. With their annual impacts, these regulatory alternatives qualify as major regulations under
Executive Order 12291. Annualized costs for Scenario HI range from $5 million to $93 million
per year. Again, costs for landfills that substitute for some of the combustors due to the NSPSs
are not included in these data.
Table 6-4 shows the estimated average annualized enterprise costs per Mg of solid waste
combusted by plants that have to install APCDs under a given regulatory alternative. If the
regulatory alternative does not require that a MWC plant incur costs to meet the regulation, then
the plant's waste stream is not included in the denominator used to compute cost per Mg. These
data are discussed further below.
6.3 PRICE IMPACTS
The NSPSs will increase the cost of operating most MWC plants. The amount of this
cost that is passed on in the form of higher tipping fees (prices) to waste collectors and by
collectors, in turn, to waste generators in the form of higher collection fees, is determined by the
institutional and market conditions prevailing in the MSW service area. In this section of the
report we discuss several variations in institutional and market conditions that result in very
different "price" impacts. In each instance we use enterprise costs to estimate these impacts
because these are the costs estimated for the individual firm or government entity.
In circumstances in which different disposal technologies compete with one another
without institutional constraints, the price of waste disposal will be established by the least-cost
disposal option. For a privately owned plant, the plant would only be built if combustion was
still the least-cost way to provide waste disposal services after the NSPSs and if the costs could
be fully passed on to waste collectors. Thus for privately owned plants in a competitive market
6-10
-------
either all control cost are passed to consumers in the form of a price or tipping fee increase or the
plant is not built and no costs of control for MWC plants are passed on. Most of this increase in
tipping fees will, in turn, be passed on to waste generators if there is a highly inelastic demand
for waste collection (as postulated in Chapter 2); that is, all collectors experience the same
increase in tipping fees and the supply of waste collection is relatively elastic. In such a case, the
price increases would, on average, be roughly equal to the cost-per-Mg values of Table 6-4.
For a publicly owned MWC plant, whether the MWC plant will be built and how the
control costs will be allocated to price is a public policy decision. Notwithstanding the limits
placed on it by the terms of any revenue bond financing arrangements, a public entity often has
the option of covering some or all of the costs of combustion using tax revenue or fees derived
from other services. If combustion is not the least-cost disposal alternative after the NSPSs, the
price of the least-cost competing disposal technologies will probably set an upper bound on the
extent to which NSPSs control costs can be passed along in the form of price increases. If
combustion is the least cost solid waste disposal alternative, then the public entity may pass all or
part of the cost of control along in the form of a price increase. Recall again, however, that
because a public entity does not necessarily have to cover costs incurred in provision of a
particular line of business, prices and costs are not necessarily linked closely with one another.
Price impacts will also vary because of the differences in costs of control for different
MWC plants. Table 6-1 shows the range in control cost per Mg from model plant to model plant.
For any given regulatory alternative, the individual model plant can have control costs from zero
to many hundreds of percent above the average cost per Mg.
To provide some perspective on potential price increases, we compare the costs per Mg
with average tipping fees reported for resource recovery facilities by Pettit (1989). The average
tipping fee for resource recovery facilities in 1988 was $42.70 per Mg (1987 dollars). Pettit
notes these important features of that figure:
• It is based on "gate fees" (municipality and/or contract waste may be charged lower
tipping fees). 3 6
• Those facilities that reported zero tipping fees were excluded from the average.
Furthermore, Pettit notes that the average tipping fee has risen sharply over the past 7 years—
68 percent since the average was first computed for 1982. This dramatic increase is likely due to
some combination of both the introduction of new, more costly plants into the sample and a
change in the pricing policies of owners and operators.
6-11
-------
Table 6-5 provides the average percentage increases in tipping fees for scenario and
regulatory alternative combinations, assuming that all the estimated increase in enterprise cost
per Mg of MSW is passed through to the waste collector. The average percentage increase in
tipping fees per Mg increases with the regulatory alternative and is significant beyond Regulatory
Alternative I. The percentage increases for individual model plants vary considerably with both
the model plant and regulatory alternative. This can also be seen by reviewing Table 6-3. There
is wide variation in the percentage increases in cost per Mg associated with each regulatory
alternative. Admittedly, much of this variation is due to the variations in the baseline cost per
Mg used, but enough is attributable to the control cost itself to indicate that the price impacts of
the NSPSs will vary widely depending on the size and technology of the MWC plants.
TABLE 6-5. PERCENTAGE PRICE INCREASES BASED ON FULL PASS THROUGH
OF ESTIMATED NSPSs ENTERPRISE COSTS OF CONTROL PER Mg
OF MUNICIPAL SOLID WASTE3
Regulatory Regulatory Regulatory Regulatory Regulatory
Alternative Alternative Alternative Alternative Alternative
Scenario I HA KB HI IV
Scenario I
Scenario III
0.86
0.86
15
15
17
17
22
21
23
23
a Based on average resource recovery facility tipping fee for 1988 of $43.96 per Mg (Pettit,
1989), converted to last quarter 1987$ to an average $42.70 per Mg.
In situations where there is little competition among waste disposal technologies or waste
disposal firms, attempts to translate cost changes into price changes become even more
problematic. Usually, disposal of solid waste in such non-competitive situations rests with a
public entity. In this case the price increase will more than ever depend on how the public entity
allocates costs.
6.4 SOCIAL COSTS
To calculate the social cost of a regulation, we recast the private cost data and insert them
into a social opportunity cost framework. While, in principle, a multiplicity of issues account for
differences between private and social costs, these issues usually center on selecting an
appropriate discount rate and measuring social losses due to quantity adjustments.
6-12
-------
Over the past several decades, views have ranged widely regarding the appropriate
discount rate to use when evaluating a public project or estimating the economic impact of a
government program. A recent discussion of the issues involved is presented in Lind (1982).
While no particular approach commands the complete support of economists, a number of
prominent economists agree that a recently introduced set of principles for discounting when
seeking social measures of costs or benefits is a step in the right direction. This position is
represented in two recent papers (Kolb and Scheraga, 1988; Arnold, 1986). In summary, these
papers direct the analyst to employ a "two stage" or two discount rate procedure. A capital rate
of discount is used to annualize any capital expense, and a consumption rate of interest (discount)
is used to determine the present value of the annual expenditures, as well as the annualized
capital cost. This procedure is used to estimate national social costs in this chapter. Because the
procedure and the discount rates employed are still controversial, Chapter 9 includes results
using single discount rates of 10 and 3 percent.
The papers provide only modest guidance on choosing the appropriate rates of
consumption and capital discount. The general consensus is that the consumption rate of interest
is lower than the capital discount rate. The consumption rate of interest has been estimated by
various authors to be in the 0 to 6 percent range; the capital discount rate in the 8 to 13 percent
range. This analysis uses a consumption rate of interest of 3 percent and a capital discount rate
of 10 percent. The resulting social cost estimates for each scenario and regulatory alternative are
shown in Table 6-6.
Because of the discounting procedure, the annualized and per-Mg social costs are 10 to
20 percent higher than comparable enterprise costs. The annualized cost of Regulatory
Alternative IIA under Scenario I is now nearly $100 million and Regulatory Alternative IV under
Scenario III is a "major" regulation using national social costs.
6-13
-------
TABLE 6-6. NSPSs NATIONAL COST IMPACTS (1987 $)
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
Scenario HI
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative IE
Regulatory Alternative IV
PVof
Social
Capital
Costs
37.8
227
268
638
676
36.9
150
185
398
430
Annualized
Social Costsb
($10«/yr)
6.41
97.2
115
150
168
6.26
67.6
82.0
93.8
107
Annualized
Social Costs
per Mg
MSWM
($/Mg)
0.46
6.99
7.70
10.80
11.20
0.46
6.86
7.69
10.60
11.10
Annualized
Enterprise
Costs per
Mg MSWc>d
($/Mg)
0.37
6.45
7.10
9.23
9.69
0.37
6.35
7.11
9.09
9.65
' ' ""-- "" •' """' "" ' • gi - ,_
a Present value of capital costs are based on 2 consecutive, 15-year life cycles for APCD
equipment over the 30-year plant life. These assumptions make no difference in the
annualized cost impacts (Robson, 1989).
b Annualized social costs are the sum of capital costs, annualized at 10 percent, and annual
operating costs.
c Annualized public enterprise costs are the sum of capital costs, annualized at 4 percent and
annual operating costs. '
d Computed by dividing total annualized cost by the estimated amount of MSW processed oer
year. * v
6-14
-------
CHAPTER 7
EMISSION REDUCTIONS AND COST-EFFECTIVENESS
MWC air emissions are changed, usually reduced, with the increased costs MWCs incur
with the application of air pollution control devices (APCDs). For Scenario I the emission
reductions are computed as the difference between estimates of baseline emissions from MWC
plants and estimates of emissions after the installation of the APCDs on those plants. Emission
reductions for Scenario III are computed differently because, as the controls called for by the
NSPSs become more costly, fewer new MWC plants are built. The emission reductions for
Scenario III computed in this chapter are the difference between baseline emissions from MWCs
that will be built, which will be fewer than planned, and emissions from the same plants after the
NSPSs. Two alternative ways of computing emission reductions for Scenario III are to compute
them relative to the Scenario I baseline (in which case they would be larger than reported here)
or, for a more comprehensive perspective, to compute them relative to the Scenario I baseline but
to add back in emissions from landfills that substitute for the new MWC plants in Scenario III.
This chapter provides estimates of these reductions for six air emissions and solid waste,
describes how they might be combined with cost data to obtain cost-effectiveness estimates, and
notes the problems in comparing the cost-effectiveness of regulatory alternatives when emissions
are denominated in different units (e.g., Mg of SO2 vs Mg of HC1).
Scenario I energy impacts are also presented in this chapter. Energy impacts are
computed as the difference between estimated baseline energy usage and energy usage after the
regulation. Electrical energy usage is generally higher after the installation of APCD equipment.
No change in gas usage is projected for any of the regulatory alternatives.
7.1 EMISSION REDUCTIONS AND ENERGY IMPACTS
Radian (EPA, 1989c) estimated emission reductions and energy impacts associated with
use of APCDs for each model plant. APCDs control several pollutants both in the baseline and
under the regulatory alternatives, including
• polychlorinated dibenzo-/?-dioxins and dibenzofurans (CDD/CDF),
• carbon monoxide (CO),
• paniculate matter (PM),
• hydrogen chloride (HC1),
• sulfur dioxide (S02),
• lead (Pb), and
• ash.
7-1
-------
Table 7-1 presents estimated changes in national emissions for each pollutant associated with
each of the regulatory alternatives under Scenarios I and III. Table 7-2 presents energy usage
impacts associated with each of the regulatory alternatives under Scenario I. We compute these
data using the same strategy of scaling model plant emissions, emission reductions, and energy
impacts to national emissions, emission reductions, and energy impacts as that employed in
estimating the national costs for these regulatory alternatives.
TABLE 7-1. NSPSs NATIONAL BASELINE EMISSIONS AND EMISSIONS
REDUCTIONS (Mg per Yr.)
Scenarios and
Regulatory CDD/CDF CO PM SO2 HC1 Pb Ash«>
Alternatives
Scenario I
Baseline Emissions 0.0152 5,470 7,540 42,000 49,300 127 3,700,000
Emissions Reductions
Regulatory Alternative I 0 0 5,220 0 0 88.5 -154,000
Regulatory Alternative HA 0.0107 0 5,220 18,300 36,700 108 -383,000
Regulatory Alternative HB 0.0115 0 5,960 19,300 39,500 124 -401,000
Regulatory Alternative HI 0.0139 0 5,220 35,400 44,400 108 -314,'oOO
Regulatory Alternative IV 0.0146 0 5,960 36,400 47,200 124 -332,'oOO
Scenario III
Baseline Emissions^
Emissions Reductions
Regulatory Alternative I 0 0 5,120 0 0 86.3 -138,000
Regulatory Alternative HA 0.00781 0 3,440 13,700 25,700 77.9 -30s!oOO
Regulatory Alternative HE 0.00841 0 4,040 14,500 27,900 90.3 -322,'oOO
Regulatory Alternative HI 0.00916 0 2,980 22,400 27,700 68.6 -232/XX)
Regulatory Alternative IV 0.00974 0 3,550 23,100 29,900 81.6 -246,000
a Scenario IE baseline emissions vary with each regulatory alternative because the level of total
waste flows handled by MWC varies with each regulatory alternative under Scenario HI.
b Includes bottom ash and fly ash with some residual quench water. Negative values reflect
increases in ash emissions relative to the baseline.
KEY: polychlorinated dibenzo-p-dioxins and dibenzofurans (CDD/CDF), carbon monoxide
(CO), paniculate matter (PM), sulfur dioxide (SO2), hydrogen chloride (HC1), and lead (Pb).
7-2
-------
TABLE 7-2. NSPSs NATIONAL ENERGY IMPACTS3
Scenario and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative I1A
Regulatory Alternative IIB
Regulatory Alternative III
Regulatory Alternative IV
Electrical
Use
(Tj/yr)
52.9
501
570
687
755
Gas
Use
(Tj/yr)
0
0
0
0
0
a Energy impacts use refer to air pollution control only.
7.2 COST-EFFECTIVENESS
The cost-effectiveness of regulatory alternatives provides a measure of the cost per unit
of emission reduction associated with each regulatory alternative. These ratios are meant to
measure the marginal effectiveness of each regulatory alternative. So, first they must be
arranged in order of increasing levels of emission reductions. Then, both the numerator and
denominator of the cost-effectiveness ratio are calculated as the difference between costs (or
emissions) of the regulatory alternative being evaluated and the previously most stringent
alternative. To highlight the nature of the resulting cost-effectiveness ratios, some analysts refer
to them as "incremental" cost-effectiveness measures. In mathematical terms, the cost-
effectiveness ratio for regulatory alternative i and emission j is
,j.ij (7.1)
where C is cost and E is emissions.
This relationship has a number of other notable features. First, the presumption is made
that costs increase and emissions decrease as the regulatory alternatives become more stringent.
Thus, the level of costs under the next most stringent regulatory alternative is subtracted from the
level of costs under the regulatory alternative of interest and vice versa for emissions. The cost-
effectiveness measure, therefore, is positive under normal circumstances. The cost-effectiveness
measure is the cost per unit of emission reduction for that regulatory alternative— for example,
dollars per Mg of sulfur dioxide reduction.
Second, while this relationship can be applied to many different emissions, it has no
inherent ability to identify which cost is assignable to which pollutant. To do this, either the
7-3
-------
regulatory alternatives must be redefined and narrowed to apply to only one emission of interest
(all other emissions constant) or some way must be devised to find a common measure for
valuing the emissions. Applying the cost-effectiveness measure to each of the pollutants when
the APCDs are responsible for simultaneously controlling a number of pollutants, as in the case
of MWCs, assigns all the cost for control to the particular pollutant for which the cost-
effectiveness measure is computed. In this such circumstance, the cost-effectiveness measure
overestimates the cost-effectiveness (i.e., implies emission reduction is more costly than in fact it
is).
This obviously argues for a more comprehensive measure of cost-effectiveness—one that
addresses the joint effectiveness of control on multiple emissions. Economists argue that such a
comprehensive measure would be denominated in dollars, since the value of a commodity for
purposes of trade is denominated in dollars. Put another way, if the additional benefit of a given
emission reduction were measured in dollars, then the joint benefit of all emission reductions
would be the sum of benefits of the individual emission reductions. Unfortunately, such benefit
estimates are not available for many of the pollutants reduced by the NSPSs' regulatory
alternatives. For this reason the cost-effectiveness estimates for individual pollutants are not
presented at this time.
7-4
-------
CHAPTER 8
ECONOMIC IMPACT ON SECTORS OF THE ECONOMY
The costs and emission reductions estimated for each of the regulatory alternatives will
affect firms, households, and government units. This chapter examines in more detail how the
costs associated with the regulatory alternatives will affect these sectors of the economy. In
particular, we are interested in whether these costs are systematically larger for smaller entities or
service areas and whether these impacts are severe under some of the criteria adopted in other
analyses.
8.1 REGULATORY FLEXIBILITY ANALYSIS
The impact of government regulation on small entities (non-profit organizations,
governmental jurisdictions, and businesses) is a special social concern as demonstrated by the
Regulatory Flexibility Act of 1980. Among other things, the Act requires that federal agencies
consider whether regulations they develop will have "a significant economic impact on a
substantial number of small entities" (U.S. Small Business Administration, 1982).
Small government jurisdictions are identified in the Act as those with populations less
than 50,000. Small businesses are identified by the Small Business Association general size
standard definitions. These vary by Standard Industrial Classification (SIC) code. For SIC code
4953, Refuse Systems, small business concerns are those receiving less than $6 million dollars
per year averaged over the most recent 3 fiscal years. These definitions are not, however, fixed
for all regulatory actions. According to both SBA (1982) and EPA (1982) guidelines, with
appropriate justification these definitions can be modified by the regulatory agency.
EPA (1982) provides guidelines for determining when a "substantial number" of these
small entities have been "significantly impacted." Impacts may be considered significant if:
1. compliance costs are greater than five percent of production costs,
2. compliance costs, as a percent of sales, are at least 10 percent higher for small entities
than for other entities,
3. capital costs of compliance are a significant portion of capital available, or
4. the requirements are likely to result in closures of small entities.
Three of these criteria apply absolute measures, but the second measure determines the adversity
of the impact on small entities relative to other, larger entities. In its guidance, EPA also notes
8-1
-------
that these criteria, as well as the criterion suggested for determination of a "substantial number"
are only guidelines, and that lead Offices may adopt other criteria as appropriate.
In its guidance EPA suggests that a "substantial number" is "more than 20 percent of
these (small entities)...affected for each industry the proposed rule would cover." This criterion
is open to a certain amount of interpretation as to the scope of the industry and "affected"
industry segments, In this analysis we choose to address the question of "a substantial number"
by treating public entities and private businesses separately. We assume that any entity that
operates a MWC plant will be "affected," and determined that a "substantial number" of small
entities are affected if small entities are more than nominally represented in the industry segment.
8.2 PRIVATE BUSINESS IMPACTS
Impacts of the regulation on private firms may be direct or indirect in nature. Owners
who must purchase and install control equipment, train employees, or change operating practices
will be directly impacted. On the other hand, firms that supply services or equipment but do not
own a plant will be indirectly impacted, and may actually benefit from the regulation as demand
for air pollution control technology and equipment increases. The extent of impacts for a
specific MWC plant owner or supplier is dependent on local market conditions and contractual
arrangements, size of the MWC plant, and financial status of the firm.
Many of the privately owned "merchant" MWC plants are large and were built (or are
about to be built) under much more favorable tax and financing conditions than would likely
apply to control equipment used to meet the NSPSs. Based on the cash flow analysis of Chapter
6, the NSPSs will result in higher control costs for private owners than for public owners of
MWC plants. For private MWC plants that have long-term contracts to dispose of waste that
include escalator provisions to cover contingencies such as pollution control equipment, these
higher costs can be passed on to waste collectors and generators. Private MWC plants generally
have such arrangements.
In contrast, privately owned plants that don't have such long-term contracts will be
adversely affected by the NSPSs. How adverse the effect will be depends on the cost of
production of the private MWC plant relative to other local means of solid waste disposal. If,
after the NSPSs, the private plant still has relatively low costs of production, the NSPSs will
reduce expected profits but will not force sale or closure of the plant. If, however, control costs
are large enough to increase cost of production beyond prevailing tipping fees at landfills or
public MWC plants, the private MWC plant may have to close or operate at a loss in the hope
that tipping fees increase in the future. A local cost and market analysis for each such private
MWC plant would be required to determine the severity of the NSPSs' impacts.
8-2
-------
CHAPTER 8
ECONOMIC IMPACT ON SECTORS OF THE ECONOMY
The costs and emission reductions estimated for each of the regulatory alternatives will
affect firms, households, and government units. This chapter examines in more detail how the
costs associated with the regulatory alternatives will affect these sectors of the economy. In
particular, we are interested in whether these costs are systematically larger for smaller entities or
service areas and whether these impacts are severe under some of the criteria adopted in other
analyses.
8.1 REGULATORY FLEXIBILITY ANALYSIS
The impact of government regulation on small entities (non-profit organizations,
governmental jurisdictions, and businesses) is a special social concern as demonstrated by the
Regulatory Flexibility Act of 1980. Among other things, the Act requires that federal agencies
consider whether regulations they develop will have "a significant economic impact on a
substantial number of small entities" (U.S. Small Business Administration, 1982).
Small government jurisdictions are identified in the Act as those with populations less
than 50,000. Small businesses are identified by the Small Business Association general size
standard definitions. These vary by Standard Industrial Classification (SIC) code. For SIC code
4953, Refuse Systems, small business concerns are those receiving less than $6 million dollars
per year averaged over the most recent 3 fiscal years. These definitions are not, however, fixed
for all regulatory actions. According to both SBA (1982) and EPA (1982) guidelines, with
appropriate justification these definitions can be modified by the regulatory agency.
EPA (1982) provides guidelines for determining when a "substantial number" of these
small entities have been "significantly impacted." Impacts may be considered significant if:
1. compliance costs are greater than five percent of production costs,
2. compliance costs, as a percent of sales, are at least 10 percent higher for small entities
than for other entities,
3. capital costs of compliance are a significant portion of capital available, or
4. the requirements are likely to result in closures of small entities.
Three of these criteria apply absolute measures, but the second measure determines the adversity
of the impact on small entities relative to other, larger entities. In its guidance, EPA also notes
8-1
-------
that these criteria, as well as the criterion suggested for determination of a "substantial number"
are only guidelines, and that lead Offices may adopt other criteria as appropriate.
In its guidance EPA suggests that a "substantial number" is "more than 20 percent of
these (small entities)...affected for each industry the proposed rule would cover." This criterion
is open to a certain amount of interpretation as to the scope of the industry and "affected"
industry segments. In this analysis we choose to address the question of "a substantial number-
by treating public entities and private businesses separately. We assume that any entity that
operates a MWC plant will be "affected," and determined that a "substantial number" of small
entities are affected if small entities are more than nominally represented in the industry segment.
8.2 PRIVATE BUSINESS IMPACTS
Impacts of the regulation on private firms may be direct or indirect in nature. Owners
who must purchase and install control equipment, train employees, or change operating practices
will be directly impacted. On the other hand, firms that supply services or equipment but do not
own a plant will be indirectly impacted, and may actually benefit from the regulation as demand
for air pollution control technology and equipment increases. The extent of impacts for a
specific MWC plant owner or supplier is dependent on local market conditions and contractual
arrangements, size of the MWC plant, and financial status of the firm.
Many of the privately owned "merchant" MWC plants are large and were built (or are
about to be built) under much more favorable tax and financing conditions than would likely
apply to control equipment used to meet the NSPSs. Based on the cash flow analysis of Chapter
6, the NSPSs will result in higher control costs for private owners than for public owners of
MWC plants. For private MWC plants that have long-term contracts to dispose of waste that
include escalator provisions to cover contingencies such as pollution control equipment, these
higher costs can be passed on to waste collectors and generators. Private MWC plants generally
have such arrangements.
In contrast, privately owned plants that don't have such long-term contracts will be
adversely affected by the NSPSs. How adverse the effect will be depends on the cost of
production of the private MWC plant relative to other local means of solid waste disposal. If,
after the NSPSs, the private plant still has relatively low costs of production, the NSPSs will '
reduce expected profits but will not force sale or closure of the plant. If, however, control costs
are large enough to increase cost of production beyond prevailing tipping fees at landfills or
public MWC plants, the private MWC plant may have to close or operate at a loss in the hope
that tipping fees increase in the future. A local cost and market analysis for each such private
MWC plant would be required to determine the severity of the NSPSs' impacts.
8-2
-------
Current information on significant events in the waste management industry highlight the
rapidly changing market conditions for private firms. Several firms have been involved in recent
mergers or acquisitions, including the following high-profile transactions:
• Wheelabrator Technologies, Inc. and its majority shareholder, The Wheelabrator
Group, Inc. plan to merge and create a combined company Wheelabrator Technologies,
lnc.(Waste-to-Energy Report, June 28, 1989, p.l).
• Environmental Systems Company (Ensco) acquired almost 1.7 million shares of
Consumat, giving Ensco 54 percent ownership of Consumat (Waste Age, May 1989,
p. 108).
• Bramble Industries, Ltd., an Australian firm, has the potential to convert $60 million in
recently purchased Ensco securities into near-controlling interest in the company
(Waste Age, May 1989, p. 110).
• Joy Technologies purchased Ecolaire for $1.5 million cash (Waste Age, May 1989,
p. 114).
• Waste Management recently traded its combustor plants and a wastewater treatment
plant to Wheelabrator for a 22 percent equity position in Wheelabrator (Waste Age,
June 1989, pp. 74-75).
8.2.1. Private Owner Profile
Historical data were used to project impacts for plants subject to NSPSs. We assumed
that the profile of plant ownership under the NSPSs could be projected using the information on
owners of existing plants. We compiled a list of firms that own MWC plants using the 1988-89
Resource Recovery Yearbook (Gould, 1988) and the data gathered under Section 114 of the
CAA (EPA, 1988a). This list includes owners of plants currently in operation as well as owners
of plants now in planning stages and plants under construction. We then obtained annual sales
data on those firms included in Moody's Industrial Manual (1988), Standard and Poors Register
of Corporations, business periodicals (Fortune, April, 1989; Business Week, Special Edition,
1989), and Waste Age (May, 1989 and June, 1989).
Telephone contacts were made with firms not included in these sources to obtain annual
sales figures and to confirm information about the firm's line of business and organizational
structure. It should be noted that financial data for many of the firms initially identified are
unavailable due to difficulty contacting the firm, or reluctance on the part of the firm to release
the information. In addition, this list does not necessarily represent a complete listing of private
owners of MWCs since recent mergers, acquisitions, or plant closures may not be reflected in the
data. Table 8-1 lists the firms we were able to identify as owners of MWC plants, their annual
sales in millions of dollars, and a brief description of the firms' activities.
8-3
-------
TABLE 8-1. MWC PRIVATE OWNER PROFILE
Firms
oo
Annual
Sales3
($106)
General Motors Corp.
Ford Motor Co.
Occidental Chemical Corp.
Westinghouse
Ford Motor Credit Corp.
James River Corp.
Southern California Edison
General Electric Capital
Waste Management
Combustion Engineering
Northern States Power (Elk River, MN)
Wheelabrator
Ogden-Martin
Foster Wheeler
Joy Technologies'5
Blount Energy Resources
Zurn Industries
Research-Cottrell
Katy-Seghers, Inc.
Dravo
Environmental Systems Co.
Reuter, Inc.
Consumatb
KTI Energy, Inc.
Maine Energy Recovery Co. (MERC)
Penobscot Energy Recovery Co. (PERC)
Vicon Recovery Industries
American Ref-Fuel
Channel Sanitation Corp.
Waste Resources Association
101,780
92,446
19,417
12,500
5,850
5,623
5,490
3,600
3,566
3,484
1,770
1,205
1,088
1,054
500
460
406
348
261
248
66.4
30.2
14.4
>6
>6
>6
1
N/A
N/A
N/A
Line of Business (Other Than MWC Ownership)
Auto, truck, bus, locomotive, aircraft manufacturer
Auto, truck, tractor & implement manufacturer
Chemical manufacturer
Electrical products, construction, financial services
Finance company
Paper and disposable packaging manufacturer
Electric utilities services
Finance company
Recycling, medical wastes, chemical wastes
Hazardous waste systems, consulting, mass transit engineering
Electric, gas, steam, telephone utilities
Environmental services & consulting
Environmental engineering & design services/mancial services
Sludge processing, hazardous waste systems
Air pollution control equipment
General contractor
Air pollution control equipment, energy recovery systems
Air pollution control equipment, energy recovery systems
Oil field equipment, bearings
Engineering and consulting services
Hazardous waste management services
Plastic refuse container, recycling, composting
Modular systems manufacturer
Continued on next
-------
TABLE 8-1. MWC PRIVATE OWNER PROFILE (continued)
Firms for Which Financial Data Were Unavailable in Business Publications
American RR Inc. Refuse Energy Systems Co.
Bridgeport RESCO Co. Rhode Island SW MgmL
Camden Co. Energy Res. Assoc. Richards Asphalt
Catalyst W-T-E Corp. Savannah Energy Systems Co.
Channel Landfill, Inc. SEMASS Partnership
Flour RR of Mass, Ltd. Prt. SES Claremont Co. Ltd. Ptn.
Mass REFUSETECH Inc. Signal Environmental Systems
New England Trust Company Southland Exchange, Inc.
North County RR Corp. St. John's University
Power Recovery Systems Truckee Meadows Ltd. Ptn.
Pulaski Co. Ltd. Ptn. Ukiah Energy Inc.
Quadrant Waste Energy Partners Ltd. Prt.
6, aAnnual sales given for the most recent year available.
Estimated using 9-month total sales volume figures.
Sources: Moody's Industrial Manual (1988), Standard and Poors Register of 'Corporations (1989), Fortune (April 1989),
Business Week (Special Edition 1989), and Waste Age (May 1989 and June 1989)
-------
Approximately one-third of the firms we idemified are publicly held corporations; the
remaining majority are privately held firms including corporations and limited partnerships.
Some ?re very large and diversified firms (e.g., General Motors Corporation). Although some
firms for which financial data are not public would not disclose the actual amount of their annual
sales when contacted by telephone, several did say that annual sales were substantially larger
than the $6 million dollar cut-off specified by the Small Business Administration as the criterion
for defining a smrll business in this industry. Only one firm was identified as having less than
$6 million in annual sales, suggesting that small businesses are only nominally represented in the
MWC segment of the industry. Therefore, under our analysis, a substantial number of small
business are not significantly impacted. These data do suggest, however, great disparity in
annual revenues between the smallest and the largest of these firms. Consequently, we examined
the relationship between impacts, plant size, and firm size under these regulatory alternatives.
Table 8-2 shows the annualized costs per Mg under the most stringent regulatory
alternative for privately owned model plants ordered by design capacity. Impacts are greater for
small model plants than for large plants. Because of this indication of a relationship between the
size of the plant and the severity of the impact, EPA has provided greater regulatory flexibility
for small plants.
TABLE 8-2. NSPSs ENTERPRISE COSTS OF CONTROL FOR PRIVATELY OWNED
MODEL PLANTS UNDER REGULATORY ALTERNATIVE IV:
ORDERED BY DESIGN CAPACITY3
Model
Plant
#
9
10
1
8
4
2
11
12
5
6
7
3
Model
Plant
Description
MOD/SA (small)
MOD/SA (mid-size)
MB/WW (small)
MOD/EA
MB/REF
MB/WW (mid-size)
FBC/BB
FBC/CB
MB/RC
RDF
RDF/CF
MB/WW (large)
Design
Capacity
(Mg/day)
45
90
180
220
450
730
820
820
950
1,810
1,810
2,040
Annualized Cost
per Mgb
($/Mg)
58.20
20.40
19.00
10.20
19.60
12.90
11.10
11.10
12.80
12.50
17.10
9.89
a Variations in annualized cost per Mg may also be due to effects other than capacity including
energy recovery capabilities, MWC technology differences, and capacity utilization.
b Based on 8 percent discount rate, 15-year equipment life, and the cash flow model for private
financing and ownership.
8-6
-------
Specific measures to address the needs of small plants include: size cut-offs built into the
regulatory structure and less stringent requirements for small plants. We have used costs for
small plants to estimate impacts for smaller firms and costs for large plants to estimate impacts
for larger firms. While there is no evidence to clearly indicate a relationship between the size of
the plant and the size of the firm that owns the plant, if small firms do generally own small
plants, the regulatory flexibility measures aimed at small plants will help mitigate these impacts.
8.2 2 Private Supplier Profile
Historical data were used to project the indirect impacts of NSPSs on private suppliers.
We assumed that the profile of private suppliers could be projected using data for existing plants.
To assess these indirect impacts, plant construction managers and technology suppliers were
identified using information found in a Kidder, Peabody Report on resource recovery (April 29,
1988). Annual sales data were then obtained using the same sources previously used to obtain
for financial data for private owners.
Table 8-3 identifies current market shares and projected capacity of firms currently acting
as plant construction managers. Only 7 firms are identified as managers of projects due to come
on line after 1991. Table 8-4 presents similar information for suppliers of services and
equipment. Together, Martin, Von Roll, and Consumat supply over one-half of the plants
included in this study. We don't expect to observe any adverse impacts on firms managing or
supplying services and equipment to owners of new MWC plants; these firms will likely benefit
from increased demand for their services. Therefore we conclude that a substantial number of
small businesses will not be adversely affected by the proposed regulation.
8.3 IMPACTS ON HOUSEHOLDS AND GOVERNMENT ENTITIES
Analyzing the economic impacts of the NSPSs on households and government entities
was difficult because of the uncertainty regarding community characteristics, special conditions
relating to ownership, contractual arrangements, and financial responsibility, and variation in
accounting practices among government entities. We introduced assumptions into our analysis
that have the tendency to overestimate impacts, thereby assuring the credibility of any "no severe
impacts" results. The assumptions we adopted to facilitate this "screening process" include:
1. assumptions that tend to underestimate the population of each service area, and
2. assumptions that tend to overestimate control costs.
8-7
-------
TABLE 8-3. MWC SYSTEMS PRIVATE SUPPLIER PROFILE: PROJECT MANAGERS
oe
I
oo
Firm
American Ref-Fuel
Babcock & Wilcox
Blount
Catalyst Energy
Combustion Engineering
Consumat
Dravo
Flour Daniel
Foster Wheeler
General Electric
Katy-Seghers
Morrison-Knudson
Ogden Martin
Pennsylvania Engineering
Riley Energy Systems
Vicon Recovery Systems
Waste Management
Westinghouse Electric
Wheelabrator Environmental
Miscellaneous
TOTAL
Operating Capacity Through 1991a
Mg/day
9,297
3,628
2,095
454
7,936
3,137
2,376
707
3,311
1,204
816
213
19,582
833
454
2,141
2,902
5,068
10,335
25,742
102,233
Percent
9.1%
3.5%
2.0%
0.4%
7.8%
3.1%
2.3%
0.7%
3.2%
1.2%
0.8%
0.2%
19.2%
0.8%
0.4%
2.1%
2.8%
5.0%
10.1%
25.2%
100.0%
•—— —^-^— ™™^— ^— ™— — ^«— «^^^^—
Operating Capacity
Mg/day
1,088
934
544
3,129
884
1,361
771
8,712
After 199ia
Percent
12.5%
10.7%
62%
\J,A* /\j
359%
•J *J * s /\j
10 2%
1 \J . A-i /(J
15.6%
8.8%
100.0%
a Based on orders.
Source: McCoy, R.W., Jr., and R.J. Sweetnam, Jr. 1988. "A Status Report on Resource Recovery." Kidder, Peabody Renort
April 29. ^
-------
TABLE 8-4. MWC SYSTEMS PRIVATE SUPPLIER PROFILE: TECHNOLOGY SUPPLIER
oo
Firm
Babcock & Wilcox/Detriot Stoker
Combustion Engineering
Combustion Engineering/ Detriot Stoker
Consumat Systems
De Bartolomeis
Deutsche Babcock
Foster Wheeler/Detriot Stoker
Keeler/Dorr-Oliver/Detriot Stoker
Martin
Riley Stoker/Detriot Stoker
Seghers Engineers
Stienmuller
Vicon Recovery Systems
Volund
Von Roll
Westinghouse O'Conner
Widmer & Ernst
TOTAL
Operating Capacity Through 1991a
Mg/day
10,079
8,344
358
3,974
680
9,297
5,451
671
25,643
1,723
816
2,376
2,348
3,116
16,303
6,182
2,095
99,957
Percent
10.1%
8.3%
0.4%
4.0%
0.7%
9.3%
5.5%
0.7%
25.7%
1.7%
0.8%
2.4%
2.8%
3.1%
16.3%
6.2%
2.1%
100.0%
I .-
Operating Capacity After 1991a
Mg/day Percent
3,129 58.2%
1,361 25.3%
884 16.5%
5,374 100.0%
~~_ - •
a Based on orders.
Source: McCoy, R.W., Jr., and R.J. Sweetnam, Jr. 1988. "A Status Repon on Resource Recovery." Kidder, Peabody Report
-------
The first step in computing geographically specific regulatory impacts is to identify the
plants to be affected by the regulation. Radian (EPA, 1989c) presents a projected distribution of
138 plants to be affected by the NSPSs based in part on a list of plants in early planning stages,
and in part on recently constructed or under-construction plants.1 However, Radian's
distribution of model plants identifies the plant capacity and the technology used for only 85
plants. It provides no information about the probable location of affected plants of each model
plant type.
To assess the impacts of the NSPSs on households and government units, the affected
plants must be associated with particular geographic locations. This assignment was made based
on a profile of the relationship between capacity and service area for existing MWC plants. For
each of the 85 affected plants, we compiled the following information:
• ownership,
• the technology used,
• the annual throughput,
* the model plant that represents the actual plant, and
• the geographic location.
Specific ownership information is available for existing plants from the Section 114
letters (EPA, 1988a) or from the 1988-89 Resource Recovery Yearbook (Gould, 1988). These
two sources, however, have no information for plants to be regulated under the NSPSs. We
therefore made the assignment of ownership to NSPSs plants based on the ownership and
throughput distribution of existing plants. Univariate statistics computed for existing plants
indicate that privately owned plants tend on average to be somewhat larger than county-owned
plants, which, in turn, are somewhat larger than municipally owned plants. We assigned
ownership to projected plants so as to mimic, to the extent possible, the size distribution by
ownership category that is observed for the existing plants. A frequency distribution of assigned
plant ownership is shown in Table 8-5.
These same sources provide information on the areas served by existing plants. Again,
however, they have no information about the plants to be covered by the NSPSs. They could
serve only the community in which they are located, several communities, the entire county in
which they are located, or several counties. Again, we assigned service areas so as to roughly
mimic the service area distribution that is observed for the existing plants. After arranging the
lfrhis projection does not conform to that used to estimate aggregate cost and emission reductions in Chapters 6
and 7. It does, however provide the basis for a technology and community profile used in this analysis.
8-10
-------
TABLE 8-5. ESTIMATED OWNERSHIP OF NSPSs PLANTS"
Owner
Private
County
Municipality
Federal
TOTAL
Number
of Plants
36
16
32
1
85
Percent
of Plants
41.9
19.7
37.2
1.2
100
aBased on statistical relationship between the size of the plant and the type of ownership for
existing MWC plants.
plants projected under our analysis by estimated throughput, we assigned the smallest 10
(11 percent) to the municipality category. For these plants, we used municipality data from the
Census of Population and Housing to assess government and household impacts of the
regulation. The remaining 75 plants were assumed to serve either the entire county in which they
were located or a multi-city or multi-county area. For these plants, we used data from the Census
of Population and Housing for only the county specified as the geographic location to compute
their household impact indices.
Each plant was assigned one county as its service area unless it was specifically
designated as serving "only one city." This procedure tends to underestimate the population base
served in those cases where the actual service area includes multiple counties and/or multiple
cities with total population greater than the designated county (see Assumption 1 above).2 Since
household and government impact indices use population data to calculate the financial base
responsible for bearing the control costs imposed by the regulations, impacts may be
overestimated.
To assess the impacts of the NSPSs on households and government entities, we first
computed plant-specific costs of compliance. Costs of the various regulatory alternatives were
assigned to actual plants based on the model plant designations of the industry profile (Radian,
1988a). When calculating control costs, we assumed that all plants are just meeting federal
2For those plants serving multiple cities with total population less than the designated county the population base
would be overestimated. However, in the follow-up telephone contacts for government entities no MWC plant
had a service area which included less than one county.
8-11
-------
standards for emissions in the baseline (see Assumption 2 above). This assumption may tend to
overestimate control costs for those plants with baseline controls that remove more pollution than
called for by current federal standards. Using cost-of-compliance information for the model
plants, we computed two measures of cost for each plant, under each of the five regulatory
alternatives. These two costs are total annualized enterprise cost of compliance and total capital
cost for both public and private forms of ownership. Total annualized enterprise cost of
compliance for a plant equals the annual waste flow (Mg/yr) for the plant multiplied by the total
annualized compliance cost per Mg for the appropriate model plant. To compute capital cost of
compliance for a plant, capital cost of compliance for the first APCD equipment cycle of the
appropriate model plant is sca'ed up or down based on the ratio of the plant's annual waste flow
tc the appropriate model plant's annual waste flow.
8.3.1 Household Impacts
All MWC plants, whatever their ownership, are likely to pass increased costs on to their
custoraers. Among these are households, which generate much of the municipal waste
incinerated annually. To assess the impacts of the NSPSs on households, 1980 Census of
Population data were collected on each of the municipalities and counties in which plants are
located. The data collected include the county population, number of households, median
household income, and per capita income. The monetary values were inflated from 1980 dollars
to fourth-quarter 1987 dollars using the GNP deflator (1988 Economic Report of the President
p. 252).
As previously described, the total annual compliance cost of the regulation was computed
by multiplying enterprise cost per Mg times each plant's estimated throughput. We calculated
different compliance costs for each model plant, depending on whether the plant is publicly or
privately owned. For purposes of computing the plant-specific compliance costs, we used the
cash flow model with private ownership for the privately owned plants, and the cash flow model
with public ownership for all other plants. This may be somewhat inaccurate for authority-
owned plants, which describe themselves as "quasi-public." In fact, their financing is probably
much like that of a public plant. After computing total annual compliance costs for each plant,
we summed the plant compliance costs over the plants within each service area to find the
compliance cost incurred in each service area.
The inherent tendency toward underestimation of service area population has the effect of
underestimating the population base responsible for financing the control costs of the plant or
plants located in each service area. Consequently, the first step in the screening procedure was
8-12
-------
designed to compensate for this bias: only those service areas whose estimated waste generation
roughly matched the capacity of the plant or plants located in that service area were included in
the analysis. In particular, only the 23 service areas with implied capacity utilization between 40
percent and 400 percent for modular plants and 60 percent and 400 percent for all other plants
were included. We combined census data associated with the 23 counties or municipalities with
control cost data for the affected plants and used the data to estimate two service-area-specific
household impact indices:
1. compliance cost per household, and
2. compliance cost per household as a percentage of median household income.
These indices had been used previously by EPA's Office of Solid Waste (OSW) in its Subtitle D
Landfill Regulatory Impact Analysis (RIA) (Temple, Barker and Sloan et al., 1987). They are a
rough measure of the household burden associated with the regulation: the former is an absolute
dollar measure per household, while the latter is a measure of the cost of the regulation relative to
the income of an average household. In the OSW Landfill RIA, a service area is defined as
having severe household impacts if either:
1. compliance cost per household exceeds $220, or
2. compliance cost per household exceeds 1 percent of median household income.
Based on either criterion, none of the 23 service areas had severe household impacts
under any of the regulatory alternatives. However, it should be noted that our sample does not
include service areas with a population less than 2,500 because of census data limitations.
Additionally, these results are based on conditions of national average waste generation per
household and the assumption that all served households share equally in paying the cost of
compliance. In practice, the impact of the regulation on individual households would depend on
actual waste generated, actual household income, and the method by which individual
jurisdictions pass on costs to their customers. While on average, impacts of compliance are not
severe, there may well be special contractual or technical conditions, especially for small
communities and service areas, where these costs, in combination with the costs of other
environmental regulations, may impose unusual hardships.
Figures 8-1 and 8-2 show the distribution of household impacts measured by each index.
Under the most stringent regulatory alternative, no service area averages control costs greater
than $89 per household per year or 0.5 percent of median household income per household per
year. Costs for 90 percent of all households average less than $30 per household per year in
Figure 8-1 and less than 0.13 percent of median household income per household per year in
8-13
-------
Number of 25 -\
Service
Areas b-c
20 -
00
15 -
10 -
5 -
$0-$10
Regulatory Alternative I
Regulatory Alternative I!A
Regulatory Alternative MB
| | Regulatory Alternative III
!}•[ Regulatory Alternative IV
$10-$20
$20-$30
$30-$40
$70-$80
$80-$90
$220
Average Cost per
Household per Year a
Costs refer to control costs only; no baseline costs are included.
} Service areas with less than 2,500 total population were not included in the sample because census data for these service areas
were not available.
' Service areas with implicit capacity utilization less than 40 percent for modular plants, less than 60 percent for other technologies, or
greater than 400 percent for all technologies were not included in the sample. See text for discussion.
Household impacts were defined as "severe" if average cost exceeds $220 per household per year.
Figure 8-1. Distribution of Household Impacts Under NSPSs by
Number of Service Areas and Regulatory Alternative: Index 1
-------
oo
Number of 25 -
Service
Areas b-c
0%-.05%
Y/A Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative I IB
[ j Regulatory Alternative III
|B| Regulatory Alternative IV
.45%-.5%
1.0%
Average Cost per Household per Year as a
Percentage of Median Household Income a
J Costs refer to control costs only; no baseline costs are included.
'Service areas with less than 2,500 total population were not included in the sample because census data for these service areas
were not available.
' Service areas with implicit capacity utilization less than 40 percent for modular plants, less than 60 percent for other technologies
t greater than 400 percent for all technologies were not included in the sample. See text for discussion. '
Household impacts were defined as "severe" if average cost exceeds 1 percent of median household income per household per year.
or
Figure 8-2. Distribution of Household Impacts Under NSPSs by
Number of Service Areas and Regulatory Alternative: Index 2
-------
Figure 8-2. The average control cost per household per year is $20, translating to 0.08 percent of
median household income per household per year.
In Figures 8-3 and 8-4 service areas are grouped in population size categories, and
median impacts in each category are presented for the five regulatory alternatives. It appears that
smaller service areas have slightly higher household impacts. In particular, greater control of
smaller combustors under Regulatory Alternatives IIB and IV more than double the household
impact for the smaller communities.
8.3.2 Governmental Impacts
To assess impacts on governmental units, we identified the plants owned by
municipalities or counties. For each of these public plants, data were collected from the 1982
Census of Governments, including the county's or municipality's annual capital expenditures,
annual total tax revenue and total revenue, annual total expenditures, and annual sewerage and
sanitation expenditures, as well as the county's or municipality's total debt outstanding and
annual interest paid on the debt. These government expenditure and revenue figures were
inflated from 1982 dollars to 1987 dollars using the State and Local Government Expenditure
Deflator (7955 Economic Report of the President, p. 253). These data were used to estimate
annual values for three government impact indices:
1. sum of the average sewerage and sanitation cost per household and the average
control cost per household as a percent of median household income,
2. sum of total current debt service and additional debt service associated with the
capital cost of control as a percent of total general revenues, and
3. control costs as a percent of total general expenditures.
The first two indices are adaptations of indices used in the Municipal Sector Study (U.S.
EPA, 1988c). Exact duplication of the indices used in that study was impossible with the data
available to us, so our analysis substituted measures of government activity similar in principle.
The Sector Study indices are designed to measure a government entity's ability to meet
additional financial obligations incurred due to the regulation. More specifically, the indices are
designed to measure each government's ability to issue revenue bonds or obtain loans to finance
the additional control costs. The third index is one used in the OSW Subtitle D Landfill RIA. It
is a measure of the additional governmental cost burden associated with the regulation relative to
the existing government commitments.
8-16
-------
oo
Median
Cost per
Household
per Year a
$220U-|
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
o -
^
.^2K
: -L
;
'
,~
"I
Regulatory Alternative I
Regulatory Alternative MA
[g£4J Regulatory Alternative MB
[ j Regulatory Alternative III
•H Regulatory Alternative IV
0-50
50-150
150-500
500-1,000
1,000 +
Population of
Service Area (1,000's) b'c
Costs refer to control costs only; no baseline costs are included.
'Service areas with less than 2,500 total population were not included in the sample because census data for these service areas
were not available.
' Service areas with implicit capacity utilization less than 40 percent for modular plants, less than 60 percent for other technologies, or
greater than 400 percent for all technologies were not included in the sample. See text for discussion.
Household impacts were defined as "severe" if average cost exceeds $220 per household per year.
Figure 8-3. Distribution of Household Impacts Under NSPSs by
Service Area Population and Regulatory Alternative: Index 1
-------
Median Cost per d
Household per 1.000%
Year as a
Percent of
Median
Household
Income a
oo
I
ex
0.200% -
0.060% -
0.040% -
0.020% -
0.000%
0-50
50-150
150-500
aCosts refer to control costs only; no baseline costs are included.
b
Regulatory Alternative I
Regulatory Alternative IIA
|/:if-j Regulatory Alternative I IB
| j Regulatory Alternative III
I^H Regulatory Alternative IV
500-1,000 1,000 +
Population of
Service Area (1,000's)b'c
Service areas with less than 2,500 total population were not included in the sample because census data for these service areas
were not available.
'Service areas with implicit capacity utilization less than 40 percent for modular plants, less than 60 percent for other technologies, or
greater than 400 percent for all technologies were not included in the sample. See text for discussion.
Household impacts were defined as "severe" if average cost exceeds 1 percent of median household income per household per year.
Figure 8-4. Distribution of Household impacts Under NSPSs by
ire Area Peculation and Regulatory Alternative: Index 2
-------
Indices of government unit impacts were only computed for some of the plants affected
by the regulation. Our analysis naturally included only MWCs that could be identified as
government owned. We further restricted our analysis to those plants for which ownership and
service area were specified as being the same. In some cases, this meant that even though
ownership was specified as one county, the plant may have actually served multiple counties,
counties and municipalities, or multiple municipalities in one or more counties. As previously
noted, assigning only one government unit to a plant may tend to overestimate government unit
impacts for those MWCs that serve several jurisdictions by underestimating the financial base of
the communities served. We thus treat this part of the analysis as a screening procedure, aimed
at identifying potential problem communities. If assigning a single government unit to the MWC
plant in our analysis does not give severe impacts, the full assignment of all relevant government
units with their greater combined resources to the plant is unlikely to give severe impacts.
Only the model plant costs for public ownership were considered in our analysis of
government impacts. The compliance costs were summed over all the public combustor plants in
each county or municipality. While we were concerned that this reduction in the number of
plants analyzed might distort the government impact analysis, the remaining 17 plants appeared
to be fairly representative of the entire set in terms of population served, median income, and
combustor technology.
To compute the impacts, we combined Census of Governments data associated with these
17 counties or municipalities with the control cost data for the affected plant. These impacts
were compared to the "severity" measures adopted from the Sector Study and the OSW Landfill
RIA. Using the Municipal Sector Study criterion, a governmental unit is defined as severely
impacted if:
• the sum of the average sewerage and sanitation cost per household and the average
control cost per household as a percent of median household income exceeds 1 percent,
and
• the sum of total current debt service and additional debt service associated with the
capital cost of compliance to the regulation as a percent of total general revenues
exceeds 15 percent.
For the OSW Landfill RIA, the government impact is likely to be severe if:
• control costs as a percent of total general expenditures exceed 1 percent.
Figure 8-5 shows the preliminary results for our analysis using each of the indices as well
as the joint criterion for the first two indices. Since the first and second criteria must be met
8-19
-------
Number of 9
Government
Units with 8 •
Severe
Impacts
6-
5-
4-
3-
2-
1 -
0
Joint'
Index
Regulatory Alternative I jpj Regulatory Alternative IIB
Regulatory Alternative IIA | | Regulatory Alternative III
Regulatory Alternative IV
Service areas with less than 10,000 total population were not included in the sample because
census data for these service areas were not available.
Index 1 is the sum of average sewerage and sanitation cost per household and the average control
cost per household as a percent of median household income. The Municipal Sector Study (U.S.
EPA, 1988d) sets 1 percent as the criterion for severe impacts under this index.
\
Index 2 is the sum of total current debt service and additional debt service associated with
compliance to the regulation as a percent of total general revenues. The Municipal Sector Study
(U.S. EPA, 1988d) sets 15 percent as the criterion for severe impacts under this index.
Using the Municipal Sector Study criteria, both index 1 and index 2 must be exceeded to indicate
severe impacts.
Index 3 measures control costs as a percent of total general expenditures. The OSW Landfill RIA
sets 1 percent as the criterion for severe impacts under this index
Figure 8-5. Distribution of Government Impacts Under NSPSs:
Preliminary Screening Results a
8-20
-------
together to indicate severe impacts, no government units had severe impacts under the joint
criterion. Under the third criterion nine governments were identified as having severe impacts in
our analysis. Of these, however, the screening procedure revealed that only 3 service areas had
estimated waste generation that roughly matched the capacity of the associated plants (based on
implied capacity utilization between 40 percent and 400 percent for modular plants and 60
percent and 400 percent for all other plants).
Given this indication that the government impacts might indeed be severe under one of
the criteria, we examined the service and financial conditions of each community we identified
as having severe impacts. This inquiry revealed that more than one government jurisdiction was
served by the plant(s) in all nine cases where severe impacts were indicated. In many cases, the
financial base in our analysis represented only a small fraction of the actual financial base served
by the combustor. Adding the resources of these government units together would result in
smaller impact measures and reduction of the index below the "severe" threshold. We note,
however, that because of data limitations, we could not compute impact indices for plants serving
government entities representing service areas with a population of 10,000 or less.
Using county and municipality population data, we divided the service areas into
population size categories. Figures 8-6, 8-7, and 8-8 examine the relationship between the
impact ratios and the county or municipality size. Some evidence of an inverse relationship
between the magnitude of the impact variable and population size appears to exist. Regulatory
Alternatives in and IV, with more stringent controls on large combustors, tend to have the
greatest impact on government units.
8-21
-------
Index 7b 1-40%
(Median
Impact) 1i20%
0.00%
Y/A Regulatory Alternative I \^\ Regulatory Alternative I IB
Regulatory Alternative MA | | Regulatory Alternative III
Population of
Service Area
(1,000's) a
Regulatory Alternative IV
'Service areas with less than 10,000 total population were not included in the sample because
census data for these service areas were not available. ^ut>«
'index 1 is the sum of average sewerage and sanitation cost per household and the average control
Fpl P1 ±T ,^ 3S 3 Percent°f median household income. The Municipal Sector Study (U S
EPA, 1983d) sets 1 percent as the criterion for severe impacts under this index '
Figure 8-6. Distribution of Government Impacts Under NSPSs by
Service Area Population and Regulatory Alternative: Index 1
8-22
-------
Index 2 b 15.0%
(Median
Impact)
3.00% •
2.50% -
2.00% •
1.50% -
1.00% •
0.50% -
0.00%
150+ Population of
N = 9 Service Area
(1,000's) a
Regulatory Alternative I
Regulatory Alternative I IB
Regulatory Alternative IV
Regulatory Alternative IIA | | Regulatory Alternative III
Service areas with less than 10,000 total population were not included in the sample because
census data for these service areas were not available.
' Index 2 is the sum of total current debt service and additional debt service associated with
compliance to the regulation as a percent of total general revenues. The Municipal Sector Study
(U.S. EPA, 1988d) sets 15 percent as the criterion for severe impacts under this index.
Figure 8-7. Distribution of Government Impacts Under NSPSs by
Service Area Population and Regulatory Alternative: Index 2
8-23
-------
Index 3 b e.00%
(Median
Impact)
150+ Population of
N = 9 Service Area
(1,000's) a
Regulatory Alternative I jfgj Regulatory Alternative MB
Regulatory Alternative IIA | | Regulatory Alternative III
Regulatory Alternative IV
Service areas with less than 10,000 total population were not included in the sample because
census data for these service areas were not available.
Index 3 measures control costs as a percent of total general expenditures. The OSW Landfill RIA
sets 1 percent as the criterion for severe impacts under this index
Figure 8-8. Distribution of Government Impacts Under NSPSs by
Service Area Population and Regulatory Alternative: Index 3
8-24
-------
CHAPTER 9
SENSITIVITY ANALYSIS
Social costs were estimated using a two-stage discounting procedure (see Table 9-1).
Annualized social costs are the sum of capital costs, annualized at 10 percent, and operating
costs. Present values are calculated by applying a 3 percent discount rate to total annualized
values. EPA has not officially adopted the use of the two-stage discount procedure. In addition,
continuing debate surrounds the appropriate discount rate to use in any procedure. To show how
alternative views of discounting affect the results of this analysis, the social costs were
recalculated using single discount rates: 10 percent in one case and 3 percent in another. The
results are shown in Tables 9-2 and 9-3, respectively.
Average capacity utilization values reported in the 1988-89 Resource Recovery Yearbook
(Gould, 1988) were adopted for the calculation of most of the model plant baseline waste flows.
Capacity utilization estimates reported by Radian (EPA, 1989c) averaged as much as 8 percent
higher than those used in this report. Table 9-4 presents scaling factors calculated using higher
capacity utilization. These factors are used to calculate national cost impacts in Table 9-5 and
emission reductions in Table 9-6. Table 9-5 reports Scenario I national social costs using the
two-step discounting procedure described above with the higher capacity utilization values.
Table 9-6 reports Scenario I baseline emissions and emissions reductions under each regulatory
alternative with the higher capacity utilization values. Using a higher capacity utilization results
in no change in Scenario I baseline emissions or emissions reductions because even though the
number of plants in each model plant category is reduced, total waste flow processed by MWC
does not change. Calculation of Scenario III costs and emissions reductions would require the
development of control costs based on higher capacity utilization and the introduction of these
costs into the spreadsheet provided by Bentley and Spitz (1989) containing choice equations.
9-1
-------
TABLE 9-1. NSPSs NATIONAL COST IMPACTS: SOCIAL COSTS USING A
TWO-STEP DISCOUNTING PROCEDURE (1987 $)*
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
Scenario HI
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
Capital
Costs
($10«)
30.5
183
216
514
548
29.7
121
149
321
347
Present
Value of
Social Costs
($106)
126
1,190
2,260
2,930
3,290
123
1,320
1,610
1,840
2,110
Annualized
Social Costs
($l(%r)
6.41
97.2
115
150
168
6.26
67.6
82.0
93.8
107
Annualized
Social Costs
per Mg MSWb
($/Mg)
0.46
6.99
7.70
10.80
11.20
0.46
6.86
7.69
10.60
11.10
a Annualized social costs are the sum of capital costs, annualized at 10 percent, and annual
operating costs. Present values are computed as the present value of these annualized costs
using a 3 percent rate of discount.
b Computed by dividing total annualized cost by the estimated amount of MSW processed per
year by MWC.
9-2
-------
TABLE 9-2. NSPSs NATIONAL COST IMPACTS: SOCIAL COSTS USING A
10 PERCENT DISCOUNT RATE (1987 $)
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative Iffi
Regulatory Alternative III
Regulatory Alternative IV
Scenario HI
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative Iffi
Regulatory Alternative in
Regulatory Alternative IV
Capital
Costs
($106)
30.5
183
216
514
548
29.7
121
149
321
347
Present
Value of
Social Costs
($106)
60.5
917
1,090
1,410
1,580
58.9
637
773
884
1,010
Annualized
Social Costs
($10«/yr)
6.41
97.2
115
150
168
6.26
67.6
82.0
93.8
107
Annualized
Social Costs
per Mg MSWa
($/Mg)
0.46
6.99
7.70
10.80
11.20
0.46
6.86
7.69
10.60
11.10
a Computed by dividing total annualized cost by the estimated amount of MSW processed per
year by MWC.
9-3
-------
TABLE 9-3. NSPSs NATIONAL COST IMPACTS: SOCIAL COSTS
3 PERCENT DISCOUNT RATE (1987 $) S
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative IIA
Regulatory Alternative IIB
Regulatory Alternative in
Regulatory Alternative IV
.....
Capital
Costs
($10«)
— • •
30.5
183
216
514
548
Present
Value of
Social Costs
($106)
..._
97.2
1,730
2,050
2,450
2,770
~— ^-^"' --'-I'- ._!_!'-•- m1^-
Annualized
Social Costs
($l(%r)
4.96
88.5
105
125
142
==:
Annualized
Social Costs
per Mg MSWa
($/Mg)
0.36
6.37
7.01
9.00
9.46
29.7
121
149
321
347
94.8
1,210
1,470
1,540
1,780
4.84
61.8
74.9
78.5
90.9
0.36
6.28
7.02
8.87
9.43
9-4
-------
TABLE 9-4. SCENARIO I SCALING FACTORS CALCULATED USING
A HIGHER CAPACITY UTILIZATION*^
Model
Plant
#
1
2
3
4
5
6
7
8
9
10
11
12
Model
Plant
Type*
MB/WB (small)
MB/WW (mid-size)
MB/WW (large)
MB/REF
MB/RC
RDF
RDF/CF
MOD/EA
MOD/SA (small)
MOD/SA (mid-size)
FBC/BB
FBC/CB
Model Plant
Capacity
(Mg/day)
180
730
2,040
450
950
1,810
1,810
220
45
90
820
820
Scenario I
Scaling
Factors0
16.81
6.75
7.88
3.00
3.00
4.88
3.00
3.00
1.80
6.38
1.88
4.13
Model Plant Total
62.51
a Scenario in scaling factors differ with each regulatory alternative.
b Tables 3-2 describes each of the model plants listed here.
° ^non SCaling fact°rs ^ based on the annual operating hours reported by Radian (EPA
1989c) with an adjustment for a model plant #7 which co-fires 50 percent wood
Capacity utilization adopted in the Radian report was in many cases greater than that
used in the baseline and scenarios of this impact analysis.
9-5
-------
TABLE 9-5. NSPSs NATIONAL COST IMPACTS: SOCIAL COSTS USING A
HIGHER CAPACITY UTILIZATION (1987 $)«
Scenario
and
Regulatory
Alternative
Scenario I
Regulatory Alternative I
Regulatory Alternative 1IA
Regulatory Alternative IIB
Regulatory Alternative IE
Regulatory Alternative IV
Capital
Costs
($10«)
28.1
168
200
472
504
Annualized
Social Costs
($l<%r)b
6.10
96.3
114
145
163
Annualized
Social Costs
per Me MSWC
($/Mg)
0.44
6.93
7.63
10.40
10.90
a Calculated based on annual hours of operation as reported by Radian (EPA, 1989c) with an
adjustment for a model plant #7 which co-fires 50 percent wood. Capacity utilization
adopted in the Radian report was in many cases greater than that used in the baseline and
scenarios of this impact analysis.
b Annualized social costs are the sum of capital costs, annualized at 10 percent, and annual
operating costs.
c Computed by dividing total annualized cost by the estimated amount of MSW processed per
year by MWC.
9-6
-------
TABLE 9-6. NSPSs NATIONAL BASELINE EMISSIONS AND EMISSION
REDUCTIONS (Mg PER YEAR): HIGHER CAPACITY UTILIZATION3
Scenarios and
Regulatory CDD/CDF CO PM SO2 HCl Pb Ash
Alternatives
Scenario I
Baseline Emissions 0.0152 5,470 7,540 42,000 49,300 127 3,700,000
Emissions Reductions
Regulatory Alternative I 0 0 5,220 0 0 88.5 -154,000
Regulatory Alternative HA 0.0107 0 5,220 18,300 36,700 108 -383,000
Regulatory Alternative HB 0.0115 0 5,960 19,300 39,500 124 -401,000
Regulatory Alternative HI 0.0139 0 5,220 35,400 44,400 108 -314,000
Regulatory Alternative IV 0.0146 0 5,960 36,400 47,200 124 -332,000
a Calculated based on annual hours of operation as reported by Radian (EPA, 1989c) with an
adjustment for a model plant #7 which co-fires 50 percent wood. Capacity utilization
adopted in the Radian report was in many cases greater than that used in the baseline and
scenarios of this impact analysis.
KEY: polychlorinated dibenzo-p-dioxins and dibenzofurans (CDD/CDF), carbon monoxide
(CO), particulate matter (PM), sulfur dioxide (SO2), hydrogen chloride (HCl), and lead (Pb).
9-7
-------
REFERENCES
Arnold, F.S. 1986. Discounting from a Social Perspective: First Principles. Prepared for the
U.S. Environmental Protection Agency. Washington, DC: ICF Incorporated.
Bentley, Jerome T. and William Spitz. 1989. A Model of the MSW Choice Decision. Prepared
for the U.S. Environmental Protection Agency. Princeton, NJ: Mathtech Incorporated.
Berry, Patricia, C. Ed Hilton, Jr., Carl Miller, Nancy McCan, John Nesbitt, Suzanne Paounck,
Blair Pollack, Richard Ragg, and Robert VanDeman. 1988. Individual telecoms to
P. Timothy Neeley, Research Triangle Institute, June and July 1988.
Bingham, T.H., and B.S. Lee. 1976. An Analysis of the Materials and Natural Resource
Requirements and Residuals Generation of Personal Consumption Expenditure Items. Final
report prepared for the U.S. Environmental Protection Agency.
Business Week.. April 15, 1988. "The Top 1000 U.S. Companies Ranked by Industry." pp. 266,
280.
Campbell, J.C., M.P. Flynn, K.R. Crosslin, R.D. Rucker, and A. Coles. 1984. Feasibility and
Design of the Proposed Nashville Refuse-Fired Cogeneration Facility. Prepared by
Nashville Electric Service and Gresham, Smith and Partners for the Electric Power Research
Institute. Palo Alto, CA.
Coal and Synfuels Technology. July 25, 1988. "WTE Plant Capacity Forecast Up." p. 8.
Cowen, W.T. 1987. "Massachusetts Fiddles as Capacity 'Burns' Away." Waste Age
(December): 118-122.
Economic Report of the President. 1988. Transmitted to Congress February 1988 together with
The Annual Report of the Council of Economic Advisors. Washington, DC: U.S.
Government Printing Office.
Environmental Law Institute. 1983. Compendium of Cost of Remedial Technologies at
Hazardous Waste Sites. Draft report prepared for the U.S. Environmental Protection Agency.
September.
Farber, Kit D. and Gary L. Rutledge. 1988. "Pollution Abatement and Control Expenditures,
1983-86." Survey of Current Business (May):22-29.
Fortune. April 24, 1989. "Fortune 500 Largest U.S. Industrial Firms." pp. 354, 358.
Franklin Associates, Ltd. 1988. Characterization of Municipal Solid Waste in the United States,
1960 to 2000. Final report prepared for U.S. Environmental Protection Agency.
Gershman, Harvey, and Jeff Foote. "The State of the Market: A Status Report on Project
Growth." Presentation by Timothy J. Bratton at "Waste-to-Energy '88: The Integrated
Market." Washington, DC. October 3-4, 1988.
Glebs, Robert T. 1988. "Landfill Costs Continue to Rise." Waste Age (March):84-90.
R-l
-------
Goddard, Haynes C. 1975. Managing Solid Waste. New York: Praeger Publishers.
Gould Robert, M.S., M.P.H ed. 1988. 1988-89 Resource Recovery Yearbook Directors and
Guide. Governmental Advisory Associates. wroooK, uirectory and
Luke"R-"'^-es Incinerator Ash Bill With Liner
Solid Waste Finance.
Ibotson Associates. 1986. Stocks, Bonds, Bills, and Inflation: 1986 Yearbook.
Jansen.G.R. 1986. The Economics of Landfill Gas Projects. CERCLA, March
Industrial
^"ington, DC:
"-very.
Merril-Lynch Bond Index. January, 14, 1988.
Moody's Handbook of Common Stocks. Winter 1987-88.
Up Morc ma" 30% In Annual
'S "* Was Bi^s. Ever." WflBe Age
R-2
-------
Radian Corporation. 1988a. Municipal V/aste Combustion Industry Profile—Facilities Subject
to Section lll(d) Guidelines. Final report prepared for the U.S. Environmental Protection
Agency. September.
Radian Corporation. 1988b. Memorandum to U. S. Environmental Protection Aeencv
August 30. 6 *'
Radian Corporation. 1988c. Working papers on energy revenue calculation prepared by David
White.
Research Triangle Institute. 1988. Economic Impact Analysis of Municipal Solid Waste Landfill
New Source Performance Standards and Clean Air Act §III(d) Guidelines. Prepared for the
U.S. Environmental Protection Agency.
Research Triangle Institute. 1988. Memo to U.S. Environmental Protection Agency on
responses to Section 114 survey letters.
Research Triangle Institute. 1987. Industry Supply, Cost and Availability of Capital, and
Closure Analysis. Prepared for U.S. Environmental Protection Agency EPA Contract No
68-02-4321. Research Triangle Park, N.C. September.
Research Triangle Institute. 1986. A Profile of the Market for Hazardous Waste Management
Services. Draft report prepared for the U.S. Environmental Protection Agency.
Robson, John. Memorandum to John Chamberlin. February 13, 1989.
Robinson, William D., ed. 1986. The Solid Waste Handbook, A Practical Guide. Wiley-
Interscience.
Rubinfeld, D.L. 1987. "The Economics of the Local Public Sector." In Handbook of Public
Economics, vol. II, edited by AJ. Auerbach and M. Feldstein. North-Holland: Elsevier
Science Publishers.
Salimando, J.A. 1989. "How the 'Big Six' Are Doing." Waste Age (June):74-80.
Salimando, J.A. 1989. "74 Companies Make Progress." Waste Age (May): 102-124.
Salimando, J.A. 1988. "Plant Vendors are Pushing Recycling!" Waste Age (July): 127-129.
Savas, E.S., and C. Niemczewski 1976. "Who Collects Solid Waste?" 7976 Municipal Year
Book, pp. 167-172. Washington, DC: International City Management Association.
Sharpe, William F. 1985. Investments, 3rd edition, Prentice-Hall.
Solid Waste Report. July 11, 1988. "Europe, U.S. Facing Similar Garbage Crisis, Report Says "
Vol. 19, no. 28, p. 221. r j •
Solid Waste Report. September 12, 1988. "Slants and Trends." Vol. 19, no. 37, p. 285.
Solid Waste Report. October 17, 1988. "Recycling Dominates Solid Waste Laws in 1988 " Vol
19, no. 42, pp. 326-327.
Standard and Poors Bond Guide. December 1987.
R-3
-------
Standard and Poors Industry Surveys. September 1987.
Standard and Poors Statistical Service. December 1987
Stevens, BJ. 1978. "Scale, Market Structure, and the Cost of Refuse Collection." The Review
of Economics and Statistics LX(3):438-448.
Temple, Barker & Sloane, Inc., ICF, Inc., and Pope-Reid Associates. 1987. Draft Regulatory
Impact Analysis of Proposed Revisions to Subtitle D Criteria for Municipal Solid Waste
Landfills. Prepared for the U.S. Environmental Protection Agency, Office of Solid Waste.
Tucker, R.E. 1988. Telecom, to G. Morris, Research Triangle Institute. April 12.
U.S. Department of Commerce, Bureau of Census. 1987. Statistical Abstracts of the U.S., 1987.
U.S. Environmental Protection Agency. 1989a. Economic Impact of Air Pollutant Emission
Guidelines for Existing Municipal Waste Combustors. August. EPA-450/3-89-005.
U.S. Environmental Protection Agency. 1989b. Municipal Waste Combustors - Background
Information for Proposed Guidelines for Existing Facilities. EPA-450/3-89-27e.
U.S. Environmental Protection Agency. 1989c. Municipal Waste Combustors - Background
Information for Proposed Standards: 11 l(b) Model Plant Description and Costs of Control.
EPA-450/3-89-27b.
U.S. Environmental Protection Agency. 1989d. The Solid Waste Dilemma: An Agenda for
Action. Final Report prepared by the Municipal Solid Waste Task Force. Office of Solid
Waste. February. EPA/530-SW-89-019.
U.S. Environmental Protection Agency. 1989e. Background document for "The Solid Waste
Dilemma: An Agenda for Action." Final Report prepared by the Municipal Solid Waste
Task Force, Office of Solid Waste. February. EPA 1530-SW-88-054A.
U.S. Environmental Protection Agency. 1988a. Survey of Municipal Waste Combustion
Facilities.
U.S. Environmental Protection Agency. 1988b. National Survey of Solid Waste (Municipal)
Landfill Facilities. Final Report. Prepared by Westat, Inc. EPA/68-01-7359.
U.S. Environmental Protection Agency. 1988c. Municipalities, Small Business, and
Agriculture: The Challenge of Meeting Environmental Responsibilities. Washington, DC:
Office of Policy Planning and Evaluation. EPA 230-09/88-037.
U.S. Environmental Protection Agency. 1987. Municipal Waste Combustion Study: Report to
Congress. Prepared by the Radian Corporation. EPA/530-SW-87-021 A.
U.S. Environmental Protection Agency. 1982. Memoranda to Administrators and Office
Directors on EPA implementation of the Regulatory Flexibility Act. February 9.
U.S. Small Business Administration. 1982. The Regulatory Flexibility Act. October.
Value Line Investment Survey. Januarys, 1988.
The Wall Street Journal. January 22, 1988
R-4
-------
The Wall Street Journal. September 4, 1987.
Waste-to-Energy Report. June 29, 1988. "Kidder, Peabody Sees 92,025-T/D of Waste-Energy
Capacity in Planning." pp. 1-2.
Wertz, K.L. 1976. "Economic Factors Influencing Households'Production of Refuse." Journal
of Environmental Economics and Management 2:263-272.
R-5
-------
APPENDIX A
ESTIMATION OF THE REAL DISCOUNT RATE FOR PRIVATE FIRMS
AND PUBLIC ENTITIES
-------
CONTENTS
Section Page
A.I Introduction A-l
A.2 Private Cost of Capital A-l
A.2.1 The Weighted Average Cost of Capital A-2
A.2.2 Cost of Equity. A-3
A.2.3 Estimation of the WACC A-4
A.3 Public Cost of Capital A-5
A.4 Inflation and the Real Cost of Capital A-6
A.5 Social Rate of Discount A-6
in
-------
TABLES
Number
Page
A-l Capital Asset Pricing Method: Data and WACC Calculations for Private Entities
in the Municipal Waste Management Industry A-8
A-2 Standard & Poors Bond Ratings and Yields A-9
A-3 "Bond Rate Plus Four" Method: WACC Calculations for Private Entities in the
Municipal Solid Waste Management Industry A-10
A-4 Dividend Growth Model Data: Data and WACC Calculations for Private Entities
in the Municipal Solid Waste Management Industry A-l 1
A-5 Estimated Real Weighted Average Cost of Capital (Percent) A-12
IV
-------
APPENDIX A
ESTIMATION OF THE REAL DISCOUNT RATE FOR PRIVATE FIRMS AND
PUBLIC ENTITIES
A.I INTRODUCTION
The cash flow model introduced in Chapter 3 is needed to determine the cost associated
with meeting an emission standard or guideline. The model is also used to predict a decision
maker's response when a number of options for meeting a regulation are available.
Two forms of the model were presented: one for a private firm and one for a public entity
like a municipality or a solid waste authority. The reason for this distinction is that these forms
of organization face very different financial conditions due to the effect of differing tax treatment
of their incomes and expenditures. Not only are the cash flow models different, but the financial
parameters used in the models are also different. In particular, the real discount rates used in the
private and public (municipal) cash flow models differ. In this appendix we describe how the
different real discount rates for private and public (municipal) entities used in this analysis were
estimated.
This section also discusses some of the problems and issues—such as inflationary
expectations and risk and liquidity premiums—associated with both estimating and applying cost
of capital values. Additional information on estimation and using cost-of-capital values for
regulatory analyses may be found in an RTT report by Anderson, Mims, and Ross (RTI, 1987).
A.2 PRIVATE COST OF CAPITAL
The private discount rate is the "time value of money" used by firms to represent the cost
to them of investing funds today in anticipation that these expenditures will yield revenues
sometime in the future. The private discount rate is therefore often referred to as the "cost of
capital."
For a private firm, the cost of capital can be the rate of return on an alternative investment
or, if the firm has access to credit or equity markets, the lending rates prevailing in those markets.
Inasmuch as a firm with access to capital funds would find it profitable to invest in any
opportunity where the rate of return exceeds the cost of capital in credit and equity markets, these
market-based values determine the cost of capital.
A-l
-------
A.2.1 The Weighted Average Cost of Capital
Most private firms finance new investments from two sources of capital: debt and equity.
The term "weighted average cost of capital" is used to denote the fact that the firm's overall cost
of capital is a weighted average of the costs of debt and equity. Debt financing is generally
carried out by issuing long-term bonds. Equity financing occurs when firms retain earnings and
issue common stock. Other sources of capital are usually insignificant in comparison to these
two sources and therefore will not be discussed here.
The before-tax cost of debt is the interest rate a firm must pay on long-term debt. This
rate is equal to the yield to maturity (YTM) on the firm's long-term bonds. YTM is based on
current bids for the bonds in the bond market. YTM must be used rather than the rate at which
past debt was issued because the firm's current cost of capital is the relevant interest rate in new
investment decisions.
Interest paid by corporations with current revenues is tax deductible; therefore the after-
tax cost of debt is the true cost of debt for the firm. The after-tax cost of debt can be calculated
by multiplying the before-tax cost of debt by one minus the combined state and federal marginal
corporate tax rate (t).
The cost of equity is much more difficult to estimate. The cost of equity is the rate of
return that is required by the holders of the firm's common stock. This rate of return must be
earned to ensure that the market price of common stock remains unchanged. Estimation of the
cost of equity is discussed in detail below.
Once the cost of debt and equity have been computed, however, they can be combined to
estimate the overall or weighted average cost of capital (WACC) for the firm. In equation form:
WACC = (1-t) rd(D/V) + re(E/V) (A.I)
where
t = marginal corporate tax rate,
rd = before-tax cost of debt,
D= value of firm's debt,
E = value of firm's equity,
A-2
-------
V = total value of firm (D + E), and
re= cost of equity.
This is the appropriate discount rate for all of the firm's investments even though any
specific project could conceivably be financed entirely by one source. The WACC formula
assumes that the firm's debt/equity ratio will remain constant over time.
Once the firm's WACC has been estimated, then the average cost of capital for the
municipal solid waste management industry can be calculated. This value for the industry can be
used in conjunction with a discounted cash flow model and estimates of expected expenditures
(including taxes) and revenues to assess the relative attractiveness of a given investment using a
net present value criterion.
A.2.2 Cost of Equity
Three methods of estimating the cost of equity capital for use by private firms are
discussed below: the Capital Asset Pricing Model (CAPM), the Bond Rate Plus Four method,
and the Dividend Growth method.
In equation form, CAPM can be expressed as follows:
re = rf+B[rm-rf] ;A.2)
re = cost of equity,
T{ = risk-free rate of return (long-term treasury bonds),
rm = rate of return in the equity market generally (i.e., S&P 500), and
B = Beta, a measure of the relative risk of the equity asset.
Beta values of stocks are readily available through several sources such as The Value Line
Investment Survey.
The second method for estimating the cost of equity is called the "bond rate plus four"
method. In this method one adds 4 percentage points to the interest rate on a firm's long-term
debt to obtain the estimated return on equity. This is an ad hoc method that provides only rough
estimations of a firm's cost of equity and should only be used in conjunction with other methods.
A-3
-------
Tied as it is to the cost of debt, however, it tends to reflect the reasonable assumption that firms
with risky debt will also have risky and high cost equity.
A third method of estimating a firm's cost of equity is the Dividend Growth Method
(DGM). DGM is based on the theory that the current price of a share of stock is equal to the
present value of expected future dividend payments. This can be expressed as:
PO=^ (A.3)
where
PQ = current price of stock,
DI = expected dividend next year,
re = firm's cost of equity, and
g = expected annual growth rate of dividends.
PO is the current price of stock as quoted in any newspaper; estimates of DI can be found
in many sources, such as The Value Line Investment Survey; and g can be estimated from
historical data on dividend growth. With these values in hand it is a simple matter to calculate
the cost of equity, using the DGM.
A.2.3 Estimation of the WACC
The data and calculations used in estimating the WACC by the CAPM are presented in
Table A-l. In these calculations, a federal corporate tax rate of 34 percent is used. As an
approximation, a state corporate tax rate of 7 percent is used to derive an effective tax rate of
approximately 39 percent. The beta values (B) used to calculate the cost of equity are the overall
betas for the firms in Table A-l. The overall company beta is a measure of a firm's overall
undiversifiable risk, not the more specific municipal risk of waste management projects; betas
specific to waste management activities of the firms are not available.
As shown in Table A-l, using the average re found by the CAPM and the average yield of
bonds as rated by Standard and Poors (Table A-2), the average WACC is 14.11 percent, with a
range for individual firms of 11.43 percent to 16.80 percent.1
1 The only difference between the methods discussed above and their application in this appendix lies in computing
the cost of debt, r(d). Direct information on the yield to maturity for the long-term instruments of the firms in
A-4
-------
The calculations using the Bond Rate Plus Four method are presented in Table A-3. As
can be seen, the average return on long-term debt for the 12 firms in our sample is 11.02 percent.
The resulting re using the Bond Rate Plus Four method is 15.02 percent. The average WACC is
12.81 percent with a range of 10.20 percent to 15.27 percent for individual firms in the sample.
Table A-4 presents WACC calculations using the DGM methodology to find the return
on equity. The current price of stock is the price of the stock on January 14, 1988. DI is the
expected annual dividend as found in The Value Line Investment Survey. The expected growth
rate in the stock was derived from The Value Line Investment Survey as the growth rate expected
over the next 12 months. As can be seen, the industry average WACC using this methodology is
estimated to be 8.81 with a range of 5.18 to 14.37 percent. The DGM therefore results in much
lower WACC estimates than the other two methods.
The three estimates of WACC are simple arithmetic means and are not weighted by
market share of each firm. In addition, the firms included in the analysis are only those that
currently own or supply municipal waste management facilities, are publicly traded, and are large
enough to have the necessary financial statistics available. The WACC values calculated using
the above methodologies are nominal and not real WACC. The effects of expected inflation and
the underlying real cost of capital are discussed below.
A.3 PUBLIC COST OF CAPITAL
Public entities do not have equity investors, and some holders of certain types of
municipal debt do not pay income tax on interest received. Thus, the cost of capital values in the
financial markets appropriate to a private firm are not the relevant cost of capital values for
municipalities.
Public entities that manage municipal wastes usually finance new projects by issuing
municipal bonds. Revenue bonds are generally used rather than general obligation bonds. Thus,
the cost of debt is slightly higher than the rate on long-term general obligation municipal bonds.
A very broad range of public entities own municipal waste management facilities (Gould,
1986). Thus, average yields on municipal revenue bonds can be used as the cost of capital for
public entities that manage municipal waste. According to the Merrill Lynch Bond Index for
Table A-1 varied greatly for debt instruments across sources because of differences in reporting periods We
used the S&P bond ratings for each firm's bonds to find the yield to maturity based on the average yield in each
rating category. The yields for bond ratings AAA through BBB are directly from Standard and Poors while the
yields for lower ratings were estimated based on a previous analyses performed by RTI (1987).
A-5
-------
January 14, 1988, the yield on long-term municipal revenue bonds is 8.40 percent. This is also
the average over the past year.
A.4 INFLATION AND THE REAL COST OF CAPITAL
The WACCs estimated above for private and public entities are nominal rates of discount.
To find real or inflation-adjusted cost of capital, we must estimate investors' expected rate of
inflation over the long term.
The expected change in the rate of inflation over the next 20 years can be approximated
from the difference in current short-term and long-term interest rates after adjustment for
investors' liquidity preferences. Liquidity preference has been estimated to be about 1 percent.
However, the current difference between short- and long-term interest rates is observed to be
1.89 percent (9.11 - 7.22). This indicates that inflation is expected to be approximately 0.89
percentage points over the current level for the next 20 years.
Current inflation expectations over the short term generally match recent actual rates of
inflation. Recent actual rates of inflation have roughly averaged 3.50 percent (see Standard &
Poors Statistical Service). Therefore the expected average rate of inflation over the next 20 years
is estimated to be 4.39 percent (0.0350+0.0089).
Another way of estimating the expected rate of inflation is to assume that the real
discount rates determined by past nominal rates of return and inflation are still being earned by
today's investors. Historically, real discount rates appear to have been in the 3 to 4 percent
range. Applying the 4 percent figure to virtually risk-free, long-term treasury bills circulating
today yields an estimate of the expected rate of inflation of 4.91 percent f(0.0911 - 0.04)/1.04].
Estimates of the private and public real WACC under different estimates of expected
inflation and the nominal WACC are shown on Table A-5. The wide range in values highlights
the uncertainty associated with estimation of a real WACC, making selection of a particular
value for use in this impact analysis somewhat arbitrary. Given this uncertainty, rounded real
WACC values of 8 and 4 were selected for private and public entities respectively. The former is
in the neighborhood of estimates provided by two of the three methods used to estimate the cost
of equity capital. The latter is the rounded value of the average of the two estimates of the public
entity real WACC in Table A-5. One should note again that these estimates are approximate and
that the real WACC will vary substantially across individual firms and public entities.
A-6
-------
A.5 SOCIAL RATE OF DISCOUNT
The WACCs calculated above are not equal to the social rate of discount used in
determining the social costs in an economic impact analysis (see Section 6.4 above). While the
cost of capital calculated above is the opportunity cost of investing in a particular project, the
social rate of discount used in calculating social costs is the time value of money on the foregone
consumption and investment opportunities of society as a whole. Differences arise due to taxes
on investment income, transaction costs, and other "wedges" that distinguish private from social
rates of discount.
A-7
-------
TABLE A-l. CAPITAL ASSET PRICING METHOD: DATA AND WACC
CALCULATIONS FOR PRIVATE ENTITIES IN THE MUNICIPAL
WASTE MANAGEMENT INDUSTRY
Firm3
Allied Signal
Ashland Oil
Boeing
Comb. Eng.
Dravo
Foster Wheeler
Katy Industries
McDermott
Ogden
United Ind.
Westinghouse
Zurn
Industry Average
B
0.90
0.85
0.95
1.05
1.30
1.15
1.25
1.05
1.00
1.05
1.30
1.20
Rating
A
A
AA
A
BBB
BBB
B
BBB
BBB
BBB
AA
BB
D/V
0.39
0.34
0.05
0.25
0.20
0.29
0.11
0.53
0.38
0.12
0.49
0.08
rd<%,
10.69
10.69
10.07
10.69
11.22
11.22
12.23
11.22
11.22
11.22
10.07
11.71
re(%)
15.50
15.15
15.85
16.55
18.30
17.25
17.95
16.55
16.20
16.55
18.30
17.60
WACC (%)
12.01
12.23
15.37
14.05
16.02
14.25
16.80
11.43
12.66
15.39
12.36
16.77
14.11
a Firms in this list are compiled from trade publication discussion of industry participants for
which the requisite financial data are available. Most of these firms build and operate
municipal waste facilities, especially combustors. Financial statistics for small private
owners and operators of waste disposal facilities are not publicly available.
Sources: Standard and Poors, Inc., December 1987; Standard and Poors, Inc., September 1987-
Moody'sine., 1987
A-8
-------
TABLE A-2. STANDARD & POORS BOND RATINGS AND YIELDS
Rating Average Yield (%)
AAA 9.71
A A 10.07
A 10.69
BBB 11.22
BB 11.71
B 12.23
CCC 12.74
Sources-Standard & Poors Statistical Service, December 1987
A-9
-------
TABLE A-3. "BOND RATE PLUS FOUR" METHOD: WACC CALCULATIONS FOR
PRIVATE ENTITIES IN THE MUNICIPAL SOLID WASTE
MANAGEMENT INDUSTRY
Firm
D/V
WACC (%)
Allied Signal
Ashland Oil
Boeing
Combust. Eng.
Dravo
Foster Wheeler
Katy Industries
McDermott
Ogden
United Indust.
Westinghouse
Zurn
Industry Average
10.69
10.69
10.07
10.69
11.22
11.22
12.23
11.22
11.22
11.22
10.07
11.71
—
14.69
14.69
14.07
14.69
15.22
15.22
16.23
15.22
15.22
15.22
14.07
15.71
.39
.34
.05
.25
.20
.29
.11
.53
.38
.12
.49
.08
— .... .—-—.,..., , ,,_^
11.52
11.93
13.68
12.66
13.55
12.80
15.27
10.80
12.05
14.22
10.20
15.03
12.81
Moody s Inc., 1987
Inc., December 1987; Standard and Poors Inc., September 1987-
'
A-10
-------
TABLE A-4. DIVIDEND GROWTH MODEL METHOD: DATA AND WACC
CALCULATIONS FOR PRIVATE ENTITIES IN THE MUNICIPAL
SOLID WASTE MANAGEMENT INDUSTRY
Firm
Allied Signal
Ashland Oil
Boeing
Combust. Eng.
Dravo
Foster Wheeler
McDermott
Ogden
United Indust.
Westinghouse
Zurn
Industry Average
rd (%)
10.69
10.69
10.07
10.69
11.22
11.22
11.22
11.22
11.22
10.07
11.71
PO
31
56
46
29
13
15
16
29
13
50
23
D,
2.20
1.00
1.75
1.00
0.50
0.44
1.00
1.20
0.64
2.15
0.72
g
5.5
9.5
11.00
5.00
1.50
2.00
-3.00
10.00
4.50
12.50
3.00
r.(*>
12.60
11.29
14.80
8.45
5.35
4.93
3.25
14.14
9.42
16.80
6.13
D/V
.39
.34
.05
.25
.20
.29
.53
.38
.12
.49
.08
WACC (%)
10.24
9.68
14.37
7.98
5.65
5.50
5.18
11.38
9.12
11.60
6.22
8.81
Sources: Standard and Poors Inc., December 1987; Standard and Poors Inc., September 1987;
Moody's Inc., 1987; The Value Line Investment Survey, January 1988
A-ll
-------
TABLE A-5. ESTIMATED REAL WEIGHTED AVERAGE COST OF CAPITAL
(PERCENT)
Weighted Average
Cost of
Capital Method
Private Entities
Capita] Asset Pricing Model
Bond Rate Plus Four Method
Dividend Growth Model
4.39%
Expected
Inflation
9.31
8.07
4.23
4.91%
Expected
Inflation
8.77
7.53
3.72
Public Entities
Long-Term Revenue Bonds
3.84
3.33
A-12
------- |