Regulatory Impact Assessment
of Proposed Effluent Guidelines
for the Pharmaceutical Manufacturing Industry
Final Report
Engineering and Analysis Division
Office of Science and Technology
Office of Water
U.S. Environmental Protection Agency
Washington, DC 20460
February 1995
-------
-------
CONTENTS
EXECUTIVE SUMMARY 1
SECTION ONE INTRODUCTION '. 1-1
1.1 Purpose 1-1
1.2 Organization of the Report 1-1
SECTION TWO BACKGROUND ; . . 2-1
2.1 Industry Overview 2-1
2.1.1 Overview of the Pharmaceuticals Industry 2-1
2.1.2 Facility, Owner Company, and Parent Company
Characteristics 2-1
2.1.3 Industry Structure and the Pharmaceutical Market 2-2
2.2 Regulatory History 2-4
2.2.1 The Clean Water ActBackground 2-4
2.2.2 Pharmaceutical Manufacturing Category 2-5
2.3 Data Sources 2-6
2.3.1 The Section 308 Pharmaceutical Survey 2-7
2.3.2 U.S. Department of Commerce Data 2-10
SECTION THREE NEED FOR REGULATION 3-1
3.1 Failure of Markets to Control Pollutants 3-1
3.2 Environmental Factors 3-3
3.3 Legal Requirements 3-3
SECTION FOUR TECHNOLOGY OPTIONS AND REGULATORY
ALTERNATIVES 4-1
4.1 Technology Components 4-1
4.1.1 Advanced Biological Treatment 4-1
4.1.2 Multimedia Filtration 4-4
-------
4.13 Polishing Pond
4.1.4 Cyanide Destruction
4.1.5 Distillation '''' * to
4.1.6 Granular Activated Carbon Adsorption 4-8
4.1.7 pH Adjustment/Neutralization 4~9
4.1.8 Equalization 4~9
Summary of Regulatory Alternatives 4"10
SECTION FIVE ECONOMIC IMPACTS AND SOCIAL COSTS 5-1
5.1 Regulatory Compliance Costs 5-1
5.2 Economic Impact Analysis Methodology 5-2
5.2.1 Facility-Level Analysis Methodology 5-7
5.2.2 Owner Company-Level Analysis Methodology 5-7
5.2.3 Employment and Community-Level Analysis Methodology ... 5-9
5.2.4 Foreign Trade Impact Analysis Methodology 5-10
5.2.5 Regulatory Flexibility Analysis Methodology 5-10
5.2.6 Distributional Analysis Methodology 5-11
5.2.7 New Source Analysis Methodology 5-11
53 Economic Impact Analysis Results 5-12
53.1 Facility-Level Analysis Results 5-12
53.2 Owner Company-Level Analysis Results 5-12
533 Community-Level Analysis Results 5-18
53.4 Foreign Trade Impact Analysis Results 5-18
53.5 Regulatory Flexibility Analysis Results 5-19
53.6 Distributional Analysis Results 5-20
53.7 Impacts on New Sources 5-21
5.4 Social Costs of Regulation 5'22
SECTION SIX POLLUTANT REDUCTION 6-1
6.1 Estimates of Reductions in Pollutants
Based on the Section 308 Survey 6-1
6.2 Estimates of Reductions in Pollutants
Based on WATER7 Model
u
-------
6.2.1
Introduction 6-3
6.2.2 WATER? Model Description.'. 6-3
6.23 WATER? .Analysis Methodology 6-4
SECTION SEVEN ASSESSMENT OF BENEFITS 7-1
7.1 Introduction 7-1
7.1.1 Overview of Benefits Assessment .7-2
7.1.2 Benefit Categories Addressed in This Benefits
Assessment 7-3
7.2 Reductions in Cancer Risk , 7.4
7.2.1 Description of Benefits 7-4
7.2.2 Valuation Methodology 7-7
7.23 Valuation of Benefits 7-9
7.2.4 Limitations 7-11
73 Reductions in Emissions of Ozone Precursors 7-14
73.1 Human Health Benefits 7-15
73.2 Welfare Benefits from Increased Agricultural
Crop Yields 7-20
7.4 Environmental Benefits 7-26
7.4.1 Description of Benefits 7-27
7.4.2 Valuation Methodology 7-28
7.43 Valuation of Benefits 7-29
7.4.4 Limitations 7-30
7.5 Effects at POTWs 7.31
7.5.1 Description of Benefits 7-31
7.6 Reductions in Systemic Risk 7-35
7.6.1 Description, of Benefits 7-35
7.7 Summary of Results , 7-36
m
-------
SECTION EIGHT COMPARISON OF BENEFITS TO COSTS 8-1
SECTION NINE REFERENCES
9-1
IV
-------
TABLES
Table
ES-1 Regulatory Options Considered in the Economic Impact Analysis 5
ES-2 Compliance Costs for Selected Regulatory Options 8
ES-3 Postcompliance Analysis 1 12
ES-4 Compliance Costs for Selected Regulatory Options 18
ES-5 Estimated Pollutant Loadings by Type of Facility 20
ES-6 Potential Economic Benefits from the Proposed Effluent Guidelines
for the Pharmaceutical Industry 26
ES-7 Comparison of Annual Etenefits and Costs for the
Pharmaceutical Rulemaking 27
4-1 Summary of Major Treatment Technologies Used in the Pharmaceutical
Manufacturing Industry 4-3
4-2 Regulatory Options Considered in the Regulatorylmpact Analysis 4-11
5-1 Compliance Costs for A/C Direct Dischargers 5-3
5-2 Compliance Costs for B/D Direct Dischargers 5-4
5-3 Compliance Costs for Indirect Dischargers 5-5
5-4 Compliance Costs for Selected Regulatory Options 5-6
5-5 Postcompliance Analysis 1 5-13
5-6 Postcompliance Analysis 2 5-14
5-7 Postcompliance Analysis 3 5-16
5-8 Profitability Analysis - Percentage Decline in ROA, by Type of Facility
Owned Among Firms That Pass the Baseline Analysis 5-17
5-9 Compliance Costs for A/C Direct Dischargers 5-23
5-10 Compliance Costs for B/D Direct Dischargers 5-24
5-11 Compliance Costs for Indirect Dischargers 5-25
5-12 Compliance Costs for Selected Regulatory Options 5-26
6-1 Estimated Pollutant Loadings by Type of Facility 6-2
6-2 Number of Facilities Considered for Modeling 6-6
7-1 Example Framework of Benefit Categories 7-2
7-2 Estimated Annual Human Health Benefits from Cancer Risk Reduction 7-10
7-3 Derivation of Human Health Benefits per Megagram Reduction in VOC
Emissions in Nonattainment Areas 7-18
7-4 Estimated Annual Human Health Benefits from Reductions in VOC Emissions
hi Nonattainment Areas 7-19
7-5 Derivation of Estimated Agricultural Benefit per Megagram Reduction in
VOC Emissions 7-23
-------
Table
7-6
7-7
Estimated Annual Economic Welfare Benefits from Reductions in
Ozone-Induced impacts on Agriculture 7-25
Annual Economic Benefits from the Proposed Effluent Guidelines for the
Pharmaceutical Industry 7-37
8-1 Comparison of Annual Benefits and Costs for the
Pharmaceutical Rulemaking
8-2
VI
-------
EXECUTIVE SUMMARY
INTRODUCTION
This report has been prepared to comply with Executive Order 12866, which requires
federal agencies to assess the costs and benefits of each significant rule they propose or
promulgate. The regulations for the pharmaceutical industry, which are proposed by the U.S.
Environmental Protection Agency (EPA, or the Agency), meet the Order's definition of a
significant rule. The Agency has assessed both costs arid benefits of the proposed rule, as
presented in this Regulatory Impact Assessment (RIA).
BACKGROUND
Overview of .the Pharmaceutical Industry
More than 110,000 pharmaceutical products currently are on the market. These products
can be divided into three categories: new drugs (patented, branded drugs); generic drugs
(equivalent versions of previously patented drugs), and over-the-counter (OTC) drugs (available
without prescription). According to U.S. Department of Commerce data, 1,343 facilities involved
in pharmaceutical production existed in 1990. These faculties employed 183,000 people. Smaller
facilities (i.e., those with less than 100 employees) dominate the pharmaceutical industry. EPA
estimates that approximately 364 of the 1,343 pharmaceutical facilities are either direct or
indirect effluent dischargers and would be affected by the revised effluent regulations. The
Section 308 Survey obtained data from 244 of these establishments.
According to the U.S. Department of Commerce and other sources, the pharmaceutical
industry is considered a growth industry with above average profits, which has consistently
maintained a positive trade balance. The industry has a high concentration ratio and tends to be
vertically integrated. High R&D costs, FDA regulations, and other factors serve as barriers to
entry into the industry, although exit and entry rates into the industry are quite high.
ES-1
-------
Demand conditions vary significantly from a low level of price sensitivity for prescription
drugs to a more standard model of consumer demand for OTC drugs.. The degree of
substitutability among Pharmaceuticals ranges from reduced substitutability for many patented
drugs to substantial substitutability for OTC drugs. These factors seem to lead to price
inelasticity for Pharmaceuticals as a whole, although demand for specific drug products might be
relatively elastic. Many specific companies do appear to have sufficient market power to pass
through regulatory costs, however, this RIA uses the conservative assumption that manufacturers
cannot pass through compliance costs.
Regulatory Histoiy
The Clean Water Act (CWA),.33 U.S.C. 1251 et seq., established a comprehensive
program to "restore and maintain the chemical, physical, and biological integrity of the nation's
waters [Section 101(a)]. To implement the Act, the EPA is required to issue effluent limitations
guidelines, pretreatment standards, and new source performance standards for "categories and
classes of point sources" [Section 301(b)(2)(A)]. The Pharmaceutical Manufacturing Category is
one such category. A study of dischargers of hazardous waste indicated that the pharmaceutical
industry is indeed a major discharger of hazardous pollutants. Hence, a schedule was established
for the promulgation of effluent limitations guidelines and standards for the pharmaceutical
manufacturing industry.
Sources of Data
Prior to the data gathering for this regulation, EPA's last detailed information gathering
effort involving the pharmaceutical industry occurred in 1978. Thus EPA conducted a new
survey for this regulation, which was conducted under the authority of Section 308 of the Clean
Water Act. Through the survey, EPA obtained detailed technical and financial information from
a sample of pharmaceutical establishments that would potentially be affected by EPA's proposed
effluent guidelines.
ES-2
-------
Another major data source: used to supplement the survey data in the RIA is data from
the U.S. Department of Commerce. Commerce collects a wide range of data, such as number of
establishments, number of employees, volume of shipments, exports, imports, value added,
apparent consumption, and manufacturing costs. Other data sources used include the U.S. Food
and Drug Administration (FDA), Bureau of Labor Statistics (BLS), Dun & Bradstreet (D&B),
Robert Morris Associates (RMA), the Pharmaceutical Manufacturers Association (PMA), and
various journal articles.
NEED FOR REGULATION
Executive Order 12866 requires that the Agency identify the need for the regulation
being proposed. Hie discharge of pollutants into effluent and hence into surface water and the
emission of air pollutants pose a threat to human health and the environment. Risks from these
emissions and discharges include increases in cancer risk, other adverse noncancer health effects
on humans, and degradation of the environment.
The need for effluent guidelines for this source category arises from the failure of the
marketplace to provide the optimal level of pollution control desired by society. Correction of
such a market failure might require federal regulation. The Office of Management and Budget
defines market failures as the presence of externalities, natural monopolies, and inadequate
information. Environmental pollution is a classic example of the presence of externalities.
Furthermore, legal requirements also prevail. The regulations are proposed under the
authorities of Sections 301,304, 306,307, and 501 of the Clean Water Act (the Federal Water
Pollution Control Act Amendment of 1972, 33 U.S.C. 1251 et seq., as amended by the Clean
Water Act of 1987, Pub. L. 100-4, also referred to as the CWA or the Act) and under the
authority of Section 112 of the Clean Air Act Amendments of 1990.
ES-3
-------
TECHNOLOGY OPTIONS AND REGULATORY ALTERNATIVES
A number of alternatives are available to pharmaceutical facilities that would allow them
to meet more stringent effluent limits. Key alternatives are advanced biological treatment and
distillation. Advanced biological treatment is used in the pharmaceutical manufacturing industry
to treat biochemical oxygen demand (BODS), chemical oxygen demand (COD), and total
suspended solids (TSS) as well as to degrade various organic constituents and reduce ammonia.
They can be aerobic or, more rarely, anaerobic processes. The four most common aerobic
treatment technologies in the industry are activated sludge, aerated lagoon, trickling filter, and
rotating biological contactor (RBC). Distillation is used to remove gases and/or organic
chemicals from wastewater streams by injecting steam into a tray or packed distillation column.
Distillation is an effective treatment for a wide range of aqueous streams containing organics and
ammonia. Other approaches include multimedia filtration, polishing ponds, cyanide destruction,
granular activated carbon adsorption, pH adjustment/neutralization, and equalization.
Based on these technologies currently used in the pharmaceutical industry, EPA has
developed a set of regulatory options, which are divided into those for direct dischargers and
those for indirect dischargers. Within each discharger category, additional distinctions are made.
First, all technology options are divided between industry subcategories, with A and C industry
subcategories (representing facilities that use fermentation or biological and chemical synthesis
processes) being distinguished from B and D industry subcategories (representing facilities that
use biological or natural extractive processes or that are formulators of pharmaceutical products).
For direct dischargers, the technologies are then further broken down into BFT, BCT, BAT, and
NSPS options; for indirect dischargers, PSES and PSNS technology options are examined.
Table ES-1 presents the 37 regulatory options considered by EPA and defines the
technologies associated with each option. EPA has selected the following options for inclusion in
the regulation:
For direct discharging A/C facilities, BPT-A/C#2 is selected for conventional
pollutants and BAT-A/C#2 is required for nonconventional pollutants.
ES-4
-------
TABLE ES-1
REGULATORY OPTIONS CONSIDERED IN THE REGULATORY IMPACT ANALYSIS
Type of Option
- --.-;; w, -.-. \ . -.
' f
Best
Practicable
Technology
Best
Conventional
Technology*
Best Available
Technology
Name
*: "
BPT-A/C#1
BPT-A/C#2
BPT-A/C#3
BPT-A/C#4
BPT-A/C#5
BPT-B/D#1
BPT-B/D#2
BPT-B/D#3
BCT-A/C#1
BCT-A/C#2
BCT-A/C#3
BCT-B/D#1
BCT-B/D#2
BAT-A/C#1
BAT-A/C#2
BAT-A/C#3
BAT-A/C#4
BAT-B/D#1
BAT-B/D#2
Description
* s^
,-- l^at^*Dls«lKiigers , ^ ," ' ---", ^\z^ ^i-' :
Current biological treatment
Advanced biological treatment + cyanide destruction
Advanced biological treatment + cyanide destruction + effluent
filtration
Advanced biological treatment + cyanide destruction +
polishing pond
Advanced biological treatment 4- cyanide destruction + effluent
filtration + polishing pond
Current biological treatment
Advanced biological treatment
Advanced biological treatment 4- effluent filtration
Advanced biological treatment + effluent filtration
Advanced biological treatment + polishing pond
Advanced biological treatment + effluent filtration + polishing
pond
Advanced biological treatment
Advanced biological treatment + effluent filtration
Advanced biological treatment + cyanide destruction with
nitrification where necessary
Advanced biological treatment + cyanide destruction + in-plant
steam stripping
Advanced biological treatment 4- cyanide destruction 4- in-plant
steam stripping/distillation
Advanced biological treatment + cyanide destruction 4- in-plant
steam stripping/distillation 4- activated carbon
Advanced biological treatment
Advanced biological treatment 4- in-plant steam stripping
ES-5
-------
TABLE ES-1 (cont.)
Type of Option
Best Available
Technology
(Cont)
New Source
Performance
Standard
Name
BAT-B/D#3
BAT-B/D#4
NSPS-A/C#1
NSPS-A/C#2
NSPS-B/D#1
NSPS-B/D#2
Description
Advanced biological treatment + in-plant steam
stripping/distillation
Advanced biological treatment + in-plant steam
stripping/distillation + activated carbon
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation + activated carbon
Advanced biological treatment + in-plant steam
stripping/distillation
Advanced biological treatment + in-plant steam
stripping/distillation + activated carbon
f^^t^^^ * , """ - ''. ' ':" " " '"''""
Pretreatment
Standards for
Existing
Sources
Pretreatment
Standard for
New Sources
PSES-A/C#1
PSES-A/C#2
PSES-A/C#3
PSES-A/C#4
PSES-B/D#1
PSES-B/D#2
PSES-B/D#3
PSNS-A/C#1
PSNS-A/C#2
PSNS-A/C#3
PSNS-B/D#1
PSNS-B/D#2
In-plant steam stripping + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction +
end-of-pipe advanced biological treatment
In-plant steam stripping/distillation + cyanide destruction +
end-of-pipe advanced biological treatment + activated carbon
In-plant steam stripping
In-plant steam stripping/distillation
In-plant steam stripping/distillation + activated carbon
In-plant steam stripping/distillation + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction +
end-of-pipe advanced biological treatment
In-plant steam stripping/distillation + cyanide destruction +
end-of-pipe advanced biological treatment + activated carbon
In-plant steam stripping/distillation
In-plant steam stripping/distillation + activated carbon
*In the Development Document (EPA 1995a), BCT-A/C#1,2, and 3 in this table actually correspond
to Options 3,4, and 5, and BCT-B/D#1 and 2 in this table correspond to #2 and #3. The options not
listed in this table were never considered hi this report because they are equal to or less stringent than
the requirements of the selected BPT option, and thus no incremental costs are incurred over BPT.
ES-6
-------
For direct discharging B/D facilities, BPT-B/D#2 is selected for conventional
pollutants and BAT-B/D#1 is required for nonconventional pollutants.
NSPS-A/C#1 is selected for new A/C facilities that are direct dischargers (this
option is identical to BAT-A/C#3).
» NSPS-B/D#1 is selected for new B/D facilities that are direct dischargers (this
option is identical to BAT-B/D#3).
PSES-A/C#1 is selected for A/C facilities that are indirect dischargers.
» PSES-B/D#1 is selected for B/D facilities that are indirect dischargers.
PSNS-A/C#1 is selected for new A/C facilities that are indirect dischargers (this
option is identical to PSES-A/C#2).
PSNS-B/D#1 is selected for new B/D facilities that are indirect dischargers (this
option is identical to PSES-B/D#2).
The selected BAT options include all of the processes mandated in the selected BPT options.
ECONOMIC IMPACTS AND SOCIAL COSTS
This section presents an overview of the EIA methodology and describes the principle
models used: the cost annualization model, the facility-level model and the firm-level model.
The cost annualization model estimates the annual compliance cost to the facility of new
pollution control equipment and operations by allocating the capital investment over the lifetime
of the equipment, incorporating a cost-of-capital factor to address the costs associated with
raising or borrowing money for the investment and the tax-reducing effects of expenditures, and
including annual operating and maintenance (O&M) costs.
The annualized costs for the selected regulatory options are given in Table ES-2. The
aggregate annualized costs are $26.8 million ($30.6 million 1994 $) for BAT-A/C#2, $0.7 million
($0.8 million 1994 $) for BAT-B/D#1, $34.6 million ($39.5 million 1994 $) for PSES-A/C#1, and
$7.9 million ($9.1 million 1994 $) for PSES-B/D#1, for a total aggregate cost of $70.0 million
($80.0 million 1994 $).
ES-7
-------
TABLE ES-2
COMPLIANCE COSTS FOR SELECTED REGULATORY OPTIONS (1990 S)
Option
Number
BAT-A/C#2
BAT-B/D#1
PSES-A/C#1
PSES-B/D#1
Total
Capital Costs
$56392,127
$644,446
$70,795,915
$25,160,649
Total
O&M Costs
$35,689,088
$1,104,801
$46,441,499
$8,956,179
Total Posttax
Annualized Costs
$26,779,144
$708,758
$34,564,845
$7,922,101
Annual Cost
per Facility*
$1,115,798
$50,626
$392,782
$51,778
$152.993,137
$92.191,568
$69,974,848
$250,806
* Total Posttax Annualized Costs divided by the total number of facilities for each subcategory.
** Total number of facilities includes seven nondischarging facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. capital and operating costs estimates for pollution control
equipment.
ES-8
-------
Facility-Level Analysis
The facility-level analysis identifies facilities that are likely to close as a result of
incremental compliance cost. In the facility-level analysis, the 65 facilities (representing a total of
72 facilities) that certified in the Section 308 Pharmaceutical Survey that the regulation would
not affect them are automatically placed in the "no closure" category of the model. Additionally,
76 firm/facilities also are placed in the "no closure" category. These are firms that indicated in
the Section 308 survey that they and their facility were the same entity (i.e., the firm owns only
one facility). Impacts on firm/facilities are evaluated in the firm-level analysis. The facility
closure model thus evaluates the remaining 134 of the 282 facilities in the survey universe.
Facility closures are estimated by comparing the facility's "salvage value" (the expected
amount of cash the owner would receive if the facility were closed permanently and liquidated)
to the present value of its future earnings (the value in current dollars of the expected stream of
earnings that the facility can generate over a specified period of time). If the salvage value is
greater than what the facility is expected to generate in earnings, then it is assumed that the
owner would liquidate the facility. Salvage value includes the value of current (i.e., short-term)
assets and fixed (i.e., long-term) assets. Data for the facility-level analysis is either taken directly
from the 308 Pharmaceutical Survey or estimated based on data that was provided by other
facilities. Only those facilities estimated to remain open without any incremental regulatory costs
(i.e., remain open in the baseline analysis) are considered in the analysis to identify
postregulatory impacts.
Based on the methodology outlined above, it is estimated that none of the selected
regulatory options is expected to result in any facility closures.
Firm-Level Analysis
The firm-level analysis evaluates the effects of regulatory compliance on companies
owning one or more affected pharmaceutical facilities and identifies other impacts not captured
in the facility analysis. The analysis assesses the impacts of facility closures on each firm and the
ES-9
-------
impact of compliance costs at all facilities owned by the firm that do not close. These impacts
are assessed using ratio analysis, which employs two indicators of financial viability: the rate of
return on assets (ROA)1 and the interest coverage ratio (ICR)2. The ratio analysis simulates the
analysis an investor and/or creditor would employ in deciding whether to finance a treatment
system, or make any other investment in the firm. In the baseline ratio analysis, the company's
financial viability (before any regulatory costs are considered) is evaluated after the projected
baseline fecflity closures are accounted for. In the postcompliance analysis, those firms
determined to be viable in the baseline analysis are investigated to determine the firms' financial
condition following compliance with each option individually as well as the selected options in
combination. Data from the Section 308 survey and engineering cost estimates are used to
calculate baseline and postcompliance ROAs and ICRs. Postcompliance ROAs and ICRs are
adjusted to reflect annual compliance costs estimated at the facility level as well as losses in
income and liquidation of assets necessitated by facility closures, if any.
To evaluate the baseline and postcompliance viability of the companies analyzed, the
baseline ROA and ICR values are compared against the lowest quartile (25th percentile) values
for the pharmaceutical sector (SIC 283). Those companies for which the value of either the
ROA or the ICR is less than the first quartile value are judged to be vulnerable to financial
failure.
The standard postcompliance analysis, referred to as Postcorapliance Analysis 1, evaluates
impacts on firms that are not found to be vulnerable in the baseline analysis3. For these
healthier companies, if either of the postcompliance ROA and ICR values fall below the quartile
benchmarks, then the company is judged to be vulnerable to financial failure as a consequence of
regulatory compliance; these companies are determined to sustain a "significant impact" as a
xNet income divided by total assets.
2Eamings before interest and taxes divided by interest payments.
Sensitivity analyses were run to determine the potential for impacts on firms estimated likely
to fail in the baseline. These analyses determined the magnitude of change in ROA, ICR, or net
income, should the firm not fail as predicted. Most such firms showed no substantial change in these
variables.
ES-10
-------
result of the regulation. Table ES-3, which presents the results of Postcompliance Analysis 1
under the selected regulatory options, shows that only two firms with A/C indirect discharging
facilities and one firm with B/D indirect discharging facilities are expected to experience
significant impacts as a result of compliance costs. Overall, these indirect discharging firms
represent 2.3 percent of all affected pharmaceutical firms that are not estimated to fail in the
baseline analysis.
Finally, a profitability analysis was undertaken to determine impacts on profitability
among firms analyzed in Postcompliance Analysis 1 using percentage change in ROA under the
selected regulatory options to assess impacts on profitability (a change of more than 5 percent is
considered a major impact). Fifteen firms are estimated to experience major impacts, although
only one firm will have impacts of greater than 50 percent. Including the firms that certified
they would experience no impacts from the effluent guideline, only 11 percent of firms in the
postcompliance analysis are expected to experience major impacts short of firm failure. Note
that these impacts would be much less if it was assumed that firms could pass through some of
their compliance costs in the form of price increases. Additionally these 15 firms have a very
high baseline median ROA, so many would continue to have adequate to good returns.
Employment and Community-Level Analysis
The employment and community-level analysis investigates employment losses,
community-level impacts from these losses, and employment gains resulting from compliance
with the effluent guidelines (Baseline employment losses are determined and subtracted from
current employment before incremental employment losses are calculated). Primary and
secondary employment losses, which are the primary indicator of community-level impacts, are
measured as a direct result of facility and firm closures. Primary employment losses are based
on employee layoffs associated with the facility closures and firm failures estimated in the
facility-level and firm-level analyses. The significance of facility employment losses on the
community then is measured by their impact on the community's overall level of employment.
An increase in the community unemployment rate equal to or greater than 1 percent is
considered significant. Secondary impacts are assessed through multiplier analysis; which
ES-11
-------
TABLE ES-3
POSTCOMPLIANCE ANALYSIS 1*
SELECTED REGULATORY OPTIONS
Finns with A/C Direct Facilities
Finns with B/D Direct Facilities
Firms with A/C Indirect Facilities
Finns with B/D Indirect Facilities
Total
Number
of Firms
15
7
53
72
No Significant
Impact
#of
Firms
15
7
51
71
%of
Group
100.0%
100.0%
96.2%
98.6%
Impact
#of
Firms
0
0
2
1
133
130
97.7%
3
%of
Group
0.0%
0.0%
3.8%
1.4%
% of All
Firms**
0.0%
0.0%
1.5%
0.7%
2.3% 1 2.3%
* This scenario analyzes impacts from regulating A/C Direct facilities under options BAT-A/C#2A
and BPT-A/C#2, B/D Direct facilities .under options BAT-B/D#1 and BPT-B/D#2, A/C Indirect
facilities under option PSES-A/C#1A, and B/D Indirect facilities under option PSES-B/D#1A.
** Out of all firms in the postcompliance analysis (133 firms).
+Number of firms for All Finns might be less than the total firms by subcategory because some
firms have more man one type of facility. Total number of All Firms includes firms mat have
nondischarging facilities
Note: Analysis excludes three firms because of lack of financial data.
Source: ERG estimates.
ES-12
-------
measures the extent of impacts in other industries as a function of impacts in the primary
industry. Employment gains are calculated for three areas: manufacture of the compliance
equipment, installation of the equipment, and operation and maintenance of the equipment.
Hie baseline impacts from the analysis on primary employment before any compliance
costs are incurred total 14,381 jobs estimated to be lost, out of a total employment of 147,804
workers4 (9.7 percent of total employment). These losses are associated with 38 facility closures,
21 firm/facility failures, and 33 firm failures. The baseline analysis predicts that secondary job
losses will total 85,567.
No employment losses were projected to occur as a result of regulatory options for direct
dischargers. For indirect dischargers, however, total projected primary employment losses
resulting from the selected regulatory options were 78 full-time equivalent (FTE) positions
among A/C indirects and 13 FTEs among B/D indirects, for a total of 91 FTEs or 0.07 percent of
total employment for the affected portion of the industry. Secondary losses were predicted to be
541 FTEs.
None of these losses is expected to result in a change of employment rates of more than
1 percent in the affected communities.
The sum of primary and secondary employment gains is calculated to range from 218
FTEs to 2,890 FTEs. Net gains and losses thus range from a loss of 323 FTEs to a gain of 2,349
FTEs.
Foreign Tirade Impacts
Pharmaceutical products are traded in an international market, with producers and buyers
located worldwide. Changes in domestic pharmaceutical production due to the effluent
4In the affected portion of the pharmaceutical industry. Employment at other pharmaceutical
firms not covered by the proposed effluent guidelines is not counted here.
ES-13
-------
guidelines can therefore affect the balance of trade. To estimate impacts on trade, the value of
1990 pharmaceutical exports is estimated for facilities expected to close under the selected
regulatory options. These values are summed across facilities to obtain an estimate of the total
value of U.S. pharmaceutical exports that would no longer be produced. This value is then
compared to the total value of U.S. pharmaceutical exports produced in 1990.
The resulting impact of effluent guidelines on pharmaceutical exports and the U.S.
balance of trade is negligible under the selected regulatory options. The one firm/facility that is
predicted to close as a result of the effluent guidelines has pharmaceutical exports totaling $0.08
million ($0.09 million 1994 $). The loss of these exports will have virtually no effect on U.S.
pharmaceutical exports, which, according to the U.S. Department of Commerce, totalled $5.7
bfflion in 1991.
Regulatory Flexibility Analysis
A regulatory flexibility analysis has been conducted to ensure that small entities
potentially affected by the new effluent guidelines will not be disproportionately burdened by the
regulation.
Small firms make -up 76 percent of the 190 firms in the survey universe. The largest
percentage of firms are in the 100-499 employees size group (37 percent of all firms in the survey
universe).
The proposed effluent guidelines for the pharmaceutical industry are revisions to existing
effluent guidelines, thus most recordkeeping and reporting requirements are not incremental to
existing guidelines. The exception is new monitoring requirements. Monitoring costs total $9.0
million ($103 million 1994 $) annually, and are 15 percent of the total annual compliance cost
for the selected options. Large firms incur the largest proportion of monitoring costs (61 percent
of total monitoring costs).
ES-14
-------
No significant alternatives to the proposed rule will substantially reduce impacts on small
entities, thus the Agency believes the stated objectives of the Clean Water Act are met with this
proposed rule and the impacts to small firms have been considered, where possible.
Impacts on small firms measured as firm failure are as follows. Two of the three firms
that were projected to fail in the firm-level analysis under the selected regulatory options have
fewer than 750 employees, although only 2 percent of small firms are affected in this manner. In
addition, 14 of 15 firms found to experience a significant decline in ROA (over 5 percent) have
fewer than 750 employees. These firms represent about 14 percent of all small firms.
When cash flow is analyzed, however, impacts seem less disproportionate. Except in the
19 to 99 employees size category, the total present value of compliance costs as a percentage of
the present value of net income is smaller among small firms than among large firms. Over all
firms, the present value of compliance costs is less than 1 percent of the present value of net
income.
The above analyses indicate that although small firms do bear a large portion of the
impacts such as firm failures, these impacts are felt by a very small percentage of all small firms.
Additionally, the percentages of the present value of compliance costs to the present value of net
income are expected to be smaller, on average, among small firms than among large firms; thus,
impacts to small firms are not expected to be disproportionate to those for large firms.
Projected Distributional Impacts
For the distributional analysis, the zero cost passthrough assumption is not used. Instead,
it is assumed that manufacturers will raise pharmaceutical prices in response to increased
regulatory costs. To determine upper bound impacts, it is further assumed that all cost increases
can be passed through to consumers.
ES-15
-------
The extent to which drag prices can rise assuming perfectly inelastic demand is
determined as the ratio of total compliance costs to total cost of pharmaceutical production in
the affected facilities and in the pharmaceutical industry as a whole.
For all the selected regulatory options, the ratio of compliance costs to total
pharmaceutical costs averaged 1.6 percent Most facilities would incur compliance costs less than
1 percent of total pharmaceutical costs. Only three facilities (1 percent of all facilities) would
incur compliance costs greater than 10 percent of total pharmaceutical costs.
When possible uses for products produced by a sampling of highly affected facilities
(those where compliance costs exceed 10 percent of total pharmaceutical costs) were
investigated, it appeared that children, women, and the elderly were likely to be the major
consumers of many of these products. It was further determined that individuals who lack any
health insurance, those who are covered by government insurance, and those who are covered by
nonwork-related medical insurance might be least likely to have drug coverage. These groups
include Hispanics, young adults, African Americans, young children, and the elderly. Thus,
young adult women, children, and the elderly are likely to be the most heavily affected by
potential cost increases, if such increases can be passed through to consumers.
Because on average any potential price increases are likely to be very low (1.6 percent),
impacts on mass consumers of drugs such as HMOs, governments, and, indirectly, third-party
insurers should be minimal.
Impacts on New Sources
The selected options for new sources are NSPS-A/C#1, NSPS-B/D#1, PSNS-A/G#1, and
PSNS-B/D#1. In all cases, the requirements for new sources are more stringent than those for
existing sources. However, the difference in cost between new source requirements and existing
source requirements for typical facilities are relatively small when compared to the average
facility costs of production. In most cases, existing facilities would be required to retrofit in-plant
steam stripping systems, whereas new sources would have to install in-plant steam
ES-16
-------
stripping/distillation systems. Because designing in pollution control equipment in a new source
is typically less expensive than retrofitting the same equipment in an existing source, the cost
differential between the selected requirements for existing sources and those higher existing
source options that are technically equivalent to new source requirements should be an upper
limit on the differential annual cosfc faced by new sources. Where this differential is not
substantial relative to the typical costs of doing business in this industry, no significant barrier to
entry is likely to exist.
The average per-facility compliance costs were investigated to determine what the cost
differentials would be between proposed new source and existing source requirements. The
average per-facility cost differentials ranged from about a $34 thousand to a $590 thousand ($674
thousand 1994 $) difference (for A'C direct dischargers), depending on the type of facility. The
maximum $590 thousand differences generates the highest percentage of compliance cost
differential to pharmaceuticals manufacturing costabout 1.4 percent of total manufacturing
costs and about 3.0 percent of pharmaceutical manufacturing costs. Since this cost differential is
likely to be less than that assumed here, this small premium estimated to be paid by new sources
is not likely to have much impact on the decision to enter the market. Furthermore, these same
options, when applied to existing sources, were found to have nearly identical impacts on existing
sources as the selected options for existing sources. Thus no significant barriers to entry are
estimated to result from the proposed new source requirements.
Social Costs of Regulation
A major component of social cost (beyond the cost to industry of compliance) is the cost
to government of providing these tax savings to industry. In addition, there are other monetary
and nonmonetary outlays made by government. Government administrative costs and costs of
reallocating displaced workers are two additional monetary costs. Nonmonetary costs include
losses in consumers' or producers' surpluses in product markets, discomfort or inconvenience,
loss of time, and slowing the rate of innovation. The social costs estimated here, which include
compliance costs to industry and the costs of government tax subsidies), therefore, are a very
ES-17
-------
large portion of, but not the true total social cost of the proposed regulation. The costs reported
here are thus only a close estimate of this true cost.
To model the social costs, the entire annual pre-tax cost of the proposed effluent
guideline is estimated, using the same time period (16 years) and the discount rate advocated by
OMB as appropriate for annualizing social costs of 7 percent (rather than the higher, average
discount rate of 11.4 percent reported by Section 308 respondents).
The estimate of total annual social costs for all selected options is shown in Table ES-4.
Total social costs resulting from the proposed effluent guideline are estimated to be $108.4
million ($123.9 million 1994 $) per year.
POLLUTION REDUCTION
EPA has established final raw, current, and proposed loadings for nonconyentional
.constituents that currently are candidates for regulation in the pharmaceutical industry. EPA
calculated these final loadings using data from the Section 308 Survey database and Radian's
WATER.DBF, STEAM.DBF, and AIR.DBF treatability databases.
Because the Agency believes, however, that more air emissions are occurring from these
facilities than were reported, EPA developed an independent estimate of these loadings using the
EPA's Office of Air Quality Planning and Standards (OAQPS) WATER? model. The WATER?
model evaluates several pollutant pathways including volatilization, biodegradation, and
adsorption onto solids for individual waste constituents from a model wastewater treatment train.
Table ES-5 presents estimates of water and air pollutant loadings and reductions for the
selected regulatory scenario.
ES-18
-------
TABLE ES-4
COMPLIANCE COSTS FOR SELECTED REGULATORY OPTIONS (1990 $)
Option
Number
BAT-A/C#2
BAT-B/D#1
PSES-A/C#1
PSES-B/D#1
Total
Capital Costs
$56,392,127
$644,446
$70,795,915
$25,160,649
Total
O&M Costs
$35,689,088
$1,104,801
$46,441,499
$8,956,179
Total Pretax
Annualized Costs
$41,658,626
$1,173,021
$53,935,789
$11,619,626
Average
Annual Cost
per Facility*
$1,735,776
$83,787
$612,907
$75,945
Total**
$152,993,1371 $92,191,568
$108.387.062
$388.484
* Total Pretax Annualized Costs divided by the total number of facilities for each subcategory.
** Total number of facilities includes seven nondischarging facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. estimates of capital and operating costs for pollution control
equipment
ES-19
-------
TABLE ES-5
ESTIMATED POLLUTANT LOADINGS BY TYPE OF FACILITY
Regulatory
Option
Raw
Load
db)
Current
Load
Ob)
Total Load
Reduction
flb)
A/C Direct Dischargers
BAT-A/C#1
BAT-A/C#2
BAT-A/C#3
77,906,785
5,133,649
4,894,776
18,430,143
18,443,911
B/D Direct Dischargers
BAT-B/D#1
BAT-B/D82
BAT-B/D#3
109,252
23,230
22,610
45,152
45,204
A/C Indirect Dischargers
PSES-A/C#1
PSES-A/C#2
PSES-A/C#3
90,181,808
33,181,762
38,331,249
50,981,759
51,807,047
B/D Indirect Dischargers
PSES-B/D#1
PSES-B/D#2
7,424,870
1,934,646
4,378,836
4.990,451
Source: Section 308 Survey data and Radian Corp. estimates.
ES-20
-------
ASSESSMENT OF BENEFITS
This benefits assessment considers the human health, environmental, and economic
benefits from reductions in effluent loadings and air emissions expected to result from the
proposed rule. A variety of human health, environmental, and economic benefits might result
from these reductions in effluent loadings and VOC emissions. In particular, the benefits
assessment addresses the following benefit categories:
Human health and agricultural benefits due to reductions in emissions of ozone
precursors (i.e., reductions in VOC emissions)
Human health benefits due to reductions in excess cancer risk
Ecological and recreational benefits due to improved water quality
Benefits from reductions in Interference problems and improvements in worker
health and safety at publicly owned treatment works (POTWs)
Human health benefits due to reductions in systemic risk
Estimated benefits are monetized for the first two benefit categories, but the dollar magnitude of
benefits for the three other benefit categories could not be quantified.
Benefits from VOC Emissions Reductions
The largest category of monetized benefits from the proposed guidelines results from
reductions in ambient ozone concentrations due to reductions in VOC emissions. Controlling
VOC emissions is beneficial because VOCs are precursors to ozone, which negatively affects
human health and the environment. Studies have demonstrated that short-term exposure to
elevated ozone concentrations results in acute effects on human health. Ozone also is believed
to have chronic effects on human health. The annual human health benefits resulting from
reductions in VOC emissions due to the proposed rule ranges from $27 thousand to $1.7 million
in 1990 dollars ($31 thousand to $1.9 million 1994 $).
ES-21
-------
In addition to health effects, studies of the relationship between ambient ozone
concentrations and greenhouse-controlled ozone concentrations and agricultural crop yields
demonstrate that ozone negatively affects crop yields. Reductions in crop yields can in turn
affect agricultural production, crop prices, and incomes of agricultural producers, and thus can
affect social welfare. Thus, reductions in ozone concentrations that lead to improved crop yields
will generate welfare benefits. Furthermore, ozone-induced crop yield changes can have
secondary effects due to the responses of the agricultural community to the yield changes,
including increased fertilizer and pesticide use, loss of wildlife habitat, increased soil erosion, and
increased surface water and groundwater pollution. The annual agricultural-related economic
welfare benefits from reductions in VOC emissions are estimated to range from $163 thousand
to $276 thousand in 1990 dollars ($186 thousand to $315 thousand 1994 $).
Cancer Risk Reduction Benefits
The benefits from the proposed rule include human health benefits from reductions in
excess cancer risk. These benefits result from reduced human exposure to carcinogenic
contaminants through inhalation. The proposed regulations on the effluent discharges of
pharmaceutical facilities are expected to remove many toxic substances that otherwise would
volatilize and pose cancer risk to humans. Reductions in emissions of these carcinogens are
expected to result in reductions in excess cancer risk in exposed populations. Based on the
cancer risk assessment conducted for the RIA, the proposed guidelines are expected to result in
0.02 to 035 excess cancer cases avoided per year nationwide. The estimated value of the human
health benefits from the cancer risk reductions associated with this rule ranges from $12
thousand to $4.7 million in 1990 dollars ($14 thousand to $5.4 million 1994 $) annually.
Human Health Benefits from Reductions in Noncarcinogenic Risk
Exposure to toxic substances poses risk of systemic and other effects to humans, including
effects on the circulatory, respiratory or digestive systems and neurological and developmental
ES-22
-------
effects. The proposed rule might generate human health benefits by reducing exposure to these
substances, thus reducing the risks of these associated effects.
Reductions in air emissions are expected to result in reduced systemic risk, with benefits
ranging from reduced risk to zero to 126,000 individuals due to reduced exposure to two toxic
pollutants. No systemic risk reductions are expected to result from reduced exposure to
contaminated fish tissue or drinking water because estimated concentration levels under current
conditions are below human health criteria or toxic effect levels. Sufficient data to quantify these
benefits further are not available.
Ecological and Recreational Benefits due to Improved Water Quality
EPA expects the proposed effluent guidelines to generate environmental benefits by
improving water quality. There are a wide range of benefits associated with the maintenance and
improvement of water quality. These benefits include use values (e.g., recreational fishing),
ecological values (e.g., provision of habitat), and passive use values. For example, water
pollution might affect the quality of the fish and wildlife habitat provided by water resources,
thus affecting the species using these resources. This in turn might affect the quality of
recreational experiences of users, such as anglers fishing in the affected streams. In the RIA,
EPA considers the value of the recreational benefits resulting from the proposed rule, but does
not evaluate the other types of ecological and environmental benefits due to data limitations.
The projected reductions in toxic loadings to surface waters are significant. Pollutant
loadings are estimated to decline by 57 percent, from 39.9 million pounds per year under current
conditions to 17.1 million pounds per year under the proposed rule. The analysis comparing
instream concentration levels to aquatic life water quality criteria estimates that current discharge
loadings result in excursions of aquatic water quality criteria at two locations. The analysis also
indicates that no excursions are expected to occur at these two sites under the proposed rule.
EPA estimates that the annual recreational benefits associated with the expected changes
in water quality are on the order of thousands of dollars. EPA evaluates these recreational
ES-23
-------
benefits, applying a simple model that considers the change in consumer welfare likely to result
from improved catch rates by recreational anglers at these two sites. EPA assumes that catch
rates improve due to larger fish populations that are assumed to result from improved water
quality.
Benefits from Reductions in Loadings Dischaiged to POTWs
The RIA considers three potential sources of benefits to POTWs from the proposed
regulation: reductions in the likelihood of interference, pass through, and sewage sludge
contamination problems; reductions in health and safety risks to POTW workers; and reductions
in costs potentially incurred by POTWs in analyzing toxic pollutants and determining whether,
and the appropriate level at which, to set local limits. Although the benefits from reducing these
effects at POTWs might be substantial, the RIA does not quantify these benefits due to data
limitations.
Toxic pollutants contained in the effluent loadings of pharmaceutical plants and
discharged to POTWs might cause interference problems and/or pass through a POTW's
treatment system and potentially affect water quality or contaminate sewage sludges. These
problems might affect POTWs directly to the extent that they prevent POTWs from meeting
their permits or sewage sludge criteria and they might affect surface water quality. The proposed
rule is expected to help reduce these problems by reducing toxic loadings in the industry's
effluent Furthermore, the proposed rule might help to reduce shock releases (i.e., unexpected
releases that contain high concentrations of toxic pollutants) from pharmaceutical facilities and
thus reduce the likelihood that these releases will cause interference, passthrough and sewage
sludge contamination problems at POTWs.
Anecdotal evidence from POTW responses to an EPA survey and analytic results indicate
that such effects can occur. In addition, based on an analysis comparing POTW influent levels to
available data on inhibition levels, inhibition problems are projected to occur at six POTWs for
seven pollutants under current conditions. Inhibition problems are projected to occur at five
ES-24
-------
POTWs for three pollutants after the proposed rule. Sufficient data are not available to further
quantify this benefit category.
Toxic substances in effluent discharges to POTWs pose health risks to POTW workers.
For example, volatilization of toxics from POTW influent can pose a cancer risk to POTW
workers or increase the risk of explosion at the plant. The proposed rule is expected to reduce
these risks, thus generating human health benefits. Based on the assessment of the risk posed to
POTW workers from exposure to toxic pollutants, the proposed rule is estimated to reduce
occupational risk at six POTWs. Data are not available to monetize this benefit category.
Benefits from Reductions in Analytical Costs
In implementing local programs to control pollutants discharged to their systems,
authorized POTWs often must set numerical limits on toxic loadings in discharges to the POTW,
based on national categorical pretreatment standards or local limits determined by the POTW.
In setting these local limits, POTWs sometimes need to undertake analyses to determine which
pollutants warrant local limits and at what numerical level. Conducting these analyses is
expensive, costing on the order of hundreds of thousands of dollars. Several POTWs contacted
as part of EPA's survey of POTWs indicated that they will benefit from the establishment of
national pretreatment standards by avoiding these analytical costs. In addition, they indicated
that the pretreatment standards will; bolster the legal authority of the limits they set.
Summary of Benefits
EPA estimates that the annual benefits resulting from the proposed rule will range from
$202 thousand to $6.7 million in 1990 dollars ($231 thousand to $7.6 million 1994 $). Table
ES-6 summarizes these benefits, by category. The range reflects the uncertainty in evaluating the
effects of the proposed rule and in placing a dollar value on these effects. As indicated in the
table, these benefits ranges do not reflect many of the benefit categories expected to result under
the proposed rule, including human health benefits associated with potential reductions in
ES-25
-------
TABLE ES-6
POTENTIAL ECONOMIC BENEFITS FROM THE PROPOSED EFFLUENT GUIDELINES
FOR THE PHARMACEUTICAL INDUSTRY
Benefit Categoiy
Reductions in Emissions of Ozone
Precursors:*
Human Health
Agricultural
Cancer Risk Reductions
Environmental
Avoidance of Interference Problems
and Improvements in Worker
Health and Safety at POTWs
Systemic Risk Reductions
Total
Thousands of 1990
dollars per year
$27 - $1,688
$163 - $276
$12 - $4,725
Unknown
Unknown
Unknown
$202 - 6,689
*The estimates presented only include benefits associated with reductions in acute health
effects and agricultural-related welfare benefits in nonattainment areas. Potential welfare
benefits associated with forest yield, materials damage, and visibility are not addressed in this
analysis.
ES-26
-------
chronic effects from ozone exposure, human health benefits associated with reductions hi acute
effects from ozone exposure hi attainment areas, agricultural-related economic benefits from
reductions hi emissions of ozone precursors in attainment areas, ecological and recreational
benefits from improvements in water quality, benefits from avoided interference and passthrough
problems and unproved worker health and safety at POTWs, and human health benefits from
potential reductions in systemic risk. Therefore the reported benefit estimate understates the
total benefits of the proposed rule.
COMPARISON OF COSTS TO BENEFITS
In this section, the estimate of the annual social costs of the regulation is compared to
the estimate of the total annual benefits. Table ES-7 presents the annual benefits and annual
costs of the proposed guideline in 1990 dollars. Benefits range from $202 thousand to $6.7
million ($231 thousand to $7.6 million 1994 $) annually. Costs total $108.4 million ($123.9
million 1994 $) annually.
ES-27
-------
TABLE ES-7
COMPARISON OF ANNUAL BENEFITS AND COSTS
FOR THE PHARMACEUTICAL RULEMAK1NG
(in thousands of 1990 dollars)
Benefits
Cancer risk reductions
Reductions in emissions of ozone precursors
Human health
Agricultural benefits
Total quantifiable benefits
$12 - $4,725
$27 - $1,688
$163 - $276
$202 -$6,689
B Costs
Total Annual Costs to Industry
Total Annual- Social Costs
$70,000
$108,400
ES-28
-------
SECTION ONE
INTRODUCTION
1.1 PURPOSE
This report has been prepared to comply with Executive Order 12866, which requires
federal agencies to assess the costs and benefits of each significant rule they propose or
promulgate. The regulations for the pharmaceutical industry, which are proposed by the U.S.
Environmental Protection Agency (EPA, or the Agency), meet the Order's definition of a
significant rule. The Agency has assessed both costs and benefits of the proposed rule, as
presented in this Regulatory Impact Assessment (RIA).1
1.2 ORGANIZATION OF THE REPORT
The principal requirements of the Executive Order are that the Agency perform an
analysis comparing the benefits of the regulation to the costs that the regulation imposes, that
the Agency analyze alternative approaches to the rule, and that the need for the regulation be
identified. Wherever possible, the costs and benefits of the rule are to be expressed in monetary
terms. To address the analytical requirements of the Executive Order, this RIA is organized into
seven major sections:
Background
Need for the Regulation
Technology Options and Regulatory Alternatives
Economic Impacts aind Social Costs
lrThe reader is referred to the Economic Impact Analysis (U.S. EPA, 1995b) for more detailed
information on economic impacts, the profile of the pharmaceutical industry, and other sources of
data.
1-1
-------
Pollutant Reduction
Assessment of Benefits
Comparison of Benefits to Costs
Section Two (Background) presents an overview of the pharmaceutical industry and
describes the regulatory history of the proposed rulemakmg.
Section Three (Need for the Regulation) briefly explains the marketplace failures that the
proposed pollution control regulations are intended to correct. In addition, this section discusses
the environmental factors necessitating the development of the rulemaking. Finally, the Agency's
legal mandate for developing the regulation is summarized.
Section Four (Technology Options and Regulatory Alternatives) describes the options
considered in the development of the proposed effluent guidelines and emission standards.
Section Five (Economic Impacts and Social Costs) presents the costs of compliance with
the proposed regulations, the methodology and results of the economic impact analysis, and
estimates of the social costs associated with the proposed effluent guidelines and emissions
standards.
Section Six (Pollutant Reduction) presents proposed reductions in conventional and
nonconventional constituents (including priority pollutants) released, by the pharmaceutical
industry. Estimates are based on current loading as reported by each survey respondent in the
Section 308 Pharmaceutical Industry Survey. These loadings also are compared to independently
estimated loadings generated through computer modeling.
Section Seven (Assessment of Benefits) presents qualitative and quantitative estimates of
the human health and air and water quality benefits of the proposed rule.
Section Eight (Comparison of Benefits to Costs) compares annualized benefits and costs
and discusses the context within which these results should be interpreted.
References are provided in Section Nine.
1-2
-------
SECTION TWO
BACKGROUND
2.1 INDUSTRY OVERVIEW
This section presents a profile of the pharmaceutical industry that covers the statistical
and descriptive information used to develop the methodology for the RIA. The profile includes
an overview of the industry; a variety of statistics at the facility, owner company, and parent
company level; and information on market structure and demand.
2.1.1 Overview of the Pharmaceuticals Industry
More than 110,000 pharmaceutical products currently are on the market. These products
can be divided into three categories: new drugs (patented, branded drugs); generic drugs
(equivalent versions of previously patented drugs), and over-the-counter (OTC) drugs (available
without prescription). Drugs are manufactured using an array of complex batch-type processes
and technologies that occur in three main stages: research and development (R&D);
fermentation, extraction, and chemical synthesis, which covers the conversion of organic and
chemical substances into bulk active ingredients; and formulation, which refers to the combining
of bulk active ingredients with other substances to produce proper dosages.
2.1.2 Facility, Owner Company, and Parent Company Characteristics
According to U.S. Department of Commerce data, 1,343 facilities involved in
pharmaceutical production existed in 1990. These facilities employed 183,000 people. Smaller
facilities (i.e., those with less than 100 employees) dominate the pharmaceutical industry,
although a higher percentage of facilities in the pharmaceutical industry have more than 250
employees than in the manufacturing sector overall. EPA estimates that approximately 286 of
2-1
-------
the 1^43 pharmaceutical facilities are either direct or indirect effluent dischargers and would be
affected by the revised effluent regulations. The Section 308 Survey obtained data from 244 of
these establishments.
U.S. Department of Commerce data indicate that the value of shipments for the drug
industry were $64.1 billion in 1992. In real terms, growth has averaged 2 to 4 percent annually
for the pharmaceutical industry. The Section 308 Survey data indicate that pharmaceutical
facility revenues average approximately $100 million per fecility per year, while average revenues
for owner companies are approximately $600 million. The U.S. pharmaceutical industry also has
consistently maintained a positive balance of trade, with a trade surplus of $961 million in 1991.
According to the Section 308 Survey, the mean pharmaceutical export rate for sample facilities
was 8.8 percent in 1990.
Manufacturing costs for the pharmaceutical industry from 1988 to 1990 rose from $7.4
billion to $9.6 billion at the facility level, from $58.7 billion to $63.8 billion at the owner-company
level, and from $149.1 billion to $177.3 billion at the parent company level. In addition, the
research and development expenditures for the Pharmaceuticals industry are more than 16
percent of sales, one of the highest proportions for any U.S. industry, while promotional
expenditures account for approximately 22 percent of the industry's revenues.
Data from the Section 308 Survey indicate that the median rate of return on assets by
asset site groups facilities level from 1988 to 1990 ranged from approximately 5 percent to 7
percent. The median interest coverage ratios by asset size groups vary from approximately 464
percent to 2,043 percent. In addition, the profitability of the pharmaceutical industry appears to
be above average among U.S. industries.
2.L3 Industry Structure and the Pharmaceutical Market
Although the number of pharmaceutical facilities has grown over the past several
decades, it is likely that competition would have been greater in the industry if high R&D costs,
EDA regulations, and other factors did not serve as barriers to entry into the industry. In
2-2
-------
addition, concentration ratios in the pharmaceutical industry, as well as exit and entry into the
industry, are quite high. There also is some indication that pharmaceutical companies are
vertically integrated. These factors all affect entry of new firms into the pharmaceutical market.
Demand conditions vary significantly among specific drug markets. In the prescription
drug market, demand is complicated by the role of health care providers and the presence of
health insurance, which reduce the competitive nature of the market. The lack of price
sensitivity among consumers, however, is partly offset by increasing sensitivity among insurers.
Demand for OTC drugs, on the other hand, conforms more readily to standard models of
consumer demand.
The degree of substitutability among Pharmaceuticals varies. Patented drugs in the
United States enjoy ostensible protection from bioequivalent drugs for a number of years, which
obviously reduces direct substitutability. The increase in generic drugs, however, increases
substitutability once the patent for a drug expires. For OTC drugs, the market is much like
other competitive commodity markets, with a high degree of substitutability causing demand to
be relatively sensitive to price changes. In addition, Pharmaceuticals are not a very close
substitute for most other forms of medical treatments, although they might act as complements.
These factors seem to lead to price inelasticity for Pharmaceuticals as a whole. Available
studies indicate that the demand for Pharmaceuticals as a group may be quite inelastic (i.e.,
between 0 and -1.0). Demand for specific drug products, however, may be relatively elastic (i.e.,
less than -1.0). The absence of close substitutes for drug therapies in general and the presence
of health insurance probably explains that inelasticity of demand for Pharmaceuticals. The
existence of close substitutes for individual drugs and the pressure to control health care costs,
on the other hand, probably explains the relative elasticity of demand for specific drugs.
Because regulatory costs associated with new effluent standards can affect a large portion
of the pharmaceutical industry, the industry as a whole might be able to pass through regulatory
costs to consumers in the form of higher drug prices. Individual companies, however, will have
less latitude in passing through costs, although many specific companies do appear to have
sufficient market power to pass through regulatory costs. Throughout the RIA, however, the
2-3
-------
conservative assumption that manufacturers cannot pass through compliance costs is used, except
where impacts on consumers are analyzed {100 percent cost passthrough is assumed for this
analysis).
22 REGULATORY HISTORY
The Clean Water Act Background
The Clean Water Act (CWA), 33 U.S.C. 1251 et seq., established a comprehensive
program to "restore and maintain the chemical, physical, and biological integrity of the nation's
waters [Section 101(a)]. To implement the Act, the EPA is required to issue effluent limitations
guidelines, pretreatment standards, and new source performance standards for "categories and
classes of point sources" [Section 301(b)(2)(A)]. The Pharmaceutical Manufacturing Category is
one such category. The Act calls for limitations to be based on Best Practicable Technology
(BPT), Best Available Technology Economically Achievable (BAT), Best Conventional Pollutant
Control Technology (BCT), New Source Performance Standards (NSPS), Pretreatment Standards
for Existing Sources (PSES), and Pretreatment Standards for New Sources (PSNS).
The CWA establishes a BAT standard for the control of any of 65 "priority" (toxic)
pollutants or classes of pollutants, as well as nonconventional pollutants directly discharged by an
industrial category to navigable waters. In addition to considering economic achievabiliry, the
EPA, in assessing BAT, is to consider factors including the following: the age of the equipment
and facilities involved, the process employed, process changes, and nonwater quality
environmental impacts [Section 304(b)(2)(B)]. A BCT standard replaces BAT for the discharge
of conventional pollutants from existing sources. Along with other factors specified in Section
304(b)(4)(B), the Act requires BCT limitations to be assessed in light of a two-part "cost-
reasonableness" test In addition, in setting new source performance standards, EPA is directed
by Section 306 to consider, among other things, the cost of achieving effluent reductions.
Section 304(m) of the CWA, added by the Water Quality Act of 1987 (P.L. 100-4,
February 4, 1987), requires EPA to review existing effluents limitations guidelines and standards
2-4
-------
and to promulgate new or revised effluent guidelines and standards as necessary. On January 2,
1990, EPA published its Effluent Guidelines Plan in the Federal Register (55 ER 80) in which a
schedule was established for the promulgation of effluent limitations guidelines and standards for
several industry categories, including the pharmaceutical manufacturing industry.
Pharmaceutical Manufacturing Category
On October 27,1983, EPA published final BPT, BAT, PSES, and NSPS limitations and
standards for the Pharmaceutical Manufacturing Category (40 CFR Part 439). On December 16,
1986, EPA promulgated final BCT limitations essentially equivalent to the existing BPT
limitations for this category. These final limitations and standards did not include any controls
on the discharge of toxic volatile organic compounds (VOCs). On September 9,1985 EPA
stated in a Federal Register notice (50 FR 36638) that it was considering controls on the discharge
of methylene chloride and other volatile VOCs from pharmaceutical point sources. In the 1986
Domestic Sewage Study (DSS) Report to Congress (EPA, 1986),1 the Agency indicated that the
Pharmaceutical Industry was a major source of hazardous pollutant discharges to the nation's
Publicly Owned Treatment Works (POTWs). These discharges include both priority and
nonconventional VOCs.
/
One of the recommendations of the DSS was that industry categories that were shown to
be significant dischargers of hazardous waste to the nation's POTWs should undergo additional
study. The purpose of these studies was to determine whether pretreatment standards
controlling the discharge of hazardous pollutants should be promulgated for specific categories.
As stated above, pharmaceutical manufacturing category was judged to be a major discharger of
hazardous pollutants by the DSS. Consequently, the Industrial Technology Division (ITD) of
EPA conducted a sampling and data gathering effort mainly among indirect discharging
pharmaceutical manufacturing plants (some direct dischargers were also sampled) in order to
obtain more information about hazardous pollutant discharges from pharmaceutical facilities and
1The study was conducted pursuant to the 1984 amendments to the Resource Conservation
and Recovery Act (RCRA).
2-5
-------
to determine if additional pretreatment standards designed to control the discharge of hazardous
pollutants should be promulgated. The pharmaceutical sampling/study program was part of an
overall effort known as the ITD/Resource Conservation and Recovery Act (RCRA) program and
was carried out within ITD in response to the DSS.
The preliminary results of this study indicated that the pharmaceutical industry is indeed
a major discharger of hazardous pollutants. Hence, as discussed in the last paragraph of Section
2.2.1, a schedule was established for the promulgation of effluent limitations guidelines and
standards for the pharmaceutical manufacturing industry.
Prior to the data gathering for this regulation, EPA's last detailed information gathering
effort involving the pharmaceutical industry occurred in 1978 and utilized a questionnaire that
requested information about water use, wastewater characteristics, raw materials usage, treatment
practices, and discharge status. The information obtained from the census was an essential part
of the database used to promulgate the 1983 and 1986 regulations. However, in the course of
conducting the recent ITD/RCRA sampling program, the Agency learned that the industry had
changed in many respects since 1978, and much of their information was out of date.
EPA, therefore, conducted a new survey for this regulation. The Agency conducted its
information collection activity via a two step process. The first step involved a screener
questionnaire, which was sent to about 1,130 known pharmaceutical manufacturing facilities.
This screener was administered in 1989 (OMB No. 2040-0124). The results of this screener were
used to identify the pharmaceutical manufacturers that were to be a part of the second step, the
Section 308 Pharmaceutical Survey, as discussed in the following section.
23 DATA SOURCES
The EIA (U.S. EPA 1995b), the basis of much of this RIA, relies on a variety of data
sources including the Section 308 Pharmaceutical Survey conducted specifically for this regulatory
development effort, the U.S. Department of Commerce, the U.S. Food and Drug Administration
(FDA), Bureau of Labor Statistics (BLS), Dun and Bradstreet (D&B), Robert Morris Associates
2-6
-------
(RMA), the Pharmaceutical Manufacturers Association (PMA), and various journal articles.
Most of the analysis conducted in Chapters Four through Eleven make extensive use of the data
collected from the Section 308 Pharmaceutical Survey. Other data sources were used primarily
in the development of the industry profile in Chapter Three. Data gathered in the profile,
however, provides the foundation for much of the analysis in later Sections.
The following sections describe the two principal data sources for this EIA: the Section
308 Pharmaceutical Survey and sources available through the U.S. Department of Commerce.
Other data sources are described, as necessary, as they are used to support the analyses in
subsequent sections.
2.3.1 The Section 308 Pharmaceutical Survey
2.3.1.1 Background
The Section 308 Pharmaceutical Survey obtained detailed technical and financial
information from a sample pharmaceutical establishments potentially affected by EPA's proposed
effluent guidelines. EPA stratified the industry into five groups based on type of operation:
A) Fermentation
B) Biological and natural extraction
C) Chemical synthesis
D) Formulation and mixing/compounding
E) Research
The stratification permitted EPA to census (i.e., survey all facilities) facilities within some
subcategories and sample facilities within others. EPA took a census of all facilities that
manufacture active ingredients (subcategories A, B, C) and discharge process wastewater and
formulating and mixing/compounding (subcategory D) facilities that have direct discharges or a
2-7
-------
combination of direct and indirect discharges. EPA judged that a census of these facilities was
necessary because the overall universe was small, few facilities were in the same combination of
subcategories, and each facility was expected to have wastewater generated by proprietary
processes which would make their effluent significantly different from other facilities in the same
subcategory. Overall, EPA conducted a census of 202 facilities in these three subcategories
(EPA, 1990).
EPA also censused subcategory D stand-alone facilities that use solvents and discharge
indirectly and subcategory D facilities with onsite research facilities (i.e., subcategory D/E) that
use solvents and discharge indirectly, and have less than 19 employees or more than 747
employees. For subcategory D indirect discharging facilities with between 19 and 168 employees
and between 169 and 747 employees, EPA used a sampling methodology. Tlie sampling
methodology stratified these facilities by employee size based on a linear regression between the
log of the number of employees and log of the flow rate. Employee and flow rate data were
available from EPA's Development Document for Effluent Limitations Guidelines and Standards for
the Pharmaceutical Point Source Category (1982). Overall, EPA sampled 42 pharmaceutical
facilities in subcategories D and D/E. Survey results used throughout the EIA are weighted
according to the sampling plan. Subcategory D and D/E facilities with between 19 and 747
employees received a sample weight of 2. All other censused facilities received a weight of 1.
The coefficient of variation in any particular strata, i.e., employment size, is no greater than 15
percent All subcategory D facilities are grouped with subcategory B facilities for the purpose of
this analysis, which is discussed in Section Four.
EPA determined that no information was needed from three groups of pharmaceutical
facilities:
Facilities that do not discharge wastewater
2-8
-------
Facilities that do not use solvents and whose only source of process wastewater is
from formulation and mixing/compounding
Stand-alone research facilities
These facilities do not require effluent guidelines because their impact on water quality and
POTW operations is considered to be negligible.
23.1.1 Uses of Survey Data
Certifying Facilities. All surveyed facilities were given the option to certify that the
facility would incur no significant economic impact as a result of the effluent guidelines. These
facilities gave up their right to challenge aspects of the effluent guidelines based on economic
achievability so long as the cost of compliance of the guidelines ultimately promulgated by EPA
does not exceed the compliance cost estimated in the survey. Certifying facilities were excused
from completing the bulk of the financial questionnaire. Sixty-five of the surveyed facilities
.certified no significant economic impact and thus did not provide financial data.
Responding Firms and Facilities. The survey data from firms and reporting facilities were
used extensively in the development of BPT, BCT, BAT, NSPS, PSES, and PSNS regulations for
the industry. Surveyed facilities provided technical information on pharmaceutical products;
compound and chemical usage and disposition; waste minimization and pollution prevention
activities; wastewater generation, collection, and conservation; wastewater treatment; steam
stripping; and wastewater characteristics. The survey also collected financial data such as number
of employees; ownership structure; discount rate; market value of land, buildings, and equipment;
value of shipments; manufacturing costs; assets; liabilities; and net income. Financial data were
collected at the facility, owner-company, and parent company levels.
2-9
-------
233. U.S. Department of Commerce Data
The EIA supplements financial data collected in the Section 308 Pharmaceutical Survey
with data from the U.S. Department of Commerce. Commerce divides the pharmaceutical
industry into four major Standard Industrial Classifications (SICs):
SIC 2833 Medicinal and Botanical. Establishment primarily engaged in : (1)
manufacturing bulk organic and inorganic medicinal chemicals and their
derivatives and (2) processing bulk botanical drugs and herbs.
SIC 2834 Pharmaceutical Preparations. Establishments primarily engaged in
manufacturing, fabricating, or processing drugs in pharmaceutical preparations for
human or veterinary use. The greater part of the products of these establishments
are finished in the form intended for final consumption, such as tablets, capsules,
liquids, etc. These pharmaceutical preparations are promoted to the medical
profession (prescription drugs) and the general public [over-the-counter (OTC)].
,SIC 2835 In Vitro and In Vivo Diagnostic Substances. Establishments engaged in
the manufacturing chemical, biological, and radioactive substances used in
diagnosing or monitoring human and animal health by identifying and measuring
normal and abnormal constituents of body fluids or tissues.
SIC 2836 Biological Products, Except Diagnostic Substances. Establishments
engaged primarily in the production of bacterial and virus vaccines, toxoids and
analogous products, serums, plasmas, and other blood derivatives for human and
veterinary use.
Commerce collects a wide range of data at the 4-digit SIC level including number of
establishments, number of employees, volume of shipments, exports, imports, value added,
apparent consumption, manufacturing costs, and .other data. Commerce further segments the
pharmaceutical into 14 5-digjt and hundreds of 7-digit SIC codes. Comprehensive financial data
at the 5 and 7-digit levels, however, is available only under SIC 2834 Pharmaceutical
Preparations. Commerce data are reported in publications such as the Census of Manufactures,
County Business Patterns, and U.S. Industrial Outlook. The EIA uses the most current available
data from these sources in the development of the industry profile.
2-10
-------
Numerous other data sources employed by the EIA also are organized by SIC code. For
example, price indices generated by BLS are reported according to SIC code. Financial ratio
data reported by D&B and RMA also are organized by SIC.
A major difficulty with using data organized by SIC, however, is its inability to capture all
establishments engaged in the production of Pharmaceuticals. Commerce classifies facilities by
their primary line of business. Thus, only establishments that garner at least 50 percent of then-
revenues from pharmaceutical-related business are classified in the four pharmaceutical SIC
codes. Facilities that manufacture pharmaceuticals but list some other line of business (e.g.,
chemical production) as their primary SIC are not captured in the four pharmaceutical SICs.
facilities that manufacture pharmaceuticals but whose primary business is classified in some other
SIC code. Thus, Commerce data do not provide a complete picture of the U.S. pharmaceutical
industry.
The Section 308 Pharmaceutical Survey data cover only a subset of the pharmaceutical
industry. The five categories used to segment the pharmaceutical industry in the survey do not
correspond with the four pharmaceutical SICs. Moreover, surveyed facilities were not asked to
report their SIC. Thus, no direct comparison can be made between Commerce and survey data.
2-11
-------
-------
SECTION THREE
NEED FOR REGULATION
Executive Order 12866 requires that the Agency identity the need for the regulation
being proposed. The discharge of pollutants into effluent and hence into surface water and the
emission of air pollutants pose a threat to human health and the environment. Risks from these
emissions and discharges include increases in cancer risk, other adverse noncancer health effects
on humans, and degradation of the environment. This section discusses: (1) the reasons the
marketplace does not provide for adequate pollution control absent appropriate incentives or
standards; (2) the environmental factors that indicate the need for additional pollution controls
for this source category; and (3) the legal requirements that dictate the necessity for and timing
of this regulation.
3.1 FAILURE OF MARKETS TO CONTROL POLLUTANTS
The need for effluent guidelines for this source category arises from the failure of the
marketplace to provide the optimal level of pollution control desired by society. Correction of
such a market failure can require federal regulation. The Office of Management and Budget
defines market failures as the presence of externalities, natural monopolies, and inadequate
information (U.S. Office of Management and Budget, 1989). This section addresses the category
of externalities, which is the category of market failure most relevant to the general case of
environmental pollution.
The concept of externalities partially explains the discrepancy between the supply of
pollution control provided by owners and operators of pollution sources and the level of
environmental quality desired by the general population. The case of environmental pollution
can be classified as a negative externality because it is an unintended byproduct of production
that creates undesirable effects on human health and the environment.
3-1
-------
In making production decisions, owners and operators will only consider those costs and
benefits that accrue to them personally (i.e., internalized costs and benefits). However, the cost
of environmental pollution is not borne solely by the creators of the pollution because all
individuals in the polluted area must share the social cost of exposure to the pollution, even if
they had no part in creating the pollution. Therefore, although owners and operators might be
the creators of pollution, they do not necessarily bear the full costs of the pollution.
Government regulation is an attempt to internalize the costs of pollution.
If the people affected by a particular pollution source could negotiate with the party
responsible for that source, the parties could negotiate among themselves to reach an
economically efficient solution. The solution would be efficient because it would involve only
those individuals who are affected by the pollution. In effect, the solution would involve the
trading of pollution and compensation among the owner or operator and the people affected by
the pollution.
Individual negotiation often does not occur in an unregulated market, however, because
of high transactions costs, even if trade among the affected parties would be beneficial to all
parties involved. For the majority of environmental pollution cases, the costs of identifying all
the affected individuals and negotiating an agreement among those individuals is prohibitively
high. Another problem preventing negotiations from taking place is that our current market
system does not clearly define liability for the effects of pollution.
In the case of environmental quality, an additional problem is the public nature of this
"good." Environmental quality is a public good because it is predominantly nonexcludable and
nonrfval. Individuals who willingly pay for reduced pollution cannot exclude others who have not
paid from also enjoying the benefits of a less polluted environment. Because many
environmental amenities are nonexcludable, individuals utilize but do not assume ownership of
these goods, and therefore will not invest adequate resources in their protection. The result is
that in the absence of government intervention, the free market will not provide public goods,
such as a clean environment, at the optimal quantity and quality desired by the general public.
3-2
-------
3.2 ENVIRONMENTAL FACTORS
In the case of the pharmaceutical industry, the result of the market's failure to promote
water and air pollution control is that pollution of the nation's surface water, ground water, and
air is not controlled to the optimal level. This industry releases significant amounts of pollutants
to surface waters, wastewater treatment plants, wastewater treatment sludges, and air. Despite
state and local regulatory programs, many areas are still adversely affected by pollutant
discharges and emissions by this industry. Section Six discusses in detail water and air quality
impacts of the regulations.
3.3 LEGAL REQUIREMENTS
The regulations are proposed under the authorities of Sections 301,304,306,307, and
501 of the Clean Water Act (the Federal Water Pollution Control Act Amendment of 1972, 33
U.S.C. 1251 et seq., as amended by the Clean Water Act of 1987, Pub. L. 100-4, also referred to
as the CWA or the Act) and under the authority of Section 112 of the Clean Air Act
Amendments of 1990.
3-3
-------
-------
SECTION FOUR
TECHNOLOGY OPTIONS AND REGULATORY ALTERNATIVES
A number of alternatives are available to pharmaceutical facilities that would allow them
to meet more stringent effluent limits. Key alternatives are advanced biological treatment and
distillation. Other approaches include multimedia filtration, polishing ponds, cyanide destruction,
granular activated carbon adsorption, pH adjustment/neutralization, and equalization. The
following sections briefly describe these alternatives, all of which were considered by EPA for
developing effluent guidelines for the pharmaceutical industry.
4.1 TECHNOLOGY COMPONENTS1
4.1.1 Advanced Biological Treatment
Advanced biological treatment is used in the pharmaceutical manufacturing industry to
treat biochemical oxygen demand (BODS), chemical oxygen demand (COD), and total suspended
solids (TSS) as well as to degrade various organic constituents. Among facilities with advanced
biological treatment technologies that provide reduction of ammonia in the wastewater through
nitrification, "best performance" has been defined as referring to systems that achieve, on a long-
term basis, 90 percent BOD5 reduction and 74 percent COD reduction in pharmaceutical
manufacturing wastewater, under the existing BPT effluent limitations guidelines (40 CFR Part.
439).
Biological systems can be divided into two basic types: aerobic (treatment takes place in
the presence of oxygen) and anaerobic (treatment takes place in the absence of oxygen).
According to the Section 308 Survey, only two pharmaceutical manufacturing facilities reported
using anaerobic biological treatment systems. The four most common aerobic treatment
1This section is summarized from EPA's Development Document for this regulation (U.S.
EPA, 1995a).
4-1
-------
technologies in the industry are activated sludge, aerated lagoon, trickling filter, and rotating
biological contactor (RBC).
In aerobic biological treatment processes, oxygen-requiring microorganisms decompose
organic and nonmetallic inorganic constituents into carbon dioxide, water, nitrates, sulfates,
organic byproducts, and cellular biomass. The microorganisms are maintained by adding oxygen
and nutrients (usually nitrogen and phosphorus) to the system. Activated sludge and aerated
lagoon processes are suspended-growth processes in which the microorganisms are maintained in
suspension within the liquid being treated. The trickling filter and RBC processes are attached-
growth processes in which microorganisms grow on an inert medium (e.g., rock, wood, plastic).
Three types of activated sludge processes were listed as choices in the Section 308 Survey:
single, two-stage, and oxygen-activated sludge. Table 4-1 shows that the majority of biological
treatment systems used in the pharmaceutical industry involve the activated sludge approach.
An activated sludge treatment system normally consists of an equalization basin, a settling
tank (primary clarifier), an aeration basin, a secondary clarifier, and a sludge recycle line. Sludge
produced by these systems generally consists of biological waste products and expired
microorganisms. Because the sludge can accumulate under certain operating conditions, periodic
removal of sludge from the aeration basin might be necessary.
Generated sludge will require some type of storage, handling, and disposal. Biological
sludges are normally treated in a two-step process prior to disposal: thickening followed by
dewatering. Other sludge treatment can also be performed, but these two are the most
significant processes. The goal for each of these operations is to decrease the overall volume of
sludge.
Ammonia treatment by nitrification is achieved in biological treatment units by
incorporating two additional sets of autotrophic microorganisms. The first set of microorganisms
converts ammonia to nitrites and the second set converts nitrites to nitrates.
Some key design parameters for activated sludge systems include nutrient-to-
microorganism ratio, mixed liquor suspended solids (MLSS), sludge retention time, oxygen
4-2
-------
TABLE 4-1
SUMMARY OF MAJOR TREATMENT TECHNOLOGIES
USED IN THE PHARMACEUTICAL MANUFACTURING INDUSTRY
Technology
pH Adjustment/Neutralization
Equalization
Biological Treatment
Single-Stage Activated Sludge
Two-Stage Activated Sludge
Oxygen Activated Sludge
Aerated Lagoons
Trickling Filters
Rotating Biological Contactors
Multi-Media Filtration
Cyanide Destruction
Alkaline Chlorination
H202 Oxidation
Hydrolysis
Distillation Technologies
Solvent Recovery
Distillation
Distillation with reflux
Rectification
Wastewater treatment
Steam stripping
Carbon Adsorption
Polishing Pond
Air Stripping
Number of Facilities Using
the Technology
Subcategory A/C Subcategoiy B/D
81 45
44 26
31 21
2 2
1 1
7 5
4 1
2 1
3 3
6 0
3 0
1 0
12 3
28 5
12 1
4 0
6. 4
2 6
2 0
Source: Data based on responses from the 1990 Section 308 Survey (244 responding facilities).
4-3
-------
requirements, nutrient requirements, sludge production, substrate removal rate constant (K), and
percent BOD of effluent TSS.
4.1.2 Multimedia Filtration
Multimedia filtration is used in the pharmaceutical manufacturing industry to reduce TSS
in wastewater. This technology also can be used to treat BOD in wastewater by removing
particulate BOD. Multimedia filtration is performed by introducing a wastewater to a fixed bed
of inert granular media. Suspended solids are removed from the wastewater by one or more of
the following processes: straining, interception, impaction, sedimentation, and adsorption. This
operation is continued until either solids "breakthrough" occurs (i.e., solids concentration
increases to an unacceptable level in the discharge from the bed) or the head loss across the bed
becomes too great (due to trapped solids) to operate the bed efficiently.
In multimedia filtration, a series of layers, each with a progressively smaller grain size
medium (traveling from inflow to outflow of the bed), are used in the filtration bed. This design
allows solids to penetrate deeper into the bed before becoming fixed, thus increasing the capacity
of the bed and decreasing the buildup of head loss in the unit. Typical filtration media include
garnet, crushed anthracite coal, resin beads, and sand. Though downflow (gravity flow) systems
are the most common, upflow and biflow (influent is introduced above and below the filter
medium, and the effluent discharges from the center of the filter medium) filtration units can
also be used.
Some key design parameters associated with multimedia filtration units include
wastewater flow rate, hydraulic loading rate, and filter medium depth.
4-4
-------
4.1.3 Polishing Pond
Polishing ponds are used in the pharmaceutical manufacturing industry to remove TSS
from wastewater using gravity settling. Some BOD removal associated with the settling of
suspended solids can also occur.
The wastewater is introduced at one end of the pond and ultimately flows out the other
end of the pond. The pond is designed such that the water retention time is high enough and
the water is still enough to allow solids to fall out of suspension. If the flow is too fast or other
mixing is added to the system, solids can be maintained in suspension and discharged from the
pond.
To avoid anaerobic conditions in the bottom portion of the pond, these units must be
designed to be shallow, which might require a large land area if flow to the unit is high. Depths
of polishing ponds currently used in the industry range from 2.5 to 14 feet. Retention times
range from 0.2 days to 14.6 days. In the past, polishing ponds have been designed with an
earthen liner only; however, current regulations require installation of a minimum of two liners
and a leak detection system (40 CFR 261.221) for most new applications. Polishing ponds will
accumulate solids over time and therefore will require periodic cleanout.
4.1.4 Cyanide Destruction
Several cyanide destruction treatment technologies are currently used in the
pharmaceutical manufacturing industry, including alkaline chlorination, hydrogen peroxide
oxidation, and base hydrolysis. The alkaline chlorination treatment process involves reacting
cyanide with elemental chlorine or hypochlorite to form nitrogen and carbon dioxide. The
reaction is a two-step process and is normally performed separately in two reactor vessels.
Because treatment is normally performed in batches, it is necessary to use an additional
equalization tank to store accumulated wastewater during treatment. The reactors need to be
equipped with agitators, and both reaction steps require close monitoring of pH and
4-5
-------
oxidation/reduction potential (ORP). These reactions are normally performed at ambient
temperatures.
Hydrogen peroxide treatment involves adding hydrogen peroxide to cyanide-bearing
wastewater to convert free cyanide to ammonia and carbonate ions. This treatment is normally
performed in batches in a reaction vessel or vessels. The treatment process consists of heating
the wastewater to around 125°F and adjusting the pH in the reaction vessel to approximately 11.
Hydrogen peroxide is added to the vessel and is allowed to react for approximately 1 hour.
Hydrolysis treatment involves reacting free cyanide with water under basic conditions to
produce formate and ammonia. This process requires approximately 1 hour to proceed and is
typically performed at a temperature between 170° and 25Q°C, and at a pH between 9 and 12.
4.1.5 Steam Stripping/Distillation
Steam stripping/distillation is used both in industrial chemical production (for chemical
recovery and/or recycling) and in industrial waste treatment to remove gases and/or organic
chemicals from wastewater streams by injecting steam into a tray or packed distillation column.
In most cases, the organic components removed by steam stripping/distillation are water soluble.
Steam stripping/distillation is an effective treatment for a wide range of aqueous streams
containing organics and ammonia. Appropriately designed and operated distillation columns can
treat a variety of waste streams ranging from wastewaters containing a single highly volatile
constituent to complex organic/inorganic mixtures. Steam stripping/distillation can be used both
as an in-plant process to recover concentrated organics from aqueous streams and as an end-of-
pipe treatment to remove organics from wastewaters prior to discharge or recycle.
Steam stripping/distillation is usually conducted as a continuous operation in a packed
tower or tray tower (sieve tray or bubble cap) with more than one stage of vapor-liquid contact.
The wastewater enters near the top of the column and then flows downward by gravity,
countercurrent to the steam, which is introduced at the bottom of the column. Steam can be
either directly injected or reboiled, although direct injection is more common. The steam strips
4-6
-------
volatile organics from the wastewater, which are then included in the upward vapor flow. As a
result, the wastewater contains progressively less volatile organic compounds as it moves toward
the bottom of the column. The extent of separation is governed by physical properties of the
VOCs being stripped, the temperature and pressure at which the column is operated, and the
arrangement and type of equipment used.
Steam stripping columns are used for stripping only and typically will not have any
rectifying stages. Distillation columns can be used for stripping, rectification, or both. The
difference between stripping columns and rectifying columns is the location of the feed stream.
Stripping columns have a feed stream located near the top of the column while rectifying
columns have a feed stream located near the bottom of the column. Pollutants that are phase-
separated can usually be stripped from the wastewater in a steam stripper. Pollutants that are
not phase-separable, such as methanol, will be more easily removed in a column with rectifying
stages. Reflux, recycling of the overhead stream back to the column, can also be used to help
pollutants which are difficult to strip to achieve a high concentration in the overhead steam.
The ancillary equipment used in conjunction with steam stripping/distillation columns
includes a condenser and subcooler, pumps for the feed and reflux streams, a feed preheater and
bottoms cooler, a decanter, a storage tank, and an air pollution control device to contain any
vapors from the condenser.
The majority of pharmaceutical manufacturing facilities that currently use steam
stripping/distillation columns to treat their wastewater use stainless steel. Salts and other
pollutants can contribute to scaling and corrosion inside the column. Thus, timely maintenance
and adequate labor should be provided to deter scaling problems.
The key design parameters for steam stripping/distillation columns are the steam-to-feed
ratio and the number of trays or equilibrium stages in packed columns. These parameters are
determined by the equilibrium ratio of the least strippable contaminant in the wastewater stream
and the removal efficiency required to treat the contaminant to the desired concentration.
4-7
-------
4.1.6 Granular Activated Carbon Adsorption
Granular activated carbon (GAC) adsorption is used in the pharmaceutical manufacturing
industry to treat BOD, COD, or organic constituents in wastewater. Adsorption is a process in
which soluble or suspended materials in water are bonded onto the surface of a solid medium.
Activated carbon is an excellent medium for this process because of its high internal surface area,
high attraction to most adsorbates (i.e., the constituents to be treated), and the fact that it is
hydrophobic (i.e., water will not occupy bonding sites and interfere with the adsorption process).
Constituents in the wastewater bond onto the GAC grains until all surface bonding sites are
occupied. At this point the carbon is considered to be "spent," and it requires regeneration,
cleaning, or disposal.
Activated carbon is normally produced in two standard grain sizes: powdered activated
carbon (PAC) with diameters less than a 200 mesh, and GAC with diameters greater than 0.1
mm. PAC is generally added to the wastewater, whereas GAC is normally used in flow-through
fixed-bed units.
For treatment units, GAC is packed into one or more beds or columns. Multiple beds
are more common and are normally operated in series because this design allows for monitoring
between beds, and therefore minimizes the risk of discharging wastewater from the system with
concentrations above acceptable levels. Wastewater flows through a bed and is allowed to come
in contact with all portions of the GAC. The GAC in the upper layers of the bed is spent first
as bonding sites are occupied, and the GAC in progressively lower regions is spent over time as
the adsorption zone moves down through the unit. When contaminant concentrations at the
bottom of the bed begin to increase above acceptable levels, the bed is considered to be spent
and must be removed. This description assumes that beds are operated in downflow mode;
however, it is also possible to use an upflow design for GAC systems.
Once a bed is spent, the carbon can be treated in three ways: regeneration, backwash, or
disposal. Normally, it is possible to use high heat (1,500 to 1,700° F), steam, or chemical
treatment to regenerate the spent carbon.
4-8
-------
The performance of GAC treatment units can be affected by several factors. Three
important design criteria are saturation loading, wastewater TSS concentration, and hydraulic
loading.
4.1.7 pH Adjustment/Neutrsilization
Because many treatment technologies used in the pharmaceutical manufacturing industry
are sensitive to pH fluctuations, pH adjustment, or neutralization, can be required as a part of an
effective treatment system. A pH adjustment system normally consists of a small tank (10 to 30
minutes retention time) with mixing and a chemical addition system. To adjust pH to a desired
value, either acids or caustics can be added In the mixing tank. Some treatment technologies
require a high or low pH to effectively perform treatment (e.g., air stripping of ammonia requires
a pH around 10 or 11). pH is generally adjusted to between 6 and 9 prior to final discharge.
4.1.8 Equalization
Because many of the treatment technologies listed in this section are performed
continuously and some are sensitive to spikes of high flow or high contaminant concentrations, it
is necessary to include equalization as a part of most treatment systems. Equalization is normally
performed in large tanks or basins designed to hold a certain percentage of a facility's daily
wastewater flow. Equalization will equalize high- and low-flow portions of a typical production
day by allowing wastewater to be discharged to downstream treatment operations at a constant
flow rate. Equalization can provide a continuous wastewater feed to operations such as
biological treatment that perform more effectively under continuous load conditions.
The mixing that occurs in an equalization basin will help to minimize spikes of various
contaminants in the discharged wastewater. This equalization will prevent loss of treatment
effectiveness or treatment system failures associated with these spikes.
4-9
-------
4.2 SIMMARYOFIffiGUIATORYALTEIWA'nVES
EPA has developed a set of regulatory options, which are divided into those for direct
dischargers and those for indirect dischargers. Within each discharger category, additional
distinctions are made. First, all technology options are divided between industry subcategories,
with A and C industry subcategories (representing facilities that use fermentation or biological
and chemical synthesis processes) being distinguished from B and D industry subcategories
(representing facilities that use biological and natural extractive processes or that are formulators
of pharmaceutical products). For direct dischargers, the technologies are then further broken
down into BPT, BCT, BAT, and NSPS options; for indirect dischargers, PSES and PSNS
technology options are examined.
Table 4-2 presents the 37 regulatory options considered by EPA and defines the
technologies associated with each option. EPA has selected the following options for inclusion in
the regulation:
For direct discharging A/C facilities, BPT-A/C#2 is selected for conventional
pollutants and BAT-A/C#2 is selected for nonconventional pollutants.
For direct discharging B/D facilities, BPT-B/D#2 is selected for conventional
pollutants and BAT-B/D#1 is selected for nonconventional pollutants.
NSPS-A/C#1 is selected for new A/C facilities that are direct dischargers (this
option is identical to BAT-A/C-#3).
NSPS-B/D#1 is selected for new B/D facilities that are direct dischargers (this
option is identical to BAT-B/D#3).
PSES-A/C#1 is selected for A/C facilities that are indirect dischargers.
PSES-B/D#1 is selected for B/D facilities that are indirect dischargers.
PSNS-A/C#2 is selected for new A/C facilities that are indirect dischargers (this
option is identical to PSES-A/C#2).
PSNS-B/D#1 is selected for new B/D facilities that are indirect dischargers (this
option is identical to PSES-B/D#2).
The selected BAT options include all of the processes mandated in the selected BPT options.
4-10
-------
TABLE 4-2
REGULATORY OPTIONS CONSIDERED IN THE REGULATORY IMPACT ANALYSIS
Type of
Option
Name
Description
SVVS''% ^ ,,:,,"""' ^ -."."£*. ' IMfew^-iE>»5cIi»i^erS'; - - ^ \" ,', ,,,".
Best
Practicable
Technology
Best
Conventional
Technology*
Best
Available
Technology
BPT-A/C#1
BPT-A/C#2
BPT-A/C#3
BPT-A/C#4
BPT-A/C#5
BPT-B/D#1
BPT-B/D#2
BPT-B/D#3
BCT-A/C#1
BCT-A/C#2
BCT-A/C#3
BCT-B/D#1
BCT-B/D#2
BAT-A/C#1
BAT-A/C#2
BAT-A/C#3
BAT-A/C#4
BAT-B/D#1
BAT-B/D#2
Current biological treatment
Advanced biological treatment + cyanide destruction
Advanced biological treatment + cyanide destruction + effluent
filtration
Advanced biological treatment + cyanide destruction + polishing
pond
Advanced biological treatment + cyanide destruction + effluent
filtration + polishing pond
Current biological treatment
Advanced biological treatment
Advanced biological treatment + effluent filtration
Advanced biological treatment + effluent filtration
Advanced biological treatment + polishing pond
Advanced biological treatment + effluent filtration + polishing
pond
Advanced biological treatment
Advanced biological treatment + effluent filtration
Advanced biological treatment + cyanide destruction with
nitrification where necessary
Advanced biological treatment + cyanide destruction + in-plant
steam stripping
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation + activated carbon
Advanced biological treatment
Advanced biological treatment + in-plant steam stripping
4-11
-------
TABLE 4-2 (cont.)
Type of
Option
Best
Available
Technology
(Cont.)
New Source
Performance
Standard
Name
BAT-B/D#3
BAT-B/D#4
NSPS-A/C#1
NSPS-A/C#2
NSPS-B/D#1
NSPS-B/D#2
Description
Advanced biological treatment + in-plant steam
stripping/distillation
Advanced biological treatment + in-plant steam
stripping/distillation + activated carbon
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation
Advanced biological treatment + cyanide destruction + in-plant
steam stripping/distillation + activated carbon
Advanced biological treatment + in-plant steam
stripping/distillation
Advanced biological treatment + in-plant steam
stripping/distillation + activated carbon
s" :C'-'^;:"-Cvrr' :\*';tfi&^JM8dtM&u*'* '"*,!>" '-""-- * ' - */'
Pretreatment
Standards for
Existing
Sources
Pretreatment
Standard for
New Sources
PSES-A/C#1
PSES-A/C#2
PSES-A/C#3
PSES-A/C#4
PSES-B/D#1
PSES-B/D#2
PSES-B/D#3
PSNS-A/C#1
PSNS-A/C#2
PSNS-A/C#3
PSNS-B/D#1
PSNS-B/D#2
In-plant steam stripping + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction + end-
of-pipe advanced biological treatment
In-plant steam stripping/distillation + cyanide destruction + end-
of-pipe advanced biological treatment 4- activated carbon
In-plant steam stripping
In-plant steam stripping/distillation
In-plant steam stripping/distillation + activated carbon
In-plant steam stripping/distillation + cyanide destruction
In-plant steam stripping/distillation + cyanide destruction + end-
of-pipe advanced biological treatment
In-plant steam stripping/distillation + cyanide destruction + end-
of-pipe advanced biological treatment + activated carbon
In-plant steam stripping/distillation
In-plant steam stripping/distillation + activated carbon
*In the Development Document (EPA, 1995a), BCT-A/C#1,2, and 3 in this table actually correspond
to Options 3,4, and 5, and BCT-B/D#1 and 2 in this table correspond to #2 and #3. The options not
listed in this table were never considered in this report because they are equal to or less stringent than
the requirements of the selected BPT option, and thus no incremental costs are incurred over BPT.
4-12
-------
SECTION FIVE
ECONOMIC IMPACTS AND SOCIAL COSTS
The Agency evaluates the costs and economic impacts of pollution control standards in
order to assess the potential impact of the standards on the nation's economy in terms of facility
closures, job losses, and market disruptions. These impacts can be translated into measures of
the social cost of the regulation, which is the monetary value of these disturbances. This
information, compared to the potential benefits of the proposed standards, is useful for policy
decisions concerning the stringency of the standard.
5.1 REGULATORY COMPLIANCE COSTS
The regulatory compliance costs considered in this section do not represent full social
costs. Instead, they represent the costs that will be experienced by the regulated industry.
Section 5.4 discusses social costs, including costs to the federal government for providing tax
subsidies for pollution control equipment.
EPA used a cost annualization model to estimate the annual compliance cost for each
pharmaceutical facility of pollution control equipment and operations needed to meet the
regulation1. This model provides the data necessary for a facility-level analysis. Annualizing
costs is a technique that allocates the capital investment over the lifetime of the equipment,
incorporates a cost-of-capital factor to address the costs associated with raising or borrowing
money for the investment and the tax-reducing effects of expenditures (i.e., depreciation
allowances on corporate income tax filings), and includes annual operating and maintenance
(O&M) costs. The resulting annualized cost represents the average annual payment that a given
company will need to make to upgrade its facility.
that some facilities incur no costs for certain options because EPA has determined
their current pollution control systems meet the regulatory requirements of those options.
5-1
-------
The annualized costs for all regulatory options for direct dischargers are given in Tables
5-1 and 5-2. The average annualized costs per facility for BPT options range from $0 to
$764,975 for A/C facilities and $0 to $54,346 for B/D facilities. The aggregate annualized costs
for BPT options range from $0 to $18.4 million for A/C facilities and $0 to $0.8 million for B/D
facilities. For BCT options, the average annualized costs per facility range from $147,972 to
$637,021 for A/C facilities and $22,888 to $51,047 for B/D facilities. For these same facility
types, the aggregate annualized costs for BCT options range from $3.6 million to $15.3 million
and from $03 million to $0.7 million, respectively. For the BAT options, the average annualized
costs per facility range from $274,188 to $3,172,654 for A/C facilities and $50,626 to $206,634 for
B/D facilities. The aggregate annualized costs for these facilities range from $6.6 million to $76.1
million and $0.7 million to $2.9 million, respectively.
The annualized costs for regulatory options for indirect dischargers are given in
Table 5-3. For A/C facilities, the average annualized costs per facility of the PSES options range
from $392,782 to $1,398,273. The aggregate annualized costs for these facilities range from $34.6
million to $123.0 million. For B/D facilities, the average annualized costs per facility of the
.PSES options range from $51,778 to $414,791. The aggregate annualized costs for these facilities
range from $7.9 million to $63.5 million.
The annualized costs for the selected options for this proposed rulemaking are shown in
Table 5-4. The average annualized costs per facility for the selected options are $1.1 million for
BAT-A/C#2, $50,626 for BAT-B/D#1, $0.4 million for PSES-A/C#1, and $51S778 for PSES-
B/D#1. The aggregate annualized costs are $26.8 million for BAT-A/C#2, $0.7 million for BAT-
B/D#1, $34.6 million for PSES-A/C#1, and $7.9 million for PSES-B/D#1, for a total aggregate
cost of $70.0 million.
5.2 ECONOMIC IMPACT ANALYSIS METHODOLOGY
EPA performed several analyses to determine the economic impacts of the proposed
effluent guidelines for pharmaceutical facilities. These analyses included a facility-level analysis,
an owner company-level analysis, an employment and community-level analysis, a foreign trade
5-2
-------
TABLE 5-1
COMPLIANCE COSTS FOR A/C DIRECT DISCHARGERS (1990 $)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Posttax
Annualized Costs
Average
Annual Cost
per Facility*
BPT Option Costs
BPT-A/C#1
BPT-A/OP2
BPT-A/C#3
BPT-A/CM
BPT-AO5
$0
$14,742,689
$21,891,929
$37,455,760
$44,204,216
$0
$7,046,870
$7,488,423
$21,764,186
$23,420,779
$0
$5,681,474
$6,717,116
$16,665,409
$18,359,400
$0
$236,728
$279,880
$694,392
$764,975
BCT Ootion Costs
BCT-A/C#1
BCT-A/C#2
BCT-A/C#3
$16,875,845
$32,439,676
$39,188,132
$2,957,486
$16,545,942
$19,054,074
$3,551,327
$13,102,463
$15,288,512
$147,972
$545,936
$637,021
BAT Option Costs
BAT-A/C#1
BAT-A/C#2
BAT-AO3
BAT-AO4
$15,050,112
$56,392,127
$68,035,029
$92 851 663
$8,544,621
$35,689,088
$57,980,678
$114,229,651
$6,580,502
$26,779,144
$40,93 1,284
$76.143.696
$274,188
$1,115,798
$1,705,470
$3,172.654
Total Posttax Annualized Costs divided by the total number of A/C direct discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. capital and operating costs estimates for pollution control
equipment.
5-3
-------
TABLE 5-2
COMPLIANCE COSTS FORB/D DIRECT DISCHARGERS (1990 $)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Posttax
Annualized Costs
Average
Annual Cost
per Facility*
BPT Option Costs
BPT-B/D#1
BPT-B/D82
BPT-B/D#3
$0
$605,700
$2,976,515
$0
$519,349
$754,333
$0
$366,228
$760,837
$0
$26,159
$54,346
BCl Option Costs
BCT-B/D#1
BCT-B/D#2
$559,015
$2,929,830
$448,905
$683,889
$320,426
$715,035
$22,888
$51,074
BAT Option Costs
BAT-B/D#1
BAT-B/D#2
BAT-B/D#3
BAT-B/D#4
$644,446
$1,741,330
$3,002,607
$10,310,180
$1,104,801
$937,108
$1,950,161
$3.058,423
$708,758
$731,606
$1,454,688
$2.892,869
$50,626
$52,258
$103,906
$206,634
*Total Posttax Annualized Costs divided by the total number of B/D direct discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report
Source: ERG estimates based on Radian Corp. capital and operating costs estimates for pollution control
equipment
5-4
-------
TABLE 5-3
COMPLIANCE COSTS FOR INDIRECT DISCHARGERS (1990 S)
(PSES)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Posttax
Annualized Costs
Average
Annual Cost
per Facility*
A/C Facilities
PSES-A/C#1
PSES-AO2
PSES-A/C#3
PSES-A/C#4
$70,795,915
$90,082,486
$143,989,655
$186,990,945
$46,441,499
$81,860,584
$105,781,635
$177,615,256
$34,564,845
$57,137,102
$76,844,867
$123,048,025
$392,782
$649,285
$873,237
$1,398,273
B/D Facilities
PSES-B/Dtfl
PSES-B/D#2
PSES-B/D23
$25,160,649
$30,429,899
$61,970,107
$8,956,179
$16,986,223
$98,119,347
$7,922,101
$13,137,467
$63,463,066
$51,778
$85,866
$414,791
"Total Posttax Annualized Costs divided by the total number of indirect discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report
Source: ERG estimates based on Radian Corp. capital and operating costs estimates for pollution control
equipment.
5-5
-------
TABLE 5-4
COMPLIANCE COSTS FOR SELECTED REGULATORY OPTIONS (1990 $)*
Option
Number
BAT-A/C#2
BAT-B/D31
PSES-A/C#1
PSES-B/D#1
Total
Capital Costs
$56,392,127
$644,446
$70,795,915
$25,160,649
Total
O&M Costs
$35,689,088
$1,104,801
$46,441,499
$8,956,179
Total Posttai
Annualized Costs
$26,779,144
$708,758
$34,564,845
$7,922,101
Annual Cost
per Facility**
$1,115,798
. $50,626
$392,782
$51,778
Total+
$152.993,1371 $92.191.568
$69.974.848
$250,806
* The cost of BAT options include the costs of meeting the selected BPT options.
** Total Posttax Annualized Costs divided by the total number of facilities for each subcategory.
+Total number of facilities includes seven nondischarging facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. capital and operating costs estimates for pollution control
equipment.
5-6
-------
impact analysis, a regulatory flexibility analysis, a distributional analysis, and an analysis of the
impacts of the guidelines on new sources. The methodologies for these analyses are described
below.
5.2.1 Facility-Level Analysis Methodology
The facility-level analysis identified facilities that are likely to close as a result of
incremental compliance costs. In the facility-level analysis, the 65 facilities that certified in the
Section 308 Pharmaceutical Survey that the regulation would not affect them were automatically
placed in the "no closure" category of the model. These 65 facilities represent 72 facilities in the
survey universe. The 76 firms operating only as one facility, i.e., owner-operated facilities or
"firm/facilities," were counted in this analysis, but were assumed to pass through this analysis
without closing. Impacts on these fiirm/facilities were measured in the company-level analysis.
The facility closure model evaluated the remaining 134 of the 282 facilities in the survey
universe. Facility closures were estimated by comparing the facility's "salvage value" (the
expected amount of cash the owner would receive if the facility were closed permanently and
liquidated) to the present value of its future earnings (the value in current dollars of the
expected stream of earnings that the facility can generate over a specified period of time). If the
salvage value was greater than what the facility is expected to generate in earnings, then it was
assumed that the owner would liquidate the facility. Salvage value includes the value of current
(i.e., short-term) assets and fixed (i.e., long-term) assets. Data for the facility-level analysis was
either taken directly from the Section 308 Survey or estimated based on data that was provided
by other facilities.
5.2.2 Owner Company-Level Analysis Methodology
The company-level analysis evaluated the effects of regulatory compliance on companies
owning one or more affected pharmaceutical manufacturing facilities and identified other impacts
not captured in the facility analysis. The analysis assessed the impacts of facility closures on each
firm and the impact of compliance costs at all facilities owned by the firm that do not close.
5-7
-------
These impacts were assessed using ratio analysis, which employs two indicators of financial
viability: the rate of return on assets (ROA) and the interest coverage ratio (ICR). The ratio
analysis simulates the analysis an investor and/or creditor would employ in deciding whether to
finance a treatment system or make any other investment in the firm. Data from the Section 308
Survey and engineering costs analyses were used to calculate baseline and postcompliance ROAs
andlCRs.
In the baseline ratio analysis, the company's financial viability before the investment was
evaluated. ROA and ICR were computed with the survey data. To evaluate the baseline
viability of the companies analyzed, the baseline ROA and ICR values were compared against
the lowest quartile (25th percentile) values for the pharmaceutical sector (SIC 283). Those
companies for which the value of either the ROA or the ICR was less than the first quartile
value from Robert Morris Associates (RMA) and Dun & Bradstreet (D&B) were judged to be
vulnerable to financial failure regardless of compliance costs.
In the postcompliance analysis, the company's financial condition subsequent to the
investment was predicted. The relevant survey data (net income, EBIT [net income and earnings
before interest and taxes], total assets, and interest expenses) were adjusted to reflect annual
compliance costs estimated at the facility level as well as losses in income caused by facility
closures, if any. The standard postcompliance analysis, referred to as Postcompliance
Analysis 1, evaluated impacts on companies that were not found to be vulnerable in the baseline
analysis. For these healthier companies, if either of the postcompliance ROA and ICR values
fell below the quartile benchmarks, then the company was judged to be vulnerable to financial
failure as a consequence of regulatory compliance; these companies were determined to sustain a
"significant impact" as a result of the regulation.
Postcompliance Analysis 1 incorporates EPA's standard methodology for judging impact.
Because of EPA's concerns that pharmaceutical industry investors are more tolerant of extended
periods of loss during product development phases, two additional analyses investigating firms
that are projected to fall in the baseline case have been undertaken. These two analyses are
discussed below as Postcompliance Analysis 2 and 3. These analyses are presented as sensitivity
analyses.
5-8
-------
Postcompliance Analysis 2 examined the relative percentage change in ROA or ICR as a
result of compliance costs or facility closures for firms that have positive net income and/or
EDIT, but whose ROA or ICR fall below benchmarks that is the firm is considered a baseline
failure. This analysis determined the severity of impact, assuming these firms do not close in the
baseline. A percentage change in ROA or ICR of more than 5 percent was considered a major
impact.
Postcompliance Analysis 3 evaluated firms with negative.ROA or ICR ratios that were
projected to fail in the baseline case. Although changes in ROA or ICR ratios that already are
negative are difficult to present meaningfully, the proportion of the post-compliance net income
or EBIT loss attributable to complian.ce costs provided a qualitative sense of impact. As in
Postcompliance Analysis 2, a change of more than 5 percent was considered a major impact.
When changes in ROA, ICR, or net income among these firms in postcompliance
Analysis 2 and 3 exceeded 5 percent, the firms in question were evaluated in more detail to
determine if any unusual circumstances might indicate they were not likely to fail in the baseline.
Finally, the Profitability Analysis determined impacts on profitability among firms
estimated to have no significant impact from compliance costs in postcompliance Analysis 1.
This analysis investigated the percentage change in ROA among the healthy firms to assess
impacts on profitability. Again, a change of more than 5 percent was considered a major impact.
5.23 Employment and Community-Level Analysis Methodology
The employment loss and community-level impact analysis investigated employment losses
and resulting community-level impacts from compliance with the effluent guidelines. Primary
employment losses are those losses occurring only within the pharmaceutical manufacturing
industry and stem from facility closures and firm failures. Secondary impacts include
employment losses in other industries that provide input to the pharmaceutical manufacturing
process and other supporting industries and are based on the use of a multiplier. Primary and
secondary employment losses were summed to obtain the total impact on community
5-9
-------
employment levels resulting from implementation of the effluent guidelines. A change in the
community employment rate of more than 1 percent is considered major.
Employment losses are offset to some extent by the need to hire workers to manufacture,
install, and maintain the pollution control equipment. Direct employment effects associated with
the manufacture, installation, and operation of the pharmaceutical industry compliance
equipment were estimated. Then the additional employment effects that might occur through
the indirect and induced effect mechanisms (secondary effects) were considered, using a range of
multipliers. Primary and secondary employment losses were then subtracted to produce the net
gain or loss associated with the proposed effluent guideline.
5.2.4 Foreign Trade Impact Analysis Methodology
Pharmaceutical products are traded in an international market, with producers and buyers
located worldwide. Changes in domestic pharmaceutical production due to the effluent
guidelines can therefore affect the balance of trade. Consequently, EPA conducted a foreign
trade impact analysis. Using data from the Section 308 Survey, the value of 1990 pharmaceutical
exports was estimated for facilities expected to close. These values were summed across facilities
to obtain an estimate of the total value of U.S. pharmaceutical exports that would no longer be
produced. This value was then compared to the total value of U.S. pharmaceutical exports
produced in 1990.
5.25 Regulatory Flexibility Analysis Methodology
EPA conducted a regulatory flexibility analysis to ensure that small entities potentially
affected by the new effluent guidelines will not be disproportionately burdened by the regulation.
The affected population of small businesses was defined and estimated. For simplicity, the
regulatory flexibility analysis defined all pharmaceutical firms as small if they employ fewer than
750 persons. Impacts stemming from recordkeeping and reporting requirements and from
significant alternatives to the proposed rule are discussed. Finally, firm failures and impacts on
5-10
-------
profitability and the present value of net income among small firms resulting from the selected
regulatory options are discussed by size category and compared to impacts among large firms to
determine if small firms are disproportionately affected by the proposed effluent guidelines.
5.2.6 Distributional Analysis Methodology
For the distributional analysis, the zero cost passthrough assumption used in the other
analyses was not used. Instead, it was assumed that manufacturers will raise pharmaceutical
prices in response to increased regulatory costs. To determine upper bound impacts, it was
further assumed that all cost increases can be passed through to consumers. The extent to which
drug prices can rise assuming perfectly inelastic demand was determined as the ratio of total
compliance costs to total cost of pharmaceutical production in the affected facilities. First, the
extent to which drug prices could rise was determined as the ratio of total compliance costs to
total cost of pharmaceutical production in the affected facilities. The analysis then investigated
the impacts of increased drug prices on various demographic groups, looking at a sampling of
products among some of the more highly affected facilities (those facilities where compliance
costs as a proportion of total costs of production are greater than 10 percent) to determine their
likely consumers.
5.2.7 New Source Analysis Methodology
The methodology for analyzing impacts on new sources extrapolates qualitatively from
impacts on existing sources to those on new sources (see Section 5.3.7). NSPS and PSNS options
are more stringent than BAT options. Thus, the difference in cost to an average facility to meet
the more stringent requirements is compared to average total and pharmaceutical manufacturing
costs. If the cost difference is less than 1 percent of pharmaceutical manufacturing cost, barriers
to entry are considered negligible.
5-11
-------
5.3 ECONOMIC IMPACT ANALYSIS RESULTS
The results for the facility-level analysis, owner company-level analysis, employment and
community-level analysis, foreign trade impact analysis, regulatory flexibility analysis,
distributional analysis, and the analysis on the impacts of the guidelines on new sources are
presented below. As in Section 5.1, the regulatory compliance costs considered in this section do
not represent full social costs, which are discussed in Section 5.4.
5.3.1 Facility-Level Analysis Results
The baseline facility-level analysis indicated that 38 facilities, or 13 percent of the total
number of facilities, will close in the baseline. These facilities projected to close in the baseline
case are not owner-operated facilities. The postcompliance analysis predicts that no facilities will
close as a result of the selected regulatory options, although one A/C direct facility and one B/D
indirect facility are estimated to close under the most stringent options for these two groups of
facilities.
533, Owner Company-Level Analysis Results
The baseline analysis indicated that out of 187 firms, 54, or 29 percent, are likely to fail
even before the impact of the effluent guideline requirements are considered. The results of
Postcompliance Analysis 1 are presented in Table 5-5. Postcompliance Analysis 1 indicates that
under the selected regulatory options only two firms with A/C indirect discharging facilities and
one firm/facility, a B/D indirect discharger, are expected to experience significant impacts as a
result of compliance costs. Overall, these firms represent 2.3 percent of all regulated firms that
do not fail in the baseline.
The results of the Postcompliance Analysis 2, which are presented in Table 5-6, indicate
that a total of nine firms, or about 31 percent of marginal firms with positive EBIT, are expected
to incur substantial impacts (i.e., greater than 5 percent change in ICR) if they do not fail for
5-12
-------
TABLE 5-5
POSTCOMPLIANCE ANALYSIS 1*
Firms with A/C Direct Facilities
Firms with B/D Direct Facilities
Firms with A/C Indirect Facilities
Firms with B/D Indirect Facilities
Total
Number
of Firms
15
7
53
72
No Significant
Impact
#of
Firms
15
7
51
71
%of
Group
100.0%
100.0%
96.2%
98.6%
Significant
Impact
#of
Firms
0
0
2
1
%of
Group
0.0%
0.0%
3.8%
1.4%
% of All
Firms**
0.0%
0.0%
1.5%
0.7%
AllFirms+
133
1301 97.7%
3
2.3% 1 2.3%
* This scenario analyzes impacts from regulating A/C Direct facilities under options BAT-A/C#2
and BPT-A/C#2, B/D Direct facilities under options BAT-B/D#1 and BPT-B/D#2, A/C Indirect
facilities under option PSES-A/C#1, and B/D Indirect facilities under option PSES-B/D#1.
** Out of all firms in the postcompliance analysis (133 firms).
+Number of firms for All Firms might be less than the total firms by subcategory because some
firms have more than one type of facility- Total number of All Firms includes firms that have
nondischarging facilities
Note: Analysis excludes three firms because of lack of financial data.
Source: ERG estimates.
5-13
-------
TABLE 5-6
POSTCOMPLIANCE ANALYSIS 2
(FIRMS PROJECTED TO FAIL IN THE BASELINE ANALYSIS)
Range of
Change
0
X)-<=5
>5-<-10
>10-<^=20
>20-«^50
>50
Percent Change
inlCR*
# Finns
9
11
2
0
5
2
% of Total
31.0%
37.9%
6.9%
0.0%
17.2%
6.9%
Percent Change
in ROA**
# Firms
6
3
3
2
2
4
% of Total
30.0%
15.0%
15.0%
10.0%
10.0%
20.0%
Total # Firms
29 1 100.0%
20 1 100.0%
* Finns that failed the baseline analysis are analyzed here if [(base EBIT and net
income>0)and(base ICR or ROA0)].
Because only firms -with positive EBIT can be analyzed here, those with negative
EBIT are analyzed for percent decline in EBIT in Table 5-7
** Firms that failed the baseline analysis are analyzed here if [(base EBIT and net
income>0)and(base ICR or ROA«3)ench)]or[(base EBIT<=0)and(base net income>0)].
Because only firms with positive net income can be analyzed here, those with negative
net income are analyzed for percent decline in net income in Table 5-7
Source: ERG estimates.
5-14
_
-------
other reasons. In addition, 11 firms (55 percent of the marginal firms with positive ROA), are
expected to incur substantial impacts (i.e., greater than 5 percent change in ROA) if these firms
do not fail for other reasons.
The Postcompliance Analysis 3 results, which are presented in Table 5-7, indicate that six
firms with negative EBIT (24 percent) will incur substantial impacts if they do not fail for other
reasons. For firms with negative net income in the baseline, only five firms (or about 15 percent)
with negative net income are expected to incur substantial impacts if they do not fail for other
reasons.
The more in-depth analysis of the viability of these firms noted as potentially highly
affected in the post-compliance analysis if they do not fail in the baseline revealed that there
appeared to be no unusual circumstances that might indicate these firms would not fail, with one
exception. The one firm that might not fail showed unusually strong growth over the 3-year
survey period. However, when only 1990 data were used (i.e., it is assumed that 1990 data are
more representative of future health than the 3-year average), this firm was shown to be able to
absorb the compliance costs without undue effect on ROA or ICR. Thus the initial
Postcompliance Analysis 1 is considered a reliable estimate of postcompliance impact.
Table 5-8 presents results from the Profitability Analysis, which indicated that 15 firms-
(15 percent of all firms in the analysis) will have significant impacts on returns, although only one
firm will have impacts of greater than 50 percent decline on returns. When the firms that
certified that they would experience no impacts from any effluent guideline also were considered,
only 11 percent of firms in the postcompliance analysis were considered likely to experience
major impacts short of firm failure. Note that this impact would be less if it was assumed that
firms could pass through some of their compliance costs in the form of higher prices. It is also
useful to note that median ROA of this group is quite high and thus many of these firms could
sustain this level of impact while still showing adequate to good returns.
5-15
-------
TABLES-?
POSTCOMPLIANCE ANALYSIS 3
(FIRMS PROJECTED TO FAIL IN THE BASELINE ANALYSIS)
Range of
Change
0
>0-<=5
>5-<=10
>10-<=20
>20-50
Percent Change
inEBIT*
# Firms
12
7
1
1
1
3
% of Total
48.0%
28.0%
4.0%
4.0%
4.0%
12.0%
Percent Change
in Net Income**
# Firms
15
14
1
2
1
1
% of Total
44.1%
41.2%
2.9%
5.9%
2.9%
2.9%
Tnfal # Firms
25
100.0%l 341 100.0%
* Firms that failed the baseline analysis are analyzed here if (base EBIT<=0).
** Firms that failed the baseline analysis are analyzed here if (base net income<=0).
Source: ERG estimates.
5-16
-------
op
A
S2
VI
w 5
11
Eti
e
06
gi"
& ^i
W3 fe fiw
$ot
03
H S3
I
H
O
s
r
a
o o o
V3
00
fc.
00
VO
VD
CO
ON
S
o
J
s.-g
II
8 §
P
8.-S
t?ts
13 .5
o **
.o w
J-S
"o
3-2
11
je
*> c
l:f
^a
1
§
vs
1
1-
Ǥ
I 1
O
-------
5.3.3 Community-Level Analysis Results
The baseline impacts from the analysis on primary employment before any compliance
costs are incurred total 14,381 jobs estimated to be lost, out of a total employment of 147,804
workers2 (9.7 percent of total employment). These losses are associated with 38 facility closures,
21 firm/facility failures, and 33 firm failures. The baseline analysis predicts that secondary job
losses will total 85,567.
No employment losses were projected to occur as a result of regulatory options for direct
dischargers. For indirect dischargers, however, total projected primary employment losses
resulting from the selected regulatory options were 78 full-time equivalent (FTE) positions
among A/C indirects and 13 FTEs among B/D indirects, for a total of 91 FTEs or 0.07 percent of
total employment for the affected portion of the industry. Secondary losses were predicted to be
541 FTEs.
None of these losses is expected to result in a change of employment rates of more than
1 percent in the affected communities.
The sum of primary and secondary employment gains is calculated to range from 218
FTEs to 2,890 FTEs. Net gains and losses thus range from a loss of 323 FTEs to a gain of 2,349
FTEs.
5.3.4 Foreign Trade Impact Analysis Results
The impact of effluent guidelines on pharmaceutical exports and the U.S. balance of
trade was found to be negligible. The one firm/facility predicted to close as a result of the
effluent guidelines has pharmaceutical exports totaling $0.08 million. The loss of these exports
the affected portion of the pharmaceutical industry. Employment at other pharmaceutical
firms not covered by the proposed effluent guidelines is not counted here.
5-18
-------
will have virtually no effect on U.S. pharmaceutical exports, which, according to the U.S.
Department of Commerce, totalled $5.7 billion in 1991.
5.3.5 Regulatory Flexibility Analysiis Results
Small firms make up 76 percent of the 190 firms in the survey universe. The largest
percentage of firms are in the 100-499 employees size group (37 percent of all firms in the survey
universe).
The proposed effluent guidelines for the pharmaceutical industry are revisions to existing
effluent guidelines, thus most recordkeeping and reporting requirements are not incremental to
existing guidelines. The exception is new monitoring requirements. Monitoring costs total $9.0
million annually and are 15 percent of the total annual compliance cost for the selected options.
Large firms incur the largest proportion of monitoring costs (61 percent of total monitoring
costs).
No significant alternatives to the proposed rule will substantially reduce impacts on small
entities, thus the Agency believes the stated objectives of the Clean Water Act are met with this
proposed rule and the impacts to small firms have been considered, where possible.
Impacts on small firms measured as firm failure are as follows. Two of the three firms
that were projected to foil in the film-level analysis under the selected regulatory options have
fewer than 750 employees, although only 2 percent of small firms in the postcompliance analysis
are affected in this manner. In addition, 14 of 15 firms found to experience a significant decline
in ROA (over 5 percent) have fewer than 750 employees. These firms represent about 14
percent of all small firms. Note, however, that impacts would be less if firms can pass through
some of their compliance costs. Also, even though returns drop substantially, the median
baseline ROA in this group is quite high.
When cash flow is analyzed, impacts seem less disproportionate. Except in the smallest
size category (0 to 18 employees) the total present value of compliance costs as a percentage of
5-19
-------
the present value of net income is smaller among small firms than among large firms. Over all
firms, the present value of compliance costs is less than 1 percent of the present value of net
income.
The above analyses indicate that although small firms do bear a large portion of the
impacts such as firm failures, these impacts are felt by .a very small percentage of all small firms.
Additionally, the percentages of the present value of compliance costs to the present value of net
income are expected to be smaller, on average, among small firms than among large firms; thus,
impacts to small firms are not expected to be disproportionate to those for large firms.
5.3.6 Distributional Analysis Results
For all the selected regulatory options, the ratio of compliance costs to total
pharmaceutical costs averaged 1.6 percent. Most facilities would incur compliance costs less than
1 percent of total pharmaceutical costs. Only three facilities (2 percent of all facilities) would
incur compliance costs greater than 10 percent of total pharmaceutical costs.
When possible uses for products produced by a sampling of highly affected facilities
(those where compliance costs exceed 10 percent of total pharmaceutical costs) were
investigated, it appeared that children, women, and the elderly were likely to be the major
consumers of many of these products. It was further determined that individuals who lack any
health insurance, those who are covered by government insurance, and those who are covered by
nonwork-related medical insurance might be least likely to have drug coverage. These groups
include Hispanics, young adults, African Americans, young children, and the elderly. Thus,
young adult women, children, and the elderly are likely to be the most heavily affected by
potential cost increases, if such increases can be passed through to consumers.
Because on average any potential price increases are likely to be very low (1.6 percent),
impacts on mass consumers of drugs such as HMOs, governments, and, indirectly, third-party
insurers should be minimal.
5-20
-------
5.3.7 Impacts on New Sources
The selected options for new sources are NSPS-A/C#1, NSPS-B/D#1, PSNS-A/C#1, and
PSNS-B/D#1. In all cases, the requirements for new sources are more stringent than those for
existing sources. However, the difference in cost between new source requirements and existing
source requirements for typical facilities are relatively small when compared to the average
facility costs of production. In most cases, existing facilities would be required to retrofit in-plant
steam stripping systems, whereas new sources would have to install in-plant steam
stripping/distillation systems. Because designing in pollution control equipment in a new source
is typically less expensive than retrofitting the same equipment in an existing source, the cost
differential between the selected requirements for existing sources and those higher existing
source options that are technically equivalent to new source requirements should be an upper
limit on the differential annual cost faced by new sources. Where this differential is not
substantial relative to the typical costs of doing business in this industry, ho significant barrier to
entry is likely to exist.
The average per-facility compliance costs were investigated to determine what the cost
differentials would be between proposed new source and existing source requirements. The
average per-facility cost differentials ranged from about a $34 thousand to a $590 thousand
difference (for A/C direct dischargers), depending on the type of facility. The maximum $590
thousand difference generates the highest percentage of compliance cost differential to
Pharmaceuticals manufacturing costabout 1.4 percent of total manufacturing costs and about
3.0 percent of pharmaceutical manufacturing costs. Since this cost differential is likely to be less
than that assumed here, this small premium estimated to be paid by new sources is not likely to
have much impact on the decision to enter the market. Furthermore, these same options, when
applied to existing sources, were found to have nearly identical impacts on existing sources as the
selected options for existing sources. Thus no significant barriers to entry are estimated to result
from the proposed new source requirements.
5-21
-------
5.4 SOCIAL, COSTS OF REGULATION
Costs reported earlier in this section are those imposed on industry. However, these
costs are reduced by the tax savings realized by industry because of depreciation allowances on
the capital and operating expenditures for the applicable pollution control equipment. Thus
these costs do not reflect the true social costs of installing and operating the equipment. A
major component of social cost is the cost to government of providing these tax savings to
industry.
In addition to compliance costs and costs to government for depreciation allowances,
other monetary and nonmonetary outlays are made by government. Government administrative
costs and costs of reallocating displaced workers are two additional monetary costs.
Nonmonetary costs include losses in consumers' or producers' surpluses in product markets,
discomfort or inconvenience, loss of time, and slowing the rate of innovation. The social costs
estimated here, therefore, are a very large portion of, but not the true total social cost of the
proposed regulation. The costs reported here are thus only a close estimate of this true cost.
To model the social costs, the entire annual pre-tax cost of the proposed effluent
guideline is estimated, using the same time period (16 years) and the discount rate advocated by
OMB as appropriate for annualizing social costs of 7 percent (rather than the higher-average
discount rate of 11.4 percent reported by Section 308 respondents).
The estimated social costs of the proposed regulation are shown in Tables 5-9 to 5-12.
Tables 5-9,5-10, and 5-11 present an estimate of the social costs of regulating the A/C direct
discharges, B/D direct dischargers, and the indirect dischargers, respectively, by option and are
reported in 1990 dollars. Aggregate annualized costs range from $0 per year to $28.1 million per
year for BPT options for A/C direct discharges and from $0 to $1.1 million for B/D direct
dischargers. Aggregate annualized costs of BAT options range from $10.1 million to $124.1
million for A/C direct dischargers and range from $1.2 million to $4.1 million for B/D direct
dischargers. Aggregate annualized costs for PSES options for A/C indirect dischargers range
from $53.9 million to $197.4 million, while for B/D indirect dischargers they range from $11.6
million to $104.7 million. The estimate of total annual social costs for all selected options is
5-22
-------
TABLE 5-9
COMPLIANCE COSTS FOR A/C DIRECT DISCHARGERS (1990 $)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Pretax
Annualized Costs
Annual Cost
per Facility*
BPT Option Costs
BPT-A/C#1
BPT-A/C#2
BPT-A/C#3
BPT-A/C#4
BPT-A/C#5
$0
$14,742,689
$21,891,929
$37,455,760
$44,204,216
$0
$7,046,870
$7,488,423
$21,764,186
$23,420,779
$0
$8,607,496
$9,805,851
$25,729,165
$28,100,133
BCT Option Costs
BCT-A/C#1
BCT-A/C22
BCT-A/C33
$16,875,845
$32,439,676
$39,188,132
$2,957,486
$16,545,942
$19,054,074
$4,743,924
$19,979,929
$23,202,438
BAT Ootion Costs
BAT-A/C#1
BAT-A/C#2
BAT-A/C#3
BAT-A/C#4
, $15,050,112
$56,392,127
$68,035,029
$92,851,663
$8,544,621
$35,689,088
$57,980,678
$114.229,651
$10,137,790
$41,658,626
$65,182,706
$124,058,709
$0
$358,646
$408,577
$1,072,049
$1,170,839
$197,663
$832,497
$966,768
$422,408
$1,735,776
$2,715,946
$5,169,113
Total Pretax Annualized Costs divided by the total number of A/C direct discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. estimates of capital and operating costs for pollution control
equipment.
5-23
-------
TABLE 5-10
COMPLIANCE COSTS FOR B/D DIRECT DISCHARGERS (1990 $)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Pretax
Annualized Costs
Average
Annual Cost
per Facility*
BPT Option Costs
BPT-B/D#1
BPT-B/D#2
BPT-B/D#3
$0
$605,700
$2,976,515
$0
$519,349
$754,333
$0
$583,467
$1,069,420
$0
$41,676
$76,387
BCT Option Costs
BCT-B/D#1
BCT-B/D#2
$559,015
$2,929,830
$448,905
$683,889
$508,081
$994,034
$36,292
$71,002
BAT Option Costs
BATrB/D#l
BAT-B/D#2
BAT-B/D#3
BAT-B/D#4
$644,446
$1,741,330
$3,002,607
$10.310,180
$1,104,801
$937,108
$1,950,161
$3.058.423
$1,173,021
$1,121,441
$2,268,010
$4.149,835
$83,787
$80,103
$162,001
$296.417
*Total Pretax Annualized Costs divided by the total number of B/D direct discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. estimates of capital and operating costs for pollution control
equipment
5-24
-------
TABLE 5-11
COMPLIANCE COSTS FOR INDIRECT DISCHARGERS (1990 $)
(PSES)
Option
Number
Total
Capital Costs
Total
O&M Costs
Total Pretax
Annualized Costs
Average
Annual Cost
per Facility*
A/C Facilities
PSES-A/C#1
PSES-A/C#2
PSES-A/C#3
PSES-A/C34
$70,795,915
$90,082,486
$143,989,655
$186,990,945
$46,441,499
$81,860,584
$105,781,635
$177,615,256
$53,935,789
$91,396,504
$121,024,041
$197,409,678
$612,907
$1,038,597
$1,375,273
$2,243,292
B/D Facilities
PSES-B/D#1
PSES-B/D#2
PSES-B/D#3
$25,160,649
$30,429,899
$61,970,107
$8,956,179
$16,986,223
$98,119,347
$11,619,626
$20,207,461
$104,679,357
$75,945
$132,075
$684,179
Total Pretax Annualized Costs divided by the total number of indirect discharge facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the anafyses in this report
Source: ERG estimates based on Radian Corp. estimates of capital and operating costs for pollution control
equipment.
5-25
-------
TABLE 5-12
COMPLIANCE COSTS FOR SELECTED REGULATORY OPTIONS (1990 $)
Option
Number
BAT-A/C#2
BAT-B/D#1
PSES-A/C#1
PSES-B/D#1
Total
Capital Costs
$56,392,127
$644,446
$70,795,915
$25,160,649
Total
O&M Costs
$35,689,088
$1,104,801
$46,441,499
$8,956,179
Total Pretax
Annualized Costs
$41,658,626
$1,173,021
$53,935,789
$11,619,626
Annual Cost
per Facility*
$1,735,776
$83,787
$612,907
$75,945
$152.993.137
$92,191.5681 $108.387.062
$388,484
* Total Pretax Annualized Costs divided by the total number of facilities for each subcategory.
** Total number of facilities includes seven nondischarging facilities.
Note: These numbers are for all facilities and do not reflect closures predicted by the analyses in this report.
Source: ERG estimates based on Radian Corp. estimates of capital and operating costs for pollution control
equipment.
5-26
-------
shown in Table 5-12. Total social costs resulting from the proposed effluent guideline are
estimated to be $108.4 million per year.
5-27
-------
-------
6.1
SECTION SIX
POLLUTION REDUCTION
ESTIMATES OF REDUCTIONS IN POLLUTANTS BASED ON THE SECTION 308
SURVEY
EPA has established final raw, current, and proposed loadings for nonconventional
constituents that currently are candidates for regulation in the pharmaceutical industry and for
which treatment data are available or a data transfer could be performed to develop proposed
long-term mean (LTM) effluent concentrations. In establishing these loadings, EPA relied on
preliminary loadings developed for conventional and nonconventional pollutants in the
pharmaceutical industry. For the organic constituents given preliminary consideration, available
data were insufficient to develop proposed loadings. Raw, current, and proposed loadings for the
conventional pollutants' BOD and TSS and the nonconventional pollutants' COD have not
changed from the earlier assessment and therefore are not included here.
This section summarizes the methodologies used and assumptions made to develop final
loadings for the nonconventional pollutants that are candidates for regulation. EPA calculated
these final loadings using data from, the Section 308 Survey database and Radian's
WATER.DBF, STEAM.DBF, and AIR.DBF treatability databases, as well as some results
derived using a computer model, WATER?. More information on the methods used for
estimating water and air pollutant reductions can be found in the Development Document (EPA,
1995a).
Table 6-1 presents the loadings and reduction estimates by type of facility (i.e., A/C
direct, B/D direct, A/C indirect, and B/D indirect). Section 6.2 describes how WATER?
estimates were derived. Note that WATER? estimates were derived for both water and air
emissions but the loads and reductions calculated and used in this RIA are estimated using
WATER? for air emissions only. The Section 308 survey was used to calculate water loads and
reductions.
6-1
-------
TABLE 6-1
ESTIMATED POLLUTANT LOADINGS BY TYPE OF FACILITY
Regulatory
Option
Raw
Load
Ob)
Current
Load
Ob)
Total Load
Reduction
Ob)
A/C Direct Dischargers
BAT-A/C#1
BAT-A/C#2
BAT-A/C#3
77,906,785
5,133,649
4,894,776
18,430,143
18,443,911
B/D Direct Dischargers
BAT-B/D#1
BAT-B/D#2
BAT-B/D#3
109,252
23,230
22,610
45,152
45,204
A/C Indirect Dischargers
PSES-A/C#1
PSES-A/C#2
PSES-A/C#3
90,181,808
33,181,762
38,331,249
50,981,759
51,807,047
B/D Indirect Dischargers
PSES-B/D#1
PSES-B/D#2
7,424,870
1,934,646
4,378,836
4,990,451
Source: Section 308 Survey data and Radian Corp. estimates.
6-2
-------
63 ESTIMATES OF REDUCTIONS IN POLLUTANTS BASED ON THE WATER? MODEL
6.2.1 Introduction
As noted above, the Section 308 Survey requested information from facilities on the fate
of raw chemicals used in their manufacturing processes. Table 3-2 of the survey requested from
each facility an estimate of the quantity of raw materials going into product, emitted into air as
in-plant fugitive emissions, emitted into air from wastewater collection and treatment, degraded
or destroyed, discharged in wastewater, and otherwise disposed. After reviewing these quantities,
the Agency questioned the pollutant loads reported as emitted to air from wastewater,
biodegraded or destroyed, and discharged in wastewater. Because the Agency believes that more
air emissions are occurring from these facilities than were reported, it developed an independent
estimate of these loadings using the EPA's Office of Air Quality Planning and Standards
(OAQPS) WATER? model.
EPA's estimate of the fate of organic loads was performed using raw wastewater pollutant
load and treatment unit data from the detailed survey questionnaires collected from 244
pharmaceutical manufacturing facilities. The Agency used the total loading for pollutants emitted
to air from wastewater collection and treatment, degraded or destroyed, and discharged in
wastewater as an estimate of the raw wastewater loading. The disposition of this raw loading
from the survey responses was compared to the disposition estimated using the WATER?
computer model.
Section 6.22 provides a brief description of the WATER? model and Section 6.23
discusses the methodology followed in performing the WATER? analysis.
6.2.2 WATER? Model Description
The WATER? model evaluates several pollutant pathways including volatilization,
biodegradation, and adsorption onto solids for individual waste constituents from a model
6-3
-------
wastewater treatment train. The treatment process includes the following technologies in
sequence:
Pretreatment
Primary clarifier
Trickling filter
Equalization
Aeration 1
Aeration 2
Secondary clarifier
WATER? Analysis Methodology
To model raw wastewater load disposition, the following analysis methodology was used:
Default values were determined for many of the treatment modules. These values
were then used in conjunction with facility-specific information to run the model.
Facilities that provided survey information were considered for modeling.
Facilities were grouped based on subcategory type as either a Subcategory A/C
facility or a Subcategory B/D facility. Facilities with Subcategory A and/or C
manufacturing processes that also included some Subcategory B and/or D
manufacturing processes were grouped as Subcategory A/C facilities.
Facilities were grouped based on their discharge status as either direct or indirect
dischargers.
For each subcategory and discharge group, an assessment of which facilities to
model was conducted based on the following criteria:
If a facility did not have raw wastewater loadings of constituents that are
candidates for regulation, the facility was not modeled.
6-4
-------
If a facility did not have any in-plant or end-of-pipe treatment that could
be modeled by the WATER? treatment modules, the facility was not
modeled.
For Subcategory A/C indirect discharge facilities, end-of-pipe treatment
tram categories were developed and a subset of the facilities in each
category was modeled.
When a treatment train could not be modeled, an engineering judgment
was made concerning the accuracy of the organic load disposition reported
by the facility in the survey.
Several of these methodology steps are discussed further below.
6.23.1 Sources of Data
All pollutant loading data used in the analysis came from Table 3-2 of the Section 308
Survey; if only part of the total flow or loading went through certain treatment units, data from
Table 4-8 of the Section 308 Survey were used. Wastewater treatment system data were obtained
from Table 5-2 of the Section 308 Survey and the facility diagram. Data on the flow rate through
the treatment units were found on the facility diagram and in Table 7-1 of the Section 308
Survey.
6.233 Facility Groupings
As mentioned above, the 244 pharmaceutical manufacturing facilities that responded to
the survey were grouped according to manufacturing processes and discharge type. After
removing 7 zero-discharge facilities, the remaining 237 facilities were divided into four groups:
(1) A/C directs, (2) B/D Directs, (3) A/C Indirects, and (4) B/D Indirects. Pollutant loadings
from Table 3-2 of the Section 308 Survey were examined to determine which facilities reported
raw wastewater loadings of pollutants that are candidates for regulation. Facilities with
treatment systems and loadings of the pollutants of interest were considered for modeling.
Table 6-2 presents the number of facilities considered for modeling.
6-5
-------
TABLE 6-2
NUMBER OF FACILITIES CONSIDERED FOR MODELING
Facility Group
A/C Direct Dischargers
B/D Direct Dischargers
A/C Indirect Dischargers
B/D Indirect Dischargers
Zero Discharge
Total Number of
Facilities
24
14
88
111
7
Number of
Facilities With No
Wastewater
Treatment or
Without Pollutant
Loading Data
7
5
30
91
N/A
Number of
Facilities With
Pollutants of
Interest
17
9
58
20
N/A
Number of
Facilities
Modeled
17
6
12
18
0
Source: Section 308 Survey data and Radian Corp. estimates.
6-6
-------
Most of the Subcategory A/C and B/D direct discharge and Subcategory B/D indirect
discharge facilities were modeled individually. The largest group available for modeling, the
Subcategory A/C indirect discharge facilities, was subdivided into groups based on similarities in
treatment trains. The subgroups included facilities with:
Neutralization only
Primary treatment only
Secondary treatment only
Both primary and secondary treatment
Other treatment
The last subgroup listed is discussed below. The other subgroups were further divided
based on initial modeling results. Facilities with enclosed neutralization only were separated from
those with open neutralization units. Within both the primary only and the primary and
secondary groups, facilities with equalization without mixing and enclosed equalization with
mixing were grouped separately from those facilities with open equalization with mixing.
63,3.3 Treatment Units That Could Not Be Modeled
For facilities that showed no onsite treatment or primary treatment only, EPA reviewed
the facility's reported organic load disposition for reasonableness. Based on these reviews, EPA
either concurred with the load disposition reported by the facility or found that the load
identified as degraded/destroyed could not be justified and made adjustments to the disposition
loads accordingly.
In some cases, the information provided in the survey was unclear or included a
treatment unit for which no WATER? module has been developed (e.g., incineration or
stripping) early in the treatment train. For those facilities, the survey responses were evaluated
for reasonableness and either accepted as reported or adjusted based on engineering judgment
according to the data provided. If a treatment unit not included in the modeling, such as a
6-7
-------
trickling filter, was found near the end of a treatment train, the treatment train was modeled up
to that point and the remaining pollutant loading, if any, was assumed to be present in the
wastewater discharge.
6-8
-------
SECTION SEVEN
ASSESSMENT OF BENEFITS
7.1 INTRODUCTION
This section presents an assessment of the benefits from the proposed effluent guidelines
for the pharmaceutical industry. This assessment considers the human health, environmental,
and economic benefits from reductions in effluent loadings and air emissions expected to result
from the proposed rule considering the selected regulatory options. The assessment includes a
qualitative description of each benefit category. In addition, it provides quantitative estimates of
economic benefits for those benefit categories for which there are sufficient data to develop such
estimates.
7.1.1 Overview of Benefits Assessment
A broad range of benefits might result from environmental controls, including
improvement in or maintenance of human health, environmental quality, and economic welfare.
Table 7-1 provides one example framework for categorizing the benefits of actions to reduce
environmental pollution, including consideration of the benefits resulting from direct and indirect
use of environmental resources.
In general, assessing the benefits of environmental regulation includes three steps: (1)
evaluating the physical effects of the regulation, such as the changes in contaminant
concentration levels in effluent discharges; (2) identifying the categories of benefits expected to
result from these physical changes; and (3) evaluating the scope and magnitude of these benefits,
both qualitatively and quantitatively. Identifying and evaluating economic benefits involves
translating physical effects into changes in the provision of goods and services valued by society.
For example, reduced toxic loadings in effluent discharges might lead to human health benefits.
These benefits might stem from reductions in cancer risk and/or reductions in the systemic
7-1
-------
TABLE 7-1
EXAMPLE FRAMEWORK OF BENEFIT CATEGORIES
Categories of Benefits
Types of Benefits
Human Health
Avoided Mortality Effects
cancer
noncancer
Avoided Morbidity Effects
acute effects
chronic effects
Environmental Resources
Direct Use Values
recreational opportunities
Ecological Services
provision of habitat
Passive Use Values
Economic Welfare
Production
agriculture
commercial fishing
timber supply
Avoided Materials Damage
Maintenance and Enhancement of Water
Supply (quality and quantity)
Maintenance and Enhancement of Aesthetics
visibility
odor
7-2
-------
effects associated with these toxics. In addition, reductions in toxic loadings in effluent
discharges can generate environmental benefits by improving water quality, thus improving
recreational opportunities for the public.
Benefits assessment of a proposed rule requires consideration of the likely incremental
effect of the rule on human health and environmental quality. For example, a change in
emission levels should be measured as the difference between the level of emissions with the rule
and the baseline level of emissions in the absence of the rule. The benefits assessment presented
below uses such an incremental approach.
The next part of this section identifies the benefit categories considered in this
assessment. The remainder of the section describes the approach used to generate benefit
estimates for each of these categories.
7.1.2 Benefit Categories Addressed in This Benefits Assessment
Section Six provides estimates of reductions in effluent loadings from pharmaceutical
facilities expected to result from the proposed regulation and the resulting reductions in
emissions of VOCs. A variety of human health, environmental, and economic benefits might
result from these reductions. In particular, this assessment addresses the following benefit
categories:
Human health benefits due to reductions in excess cancer risk;
Human health and agricultural benefits due to reductions in emissions of ozone
precursors (i.e., reductions in VOC emissions);
Ecological and recreational benefits due to improved water quality;
Benefits from reductions in interference and pass through problems and
improvements in worker health and safety at publicly owned treatment works
(POTWs); and
7-3
-------
Human health benefits due to reductions in systemic and other risks, such as risk
of respiratory, digestive, circulatory, neural and developmental effects, or
individual organ toxicity.
For the first two benefit categories, sufficient information is available to monetize the
benefits of the proposed rule. Sections 7.2 and 73 address these two benefit categories. Each
section includes a qualitative description of these benefit categories, an overview of the approach
used to monetize these benefits, a presentation of the resultant economic benefit estimates, and a
review of the limitations of the valuation methods. Hie dollar magnitude of the benefits for the
other three benefit categories could not be quantified. The next parts of this section (Sections
7.4 through 7.6) discuss these benefit categories qualitatively. Section 7.7 presents a summary of
results.
7.2 REDUCTIONS IN CANCER RISK
This section describes the assessment of cancer risk reductions expected to result from
this rule due to reductions in VOC emissions and reductions in pollutant loadings in wastewaters
discharged to surface waters. Section 7.2.1 presents the results of the cancer risk assessment,
estimating the number of excess cancer cases avoided due to the proposed rule, Section 7.2.2
describes the method used for valuing these effects, and Section 7.23 presents monetized
estimates of the resulting human health benefits.
7.2.1 Description of Benefits
The analysis of the effects of the proposed rule, which were presented in Section Six,
estimates the reduction in VOC emissions expected to result from the proposed rule. Many of
these VOCs are carcinogenic and thus pose a risk to human health. Reductions in emissions of
these substances therefore will result in reduced cancer risk in the exposed populations. In
addition, the analysis estimates reductions in pollutant loadings to surface waters. These
reductions will improve water quality and reduce cancer risk to the exposed populations from
7-4
-------
ingestion of contaminated drinking water and fish tissue. The results from the cancer risk
assessments for these exposure routes are described below.
7.2.1.1 VOC Emissions Reductions
Based on the cancer risk assessment for reductions in VOC emissions, it is estimated that
the proposed effluent guidelines will result in the avoidance of 0.02 excess cancer cases per year
nationwide (Versar, 1995). This estimated decrease in cancer risk results from reductions in
emissions of 10 carcinogens, principally chloroform and methylene chloride. The number of
excess cancer cases avoided is calculated as (the difference in the estimated annual cancer
incidence in the exposed population under baseline emission conditions and after implementation
of the rule.
The Industrial Source Complex Long Term (ISCLT) air dispersion model in the
Graphical Exposure Modeling System (GEMS) Atmospheric Modeling System (ISCLT/GAMS) is
used to estimate the reduction in excess cancer risk likely to result from the proposed rule.
Specifically, 33 facility/pollutant release combinations are modeled, using data provided by the
pharmaceutical industry in response to EPA's Section 308 survey. The analysis considers cancer
risk reductions at open air settling, neutralization, equalization, or treatment tanks at facilities
directly discharging to surface waters.
Based on this analysis, it is estimated that 10 facility/pollutant release combinations
currently exhibit cancer risk levels exceeding 10"6 for a portion of the exposed population. It is
estimated that approximately 190 thousand people nationwide are exposed to carcinogens as a
result of these releases (based on 1990 population data). The estimated reductions in emissions
due to this rule are expected to result in 0.02 excess cancer cases avoided per year nationwide.
As discussed in Section Six, the emissions estimates reported in the Section 308
questionnaires might understate the quantity of VOCs released by the pharmaceutical industry
and thus might understate the beneiSts of the proposed rule. To test the sensitivity of the
resulting benefit estimates to this factor, an upper bound estimate of the level of VOC emissions
7-5
-------
is developed based on a "maximum emissions" scenario. The maximum emissions scenario uses
estimates of emissions reductions based on the WATER? model.
Based on this scenario, the proposed rule will result in the avoidance of 0.35 excess
cancer cases per year nationwide. This estimate also is developed using the ISCLT/GAMS
model, considering 43 facility/pollutant combinations. Based on this analysis, 12 facility/pollutant
combinations currently exhibit cancer risk levels exceeding lO"6. It is estimated that 2.4 million
individuals nationwide are exposed to carcinogens as a result of these pollutant releases.
7.2 J.2 Seductions in Pollutant Loadings to Surface Waters
Based on the cancer risk assessment, it is estimated that the proposed rule will result in
about 0.002 excess cancer cases avoided per year due to reductions in risk from exposure to
contaminants in fish tissue and drinking water. This estimate is small because the estimated
cancer incidence from consumption of fish tissue and drinking water potentially affected by
discharges from pharmaceutical facilities at current discharge levels is small.
The assessment of the cancer risk from consumption of contaminated drinking water and
fish tissue evaluates the risks associated with the effluent from 17 direct dischargers for 44
pollutants and 116 indirect dischargers for 55 pollutants. The analysis first uses a stream
dilution model to estimate instream pollutant concentrations based on current and projected
discharge levels (Versar, 1995). The analysis then estimates lifetime average daily doses and
individual risk levels for each pollutant discharged by a facility based on these instream pollutant
concentration levels. The estimated individual cancer risk levels are aggregated on a facility basis
and for those facilities with total risk exceeding 10^, cancer incidence in the exposed population
is evaluated. In the case of cancer risk from pollutants in drinking water, the analysis considers
exposure to the population served by drinking water utilities with intakes within 50 miles
downstream of the discharge point. In the case of contaminated fish tissue, the analysis
considers exposure to subsistence and recreational anglers and the general population.
7-6
-------
For the fish tissue analysis, the estimated number of excess annual cancer cases avoided
due to the proposed rule is less than 0.0001. At current discharge levels, total cancer risk to
subsistence anglers exceeds IQr6 due to the discharge of eight carcinogens from four facilities into
three streams. Given these risk levels and the size of the population exposed, however,
estimated cancer incidence is small. Thus, while the proposed rule is expected to reduce the risk
to acceptable levels (i.e., below 10**), the magnitude of the human health benefits is negligible.
Total cancer risk for recreational anglers and the general population is not expected to exceed
10"6 for any discharges.
Similarly for the drinking water analysis, the estimated number of excess cancer cases
avoided per year due to the proposed rule is small, at 0.002 cases. Based on this analysis, three
streams receive discharges from ten facilities of seven carcinogens at risk levels exceeding 10"6.
However, only two of the streams have drinking water utilities within 50 miles downstream of the
discharge point. Cancer incidence for these two streams is analyzed both under current
discharge levels and projected levels under the proposed rule, yielding an.estimate of the number
of excess cancer cases avoided of 0.002.
7.22 Valuation Methodology
To value the reductions in cancer risk expected to result from this rule, an estimate of the
range of the value of avoided mortality (also known as the value of a statistical life) is applied to
the number of excess cancer cases avoided. The estimated range of the value of life used in this
analysis is $0.6 million to $13.5 million, with a best estimate of $4.8 million (1990 $).1 This
range is based on a review of the value of life literature conducted for EPA's Office of Air and
Radiation (OAR).2 OAR undertook that analysis to support its retrospective assessment of the
lrrhis best estimate is the average of 26 value-of-life estimates identified as most appropriate
for policy analysis purposes, as discussed later in this chapter.
2See Unsworth et al. (1992) and Neumann and Unsworth (1993). Note that this estimate
also has been used in assessing the economic benefit of proposed environmental tobacco smoke
legislation (U.S. EPA, 1994a).
7-7
-------
costs and benefits of the Clean Air Act as mandated under Section 812 of the 1990 Clean Air
Act Amendments.
OAR's review of the value of life literature considered estimates of the value of life from
39 wage-risk, contingent valuation, and consumer market behavior studies. Wage-risk studies
estimate the value of life considering individuals' implicit trade-off between job-related mortality
risk and compensation. Contingent valuation studies provide estimates of the value of life based
on individuals' expressed willingness to pay to avoid hypothetical increments of mortality risk.
Consumer market behavior studies consider risk-dollar tradeoffs made by individuals based on
observed purchasing behavior in consumer markets (e.g., dollars spent in purchasing car options
that improve automobile safety).
The 39 studies considered in OAR's review of the literature were identified earlier in
literature reviews by Viscusi (1992,1993) and Fisher et al. (1989). These studies were screened
to identify estimates considered reasonable and reliable for policy analysis purposes. OAR
selected 26 estimates from 23 studies based on four criteria suggested by Viscusi (1992) for
.considering the applicability of value of life estimates to policy analysis. These four criteria are:
Use of appropriate risk measuresSome wage-risk studies rely on actuarial data
to determine risk levels faced by workers in making a wage-risk tradeoff. This
approach will bias the estimated value of a statistical life downwards because
actuarial data are not limited to work-related risks; they include all types of risk
faced by an individual.
Adequate sample sizeSome contingent valuation studies are based on small
sample sizes, reducing the reliability of the resulting estimates.
Model typeThe results of labor market studies that focus on the value of a life-
year or the implicit discount rate people apply to the value of a life-year yield less
robust estimates of the value of life, due to the complexity of the structural
models used in developing these estimates.
Use of appropriate willingness to pay proxiesViscusi (1992) finds that the
consumer market studies he reviewed failed to provide an unbiased estimate of
the value of life. The resulting estimates are biased because these studies had to
assume values either for the level of risk or dollars spent to reduce this risk to
estimate the value of life from the observed risk-dollar tradeoffs. These
assumptions reduce the reliability of the resulting value of life estimates.
7-8
-------
The resulting range of $0.6 million to $13 J million reflects the entire range of the 26 selected
value-of-life estimates.
This estimate of the range of the value of life is similar to other reported estimates of the
range of the value of life. For example, based on his expert judgement, Viscusi finds that most
of the reasonable estimates of the value of life are clustered in the $3.0 to $7.0 million range. In
addition, the range applied in this analysis encompasses the range applied in EPA's recent RIA
of the effluent guidelines for the pulp and paper industry (U.S. EPA, 1993). In that assessment,
EPA applied a range of the value of life of $2.0 million to $10.0 million in 1992 dollars (equal to
$2.0 million to $9.6 million in 1990 dollars). That range was based on a review of the literature
conducted by Fisher et al. (1989). A $0.6 million to $13.5 million range is applied because this
range includes estimates of the value of life from recent studies, including several published since
the Fisher et al. study was completed.
7.2.3 Valuation of Benefits
Applying this estimated range; of the value of life to the estimate of excess cancer cases
avoided based on the Section 308 data (0.02 cases avoided), the human health benefits from
reductions in cancer risk associated with this rule are valued at $12 thousand to $270, thousand
(1990 $). These benefit estimates represent annual nationwide benefits for the proposed rule
resulting from cancer risk reductions due to exposure both through inhalation and ingestion of
carcinogens. This analysis is summarized in Table 7-2.
An upper bound estimate of the cancer risk reduction benefits is generated based on the
maximum emissions scenario. Applying the value of life range to the upper bound estimate of
the number of cancer cases avoided (035 cases avoided) yields benefits on the order of $210
thousand to $4.7 million per year. This analysis also is summarized in Table 7-2.
7-9
-------
TABLE 7-2
ESTIMATED ANNUAL HUMAN HEALTH BENEFITS
FROM CANCER RISK REDUCTIONS
(1990 dollars)
-~SSS5SSSS====S==SS=======S5=
Number of Excess Cancer
Cases Avoided
Value of Life (million)
Total Benefits
Section 308 Data
Low
0.02
$0.6
$12,000
Best
Estimate
0.02
$4.8
$96,000
High
0.02
$13.5
$270,000
Maximum Emissions Data
Low
0.35
$0.6
$210,000
Best
Estimate
035
$4.8
$1,680,000
High
0.35
$13.5
$4,725,000
7-10
-------
7.2.4 Limitations
Apart from the uncertainty in the emissions data, there are several limitations to the risk
assessment methodology and valuation approach used in this benefits assessment that might
contribute to uncertainty in the resultant human health benefit estimates. These limitations are
discussed below.
73, A.I Risk Assessment Methodology Limitations
Limitations in the cancer risk assessment methods applied in this analysis introduce
uncertainty to the resulting benefits estimates. These limitations largely result from the
assumptions used in estimating exposure levels and the size of the population exposed. This
section lists some of the assumptions used in the analysis. A more complete discussion of the
assumptions used and their effects on the resulting benefit estimates appears in Versar (1995).
The cancer risk assessment incorporates many assumptions in evaluating exposure levels
from inhalation of carcinogens and consumption of contaminated fish tissue and drinking water.
For example, in modeling fugitive air emissions from pharmaceutical facilities and the resulting
exposure levels, the ISCLT/GAMS model uses a gaussian air dispersion algorithm, and relies on
assumptions related to facility characteristics (e.g., height of the stack), wind speed and direction,
and weather conditions (based on averages). In addition, the dilution model used in estimating
instream concentration levels assumes that complete mixing of discharge flow and stream flow
occurs. This mixing assumption implicitly results in the calculation of an average stream
concentration, although the actual concentration might vary across the width and depth of the
stream. Further, pollutant fate processes such as sediment adsorption, volatilization, and
hydrolysis are not considered in the model; this simplification might overstate the instream
concentrations of pollutants.
The risk assessment also applies many assumptions in determining the exposed population
and estimating the likely cancer incidence. For example the following assumptions are made in
7-11
-------
estimating risks to subsistence and recreational anglers and the general population from
consumption of contaminated fish tissue and drinking water:
Populations potentially exposed to discharges are estimated based
on the population of the state in which the facility is located;
The potentially exposed population within in each state is
estimated based on the ratio of the affected river miles (estuary
square miles) to total river miles (estuary square miles) within the
state;
Commercially or recreationally valuable species are assumed to inhabit the
waters in the vicinity of the discharges;
Five percent of the resident anglers in a state are assumed to be
subsistence anglers and the remaining 95 percent are assumed to be
recreational anglers; and
Cancer incidence in part is estimated assuming that subsistence and
recreational anglers share their catch with other members of their
households.
Finally, the assessment makes many assumptions commonly employed by EPA in evaluating
cancer risks, including use of high dose-response information to predict low dose-response
toricity in deriving cancer slope factors, extrapolation from cancer effects observed in animal
studies to effects in humans, and use of the 95 percent confidence limit of response. These
assumptions are considered conservative, and thus might overestimate risk levels.
This discussion of some of the assumptions used in the cancer risk assessment is not
meant to be comprehensive; rather it is meant to show that there is a significant level of
uncertainty associated with the benefits estimates from the analysis. Some assumptions might
result in overstatement of risk levels, while others might result in understatement of risk levels.
The overall level and direction of bias and uncertainty introduced by these assumptions is
unclear.
7-12
-------
7.2.4.2 Valuation Methodology Limitations
There is significant uncertainty associated with the valuation method used to monetize
these human health benefits. The underlying studies used to construct the range of the value of
life applied in this analysis primarily consider the risk-dollar tradeoffs observed in the labor
market. However, there are obvious differences in the nature of cancer mortality risk and job-
related acute fatality risk. For example, premature mortality risks from exposure to pollution are
experienced on an involuntary basis and generally are uncompensated, while job-related risks are
assumed by individuals who presumably have some choice of occupation and are compensated
for taking a riskier job. In addition, job-related risks tend to be more immediate, whereas cancer
effects tend to be incurred later in life, but involve some pain and suffering. Differences in the
nature of the mortality risk might imply differences in the value of these risks.
In addition, there are several issues related to transferring value-of-life estimates from the
existing literature to value benefits associated with cancer risk reductions resulting from the
proposed rule. These benefits transfer issues might bias the resulting benefits estimates, although
the direction of the bias is unclear. These issues include:
Potential differences in the duration of life lost (i.e., the age of the affected
individual). Existing studies indicate that the age of an individual influences the
value of life. Therefore, if the age distribution of the individuals affected by this
proposed rule does not match the age distribution of the individuals considered in
the studies from which the value-of-life estimates were developed, transferring
these value-of-life estimates to this analysis might bias the resulting benefit
estimates.
Potential differences fin the level of baseline risk. Individuals are likely to value
reductions in risk differently depending on their level of baseline risk. Thus, if the
level of baseline risk faced by individuals affected by the proposed rule differs
from the risk level faced by individuals considered in the existing value-of-life
studies, transferring these value-of-life estimates to this analysis might bias the
resulting benefit estimates.
Potential differences iin income levels. Based on economic theory, and as
estimated by Viscusi and Evans (1990), the value of life increases with increasing
income. Thus, if the average income of individuals affected by the cancer risk
addressed by this rule differs from the income of individuals considered in the
value-of-life literature, transferring these value-of-life estimates to this analysis
might bias the resulting benefit estimates.
7-13
-------
In this assessment, no adjustments are made to the estimated range of the value of life to
account for possible differences in the nature of risk or for these other sources of potential bias.
However, because the cancer risk reduction benefits are evaluated using a range, the resulting
human health benefit estimates might reflect the uncertainty resulting from these limitations.
7.3 REDUCTIONS IN EMISSIONS OF OZONE PRECURSORS
The proposed effluent guidelines are expected to result in a reduction in VOC
emissions.3 Controlling VOC emissions is beneficial because VOCs are precursors to ozone,
which negatively affects human health and the environment. For example, ozone has been found
to reduce lung function in humans and to reduce agricultural crop yields. Ozone formation also
might affect tree growth, cause materials damage, and affect visibility (Krupnick and Kopp,
1988).
This assessment of the benefits from reductions in emissions of ozone precursors
considers two potential categories of benefits:
Human health benefits associated with reductions in acute health effects; and
Economic welfare benefits due to increased agricultural yield.
This assessment does not address human health benefits from reductions in chronic health effects
nor does it address economic welfare benefits related to forest growth, materials damage, or
visibility. The benefits associated with these categories could be significant, and thus the benefit
estimates presented below might understate the total benefits of the proposed rule.
SThis analysis excludes two organic pollutants, methylene chloride and trichlorofluoro-
methane, from the estimated reductions in emissions of ozone precursors. These two organic
pollutants were identified for exclusion based on a final rule recently promulgated by EPA that
defines VOCs for the purposes of developing state implementation plans to attain national
ambient air quality standards (NAAQS) for ozone (U.S. EPA 1994b). This definition excludes
specified compounds that have negligible photochemical reactivity and thus do not contribute
significantly to the formation of tropospheric ozone, including methylene chloride and
trichlorofluoromethane.
7-14
-------
Reactions between VOCs and nitrogen oxides (NOJ form ozone in the presence of
sunlight. However, ozone formation is a complicated process that is not well understood. This
uncertainty prevents estimation of the specific changes in ozone concentrations that are likely to
occur due to the reductions in VOC emissions expected to result from the proposed rule. In this
analysis, the benefits from VOC emissions reductions are evaluated assuming a linear
relationship between VOC emissions and economic benefits from reductions in ozone
concentrations, as described below.
Both the human health and economic welfare benefits of reduced emissions of ozone
precursors are evaluated applying a benefits transfer approach. In both cases, estimates of the
average value per megagram (Mg) reduction in VOC emissions are applied to the estimated total
reduction in VOC emissions in nonattainment areas due to this rule. The value per Mg
estimates used in this assessment are based on the results of a study by the U.S. Congress, Office
of Technology Assessment (OTA, 1989).
The next sections present the results of the assessment of the human health and
-economic welfare benefits associated with reductions in emissions of VOCs. First, more detailed
qualitative descriptions of each benefit category are presented. Second, the methodology used to
monetize these benefits is discussed. Third, the results of the analysis are summarized. Finally,
the limitations associated with the valuation approach are described.
73.1 Human Health Benefits
7.3.1.1 Description of Benefits
Clinical and epidemiological studies have demonstrated that short-term exposure to
elevated ozone concentration results in acute effects on human health. These acute effects
include respiratory and nonrespiratory symptoms, such as shortness of breath, headachesr and
pain upon deep inhalation; minor restricted activity days; and asthma attacks. Reductions in
ambient ozone concentrations reduce the incidence of these acute human health effects, thus
generating economic benefits.
7-15
-------
In addition, ozone is believed to have chronic effects on human health. For example,
laboratory studies have observed chronic effects on animals exposed to elevated levels of ozone,
including increased susceptibility to infection, decreased pulmonary function, and some fibrotic-
like lung damage, which could lead to respiratory diseases such as chronic bronchitis (Krupnick
and Kopp, 1988). Hie link between ozone concentration and chronic health effects in humans,
however, is not well understood. Therefore, this category of human health benefit is not
considered quantitatively in this analysis.
This analysis considers human health benefits resulting from reductions in VOC emissions
in nonattainment areas. Nonattainment areas include counties that do not meet EPA's current
ozone standard. This approach is consistent with the approach used in the OTA study, which
considers human health benefits in metropolitan statistical areas with ozone concentrations
exceeding EPA's ozone standard. Consistency with that approach is important because the
values used to monetize these human health benefits are based on the OTA study, as discussed
below. However, reductions in ozone levels in attainment areas also might lead to human health
benefits. Some individuals might experience health effects due to exposure to ozone
concentrations at concentrations below the current ozone standard. These individuals would
benefit from the expected reductions in VOC emissions in attainment areas. This analysis
excludes these benefits and therefore will understate the benefits of this rule.
Valuation Methodology
To monetize the human health benefits associated with reductions in VOC emissions, a
benefits-transfer-based approach is used. Specifically, the estimated reductions in VOC
emissions in nonattainment areas are multiplied by an existing estimate of the value per Mg
reduction in VOC emissions. The VOC emission reductions expected to occur in nonattainment
areas due to the proposed rule are estimated based on the emissions reductions data described in
Section Six, combined with information on whether the affected facilities are located in
nonattainment areas.
7-16
-------
The value per Mg reduction in VOC emissions is based on the results of the OTA study
(1989). That study evaluated the annual nationwide human health benefits of controlling VOC
emissions to reduce ozone levels in nonattaiinment areas. One scenario in the OTA study
considered the human health benefits expected to result from a 35 percent reduction in 1985
VOC emission levels in nonattainment areas. Based on the results of this scenario, the benefits
associated with reduced VOC emissions would be expected to fall in the range of $19 and $1,209
per Mg (1990 $). This range represents the average value per Mg reduction hi emissions,
assuming a 3.5 million Mg reduction in VOC emissions and total human health benefits ranging
from $67 million to $4,233 million pier year (1990 $), with an average value of $2,150 million.
The resulting unit values are summarized in Table 7-3.
7.3.1.3 Valuation of Benefits
Applying the method described above, the estimated annual human health benefits
resulting from reductions in VOC emissions ranges from $27 thousand to $1.7 million, depending
on the unit value used. Based on the Section 308 data, it is estimated that the proposed
regulations will reduce VOC emissions in nonattainment areas by 1,396 Mg per year (U.S. EPA,
1995c). This reduction represents 10.1 percent of the estimated total reduction in VOC
emissions in all areas (i.e., both attainment and nonattainment). Applying the unit-value
estimates to this quantity yields estimates of the monetary value of the resulting human health
benefits ranging between $27 thousand and $1.7 million per year, with an average of $857
thousand. These results are summarized in Table 7-4.
7.3.1.4 Limitations
Several limitations to the valuation approach applied above contribute to uncertainly in
the resultant human health benefit estimates. First, the estimates might understate the human
health benefits because they do not include potential benefits from avoided chronic human
health effects. Second, the application of a range of values per Mg reduction in VOC emissions
might overstate or understate the benefits of the proposed rule because the value range used
7-17
-------
TABLE 7-3
DERIVATION OF HUMAN HEALTH BENEFITS PER MEGAGRAM
REDUCTION IN VOC EMISSIONS IN NONATTAINMENT AREAS
Total Annual Benefits Assuming a 35 Percent
Reduction in VOC Emissions (millions of
1990 $)*
Total Expected Reduction in VOC Emissions
(millions of Mg)*
Benefits per Mg Reduction in VOC Emissions
(1990 $/Mg)
Low
$67
3.5
$19
Average
$2,150
3.5
$614
High
$4,233
3.5
$1,209
*Based on OTA (1989). Values from the OTA study are converted from 1984 to 1990 dollars using
the GDP implicit price deflator.
Source: lEc estimate.
7-18
-------
TABLE 7-4
ESTIMATED ANNUAL HUMAN HEALTH BENEFITS FROM REDUCTIONS IN VOC EMISSIONS IN
NONATTAINMENT AREAS
(1990 dollars)
Value per Mg
VOC Emissions Reductions in
Nonattainment Areas (Mg)
Total Benefits
Section 308 Data
Low
$19
1,396
$27,000
Average
$614
1,396
$857,000
High
$1,209
1,396
$1,688,000
Source: BEc estimates.
7-19
-------
reflects the average value for a 35 percent reduction in VOC emissions rather than the marginal
value of the incremental reductions resulting from this rule. If the marginal benefits from
incremental reductions in VOC emissions are assumed to be higher (lower), application of this
range of values would understate (overstate) the benefits of the rule. In addition, the benefits
estimates from the OTA study (used in estimating the value per Mg reduction in VOC
emissions) are based on 1984 data on population in metropolitan areas. Thus the benefits
estimates might understate the benefits of the proposed rule because they do not take into
account population growth between 1984 and 1990.
Third, uncertainty associated with the relationship between VOC emission levels and
ozone formation contributes to uncertainty in the resultant benefit estimates. For example,
recent research suggests that the relative atmospheric concentrations of VOCs and NOX affect
ozone formation (National Research Council, 1991). Thus, the relationship between VOC
emissions reductions and changes in ozone concentrations might not be linear, as assumed in this
analysis. The direction of bias introduced by this factor is unknown.
Finally, the analysis assumes that there are no benefits associated with reduced VOC
emissions in areas in attainment with the ozone standard. If there are human health benefits
associated with such reductions, the analysis will understate the total benefits of this rule. The
magnitude of these benefits could be significant, potentially doubling the estimated benefits of
the rule.4
7.3.2 Welfare Benefits from Increased Agricultural Crop Yields
7.3.2.1 Description Of Benefits
Studies of the relationship between ambient ozone concentrations and greenhouse-
controlled ozone concentrations and agricultural crop yield demonstrate that ozone negatively
4Assuming that there are benefits from such reductions and applying the average value per
Mg reduction of $614 to the expected VOC emissions reductions in attainment areas yields a
benefit estimate ranging from $7.0 million to $7.6 million for this benefit category,
7-20
-------
affects crop yield. These studies show that exposure to ozone reduces plant growth due to
reduced photosynthesis and transport of carbohydrates hi plants. Reductions in crop yield can in
turn affect agricultural production, crop prices, and incomes of agricultural producers, and thus
can affect social welfare.
In addition, biological research has established that air pollution can affect not only the
yield, but also the quality of agricultural crops (Shortle et al., 1988). A reduction in quality
might correlate with a reduction in ithe nutritional value of the crop. In this case, the food value
of the crop would be reduced even if the yield remained constant. If qualitative attributes are a
determinant of the demand for a commodity, changes in demand will result in changes in price,
production, and economic welfare.
Ozone-induced crop yield changes might have secondary effects due to the responses of
the agricultural community to the yield change. It is a common agricultural practice to counter
decreased crop yields with increased use of fertilizer. In addition, crops suffering from the
effects of ozone are more susceptible to pestilence, prompting farmers to increase their use of
pesticides. Increased fertilizer and pesticide use represents an economic cost to agricultural
producers, thus reducing total economic surplus. Furthermore, a reduction in crop yield often
leads to an increase in the acreage of cultivated land to compensate for yield loss. Ozone-
induced decisions to increase the amount of cultivated land could lead to the loss of wildlife
habitat, increased soil erosion, and increased agriculture-related pollution. Increased soil
erosion, fertilizer use, and pesticide use will further increase agriculture's contribution to surface
and ground-water pollution.
Although the economic implications of these secondary effects of reduced crop yields
might be significant, this analysis only considers crop productivity impacts. Existing estimates of
the secondary environmental impacts of reduced crop productivity have not been identified and
thus these benefits have not been quantified. Therefore, the resulting benefit estimates will
understate the agricultural-related economic benefits of the proposed rule.
7-21
-------
Valuation Methodology
The economic welfare benefits expected to result from reductions in VOC emissions and
the resulting change in crop yields are evaluated applying a benefits-transfer-based approach.
Specifically, the estimated reductions in VOC emissions in nonattainment areas are multiplied by
an existing estimate of the value per Mg reduction in VOC emissions (based on the OTA study).
These VOC emission reduction estimates are the same as those used in the evaluation of human
health benefits in the previous section.
OTA estimates the range of economic benefits expected to result from changes in crop
yields based on two studies: Krupnick and Kopp (1988) and Adams and Glyer (1988). These.
two studies use economic models to estimate the net change in social welfare resulting from
higher crop yields that occur as a result of lower ambient ozone levels in rural areas. These
models consider the changes in production, prices, and producer costs to evaluate the change in
producer and consumer surplus.5
The estimated values from these two studies range between $117 and $198 per Mg
reduction in VOC emissions (1990 $). This range reflects differences in the models used in the
two studies. These models apply differing assumptions regarding the relationship between ozone
levels and crop yields, differing economic parameters (e.g., different crop prices), and differing
ozone concentration data, and consider differing crops. The crops considered in these models
include corn, wheat, soybean, cotton, barley, alfalfa, sorghum, oats, peanuts, grass-legume hay,
and rice. Table 7-5 shows the derivation of the benefit estimates for reductions in ambient ozone
levels based on these two studies.
This benefits assessment evaluates agricultural benefits in nonattainment areas only due
to uncertainty in the effect of reductions in VOC emissions in rural areas on ozone
Producer surplus represents the difference between what a producer receives for a good and
the minimum amount that the producer would accept rather than forego sale of the good.
Consumer surplus represents the difference between the maximum amount a consumer would
pay rather than forego consuming the good and the amount the consumer actually pays for the
good.
7-22
-------
TABLE 7-5
DERIVATION OF ESTIMATED AGRICULTURAL BENEFIT PER MEGAGRAM
REDUCTION IN VOC EMISSIONS
Total Benefits from a 10 Percent
Reduction in Ozone Concentrations**
(millions of 1990$)
Total Reduction in VOC Emissions
(millions of Mg)
Benefit per MG Reduction in VOC
Emissions (1990 $/Mg)
Low*
$269
23
$117
Average
$363
2.3
$158
High*
$456
23
$198
*The low value is based on the results of Krupnick and Kopp (1988) while the high value is based
on the results of Adams and Glyer (1988), as reported in OTA (1989). Values were converted from
1986 to 1990 dollars using the GDP implicit price deflator.
**These economic models evaluated the benefits expected to result from a reduction in ozone
concentrations by 10 percent of the difference between then-current ambient ozone concentrations
and background ozone levels.
Source: BEc estimates.
7-23
-------
concentrations in these areas. Recent research has considered the effectiveness of NOX versus
VOC control in reducing ozone levels. These studies suggest that the effectiveness of VOC
controls in reducing ozone formation depends on the ambient levels of VOCs and NO,, In
general, VOC control will be less effective in areas with a high VOCs to NOX ratio than in areas
with a low ratio. For example, the OTA study states that "[rjecent modeling studies that have
simulated typical rural conditions suggest that outside of urban and industrial plumes, reducing
NOX emissions will generally be a more effective strategy for lowering ozone than reducing VOC
emissions." (OTA 1989). This research suggests that reductions in VOC emissions in rural areas
may not have as large of an effect on ozone levels in rural areas as might be hoped.
Unfortunately it does not provide insight into the relationship between reductions in VOC
emissions and reductions in ozone levels in rural areas. To be conservative, the assessment only
considers agricultural benefits from reductions in VOC emissions in nonattainment areas,
assuming a linear relationship between VOC emissions reductions and reductions in ozone
concentrations in these areas. Excluding potential agricultural benefits from reductions in VOC
emissions in attainment areas is likely to result in understating the benefits of the proposed rule.
7.323 Valuation of Benefits
Applying the valuation method outlined above, the estimate of the agricultural benefits
from reductions in VOC emissions ranges from $163 thousand to $276 thousand per year (1990
$). A summary of these economic benefit estimates is presented in Table 7-6.
7.3.2.4 Limitations
There are several limitations to the valuation approach used in estimating the agricultural
benefits resulting from reductions in VOC emissions. One of the principal limitations involves
the benefits-transfer-based approach applied in developing the benefit estimates. This approach
might overstate expected agricultural benefits because of differences in the magnitude of the
physical effects estimated for the proposed rule and the magnitude of the effects considered by
7-24
-------
TABLE 7-6
ESTIMATED ANNUAL ECONOMIC WELFARE BENEFITS FROM
REDUCTIONS IN OZONE-INDUCED IMPACTS ON AGRICULTURE
(1990 dollars)
Value per Mg
Reductions in VOC Emissions (Mg)
Total Benefits
Section 308 Data
Low
$117
$1,396
$163,000
Average
$158
$1,396
$221,000
High
$198
$1,396
$276,000
Source: lEc estimates.
7-25
-------
OTA (1989). The. magnitude of the change hi ozone concentration evaluated by OTA is orders
of magnitude larger than that of the likely reductions due to this proposed rule.
A second limitation to this approach is that the resulting estimates might understate the
agricultural benefits because they do not include potential benefits from reductions in VOC
emissions in attainment areas. Studies have shown that ozone concentrations in excess of
background levels affect crop yields (OTA 1989, Krupnick and Kopp 1988); this finding suggests
that reductions in ozone concentrations in attainment areas might generate agricultural benefits.
Although reductions hi VOC emissions might not be as effective in reducing ozone in attainment
areas as other approaches, such VOC emissions reductions could result in some reduction in
ozone concentrations in these areas. The degree to which the resulting estimates understate the
agricultural benefits is not known and depends on the relationship between VOC levels and
ozone formation in rural areas.
A final limitation to this approach involves the assumption that for nonattainment areas,
a linear relationship exists between reductions in VOC emissions and reductions in ozone levels.
As discussed above, the relationship between VOC emissions and ozone formation is a complex
process that is not well understood. It is not clear whether the assumed linear relationship will
overstate or understate the agricultural benefits in nonattainment areas resulting from the
proposed rule.
7.4 ENVIRONMENTAL BENEFITS
The proposed effluent guidelines are expected to generate environmental benefits by
improving water quality. These improvements in water quality are expected to result from
reduced loadings of toxic substances in the effluent of the regulated facilities and reduced
interference and passthrough problems at POTWs. The environmental benefits expected to
result from the proposed rule are discussed below. Related benefits from reduced interference
and passthrough at POTWs are discussed in more detail in Section 7.5.
7-26
-------
7.4.1 Description of Benefits
A wide range of environmental benefits is associated with the maintenance and
improvement of water quality. These benefits include use values (e.g., recreational fishing),
ecological values (e.g., provision of habitat), and passive use values.6 For example, water
pollution can affect the quality of the fish and wildlife habitat provided by water resources, thus
affecting the species utilizing these resources. This effect in turn can affect the quality of
recreational experiences of users, such as anglers fishing in the affected streams. In addition,
individuals might value actions to preserve water resources even if they do not currently plan to
use these resources. These passive-use values reflect the value placed on a resource for reasons
other than direct human use, such as the value placed on knowing that the resource exists and on
knowing that the resource will be available to future generations (Krutilla, 1967; Freeman, 1993).
The potential recreational benefits from improvements in water quality expected to result
from the proposed rule are evaluated below. The analysis only develops an order-of-magnitude
estimate of these benefits due to limitations In the available data. This analysis does not evaluate
the ecological or passive-use-value benefits of the proposed rule because sufficient data to
evaluate these benefits are not available. Ecological and passive-use values of water resources
are potentially significant, however, and omitting these benefit categories from the benefit
assessment will result in understating the environmental benefits of the proposed rule.
Individuals have shown a general concern for and willingness to pay to maintain water quality
and the habitat services provided by water resources, both through donations to conservation
organizations and through responses to contingent valuation surveys. For example, public
policies protecting natural resources and public and private expenditures on preserving the
quality of water resources demonstrate that the public values these resources and is willing to pay
and make financial sacrifices to preserve them. In addition, results from contingent valuation
surveys considering willingness to pay for water quality improvements show that individuals hold
significant positive values for water quality (e.g., see Desvousges et al., 1983; Fisher and Raucher,
1984; and Carson and Mitchell, 1988). Although these studies do provide an indication of
'Human health benefits resulting from improvements in water quality, including cancer risk
reductions and systemic risk reductions are evaluated in Sections 7.2 and 7.6.
7-27
-------
individuals' concern for water resources and the order of magnitude of an individual's willingness
to pay to preserve them, these studies do not meet current guidelines for conducting contingent
valuation studies and thus do not provide data that can be used in estimating the ecological and
passive-use benefits of the proposed rule.
7.42 Valuation Methodology
To estimate the economic value of the improvements in water quality expected to result
from this rule, instream concentrations of toxic pollutants resulting from wastewater discharges
by the affected facilities are modeled (Versar, 1995). Using a simple dilution model, these
instream pollutant concentration levels are estimated for 17 facilities directly discharging 44
pollutants to 17 receiving streams and 116 facilities indirectly discharging 55 pollutants to 87
streams. These concentrations are modeled both under current conditions and under the
proposed rule. The resulting instream concentration estimates are then compared to EPA's
freshwater acute and chronic aquatic life criteria to evaluate whether these discharges pose risk
to aquatic organisms.
The projected reductions in toxic loadings to surface waters is significant. For direct
dischargers, pollutant loadings are estimated to decline by 98 percent, from 5.2 million pounds
per year under current conditions to 0.1 million pounds per year under the proposed rule.
Similarly, for indirect dischargers, pollutant loadings are estimated to decline 51 percent, from
34.8 million pounds per year under current conditions to 16.9 million pounds per year under the
proposed rule.
The analysis comparing instream concentration levels to aquatic life water quality criteria
estimates that current discharge loadings result in pollutant levels exceeding aquatic water quality
criteria at two locations. The analysis also indicates that no pollutant levels exceeding water
quality criteria are expected to occur at these two sites based on projected discharge loadings
under the proposed rule. Thus the proposed rule is expected to result in a substantive change in
water quality at these two sites. The two locations and pollutants of concern are Honey Creek,
Indiana (ammonium hydroxide and isopropanol), and the Los Angeles River, California
7-28
-------
(ethanol). Note that although improvements in water quality would be expected at other
locations, these locations did not experience pollutant levels exceeding any water quality standard
as a result of pharmaceutical industry discharges under current conditions.
The recreational benefits from the expected water quality improvements at these sites are
estimated using a simple model. The model estimates the recreational benefits from water
quality improvements based on the change in consumer welfare likely to result from improved
angler catch rates at these sites. Although consumer surplus for a day of recreational fishing is a
function of several site characteristics (including species mix, catch rate, fish size, and other site
amenities), numerous studies have applied catch rate as a proxy for site quality, under the
assumption that catch rate is an important determinant of the value that a recreational angler
receives for a fishing day (e.g., Samples and Bishop, 1985; NAPAP, 1989). The model also
assumes that these two sites are used by recreational anglers, that improvements in water quality
at these sites will increase fish populations, and that recreational catch rates will increase with
these increases in fish populations.7
7.43 Valuation of Benefits
Based on this assessment, it is estimated that the potential recreational benefits from the
proposed rule are on the order of thousands of dollars per year. To develop this estimate,
approximations of the number of trips per year at these sites, the change in fish population, and
the change in consumer welfare due to increases in catch rate (i.e., the increased value per
fishing trip) are used. Fishing pressure at these sites is estimated based on the estimated average
number of fishing trips per river and stream mile per year in these two states and length of river
or stream affected. The change in consumer welfare is estimated based on a value per trip
ranging between $20 and $40 and an elasticity of consumer surplus equal to one (i.e., a one
percent increase in catch is assumed to yield a one percent increase in consumer surplus).
'Interviews with regional resource managers indicate that recreational fishing at the Los
Angeles River site is unlikely, given a lack of access to the affected segment of this river and
given the intermittent nature of the river.
7-29
-------
7.4.4 Limitations
Limitations in the methods used in modeling instream concentration levels and in valuing
the potential effects on recreational fishing introduce uncertainty to the benefits estimates. First,
the approach used to model instream pollutant concentrations makes several assumptions,
including:
Background concentrations of each pollutant both in the receiving stream
and the POTW influent are assumed to equal zero; thus only the
ecological effects of discharges of the facilities addressed by the rule are
evaluated.
The analysis assumes complete mixing of discharge flow and stream flow.
This mixing results in an estimated average stream concentration, whereas
the actual concentration can vary across the width and depth of the
receiving water.
The pollutant load to the receiving stream is assumed to be continuous
and representative of long-term facility operations.
The analysis does not consider sediment adsorption, volatilization, and
hydrolysis; this approach can result in concentration estimates that are
higher than actual levels.
As a result of these assumptions, the model might overestimate or underestimate instream
concentrations, thus affecting the extent to which the rule eliminates the occurrence of pollutant
levels exceeding aquatic life water quality criteria. The overall effect of these assumptions is
unclear.
Second, significant uncertainty is associated with the recreational benefit estimates due to
the limited data available to assess these benefits. For example, there are no data on the
expected effects of improved water quality on fish populations at these sites. In addition, the
analyses incorporate assumptions on fishing pressure at the two sites, the length of the river
affected by the discharges, the value of a fishing day, and the change in the value of a fishing day
due to higher catch rates. For example, fishing pressure estimates for these two sites are based
on state averages, since fishing pressure data for these sites are not available. Given the data
7-30
-------
constraints, the benefits estimate for this benefit category is presented as an order of magnitude
estimate.
7.5 EFFECTS AT POTWs
The proposed rule establishes pretreatment standards for 55 pollutants discharged to
POTWs by pharmaceutical facilities. EPA identified the pollutants to be addressed by
pretreatment standards based on analyses of the quantity of wastewaters discharged by facilities,
pollutant concentration levels in these wastewaters, and the number of facilities that discharge
these pollutants. This section qualitatively describes the potential benefits to POTWs resulting
from these pretreatment standards.
7.5.1 Description of Benefits
This analysis considers three potential sources of benefits to POTWs from the proposed
pretreatment standards:
Reductions in the likelihood of interference, passthrough, and sewage
sludge contamination problems;
Reductions in health and safely risks to POTW workers; and
Reductions in costs potentially incurred by POTWs in analyzing toxic
pollutants and determining whether, and the appropriate level at which, to
set local limits.
Each of these potential benefit categories is discussed below.
7-31
-------
7JSJ..1 Seductions in Interference, Passthrough, and Sewage Sludge Contamination Problems
Toxic pollutants contained in the effluent loadings of pharmaceutical plants and
discharged to POTWs can cause interference problems and/or pass through a POTW's treatment
system and potentially affect water quality or contaminate sludges generated during treatment.
Interference is defined as the inhibition or disruption of POTW operations. Interference can
result from large quantities or high concentrations of toxic pollutants in effluent discharges that
might adversely affect the operation of a POTW, potentially affecting the treatment efficiency or
capacity of the plant. Similarly, passthrough can result when toxic pollutants in effluent
discharges are not addressed by a POTW's treatment process or if the quantity or concentration
of pollutants prevents the POTW from fully treating the wastewaters. These pollutants can
remain in the wastewaters and be discharged by the POTW to surface waters. Alternatively,
these pollutants can remain in the treatment sludges. Passthrough and sludge contamination
problems affect POTWs to the extent that they prevent POTWs from meeting their permits or
sewage sludge criteria.
Anecdotal evidence and analytic results indicate that such effects can occur. POTW
responses to an EPA survey addressing toxic substances in effluent discharges by pharmaceutical
manufacturers and the impact of these substances on POTW operations provides evidence that
these effluent loadings can cause inhibition problems at POTWs (Radian, 1993). For example,
one POTW indicated that high concentrations of volatile organics in a pharmaceutical facility's
effluent might have caused nitrification problems at the POTW. Another POTW stated that
low-level discharges of some compounds can affect treatment plant operations. For example,
releases of siloxanes affected the efficiency of the POTW's boiler and ultimately forced the plant
to replace this equipment.
As part of the analysis of the effects of pretreatment standards, POTW influent levels are
compared to available data on inhibition levels (Versar, 1995). This analysis considers the
potential impacts of effluent from 65 pharmaceutical facilities containing 51 pollutants discharged
to 96 POTWs. Under current conditions, inhibition problems are projected to occur at six
POTWs for seven pollutants. Inhibition problems are projected to occur at five POTWs for
7-32
-------
three pollutants after the proposed rale. Sufficient data are not available to monetize this
benefit category.
Limited evidence is available on the extent to which discharges from pharmaceutical
facilities cause POTWs to fail to comply with their permits or result in pollutant levels in sewage
sludges that exceed sewage sludge criteria. There are several documented incidents of large slug
loads or accidental releases from pharmaceutical facilities that have negatively affected the
environment, including fish kills, degradation of water quality, and odor problems (Versar,
1995).8 In addition, currently many pollutants are not controlled in POTW permits because
information is unavailable on the potential impacts of these pollutants on the environment.
Although discharge and failure to treat unregulated pollutants technically does not constitute
passthrough, these pollutants enter and potentially have negative effects on the environment.
Finally, an analysis comparing sewage sludge concentration levels to sewage sludge quality
criteria is not possible, since sewage sludge quality criteria for the 55 pollutants selected for
regulation under the proposed rule have not been developed.
The proposed pretreatment standards might help reduce shock releases (i.e., unexpected
releases that contain high concentrations of toxic pollutants) from pharmaceutical facilities and
thus reduce the likelihood that these releases will cause interference, passthrough, and sewage
sludge contamination problems at POTWs. The proposed pretreatment standards are expected
to reduce toxic loadings in the industry's effluent, thus reducing the toxic loadings discharged to
surface waters. The benefits from these effects, including health benefits and benefits from
improved water quality, could be significant. However, sufficient data to quantify these benefits
further are not available.
8Note that some of these releases might have been in violation of existing regulations, and
thus it might be inappropriate to attribute benefits resulting from proper control of these releases
to the proposed rule. However, if the proposed rule does reduce the likelihood of such releases,
it might be argued that such benefits are attributable to the rule.
7-33
-------
75J.3, Benefits to POTW Worker Health and Safety
Toxic substances in effluent discharges to POTWs pose health risks to POTW workers.
Volatilization of toxics from POTW influent could pose cancer risk to POTW workers or
increase the risk of explosion at the plant. For example, in the survey of POTWs, one facility
indicated that releases of sulfuric acid led to potential human health risk at the facility. The
proposed rule is expected to reduce these risks, thus generating human health benefits.
Following procedures outlined in EPA's Guidance to Protect POTW Workers from Toxic
and Reactive Gases and Vapors (U.S. EPA, 1992), risks to POTW workers from exposure to
toxics are evaluated under current conditions and under proposed pretreatment standards
(Versar, 1995).9 Occupational exposure levels at POTWs are modeled based on the mixture of
vapors that can partition out of influent water into the surrounding air. Risks to POTW workers
are evaluated comparing these estimated exposure levels to occupational Threshold Limit Values
(TLVs). For each POTW potentially affected by the proposed rule, hazard ratios (pollutant
concentration/TLV) are estimated for each pollutant in the effluent of a facility discharging to
the POTW; these ratios are aggregated to obtain an estimate of the overall hazard level. If the
sum of the hazard ratios exceeds one, then POTW workers potentially are at risk.
Applying this approach, the proposed rule is estimated to reduce occupational risk at six
POTWs. The analysis evaluates effluent discharged by 129 facilities to 85 POTWs. Based on
current conditions, 12 POTWs treating 51 pollutants are estimated to have total hazard ratios
greater than one. In contrast, six POTWs are estimated to have hazard ratios greater than one
based on projected discharge levels under the proposed rule. Benzene is estimated to have the
largest effects on POTW workers. Data are not available to monetize this benefit category.
'The analysis does not consider risks to sewer workers, assuming that these workers would not
be exposed to toxic emissions for long periods of time without using protective gear. The analysis
considers risk levels assuming distillation pretreatment rather than steam stripping pretreatment as
under the proposed rule, and thus provides an upper bound estimate of the reductions in risks to
POTW workers under the proposed rule.
7-34
-------
73J.3 Benefits from Reductions in Analytical Costs
Under the National Pretreatment Program, authorized POTWs are required to develop
and implement programs to control pollutants discharged by facilities to their systems. These
local programs set numerical limits on toxic loadings in discharges to the POTW, based on
national categorical pretreatment standards or local limits determined by the POTW. Local
limits are designed to prevent passthrough, interference, and sewage sludge contamination, taking
into account POTW-specific and effluent-specific characteristics, as well as to implement other
specific components of the National Pretreatment Program (e.g., preventing discharges that
might cause fire, explosion hazard, corrosive structural damage, or worker health and safety
problems).
In setting these local limits, POTWs might need to undertake analyses to determine
which pollutants warrant local limits and at what numerical level. Conducting these analyses is
expensiveon the order of hundreds of thousands of dollars. Thus, establishing pretreatment
standards benefits POTWs by allowing them to avoid the costs of performing these analyses. In
addition, it is more efficient to conduct such analyses at the national level, reducing the potential
for duplication of effort. Several POTWs contacted as part of the POTW survey indicated that
they will benefit from the establishment of national pretreatment standards by avoiding these
analytical costs. In addition, they indicated that the pretreatment standards will bolster the legal
authority of the limits they set.
7.6 REDUCTIONS IN SYSTEMIC RISK
7.6.1 Description of Benefits
Exposure to toxic substances might pose risk of systemic and other effects to humans,
including effects on the circulatory, respiratory or digestive systems and neurological and
developmental effects. The proposed rule might generate human health benefits by reducing
exposure to these substances, thus reducing the risks of these associated effects.
7-35
-------
As in the case of the cancer risk assessment, systemic risks are evaluated for exposure
from air emissions and consumption of contaminated fish tissue and drinking water. Modeled
pollutant concentration levels are compared to human health criteria or estimated toxic effect
levels. This analysis is performed using both the Section 308 data and the "maximum emissions"
scenario data.
Based on the Section 308 data, the proposed rule is not expected to result in reductions
in systemic risk due to reductions in air emissions because based on these data, baseline risk
levels are low. (Versar, 1995). Under the maximum emissions scenario, the proposed rule is
estimated to reduce systemic risk due to reductions in air emissions. The analysis estimates that
126,000 individuals will benefit from reductions in exposure to two toxic pollutants, triethylamine
.and 2-methoxyethanol. These two substances might have nasal and ocular effects or have
testicular effects, respectively. For the drinking water and fish tissue exposure routes, under both
data scenarios, the proposed rule is not expected to result in reductions in systemic risk because
estimated concentration levels under current conditions are below human health criteria or toxic
effect levels. Sufficient data to quantify these benefits further are not available.
7.7 SUMMARY OF RESULTS
The estimated annual benefits resulting from the proposed rule range from $202
thousand to $6.7 million (1990 $). This range reflects uncertainties in estimating the physical
effects of the proposed rule and in placing a dollar value on these effects. Table 7-7 summarizes
the results of this analysis, by benefit category.
These estimates are likely to understate the benefits of the proposed rule because they do
not include several benefit categories that could not be monetized. Categories excluded from
these estimates include ecological and recreational benefits from improvements in water quality,
human health benefits associated with potential reductions in chronic effects resulting from
ozone exposure, human health benefits associated with reductions in acute effects in attainment
areas, agricultural benefits from reductions in emissions of ozone precursors in attainment areas,
7-36
-------
OS
i
H
f
1
CQ
'i
o
a
g
g
a
i
8'
.
g»
O
i
a
1
9»
1
O
09
a
.i
03
*X
^N
el-
's
a
i
a
tS
V
GO
J
1
1
J-
A
Oft
a
«?
^
I
Benefit Category
*o
5"
1
T 1
O
O
8
1
1 i
ae-
Cancer Risk Reductions
OO
OO ^^
§B
90 c»
81
oo ^^
21 ^
oo «s
-------
benefits from avoided interference problems and improved worker health and safety at POTWs,
and human health benefits from potential reductions in systemic risk.
7-38
-------
SECTION EIGHT
COMPARISON OF BENEFITS TO COSTS
In this section, the estimate of the annual social costs of the regulation reported in
Section Five is compared to the estimate of the total annual benefits.
Table 8-1 presents the annual benefits and annual costs of the selected regulatory
options. Total social costs are $108.4 million and total quantifiable benefits range up to $7.0
million.
8-1
-------
TABLE 8-1
COMPARISON OF ANNUAL BENEFITS AND COSTS
FOR THE PHARMACEUTICAL RULEMAKING
(in thousands of 1990 dollars)
Benefits
Cancer risk reductions
Reductions in emissions of ozone precursors
Human health
. Agricultural benefits
Total quantifiable benefits
$12 - $4,725
$27 - $1,688
$163 - $276
$202 - $6,689
Costs
Total Annual Costs to Industry
Total Annual Social Costs
$70,000
$108,400
8-2
-------
SECTION NINE
REFERENCES
Adams, Richard M. and J. David Glyer. 1988. "Assessment of the Agricultural Benefits of
Tropospheric Ozone Reductions Using Office of Technology (OTA) Assumptions."
Report prepared for the Office of Technology Assessment.
Carson, Richard T. and Robert Cameron Mitchell. 1988. The Value of Clean Water: The
Public's Willingness to Pay for Beatable, Fishable, and Swimmable Quality Water."
Discussion Paper, Department of Economics, University of California, San Diego.
Desvousges, William H., V. Kerry Smith and Matthew P. McGIvney. 1983. "A Comparison of
Alternative Approaches for Estimating Recreation and Related Benefits of Water Quality
Improvements." Report to the U.S. Environmental Protection Agency, Office of Policy
Analysis.
Fisher, Ann, Lauraine G. Chestnut and Daniel M. Violette. 1989. "The Value of Reducing Risk
of Death: A Note on New Evidence." Journal of Policy Analysis and Management 8(11:
88-100.
Fisher, Ann and Robert Raucher. 1984. "Intrinsic Benefits of Improved Water Quality:
Conceptual and Empirical Perspectives." In: Advances in Applied Microeconomics,
Volume 3, V. Kerry Smith and Ann D. Witte, eds., Greenwich, CT: JAI Press.
Freeman, A Myrick IE. 1993. The Measurement of Environmental and Resource Values.
Theory and Methods. Washington, DC: Resources for the Future.
Krupnick, Alan J. and Raymond J. Kopp. 1988. "The Health and Agricultural Benefits of
Reductions in Ambient Ozone hi the United States." Resources for the Future,
Discussion Paper QE88-10.
Krutffla, John V. 1967. "Conservation Reconsidered." American Economic Review 57:777-786.
National Acid Precipitation Assessment Program (NAPAP). 1989. Methods for Valuing Acidic
Deposition and Air Pollution Effects. State-of-Science/Technology Report 27, Part B.
National Resource Council, Committee on Tropospheric Ozone Formation and Measurement.
1Q01 Rftthinldnft the Ozone Problem in Urban and Regional Air Pollution. Washington
DC: National Academy Press.
Neumann, James E. and Robert E. Unsworth. 1993. "Revisions to the Proposed Value of Life
Methodology for the Section 812 Retrospective," memorandum to Jim DeMocker, Office
of Policy Analysis and Review, Office of Air and Radiation, U.S. Environmental
Protection Agency, from Industrial Economics, Incorporated.
9-1
-------
Radian Corporation. 1993. "Summary of POTW Survey Responses." Memorandum to Frank
Hund and Ed Terry, U.S. Environmental Protection Agency, Office of Water,
Engineering and Analysis Division, from Kirsten Mahsman, Radian Corporation, 15
November.
Samples, Karl C. and Richard C. Bishop. 1985. "Estimating the Value of variations in Anglers'
Success Rates: An Application of the Multiple-Site Travel Cost Method." Marine
Resource Economics 2:(l):55-75.
Shorfle, J.S., M. Phillips and J.W. Dunn. 1988. "Economic Assessment of Crop Damages Due to
Air Pollution: The Role of Quality Effects." Environmental Pollution 53:377-385.
Unsworth, Robert R, James E. Neumann and W. Eric Browne. 1992. "Review of Existing Value
of life Estimates: Draft Valuation Document" Memorandum to Jim DeMocker, Office
of Policy Analysis and Review, Office of Air and Radiation, U.S. Environmental
Protection Agency, from Industrial Economics, Incorporated.
U.S. Congress, Office of Technology Assessment (OTA). 1989. "Catching our Breath: Next Steps
for Reducing Urban Ozone." OTA-O-412, Washington, DC: U.S. Government Printing
Office.
U.S. EPA. 1986. Report to Congress on the Discharge of Hazardous Waste to Publicly Owned
Treatment Works. EPA 530-SW-86-004. February.
U.S. EPA. 1992. Development Document for Effluent Guidelines and Standards for the
Pharmaceutical Point Source Category.
U.S. EPA- 1990. Supporting Statement for OMB Review: Detailed Questionnaire for the
Pharmaceutical Manufacturing Industry. Washington, D.C.: Office of Water Regulations
and Standards.
U.S. EPA- 1992. "Guidance to Protect POTW Workers from Toxic and Reactive Gases and
Vapors."
U.S. EPA. 1993. Regulatory Impact Assessment of Proposed Effluent Guidelines and NESHAP
for the Pulp, Paper, and Paperboard Industry. Office of Water.
U.S. EPA. 1994a. The Costs and Benefits of Smoking Restrictions: An Assessment of the
Smoke-Free Environment Act of 1993 (H.R. 3434)." Prepared at the request of
Congressman Henry Waxman, Chairman, Subcommittee on Health and the Environment,
Committee on Energy and Commerce. Office of Radiation and Indoor Air, Indoor Air
Division.
U.S. EPA. 1994b. Air Qualify: Revision to the Definition of Volatile Organic Compounds -
Exclusion of Volatile Methyl Siloxanes and Parachlorobenzotrifluoride, 40 CFR Part 51,
59 FR 50693, October 5,1994.
9-2
-------
U.S. EPA. 1995a. Development Document for Proposed Effluent Limitations Guidelines and
Standards for the Pharmaceutical Manufacturing Point Source Category. Office of
Water, February.
U.S. EPA. 1995b. Economic Impact Analysis of Proposed Effluent Limitations Guidelines and
Standards for the Pharmaceutical Manufacturing Industry, February.
U.S. EPA. 1995c. Cost-Effectiveness Analysis of Proposed Effluent Limitations and Standards
for the Pharmaceutical Manufacturing Industry, February.
U.S. Office of Management and Budget. 1989. Regulatory Impact Guidance. Appendix V of
Regulatory Program of the United States Government, April 1,1988 - March 31,1989.
Versar, Inc., 1995. "Environmental Assessment of the Pharmaceutical Manufacturing Industry."
Prepared for the U.S. Environmental Protection Agency, Office of Science and
Technology, Standards and Applied Science Division, December.
Viscusi, W. Kip. 1992. Fatal Tradeoffe: Public and Private Responsibilities for Risk. New York:
Oxford University Press.
. 1993. The Value of Risks to Life and Health." Journal of Economic Literature
31:1912-1946.
Viscusi, W. Kip and W. Evans. 1990. "Utility Functions That Are Dependent on One's Health
Status." American Economic Review. As cited in W. Kip Viscusi. 1992. Fatal
Tradeoffe: Public and Private Responsibilities for Risk. New York: Oxford University
Press.
9-3
-------
------- |