United States
Environmental Protection
Agency
Office of Water Regulations
and Standards
Washington, DC 20460
Water
Economic Impact
Analysis  of Effluent
Guidelines and Stan-
dards for the Elec-
trical  and Electronic
Components Industry
EPA 440/2-83-005
March 1983
Phase I
              QUANTITY

-------
This document is available at EPA Regional offices.
 Copies may be obtained from the National  Technical
 Information Service, Springfield, Virginia  22161.

-------
         ECONOMIC IMPACT ANALYSIS OF
  EFFLUENT LIMITATIONS AND STANDARDS FOR THE
ELECTRICAL AND ELECTRONIC COMPONENTS INDUSTRY
                   PHASE I
                Submitted to:

       Environmental Protection Agency
      Office of Analysis and Evaluation
  Office of Water Regulations and Standards
              401 M Street, S.W.
           Washington, D.C.  20460
         EPA Contract No. 68-01-6348
       JRB Project No. 2-834-03-760-14
                  March 1983

-------
     This document is an economic impact assessment of the recently-issued
effluent guidelines.  The report is being distributed to EPA Regional
Offices and state pollution control agencies and directed to the staff
responsible for writing industrial  discharge permits.  The report includes
detailed information on the costs and economic impacts of various treatment
technologies.  It should be helpful to the permit writer in evaluating
the economic impacts on an industrial facility that must comply with effluent
limitation guidelines or water quality standards.
     The report is also being distributed to EPA Regional Libraries, and
copies are available from the National Technical  Information Service (NTIS),
5282 Port Royal Road, Springfield,  Virginia  22161 (703/487-4600).
     If you have any questions about this report, or if you would like
additional information on the economic impact of the regulation, please
contact the Economic Analysis Staff in the Office of Water Regulations
and Standards at EPA Headquarters:
                       401 M Street, S.W. (WH-586)
                       Washington,  D.C.  20460
                       (202) 382-5397
The staff economist for this project is Renee Rico (202/382-5386).

-------
                                   PREFACE

     This document is a contractor's study prepared for the Office of Water
Regulations and Standards of the Environmental Protection Apency (EPA).  The
purpose of the study is to analyze the economic impact which could result from
the application of effluent standards and limitations issued under Sections 301,
304, 306, and 307 of the Clean Water Act to the electrical and electronic
components industry.

     The study supplements the technical study (EPA Development Document) sup-
porting the issuance of these regulations.  The Development Document surveys
existing and potential waste treatment control methods and technology within
particular industrial source categories and supports certain standards and
limitations based upon an analysis of the feasibility of these standards in
accordance with the requirements of the Clean Water Act.  Presented in the
Development Document are the investment and operating costs associated with
various control and treatment technologies.  The attached document supplements
this analysis by estimating the broader economic effects which might result
from the application of various control methods and technologies.  This study
investigates the effects in terms of product price increases, effects upon
employment and the continued viability of affected plants, effects upon foreign
trade, and other competitive effects.

     The study has been prepared with the supervision and review of the Office
of Water Regulations and Standards of EPA.  This report was submitted in
accordance with Contract No. 68-01-6348, Work Assignment 14, by JRB Associates
and was completed in March 1983.

-------
                              TABLE OF CONTENTS
SECTION                            TITLE
PREFACE
SUMMARY
PART I - ELECTRONIC CRYSTALS SUBCATEGORY
Chapter
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
OVERVIEW
STEP 1:
STEP 2:
STEP 3:
STEP 4:
STEP 5:
STEP 6:
STEP 7:
STEP 8:
STEP 9:
        INTRODUCTION
        1.1  PURPOSE
        1.2  SCOPE
        1.3  TECHNICAL AND ECONOMIC SUBCATEGORIZATION
        1.4  ORGANIZATION OF PART I OF THIS REPORT

        STUDY METHODOLOGY
                      DESCRIPTION OF INDUSTRY CHARACTERISTICS
                      SUPPLY-DEMAND ANALYSIS
                      COST OF COMPLIANCE ESTIMATES
                      PLANT-LEVEL SCREENING ANALYSIS
                      PLANT-LEVEL PROFITABILITY ANALYSIS
                      CAPITAL REQUIREMENTS ANALYSIS
                      PLANT CLOSURE ANALYSIS
                      OTHER IMPACTS
                       SMALL BUSINESS ANALYSIS
        INDUSTRY -DESCRIPTION
        3.1  MANUFACTURING PROCESSES
        3.2  FIRM AND PLANT CHARACTERISTICS
        3.3  PRODUCT CHARACTERISTICS AND USES
        3.4  TRENDS
        3.5  FOREIGN TRADE

        BASELINE PROJECTIONS OF INDUSTRY CONDITIONS
        4.1  U.S.  ELECTRONIC CRYSTAL SALES PROJECTIONS
        4.2  CAPITAL EXPENDITURES PROJECTIONS
        4.3  EMPLOYMENT PROJECTIONS
        4.4  SUMMARY

        COST OF COMPLIANCE
        5.1  OVERVIEW
        5.2  POLLUTANT PARAMETERS
        5.3  RECOMMENDED TREATMENT TECHNOLOGIES
        5.4  TREATMENT COST ESTIMATES
             5.4.1  Existing Sources
             5.4.2  New Sources
PAGE
 S-l
 1-1
 1-1
 1-1
 1-2
 1-2

 1-3
 1-3
 1-5
 1-5
 1-8
 1-8
 1-9
 1-10
 1-12
 1-13
 1-14

 1-16
 1-16
 1-16
 1-17
 1-21
 1-26

 1-27
 1-27
 1-29
 1-29
 1-30

 1-31
 1-31
 1-31
 1-32
 1-33
 1-34
 1-34

-------
                        TABLE OF CONTENTS (Continued)


SECTION                            TITLE                                 PAGE


PART I - ELECTRONIC CRYSTALS SUBCATEGORY (Continued)

   6    ECONOMIC IMPACTS                                                   1-39
        6.1  PRICE A1TO QUANTITY CHANGES                                    1-39
        6.2  RESULTS OF SCREENING ANALYSIS                                 1-40
        6.3  PROFIT IMPACT ANALYSIS                                        1-40
        6.4  CAPITAL REQUIREMENTS ANALYSIS                                 1-44
        6.5  POTENTIAL PLANT CLOSURES                                      1-48
        6.6  EMPLOYMENT EFFECTS                                            1-48
        6.7  SUBSTITUTION EFFECTS                                          1-48
        6.8  FOREIGN TRADE IMPACTS                                         1-48
        6.9  NEW SOURCE IMPACTS                                            1-49
        6.10 SUMMARY OF IMPACTS OF PROMULGATED REGULATIONS                 1-49

   7    SMALL BUSINESS ANALYSIS                                            1-52

   8    LIMITATIONS OF THE ANALYSIS                                        1-56
        8.1  DATA LIMITATIONS                                              1-56
        8.2  METHODOLOGY LIMITATIONS                                       1-57
             8.2.1  Price Increases Assumption                             1-57
             8.2.2  Profit Impact Threshold Assumptions                    1-58
             8.2.3  Capital Availability Threshold Assumptions             1-59
             8.2.4  Plant Closure Assessment                               1-59
        8.3  SUMMARY OF LIMITATIONS                                        1-59

APPENDIX I-A  SENSITIVITY OF THE IMPACT ESTIMATES TO ALTERNATIVE           1-60
              ESTIMATES OF MONITORING COSTS AND SOLVENT DISPOSAL COSTS

PART II - SEMICONDUCTOR SUBCATEGORY

   1    INTRODUCTION                                                       H-l
        1.1  PURPOSE                                                       II-l
        1.2  SCOPE                                                         n-2
        1.3  ORGANIZATION OF PART II OF THIS REPORT                        II-3

   2    STUDY METHODOLOGY                                                  H-4
        2.1  OVERVIEW                                                      n-4
        2.2  STEP 1:   DESCRIPTION OF INDUSTRY CHARACTERISTICS              II-6
        2.3  STEP 2:   SUPPLY-DEMAND ANALYSIS                               I1-6
        2.4  STEP 3:   COST OF COMPLIANCE ESTIMATES                         11-10
        2.5  STEP 4:   PLANT LEVEL PROFITABILITY ANALYSIS                   11-10
        2.6  STEP 5:   CAPITAL REQUIREMENTS ANALYSIS                        11-12
        2.7  STEP 6:   PLANT CLOSURE ANALYSIS                               11-13
        2.8  STEP 7:   OTHER IMPACTS                                        11-14
        2.9  STEP 8:   SMALL BUSINESS ANALYSIS                              11-15
                                      ii

-------
                        TABLE OF CONTENTS (Continued)
SECTION                            TITLE                                  PAGE

PART II - SEMICONDUCTOR SUBCATEGORY  (Continued)

   3    INDUSTRY DESCRIPTION                                               11-17
        3.1  MARKET CHARACTERISTICS                                        11-17
             3.1.1  Major Product Groups and Trends                        11-17
             3.1.2  End-Use Markets                                        11-19
        3.2  INDUSTRY STRUCTURE                                            11-21
             3.2.1  Plant Characteristics                                  11-21
             3.2.2  Industry Concentration                                 11-23
        3.3  FOREIGN TRADE                                                 11-23
        3.4  PRICING BEHAVIOR AND TRENDS                                   11-27

   4    BASECASE PROJECTIONS                                               11-31
        4.1  U.S. SEMICONDUCTOR SALES PROJECTIONS                          11-31
        4.2  CAPITAL EXPENDITURES PROJECTIONS                              11-33
        4.3  EMPLOYMENT PROJECTIONS                                        11-33

   5    COST OF COMPLIANCE                                                 11-37
        5.1  OVERVIEW                                                      11-37
        5.2  POLLUTANT PARAMETERS                                          11-38
             5.2.1  Pollution Parameters Analyzed                          11-38
             5.2.2  Pollutants to be Regulated                             11-38
        5.3  RECOMENDED TREATMENT TECHNOLOGIES                             11-38
        5.4  TREATMENT COST ESTIMATES                                      11-39
             5.4.1  Existing Sources                                       11-41
             5.4.2  New Sources                                            11-41

   6    ECONOMIC IMPACTS                                                   11-45
        6.1  PRICE AND QUANTITY IMPACTS                                    11-45
        6.2  PLANT LEVEL PROFIT IMPACT ANALYSIS                            11-46
        6.3  CAPITAL REQUIREMENTS ANALYSIS                                 11-49
        6.4  POTENTIAL PLANT CLOSURES                                      11-49
        6.5  EMPLOYMENT IMPACTS                                            11-51
        6.6  FOREIGN TRADE IMPACTS                                         11-53
        6.7  NEW SOURCE IMPACTS                                            11-53
        6.8  SUMMARY OF SEMICONDUCTOR INDUSTRY ECONOMIC  IMPACTS            11-54
        6.9  LONG-TERM IMPLICATIONS AND OTHER IMPACTS                      11-56
        6.10 SMALL BUSINESS ANALYSIS                                       11-57
             6.10.1  Definitions of  Small Business                         11-57
             6.10.2  Impacts on Small Entities                             11-59

   7    LIMITATIONS OF THE ANALYSIS                                        11-61
        7.1  DATA LIMITATIONS                                              11-61
        7.2  METHODOLOGY                                                   11-62
             7.2.1  Price Increases                                        11-62
             7.2.2  Sensitivity Analysis                                   11-63
        7.3  PLANT CLOSURES                                                11-66
        7.4  SAMPLING                                                      11-66
        7.5  SUMMARY OF LIMITATIONS                                        11-66

APPENDIX II-A  SENSITIVITY OF THE IMPACT ESTIMATES TO  ALTERNATIVE          11-67
              ESTIMATES OF MONITORING COSTS AND  SOLVENT  DISPOSAL COSTS

                                      iii

-------
                                LIST OF TABLES
NUMBER                             TITLE
 S-l    INDUSTRY COMPLIANCE COSTS OF ALTERNATIVE TREATMENT OPTIONS        S-ll
        (In Thousands of 1979 Dollars)

 S-2    INDUSTRY COMPLIANCE COSTS OF SELECTIVE TREATMENT OPTIONS          S-l 2
        (In Thousands of 1982 Dollars)

 S-3    SUMMARY OF ESTIMATED ECONOMIC IMPACTS FOR THE ELECTRONIC          S-l 3
        CRYSTALS SUBCATEGORY

 S-4    SUMMARY OF ESTIMATED ECONOMIC IMPACTS FOR THE SEMICONDUCTORS      S-l 6
        SUBCATEGORY
PART I - ELECTRONIC CRYSTALS SUBCATEGORY

  3-1   NUMBER OF ELECTRONIC CRYSTAL PLANTS IN THE UNITED STATES          1-18

  3-2   GEOGRAPHICAL DISTRIBUTION OF ELECTRONIC CRYSTAL PLANTS            1-19

  3-3   DISTRIBUTION OF PLANTS BY PLANT SIZE                              1-20

  3-4   MAJOR APPLICATIONS FOR ELECTRONIC CRYSTALS                        1-22

  3-5   U.S. SHIPMENTS OF PIEZOELECTRIC CRYSTAL DEVICES, 1973-1978        1-24

  3-6   U.S. SHIPMENTS OF SEMICONDUCTOR DEVICES 1969-1979                 1-25

  4-1   PROJECTIONS OF U.S. ELECTRONIC CRYSTAL SHIPMENTS                  1-28

  4-2   PROJECTIONS OF U.S. ELECTRONIC CRYSTAL INDUSTRY NEW CAPITAL       1-29
        EXPENDITURES

  4-3   PROJECTIONS OF EMPLOYMENT IN THE U.S.  ELECTRONIC CRYSTAL          1-30
        INDUSTRY

  5-1   PROFILE OF NON-ARSENIDE METAL CRYSTAL MODEL PLANTS                1-35

  5-2   COMPLIANCE COSTS OF INDIRECT DISCHARGER NON-ARSENIDE METAL        1-36
        CRYSTAL MODEL PLANTS

  5-3   TOTAL COMPLIANCE COSTS FOR 53 INDIRECT DISCHARGER NON-ARSENIDE    1-37
        CRYSTAL PLANTS
                                      IV

-------
                          LIST OF TABLES (Continued)


NUMBER                             TITLE                                 PAGE

PART I - ELECTRONIC CRYSTALS SUBCATEGORY (Continued)

  5-4   COMPLIANCE COSTS OF EIGHT ARSENIDE CRYSTAL PLANTS                 1-38

  6-1   SCREENING ANALYSIS                                                1-41

  6-2   SUMMARY OF PROFIT IMPACT ASSESSMENT FOR NON-ARSENIDE CRYSTAL      1-42
        MODEL PLANTS

  6-3   PROFIT IMPACT ASSESSMENT OF POTENTIAL "HIGH IMPACT" ARSENIDE      1-43
        CRYSTAL PLANTS

  6-4   COMPLIANCE CAPITAL REQUIREMENTS ANALYSIS - NON-TOXIC METAL        1-45
        CRYSTAL MODEL PLANTS

  6-5   COMPARISON OF CASH FLOW AND TOTAL CAPITAL REQUIREMENTS            1-46

  6-6   COMPLIANCE CAPITAL REQUIREMENTS ANALYSIS - ARSENIDE CRYSTAL       1-47
        PLANTS

  6-7   SUMMARY OF ESTIMATED ECONOMIC IMPACTS                             1-50

  7-1   SUMMARY OF SMALL BUSINESS ANALYSIS FOR ARSENIDE CRYSTAL PLANTS    1-53

  7-2   SUMMARY OF SMALL BUSINESS ANALYSIS FOR NON-ARSENIDE CRYSTAL       1-54
        PLANTS

  A-l   ESTIMATED COMPLIANCE COSTS BASED ON ANNUAL MONITORING COSTS       1-61
        OF $11,000 PER PLANT

  A-2   SCREENING ANALYSIS BASED ON ANNUAL MONITORING COSTS OF $11,000    1-62
        PER PLANT

  A-3   SUMMARY OF PROFIT IMPACT ASSESSMENT BASED ON ANNUAL MONITORING    1-63
        COSTS OF $11,000 PER PLANT

  A-4   ESTIMATED COMPLIANCE COSTS ASSTJMING INCURRENCE OF SOLVENT         1-65
        DISPOSAL COSTS

  A-5   SCREENING ANALYSIS ASSUMING INCURRENCE OF SOLVENT DISPOSAL COSTS  1-67

  A-6   SUMMARY OF PROFIT IMPACT ASSESSMENT ASSUMING INCRURENCE OF        1-68
        SOLVENT DISPOSAL COSTS

-------
                          LIST OF TABLES (Continued)


NUMBER                        '     TITLE                                 PAGE

PART II - SEMICONDUCTOR SUBCATEGORY

  3-1   VALUE OF SHIPMENTS OF THE U.S.  SEMICONDUCTOR  INDUSTRY 1972-1982   11-18

  3-2   DISTRIBUTION OF SEMICONDUCTOR PRODUCTS  BY MAJOR END-USERS         11-20
        (1973-1975)

  3-3   DISTRIBUTION OF SEMICONDUCTOR PLANTS  AND VALUE OF SHIPMENTS       11-22
        BY SIZE OF PLANT

  3-4   GEOGRAPHICAL DISTRIBUTION OF DOMESTIC SEMICONDUCTOR PLANTS IN     11-24
        1977

  3-5   CONCENTRATION RATIOS OF U.S. DOMESTIC SEMICONDUCTOR SHIPMENTS     11-25
        1957, 1965, AND 1972

  3-6   PRICE INDEXES FOR SEMICONDUCTOR DEVICES (1975-1980)               11-28

  4-1   U.S. SEMICONDUCTOR SHIPMENT FORECASTS                             11-32

  4-2   NEW CAPITAL EXPENDITURES FOR THE SEMICONDUCTOR INDUSTRY,          11-34
        HISTORICAL AND FORECAST VALUES

  4-3   SEMICONDUCTOR INDUSTRY EMPLOYMENT FORECASTS                       11-35

  5-1   SEMICONDUCTOR PLANT OPERATING STATISTICS                          11-42

  5-2   MODEL PLANT COMPLIANCE COSTS                                      11-43

  5-3   TOTAL INDUSTRY COMPLIANCE COSTS                                   11-44

  6-1   SALIENT STATISTICS FOR THE SEMICONDUCTOR SAMPLE PLANTS            11-47

  6-2   SEMICONDUCTORS - ANNUAL COMPLIANCE COSTS AS A PERCENTAGE OF       11-48
        REVENUES

  6-3   SEMICONDUCTORS - POLLUTION CONTROL INVESTMENT COSTS AS A          11-50
        PERCENT OF ANNUAL CAPITAL EXPENDITURES

  6-4   POTENTIAL PLANT CLOSURES AND EMPLOYMENT IMPACTS FOR               11-52
        SEMICONDUCTOR PLANTS

  6-5   SUMMARY OF ESTIMATED ECONOMIC IMPACTS                             11-55

  6-6   SEMICONDUCTOR - AVERAGE PLANT AND TOTAL INDUSTRY FLOW RATES       I1-58

  6-7   SUMMARY PROFIT AND CAPITAL IMPACT ANALYSIS BY EMPLOYEE SIZE       H-60
        CLASSIFICATION
                                       VI

-------
                          LIST OF TABLES (Continued)


NUMBER                             TITLE                                 PAGE

PART II - SEMICONDUCTOR SUBCATEGORY (Continued)

  7-1   SEMICONDUCTORS SENSITIVITY ANALYSIS - ANNUAL COMPLIANCE COSTS     11-64
        AS A PERCENTAGE OF REVENUES

  7-2   SEMICONDUCTORS SENSITIVITY ANALYSIS - POLLUTION CONTROL           11-65
        INVESTMENT COSTS AS A PERCENTAGE OF NEW CAPITAL EXPENDITURES

  A-l   MONITORING COSTS SENSITIVITY ANALYSIS - ANNUAL COMPLIANCE         11-68
        COSTS AS A PERCENTAGE OF REVENUES

  A-2   SEMICONDUCTOR INDUSTRY STATISTICS FOR SOLVENT DISPOSAL            11-70

  A-3   SOLVENT DISPOSAL SENSITIVITY ANALYSIS - ANNUAL COMPLIANCE COSTS   11-71
        AS A PERCENTAGE OF REVENUES

  A-4   SUMMARY OF SENSITIVITY ANALYSIS BY EMPLOYEE SIZE CLASSIFICATION   11-72
                                      vii

-------
                               LIST OF FIGURES
NUMBER                             TITLE                                 PAGE




PART I - ELECTRONIC CRYSTALS SUBCATEGORY




2-1     ECONOMIC ANALYSIS STUDY OVERVIEW                                   1-4
PART II - SEMICONDUCTOR SUBCATEGORY




2-1     ECONOMIC ANALYSIS STUDY OVERVIEW                                   II-5




3-1     INTEGRATED CIRCUIT LEARNING CURVE                                  11-20
                                    viii

-------
                                   SUMMARY

1.  INTRODUCTION

1.1  Purpose

     This report identifies and analyzes the economic impacts which may
result from the promulgation of EPA's effluent regulations on the Electronic
Crystals and Semiconductors subcategories of the Electrical and Electronic
Components point source category.  These regulations include effluent limita-
tions and standards based on Best Practicable Control Technology Currently
Available (BPT), Best Available Technology Economically Achievable (BAT),
Best Conventional Pollutant Control Technology (BCT), New Source Performance
Standards (NSPS), and Pretreatment Standards for New and Existing Sources
(PSNS and PSES).  These are being promulgated under authority of Section 301,
304, 306, 307,  and 501 of the Federal Water Pollution Control Act, as Amended
by the Clean Water Act of 1977 (Public Law 92-500).  The primary economic
impact variables assessed in this study include the potential for the regula-
tions to cause  plant closures, price changes, unemployment, changes in industry
profitability,  structure and competition, shifts in the balance of foreign
trade, and impacts on small businesses.

1.2  Industry Coverage

     This study is concerned with two subcategories of the Electrical and
Electronic Products point source category.  These are:

     •  Electronic crystals; and
     •  Semiconductors.
                                     S-l

-------
     The Electronic Crystals subcategory as defined for this study includes
establishments that manufacture piezoelectric,  semiconductor, and liquid
crystals.  The major products manufactured are:

     •  Piezoelectric crystals which include quartz, ceramic,
        yttrium iron garnet, and lithium niobate crystals;
     •  Semiconducting crystals which include silicon, gallium,
        arsenide, gallium phosphide, indium arsenide, indium,
        antimonide, bismuth telluride,  and sapphire; and
     •  Liquid crystals.

     For the purpose of setting effluent standards, the Electronic Crystals
subcategory was organized into the two  following product groups:

     •  Plants which fabricate gallium  arsenide and/or indium
        arsenide crystals; and
     •  Plants which fabricate all other electronic crystals.

     The Semiconductor subcategory includes establishments that manufacture
semiconductors and related devices (SIC 3674) and electronic components not
elsewhere classified (SIC 3679).  The growing and fabrication of crystals and
the assembly of finished electronic products are not included in this sub-
category.  The major product areas of the SIC 3674 industry segment include:

     •  Hybrid Integrated Circuits - thick filament, thin
        film,  and multichip devices
     •  Bipolar Integrated Circuits
     •  Metal  Oxide Silicon Devices
     •  Transistors
     •  Diodes and Rectifiers
                                     S-2

-------
     •  Selenium Rectifiers
     •  Light Sensitive and Light Emitting Devices (solar
        cells and light emitting diodes)
     •  Thyristors.

     The major product areas of the SIC 3679 industry segment include:

     •  Magnetic Bubble Memories
     •  Liquid Crystal Displays.

2.  METHODOLOGY

     The approach used to assess the economic impacts that may occur as a
result of the costs of each regulatory option is to (1) develop an opera-
tional description of the price and output behavior of the industry and (2)
assess the likely plant-specific responses to the incurrence of the compliance
costs enumerated in the body of this report.  Thus, industry conditions before
and after compliance with the final regulations are compared.  Supplemental
analyses are used to assess linkages of the Semiconductor and Electronic Crys-
tals industries' conditions to other effects such as employment, community,
and balance of trade impacts.  These analyses were performed for four regula-
tory options considered by EPA.  The methodology of the study includes nine
tasks that can be grouped into seven major steps.  Although each step is
described independently, there is considerable interdependence among them.
Because each of the two subcategories (electronic crystals and semiconductors)
were analyzed separately, specific analytical techniques used in some of the
steps differ somewhat between the two subcategories.  Specifically, the study
proceeded using the following seven steps:

Step 1;  Description of Industry Characteristics

     The first step in the analysis is to develop a description of basic
industry characteristics such as the determinants of demand, market structure,

                                     S-3

-------
the degree of intra-industry competition,  and financial performance.  The
sources for this information include government reports, trade association
data, discussions with various trade associations and industry personnel,
and an EPA survey of firms in the industry.

Step 2:  Industry Supply-Demand Analysis

     The second step in the analysis is a  determination of the likely changes
in market prices and industry production levels resulting from each regulatory
option.  The estimates of post-compliance  price and output levels are used in
the plant-level analysis (steps 4 and 5) to  determine post-compliance revenue
and profit levels for specific plants in each product group.

     The supply-demand analysis for this study assumes that in the short to
intermediate time periods, electronic crystal and semiconductor manufacturers
would attempt to absorb all of the compliance costs, and therefore, there will
be no price increase because of the regulations.  This assumption derives from
an evaluation of the structure, operations,  and trends of the two industries.

     The post-compliance market price levels (i.e., zero price increases
resulting from the regulation) are used, in  a later step, to assess the finan-
cial conditions of individual electronic crystals and semiconductor manufac-
turing facilities.

Step 3:  Compliance Cost Estimates

     Investment and annual compliance costs  for the recommended treatment
options were estimated by EPA's Effluent Guidelines Division for treatment
systems of various selected sizes.  Based  on these cost estimates, compliance
cost curves were developed and then used to  estimate plant specific compliance
costs.
                                    S-4

-------
Step 4;  Plant Level Financial Analysis

     Two basic financial characteristics are examined:  profitability and
capital requirements.  Both characteristics are examined through standard finan-
cial analysis techniques.  That is, profitability is measured by both return
on assets (crystals) and return on sales (semiconductors).  Capital requirements
of the final regulations are evaluated in terms of the amount of the initial
capital investment in relation to revenues (crystals) and in relation to average
industry levels of annual new plant and equipment expenditures (semiconductors).

     For the Semiconductors subcategory, the use of these techniques was
hampered by a lack of plant-specific data on a number of input variables,
such as value of shipments, plant asset values, and profit margins.  However,
using a number of industry-wide parameters from various published sources
in combination with the limited firm-specific data available, estimates were
developed for each of the key parameters.

Step 5:  Estimation of Plant Closures

     This step involves the assessment of the degree of impacts on individual
plants.  The decision to close a plant, like most major investment decisions,
is ultimately -judgemental because it involves a wide variety of considerations,
many of which cannot be quantified or even identified.  These assessments were
made by evaluating the above financial variables in conjunction with non-
financial and non-quantifiable factors, such as substitutability of products,
plant and firm integration, the existence of specialty markets, and expected
market growth rates.

Step 6:  Assessment of Other Impacts

     Once the assessment of plant closures and price and quantity changes are
made, other variables which flow from these are analyzed including employment,
industry structure, and imports and exports.  These impacts are assessed
through the use of industry-wide ratios calculated from the available data
sources (e.g., value of shipments per employee).

                                     S-5

-------
Step 7:  Small Business Analysis

     This analysis identifies the economic impacts which are likely to result
from the promulgation of the regulations on small businesses in the Electronic
Crystals and Semiconductors manufacturing industries.  The primary economic
variables covered are those analyzed in the general economic impact analysis
such as plant financial performance, plant closures, and unemployment and com-
munity impacts.  Most of the information and analytical techniques in the
small business analysis are drawn from the general economic impact analysis.
The specific conditions of small firms are evaluated against the background
of the general conditions in each product group's market.  In addition to
the general economic impacts discussed in previous steps, two analytical
problems are central to this portion of the analysis:  the definition of
small businesses in the Electronic Components industry and the considerations
of alternative regulatory options that might mitigate potential impacts on
small businesses.

     Size definitions were sought which would account for firm size in com-
parison to total industry size and in comparison to unit compliance costs
(unit compliance costs increase significantly in reverse proportion to plant
size).  Since available data on compliance cost and production is on a plant
basis, the individual production facility, rather than firm is used as the
basis for the analysis.  For electronic crystals manufacturing plants three
size definitions based on plant revenues were examined to provide EPA with
alternative definitions of small plants.  These size categories are: plants
with revenues under $1 million, $3 million, and $5 million, respectively.
For semiconductor manufacturing plants the number of employees was the primary
variable used to distinguish size.  Plants with fewer than 100 employes are
considered small and those with 100 to 250 employees are considered medium-
small .

     The consideration of alternative regulatory options is considered in
terms of the most extreme special consideration — the exemption of small
entities from the regulations.  The effects of these exemptions on total
                                     S-6

-------
compliance costs, the economic impacts, and the effectiveness of the regula-
tions are assessed.

3.  INDUSTRY CHARACTERISTICS

3.1  Electronic Crystals

     In 1978, U.S. production of piezoelectric crystals was estimated to be
around $110 million, while that for semiconducting and liquid crystals was
$450 million.  Shipments of peizoelectric crystals have been growing at an
annual rate of 5 percent since 1973, while shipments of semiconducting and
liquid crystals have been growing about 17 percent a year between 1969 and
1979.  U.S. production of piezoelectric crystals is projected to continue to
grow about 3 percent a year to reach $130 million and $160 million in 1983
and 1990, respectively.  Meanwhile, shipments of semiconducting crystals are
projected to reach $770 million in 1983 and then average about 13 percent a
year for the rest of the decade to reach $1.84 billion in 1990.

     EPA identified 70 plants that manufacture electronic crystals in the
United States, employing about 10,000 employees.  The majority of the plants
are concentrated in four states:  California, Ohio, Pennsylvania, and Texas.
More than half of the plants have less than $3 million in product shipments
or less than 200 employees.  Most electronic crystal producers are diversified,
producing other electronic devices in addition to crystals.

     The electronic crystals industry exhibits some characteristics of noncom-
petitive markets.  There is a significant amount of industry concentration, low
demand elasticities, and a high degree of capital intensity.  However, the
existence of the threat of foreign competition appears to be a deterent to
noncompetitive pricing behavior.  For these reasons, most firms in the indus-
try appear to have both the capability and the incentive for absorbing small
                                     S-7

-------
and moderate amounts of production cost increases in the short and intermediate
term.

3.2  Semiconductors
     The value of shipments for all semiconductor devices in the U.S.  domestic
industry was $9.5 billion in 1982.  This represented an increase of  5  percent
over the 1981 level of shipments.  The growth in this industry segment over
this period has been a result of the decreases in the prices of semiconductor
products which stimulated demand.

     Most of the largest semiconductor companies have several domestic plants
(usually in the same region of the country) as well as several foreign plants.
The semiconductor industry is comprised of a large number of very small and
medium sized plants, which are concentrated in a few states.  Approximately a
third of these plants were located in California.  The 1977 Census of Manu-
factures reports 545 establishments whose primary activity is the manufacture
of semiconductors.  Based on an EPA. industry survey, it is estimated that
257 semiconductor plants will be affected by the regulation.  Of these, 77
plants are direct dischargers and 180 are indirect dischargers.

     The principal markets for semiconductor devices include the four major
market categories :

     •  Computer
     •  Industrial
     •  Consumer
     •  Government, including military and space.

These markets,  with few exceptions, are outperforming the overall economy.
They are expected to continue to grow in the future.
                                     S-8

-------
     The specific nature of the pricing mechanism in the semiconductor indus-
try is not clear, since the available evidence show some characteristics that
are indicative of noncompetitive markets and some that are indicative of

competitive markets.  Although the market pricing mechanism is not precisely
determined, it has been shown that the industry has both the incentive and

the capability to not raise prices in response to small and moderate levels
of mandated pollution control costs.


4.  COMPLIANCE COSTS


     Based on the analysis of the potential pollutant parameters and treatment-
in-place in the Electronic Components industry, EPA identified six treatment
technologies that are most applicable for the reduction of the selected

pollutants.  These treatment technologies are:
     •  Option 1:
     •  Option 2:



     •  Option 3:



     •  Option 4:


     •  Option 5:


     •  Option 6:
Segregation and collection of spent solvents
containing toxic organics for reuse, resale,
or contract hauling (referred to as solvent
management), plus end-of-pipe treatment for
pH control.

Option 1 plus end-of-pipe precipitation/
clarification for control of fluoride, arsenic,
and suspended solids.

Option 1 plus precipitation/clarification of
concentrated fluoride wastes for control of
fluoride.

Option 2 plus recycle of treated effluent
to further reduce pollutant discharges.

Option 2 plus filtration to further reduce
pollutant discharges.

Option 5 plus activated carbon to further
reduce toxic organics.
                                     S-9

-------
     The economic impact analysis does not consider treatment options 4 and
6 because they are not technically feasible nationwide.  For this reason,
the economic impact analysis concentrated on treatment options 1, 2, 3, and
5 only.  Table S-l presents the estimated investment and annual compliance
costs for these treatment options for the Electronic Crystals and Semicon-
ductors subcategories.

     Table S-2 summarizes the compliance costs of the selected BPT, BAT,
NSPS, PSES, and PSNS limitations.  All except the selected PSES regulations
for the arsenide crystals and BAT regulations for the Semiconductors subcate-
gory will have relatively small compliance costs.  The total investment and
annual compliance costs for the final regulations are $5.7 million and $4.3
million (1982 dollars), respectively.

5.  ECONOMIC IMPACTS

     As described previously, the primary economic impact variables assessed
are industry financial performance, plant closures, unemployment community
effects, changes in imports and exports, and industry structure.

5.1  Electronic Crystals

     The estimated economic impacts for the Electronic Crystals subcategory
are summarized in Table S-3.

     Toxic Organics

     The control of toxic organics and pH under option 1 is not expected to
cause any incremental compliance costs for non-arsenide crystal direct dis-
chargers under BPT and BAT.  These plants already control their discharges to
the selected treatment level for regulation.

     Non-arsenide crystal indirect dischargers covered under this subcategory
are expected to incur monitoring costs to comply with the total toxic organics
                                     S-10

-------
                     TABLE S-l.  INDUSTRY COMPLIANCE COSTS OF ALTERNATIVE TREATMENT OPTIONS
                                         (in Thousands of 1979 Dollars)
  INDUSTRY SEGMENT
ELECTRONIC CRYSTALS
TREATMENT OPTION 1
   Inv.    Annual
        TREATMENT OPTION 2
          Inv.      Annual
                    TREATMENT OPTION 3
                      Inv.      Annual
                               TREATMENT OPTION 5
                                 Inv.     Annual
Gallium/ Indium Arsenide
Indirect Dischargers
/
Direct Dischargers3'
Subtotal
Non-Arsenide Crystals
Indirect Dischargers
Direct Dischargers
Subtotal

8
0
8

53
0
53

16
0
16

109
0
109

761
0
761

4,510
0
4,510

560
0
560

3,182
0
3,182

N/A
N/A
N/A

2,731
0
2,731

N/A
N/A
N/A

2,096
0
2,096

856
0
856

4,970
0
4,970

614
0
614

3,452
0
3,452
 Total Crystals Subcategory

SEMICONDUCTORS
   Indirect Dischargers
   Direct Dischargers

 Total Semiconductors
 Subcategory
      61
     180
      77

     257
125
369
158

527
 5,263
65,879
28,078
 3,742
49,368
21,026
93,947    70,394
 2,731
12,147
 3,604
2,096
8,167
2,452
 5,826
74,662
31,811
           15,751    10,619   106,473
 4,066
53,916
22,968

76,884
N/A = Mot applicable.  Treatment option not selected by EPA for this product group.

a' All Electronic Crystal direct dischargers are already in compliance with the regulations.

-------
     TABLE  S-2.  INDUSTRY COMPLIANCE COSTS OF  SELECTED  TREATMENT OPTIONS
                        (In Thousands of  1982 Dollars3/)
 RECOMMENDED REGULATIONS
ELECTRONIC CRYSTALS SUBCATEGORY
 SELECTED
 TREATMENT
TECHNOLOGY
  COMPLIANCE COSTS
INVESTMENT    ANNUAL
   Gallium/Indium Arsenide Crystals

     BPT                               Option 2
     BCT                               Option 2
     BAT                               Option 2
     NSPS                              Option 2
     PSES                              Option 2
     PSNS                              Option 2

   Non-Arsenide Crystals

     BPT                               Option 2
     BCT                               Option 2
     BAT                               Option 2
     NSPS                              Option 2
     PSES                              Option I
     PSNS                              Option 1
                       0
                       0
                       0
                       Ob/
                     945
                       Ob/
                       0
                       0
                       0
                       Ob/
                      66
                  0
                  0
                  0
                  ob/
                696
                  Ob/
                  0
                  0
                  0
                  ob/
                135
 SEMICONDUCTORS SUBCATEGORY
     BPT
     BCT
     BAT
     NSPS
     PSES
     PSNS
  Option 1
  Option 1
  Option 3
  Option 3
  Option 1
  Option 1
     96
      0
  4,380C/
      Ob/
    223
      Ob/
  196
    0
2,850C/
    Ob/
  458
    Ob/
TOTAL
                                                         5,710
                               4,335
a'  Adjusted to 1982 dollars using Engineering News Record Construction Cost
   indexes.

b/  NSPS/PSNS costs are defined as incremental costs from BAT/PSES.  Since
   the recommended NSPS/PSNS limitations are the same as those proposed for
   BAT/PSES, NSPS/PSNS costs are zero.

c/  Incremental costs from BPT option.
                                    S-12

-------
              TABLE S-3.  SUMMARY OF ESTIMATED ECONOMIC IMPACTS
                   FOR THE ELECTRONIC CRYSTALS SUBCATEGORY
ECONOMIC IMPACT VARIABLES

Gallium/Indium Arsenide Crystals

  Annual Compliance Costs/Revenues
    Indirect Dischargers
    Direct Dischargers
  Change in Price (%)
  Change in Quantity (%)
  Change in Profitability
    Indirect Dischargers
    Direct Dischargers
  Capital Requirements
    Indirect Dischargers
    Direct Dischargers
  Plant Closures due to Regulations
  Employment at Closed Plants
  Balance of Trade Changes
  Industry Structure Changes

Non-Arsenide Crystals

  Annual Compliance Costs/Revenues
    Indirect Dischargers
    Direct Dischargers
  Change in Price (%)
  Change in Quantity (%)
  Change in Profitability
    Indirect Dischargers
    Direct Dischargers
  Capital Requirements
    Indirect Dischargers
    Direct Dischargers
  Plant Closures due to Regulations
  Employment at Closed Plants
  Balance of Trade Changes
  Industry Structure Changes
                                   OPTION 1 OPTION 2 OPTION 3 OPTION
                                                SELECTED
                                              5   OPTION
                                      0-0.2
                                       0
                                       0
                                       0

                                      Low
                                       0
                       0-3.5
                        0
                        0
                        0

                      Moderate
                        0
 N/A
 N/A
 N/A
 N/A

 N/A
 N/A
 0-3.9
  0
  0
  0
  0-3.5
   0
   0
   0
Moderate  Moderate
  0         0
Low
0
0
0
None
None
Moderate
0
0
0
None
None
N/A
N/A
N/A
N/A
N/A
N/A
Moderate
0
0
0
None
None
Moderate
0
0
0
None
None
               0-2.1  0.8-7.6 0.2-3.1  0.8-7.6
                0000
                0000
                0000

             Moderate  Signif. Moderate Signif,
                0000
                                                                         0-2.1
                                                                          0
                                                                          0
                                                                          0

                                                                        Moderate
                                                                          0
                                      Low
                                       0
                                       0
                                       0
                                     None
                                     None
                      Moderate Moderate Moderate  Low
                         0
                         0
                         0
                       None
                       None
  0
  0
  0
None
None
  0
  0
  0
None
None
  0
  0
  0
None
None
N/A = Not applicable.
      group.
Treatment option not selected by EPA for this product
                                     S-13

-------
limitation under PSES.  These costs, totaling $66,000 in capital investment
and $135,000 in total annual costs, assume that at least the 50 percent of
the plants already in compliance with the toxic organic limitation will choose
to certify that they do not dump their solvents into the effluent instead of
monitoring.  The remaining plants are expected, on average, to monitor quar-
terly.  Since the incremental costs of solvent disposal tend to be offset by
resale, EPA did not estimate costs for solvent management.

     These monitoring costs are expected to cause moderate reductions in
profitability for small non-arsenide crystal plants, but no plant closures
are expected as a result of this level of control.

     EPA conducted a sensitivity analysis for option 1 costs in addition to
the above analysis, consisting of two parts.  First, the impact of monthly
monitoring was estimated for all facilities (some facilities may monitor as
frequently as once per month, although EPA cannot identify precisely which
ones).  In addition to this analysis, costs and impacts were also determined
for facilities which may incur "worst case" solvent disposal costs.  These
costs are developed for facilities that could fall under the requirements of
the Resource Conservation and Recovery Act (RCRA) for disposal.  The analysis
of these costs shows that while the effects on profitability are higher using
these costs, the effects are not expected to significantly increase the
impact on these facilities and are not expected to cause any plant closures.

     Arsenic

     Three arsenide crystal plants are expected to incur investment costs
totaling $945,000 and annual costs (including monitoring costs) totaling
$696,000, as a result of the control of arsenic under option 2 for PSES.  One
arsenide crystal plant shows relatively high impacts under this control.  How-
ever, this plant is owned by a large electronics company and supplies crystals
to be used in the production of light emitting diodes (LEDs) at other produc-
tion facilities of the firm.  It is expected that the firm will keep this
                                     S-14

-------
plant operating to maintain control of the source of supply for its raw
material.  Consequently, no plant closure is expected.

     New Sources

     The effluent standards and associated technologies for new sources are
identical to those for existing sources.  Consequently, the economic impacts
for new sources will mirror those of existing sources and the promulgated
regulations are not expected to foster competitive advantages or disadvantages
between new and existing sources.

5.2  Semiconductors

     The estimated economic impacts for the semiconductors subcategory are
summarized in Table S-4.

     Toxic Organics

     The control of toxic organics and pH under option 1 for BPT and PSES is
expected to cause industry compliance costs for monitoring of $319,000 in
capital investment and $654,000 in annual costs.  These costs are based on two
assumptions, as with the crystals subcategory.  First, at least the 50 percent
of the facilities currently in compliance with toxic organic limitation are
expected to choose to certify that they do not dump solvents into their
effluent instead of monitoring.  Second, the remaining plants that do monitor
will, on average, do so quarterly.  Since the incremental costs of solvent
disposal tend to be offset by resale, EPA. did not estimate costs for solvent
management.  Monitoring costs are less than 0.2 percent of annual revenues
and are not expected to cause other than minor profitability impacts.  No
plant closures are expected.

     In addition to the above analysis, EPA conducted a sensitivity analysis
of option 1 costs consisting of two parts.  First, the impact of monthly
                                     S-15

-------
                TABLE S-4.  SUMMARY OF ESTIMATED ECONOMIC IMPACTS
                       FOR THE SEMICONDUCTORS SUBCATEGORY
ECONOMIC IMPACT VARIABLES
OPTION 1  OPTION 2  OPTION 3  OPTION 5
SELECTED
 OPTION
  Annual Compliance Costs/Revenues (%)
    Indirect Dischargers               0-0.2   0.1-11.
    Direct Dischargers                 0-0.2   0.1-11,
  Change in Price (%)                  00
  Change in Quantity (%)               00
  Change in Profitability
    Indirect Dischargers               Low     Signif,
    Direct Dischargers                 Low     Signif,
  Capital Requirements
    Indirect Dischargers               Low     Signif,
    Direct Dischargers                 Low     Signif,
  Plant Closures due to Regulations
    Indirect Dischargers                 0      34
    Direct Dischargers                   0      14
  Employment at Closed Plants
    Indirect Dischargers                 0   1,666
    Direct Dischargers                   0     686
  Balance of Trade Changes             None   None
  Industry Structure Changes           None   Signif.
0.01-2
0.01-2
0
0
Low
Low
Low
Low
0
0
0
0
None
None
.0 0.1-12
.0 0.1-12
0
0
Signif.
Signif.
Signif.
Signif.
34
14
1,666
686
None
Signif.
.2
.2












                                         Low
                                         Low

                                         Low
                                         Low

                                         0
                                         0

                                         0
                                         0
                                        None
                                        None
                                       S-16

-------
monitoring was estimated for all facilities (some facilities may monitor as
frequently as once per month, although EPA cannot identify precisely which
ones).  In addition to this analysis, costs and impacts were also determined
for facilities which may incur  "worst case" solvent disposal costs.  These
costs are developed for facilities that could fall under the requirements of
the Resource Conservation and Recovery Act (RCRA) for disposal.

     The analysis of these costs shows that while the effects of these costs
are somewhat higher, they still represent less than 2.3 percent of annual
revenues.  These costs are expected  to cause some profit reductions, but
would not cause any plant closures.

     Fluoride

     EPA's limits on fluoride under  option 3 for BAT will require an estimated
52 of 77 facilities to spend a  total of $4.4 million in capital investment and
$2.9 million annually (including monitoring costs) to comply with the regula-
tions.  These costs are not expected to cause other than small reductions in
profitability.  No plant closures are expected.

     New Sources

     The effluent standards and associated technologies for new sources are
identical to those for existing sources.  Consequently, the economic impacts
for new sources will mirror those of existing sources and the promulgated
regulations are not expected to foster competitive advantages or disadvantages
between new and existing sources.

6.  SMALL BUSINESS IMPACTS

     The Regulatory Flexibility Act  (Public Law 96-354) requires EPA to deter-
mine if a significant impact on a substantial number of small entities occurs
as a result of the proposed regulations.  This analysis may be done in conjunc-
tion with or as a part of any other  analysis conducted by the Agency.  The
                                     S-17

-------
economic impact analysis contains this analysis which indicates that for the
selected options, some reduction in the profitability of small plants occurs
in both subcategories.  However, these declines in financial performance
will not be enough to cause any plant closures or job losses for any segment
of the regulated population, large or small.

     For the Semiconductor industry the pollution control treatment options 2
and 5 would cause 48 small semiconductor plants, or 19 percent of the semicon-
ductor plants identified, to close, and approximately 2,352 people may become
unemployed.  Because these options were not selected and because the selected
options 1 and 3 do not have a significant impact on small plants, a formal
regulatory flexibility analysis is not required.

7.  LIMITATIONS TO THE ANALYSIS

7.1  ELECTRONIC CRYSTALS

     The major limitations relate to the data used in the analysis.  Limited
economic data was available on a sample of plants in the industry.  However,
this data was insufficient to develop a formal discounted cash flow analysis
to assess plant closure potential.  Instead, comparative analysis of financial
ratios such as return on investment was used to estimate plant closure poten-
tial.  Analyses based on this approach do not consider the timing of cash
flows and are based on accounting income rather than cash flows.  Neverthe-
less, return on investment analysis is widely used in the financial analysis
community and can be expected to provide consistent results for industry-wide
assessments.

     A number of assumptions were necessary to implement the analysis and to
overcome the data limitations.  These are:
     •  There will be no price increase resulting from the
        regulations
                                     S-18

-------
     •  Financial performance in the survey year (1978) is
        fairly typical of what might be expected in the initial
        year for which compliance is required
     •  Baseline data was based on sample sizes which varied
        in industry representation from one variable to
        another
     •  The yield on U.S. treasury bonds is a threshold value
        for plant closures.
     To assess the potential effects of violations of these assumptions or
of errors introduced because of the aforementioned methodology limitations,
a sensitivity analysis of plus and minus 20 percent is performed on a number
of key variables, such as compliance costs and baseline profit rates.  It
is concluded that there would be no plant closures under these conditions.
Despite the above limitations, the economic impact analysis conclusions appear
to be valid within a rather liberal range of potential error.

     In addition, a sensitivity analysis is performed (reported in Appendix
I-A) to examine the sensitivity of the economic impacts to variations in
compliance costs resulting from the control of toxic organics in the effluent.
The analysis showed that although varying assumptions on costs (i.e., monthly
monitoring versus quarterly monitoring, and worst case solvent disposal
costs) did increase the potential for profitability reductions, the general
conclusions of the analysis still hold.

7.2  SEMICONDUCTORS

     As noted above for the electronic crystals industry, the major limita-
tions relate to the data used in the study.  A survey to collect plant-
specific financial and economic data was not conducted, and these types of
data are not in the public domain for this industry.  Instead, most of the
plant-specific data used are estimates derived from industry average operating
and financial ratios.  The plant-specific data must, therefore, be considered
order-of-magnitude estimates, and not actual data for specific semiconductor
                                     S-19

-------
plants.  As a result of the lack of adequate data the methodology to assess
plant closure is limited to comparisons of simple operating ratios.  A more
appropriate closure analysis would be based on more formal investment analy-
sis techniques, which rely on precise data on the long-term capital structure,
cash flow, and the profitability of plants in the semiconductor industry.

     A sensitivity analysis was performed to examine the sensitivity of
the conclusions to some of the parameters estimated for the plant closure
analysis.  The primary variables considered were compliance cost estimates,
baseline profit rates (as measured by before tax return on sales) and plant
closure threshold criteria.  A 20 percent variation in any of these variables
would not significantly change the results for the selected option.  However,
the number of plant closures due to the regulation under options 2 and 5
would increase by 50 percent, to 72 plants.

     In addition, a sensitivity analysis is performed (reported in Appendix
II-A) to examine the sensitivity of the economic impacts to variations in
compliance costs resulting from the control of toxic organics in the effluent.
The analysis showed that while varying assumptions on costs (i.e., monthly
monitoring costs versus quarterly, and worst case solvent disposal costs) did
increase the potential for profitability reductions, the general conclusions
of the analysis still hold.

     For these reasons, it is concluded that although the above assumptions
and limitations may bias the analysis somewhat,  the potential changes to the
conclusions resulting from elimination of these  biases are small.
                                     S-20

-------
             PART I
ELECTRONIC CRYSTALS SUBCATEGORY

-------
                               1.   INTRODUCTION

1.1  PURPOSE
     The purpose of this report is to identify and analyze the economic
impacts that are likely to result  from the promulgation of regulations for
Best Practicable Control Technology (BPT), Best Conventional Pollutant Control
Technology (BCT), Best Available Technology Economically Achievable (BAT), New
Source Performance Standards (NSPS) , and Pretreatment Standards for New and
Existing Sources (PSNS and PSES) on the Electronic Crystals subcategory of the
Electrical and Electronic Components Point Source Category.  The study focuses
primarily on the impacts of the proposed EPA regulations on industry profit-
ability, ability to raise investment capital, employment, plant closures,
imports and exports, and substitution.  In addition, impacts on small business
entities are also analyzed.

1.2  SCOPE
     The Electronic Crystals subcategory as defined in this study includes
establishments that grow and fabricate piezoelectric, semiconducting, and
liquid crystals.

     Piezoelectric crystals are crystals that have the ability to generate a
voltage when a mechanical force is applied, or produce a mechanical force when
a voltage is applied.  The major types of piezoelectric crystals are quartz,
ceramics, and yttrium-iron-garnet (YIG).

     Semiconducting crystals are crystals whose electrical resistivity    is
intermediate between that of metals (conductors) and insulators (noncon-
ductors) .  The major types of semiconducting crystals are silicon, gallium
arsenide, gallium phosphide, indium arsenide, indium antimonide, bismuth
telluride, sapphire, and gallium gadolinium garnet (GGG).
  Resistivity is a property of an electrically conducting material which
  expresses the relative ease (low resistivity) or difficulty  (high
  resistivity) of passage of an electric current.
                                     1-1

-------
     Liquid crystals are organic compounds that exhibit properties of  fluidity
and molecular order simultaneously over a small temperature  range.  An
electric field can disrupt the orderly arrangement  of  liquid crystal molecules
and darken the liquid enough to form visible characters in a display assembly.

1.3  TECHNICAL AND ECONOMIC SUBCATEGORIZATION
     For the purpose of setting effluent standards, EPA proposed  the following
two product groups for the Electronic Crystals subcategory:

     •  Plants that fabricate gallium arsenide and/or  indium
        arsenide crystals; and
     •  Plants that fabricate all other electronic  crystals
        ("non-arsenide" plants).

     EPA identified only eight plants in operation  in  the U.S. which produce
gallium arsenide and indium arsenide crystals.  Analyses were  performed  on
each of these plants to assess the potential impacts of the  proposed
regulations.  EPA estimated that 62 plants produce  non-arsenide crystals.   The
assessment of the economic impacts of the proposed  regulations for these
plants was based on model plants which were developed  to represent the
62-plant population.  The representative model plants  were developed from EPA
survey data and were based on plant sizes and wastewater flow  rates.

1.4  ORGANIZATION OF PART I OF THIS REPORT
     The remainder of Part I of this report is organized into  seven sections.
Section 2 outlines the study methodologies and Section 3 presents  a
description of the industry.  Section 4 projects the baseline  conditions of
the industry assuming no further water pollution control requirements.
Section 5 details the costs of the alternative treatment technologies  being
considered.  Section 6 presents the economic impacts of pollution  control
costs.  Section 7 addresses the impacts of the regulations on  small
businesses.  Finally, Section 8 discusses the assumptions and  limitations of
this study.
                                     1-2

-------
                             2.  STUDY METHODOLOGY

2.1  OVERVIEW
     Figure 2-1 shows an overview of  the analytical  approach  used  to  assess
the economic impacts likely to occur  as a  result of  the  costs  of each proposed
regulatory option.  The approach used in this study  is to  (1)  develop an
operational description of the price  and output behavior of the industry,  and
(2) assess the likely plant-specific  responses  to the  incurrence of  the
compliance costs enumerated in Chapter 5.

     The operational description of the price and output behavior  is  used,  in
conjunction with EPA compliance cost  estimates  supplied, to determine new
post-compliance industry price and production levels  for each  regulatory
option and for each of the two electronic  crystal product  groups.  Individual
plant data is then analyzed under conditions of the  post-compliance  industry
price levels, for each regulatory option,  to isolate  those plants  whose
production costs would appear to change significantly  more than the  estimated
change in their revenues.  These identified plants are subjected to  a
financial analysis that uses capital  budgeting  techniques  to  determine likely
plant closures.  The industry description  is then revised, for each  regulatory
option, to incorporate the reduced supply  into  the analysis.   Finally,  other
effects which flow from the basic price, production,  and industry  structure
changes are determined.  These include employment, community,  and  foreign
trade impacts.  Specifically, the study proceeded in  the following nine steps:

      1.   Description of industry characteristics;
      2.   Industry supply and demand  analysis;
      3.   Analysis of cost of compliance estimates;
      4.   Plant level screening analysis;
      5.   Plant level profitability analysis;
      6.   Plant level capital requirements analysis;
      7.   Assessment of plant closure potential;
      8.   Assessment of other impacts; and
      9.   Small business analysis.
                                     1-3

-------
                   EPA POLLUTION
                   CONTROL  COSTS
  INDUSTRY
SEGMENTATION

  INDUSTRY
 STRUCTURE

   MARKET
 STRUCTURE

 FINANCIAL
    DATA
   INDUSTRY
MICROECONOMIC
   ANALYSIS
PRICE INCREASE
   ANALYSIS
                                                                    COMMUNITY
                                                                   EMPLOYMENT
                                                                     EFFECTS
IDENTIFICATION
   OF  HIGH-
    IMPACT
   SEGMENTS
                        MODEL
                      FINANCIAL
                      ANALYSIS
                        PLANT
                      CLOSURES
                    FIGURE  2-1.  ECONOMIC ANALYSIS  STUDY OVERVIEW

-------
Although each of these steps  is described  separately  in  this  section,  it  is
important to realize that there are  significant  interactions  between  them, as
shown in Figure 2-1.

2.2  STEP 1:  DESCRIPTION OF  INDUSTRY CHARACTERISTICS
     The first step in the analysis  is  to  describe  the basic  industry
characteristics.  These characteristics, which include the determinants of
demand, market structure, the degree of  intra-industry competition, and
financial performance, are described in  Chapter  3 of  this report.

     The sources for this information include government reports,  trade
association data, discussions with various  trade association  representatives
and individuals associated with the  industry, and an EPA industry  survey  under
Section 308 of the Clean Water Act.

2.3  STEP 2:  SUPPLY-DEMAND ANALYSIS
     The purpose of the supply-demand analysis is to determine  the likely
changes in market prices and  industry production levels resulting  from each
regulatory option.  The estimates of post-compliance price and  output levels
are used in the plant-level analysis to  determine post-compliance  revenue and
profit levels for specific plants in each  product group.  If  prices are
successfully raised without significantly  reducing  product demand  and
companies are able to maintain their current financial status,  the potential
for plant closings will be minimal.  If  prices cannot be raised to fully
recover compliance costs because of  the  potential for a significant decline in
product demand or because of  significant intra-industry competition, the  firms
may attempt to maintain their financial  status by closing higher cost/less
efficient plants.  The supply-demand analysis was divided into  four basic
components:  description of industry structure,  determination of industry
pricing mechanism, projection of possible  changes in industry structure during
the 1980s (when the primary economic impacts of  the regulations will be felt),
and determination of plant- and firm-specific operational parameters (e.g.,
production costs, profit rates, etc.).
                                     1-5

-------
     Short-run pricing behavior depends  upon  the market  structure  of  the
industry, which can range from competitive  to oligopoly,  and  to  monopoly
situations.  Many economic  impact  studies begin by  assuming  perfect
competition.  However, the  product groups covered  in  this  study  exhibit  some
characteristics that are indicative of non-competitive  pricing mechanisms.

     The perfectly competitive market structure is  one  in  which  there are many
buyers and sellers and the  actions of any one  of these  do  not significantly
affect the market.  Firms in a competitive  market  generally  earn a "normal"
rate of return on their assets.  If it is assumed  that  (1) the market for a
competitive good is currently at equilibrium,  or will be when the  regulations
become effective, and (2) firms will attempt  to maintain  their current
financial status by passing through industry-wide  cost  increases in  the  form
of higher prices, the post-compliance equilibrium  price  and  quantity  level  can
be derived from the interaction of the elasticities of  supply and  demand.
That is, the amount of the  cost increase that  will  be passed  through  into
higher prices is
                                        s
                                    E , + E
                                     d    s

where E  is the elasticity of supply and E   is the elasticity  of  demand.
       S                                  O

     The high concentration ratios  for some  of the product  groups and  the
existence of speciality markets could cause  non-competitive  pricing  behavior.
For example, if foreign competition were not a problem,  full cost pricing
might be assumed to characterize the behavior of oligopolistic firms.   The
price could be assumed to cover average total cost (the  sum of variable cost,
average fixed cost, and return on investment) of the "price  leaders" in the
industry.  Thus, one or a few firms may have the ability,  because of their
   In a competitive market, price = marginal cost,  therefore

E  =     .  	  =   ^s . mc , where me = marginal cost, P = market  price,  and
     dP     Qs     dmc   Qs   Qs = quantity supplied.
See Levenson, Albert M., and Solon, B.S., Outline of Price Theory,  Holt,
Rinehart and Winston, Inc., pp 56-59,  1964.
                                     1-6

-------
market power, to impose  their most  desired  price  and  output  strategy on the
other firms in the market.

     Oligopolistic pricing  schemes  are  applicable for those  product  categories
which exhibit characteristics of  oligopoly  markets,  such  as  the  following:

     •  Few firms in  the product  group;
     •  High industry  concentration;
     •  Low degree of  foreign competition;
     •  Abnormally high  profitability;
     •  Low demand elasticities;
     •  Highly capital intensive;
     •  Large degree  of  integration of  production, marketing,  and
        distribution;  and
     •  Large degree  of  specialized knowledge.

     Industries which  exhibit the  first  three of  these characteristics  are
those in which the pricing  and  output  actions of  one  firm will directly affect
those of other firms  in  the industry.  While  these conditions  do not guarantee
oligopolistic behavior,  they are  necessary  conditions and good indicators that
oligopolistic behavior exists.  Abnormally  high profits  in an  industry  would,
in  time, normally attract new entrants  to the  industry,  thereby  increasing
price competition.  However, very high  profits  over  long  periods of  time which
are not explained by  such factors as excess risk, unusual amounts  of techno-
logical innovation, or firm size  may be an  indicator that an imperfect  market
structure  exists.  Such conditions may occur  when entry into an  industry is
difficult.  The  last  three of  the above points  are indicators  of difficulty of
entry into  the market.

     Although the domestic electronic  crystals  industry exhibits  some of these
characteristics of non-competitive markets, the  existence of the  threat of
foreign competition changes their pricing behavior.   That is,  if  the differen-
tial between  foreign  and domestic prices is in  equilibrium,  an increase in  the
domestic price would  result in  lost market  share.  Since the domestic producer
                                      1-7

-------
profits are above "normal" profit  rates,  they  have  the  capability  of  absorbing
the cost increases without increasing  product  prices  and  would  probably do so.

     Having described the current  industry  structure,  it  was  necessary to
determine if the key parameters would  change significantly during  the 1980s.
Projections of industry conditions begin with  a demand  forecast.   The demand
in 1990 is estimated via trend analysis and market  research analysis.  An
examination was also made of the factors which might  affect the real  cost of
manufacturing electronic crystals.  No reason  was found  to expect  the real
price of these products to increase between now and 1990.   It was  concluded
from the projections of industry conditions that only minor changes  in market
structure would occur in the base  case.  For this reason  it was concluded that
the market structure previously described can  be used  to  determine price
changes due to the regulation.

     The post-compliance market price  levels (i.e., zero  price  increases due
to the regulation) are used, in a  later step,  to assess the financial
condition of individual electronic crystals manufacturing  facilities.

2.4  STEP 3:  COST OF COMPLIANCE ESTIMATES
     Investment and annual compliance  costs for the recommended treatment
options 1, 2, 3, and 5 were estimated  by EPA's Effluent Guidelines Division
for treatment systems of various selected sizes.  Based on these cost
estimates, compliance cost curves were developed and  then  used  to  estimate
plant-specific compliance costs.  A description of  the control  and treatment
technologies and the rationale behind  these compliance cost estimates appear
in chapter 5.

2.5  STEP 4:  PLANT-LEVEL SCREENING ANALYSIS
     The screening analysis uses a basic criterion  to separate  those  plants
with obviously small impacts from  those with potentially  significant  impacts.
The criterion uses the ratio of total  annual compliance cost  to revenue  for
each plant; if this ratio is less  than one percent, the plant is considered
a "low impact" plant.  For such plants, the impacts on plant  profitabilities
                                     1-8

-------
are very small.  Although  some of  these  plants will  probably  experience  some
drop in their profitabilities, they would  not be  considered  candidates  for
closure.  Estimates of profitability  changes  for  these  plants  were  recorded,
but they are not subject to  a detailed  financial  analysis.   Plants  with
compliance costs to revenue  ratios greater than  1  percent were subjected to  a
more detailed financial analysis to determine if  they  are  likely plant
closures .

2.6  STEP 5:  PLANT-LEVEL  PROFITABILITY  ANALYSIS
     The basic measure of  financial performance  used to assess the  impact  of
the proposed regulations on  the profitabilities  of individual  plants  is  return
on investment (ROI).  The  use of this technique  involved a comparison of the
return on investment  after compliance with a minimum required  ROI.

     The return on  investment is defined as the  ratio  of annual  profits  before
taxes to the gross  book value of fixed  assets of  a plant.  It  is based on
accounting income rather than cash flows and  it  fails  to account for  the
timing of cash flows, thereby ignoring  the time  value  of money.   However,  this
technique has the virtues  of simplicity  and common usage in  comparative
analyses of the profitability of financial entities.

     The profit impact assessment  is determined by calculating the  after-
compliance ROI for  each plant and  comparing them  to  the average  yield of U.S.
Treasury bonds.  As discussed in Section 2.3, it  is  expected  that the
electronic crystals producers will absorb  the costs of  compliance and prices
will not be raised.  As the  result, the  after-compliance ROI  can be estimated
as follows:
          ROT  .   =    PROFITU -   ACC.                                    (2)
                           A. +  CCI  .
                             i         i
                          x
                                                                            (3)
where  ROI_.       =  After-compliance ROI of  plant  i
       PROFIT. .    =  Pre-compliance  profit  of plant  i
       ACC.        =  Annual compliance  cost  for  plant  i
       A.          =  Pre-compliance  capital  investment  for  plant  i
                                      1-9

-------
       CCI.         =  Compliance capital  investment  for  plant  i
       R. .         =  Pre-corapliance  revenue  of  plant  i
       PM .         =  Pre-compliance  profit margin of  plant  i

     Plants  with after-compliance ROI below 11.5  percent,  the  average  yield  of
U.S. Treasury bonds for 1980, are considered  potential plant closures.   The
underlying assumption is that plants  cannot continue to  operate  as  viable
concerns  if they are unable to generate a return  on  investment that  is  at
least equal  to the opportunity cost of other  lower risk  investment  alter-
natives, which in this case is defined as the average  yield  of the  U.S.
Treasury  bonds.

     Most of the data used  in this analysis were  estimated  from  a combination
of  publicly available information and the EPA 308 Survey.  The following are
the most  important variables:

     •  Baseline financial  profiles (i.e., revenues, assets  value,  profit
        margin) of individual plants/model plants are  obtained from the  EPA
        308 Survey; and
     •  Pollution control costs are estimated as  described in  Step  3.

2.7  STEP 6:  CAPITAL REQUIREMENTS ANALYSIS
     This section analyzes  the ability of firms  to make  the  initial  capital
investment needed to construct and install the required  treatment systems.
Some plants which are not initially identified as potential  closures  in  the
profitability analysis may  encounter  problems raising  the amount  of  capital
required  to install the necessary treatment equipment.   The  limit on  a  given
firm's ability to raise capital to finance investment  expenditures  is  quite
variable, depending upon factors such as  the  firm's  capital  structure,  profit-
ability,  future business prospects, the industry's business  climate,  the
characteristics of the financial markets  and  the  aggregate economy,  and  the
firm management's relationships with  the  financial community.  The  precise
limit, considering all these factors, is  ultimately  judgmental.   Even  with
firm-specific data, a limit on a firm's ability  (or willingness)  to  raise
                                      1-10

-------
funds for capital  investment would  be  difficult  to  estimate.   Because firm-
specific data  for  this  study is  scarce,  the  analysis  of capital  availability
was based on the following  ratio which provides  a good  indication of the
relative magnitude of the capital requirements  for  pollution  control:
                          CGI
                            R
where
     R = plant  revenues
   CCI • capital  compliance  investment

     Plants with  after-compliance  ratios  greater  than the  difference  between
the highest and lowest ratios  of annual capital expenditures  to  value of
shipments  (CE/VS) between  1970 and  1977 for  SIC 3679  (Electronic Components,
n.e.c.) as reported by the Census  of Manufactures  are considered to be
potential  plant closures.  The difference between  the highest  and lowest CE/VS
ratios is  considered  to  be an  indicator of the maximum amount  of capital
available  for pollution  control investment.  That  is, the  highest CE/VS  ratio
of 5.1 percent  represents  the  maximum capital expenditure  rate that has
historically been made available for capital investment.   It  is  assumed  that
CE/VS ratios below this  amount can  be successfully  attained again by  the firms
in the industry,  since it has  been  achieved  before.   The lowest  CE/VS ratio  of
2.8 percent represents the minimum  amount  necessary to maintain  a targeted
rate of industry growth.  If the ratio CCI/R for a  given plant is greater than
the difference between the industry's highest and  lowest CE/VS ratios,
2.3 percent, then the plant could experience some difficulties in making the
capital investment for pollution control.  Although this analytical technique
does not precisely indicate whether or not a given  firm will  be  able  to  raise
the necessary funds,  it  does provide an indication  of the  relative magnitude
of the compliance investment.

     Plants with CCI/R ratios  greater than the difference  between the highest
and lowest CE/VS ratios  were then subjected  to the  following  cash flow test.
Cash flow is defined  as  net after  tax profits plus  depreciation  charges.
                                     1-11

-------
Depreciation is assumed to be 10 percent of  fixed  assets  and  the  tax  rate is
assumed to be 40 percent.  If the CCI/R ratio  is  less  than  the  cash  flow, th<
investment may be financed out of a  single year's  cash  flow without  addition.
long-term debt burden.

     Although the above analyses provide a good  indication  of the relative
burden created by the compliance requirement,  they  do  not precisely  indicate
whether or not firms can afford to make the  investments.  If,  for example, tl
same investment requirements were placed on  a  firm  which  is already  highly
leveraged (as indicated by a high debt/equity  ratio) and  a  firm which is  not
leveraged (as indicated by a debt/equity ratio of  zero),  the  highly  leveragec
firm is likely to experience the most significant  impact.   In addition,  the
capital requirements must be evaluated together with other  factors,  such  as
profitability.  For example, a plant that is extremely  profitable would
consider the risk of more leverage or increased cost of capital resulting fr<
investment more worthwhile than would a less profitable plant.

2.8  STEP 7:  PLANT CLOSURE ANALYSIS
     The plant-level analysis examined the individual  production  units  in ea<
product group to determine the potential for plant  closures and profitability
changes.  The decision to close a plant, like most  major  investment  decisions
is ultimately judgmental.  This is because the decision involves  a wide
variety of considerations, many of which cannot be  quantified or  even identi-
fied.  Some of the most important factors are:

     •  Profitability before and after compliance;
     •  Ability to raise capital;
     •  Market and technological integration;
     •  Market growth rate;
     •  Other pending Federal, state, and local regulations;
     •  Ease of entry into market;
     •  Market share;
     •  Foreign competition;
     •  Substitutability of the product; and
                                     1-12

-------
     •  Existence of specialty markets.

     Many of these factors are highly  uncertain,  even  for  the  owners  of the
plants.  However, this analysis was  structured  to make  quantitative estimates
of the first two factors, as described  above, and to qualitatively consider
the importance of the others.  In  this  analysis,  the first  two factors  are
given the greatest amount of weight, while  the  importance  of the  other  factors
varies from plant to plant.

2.9  STEP 8:  OTHER IMPACTS
     This analysis addresses economic  impacts which  flow from  the basic price,
production, and plant level profitability changes.  These  impacts include
impacts on employment, communities,  industry structure,  and balance of  trade.

     The estimate of employment effects  flows directly  from the outputs of  the
industry level analysis and the plant  closure analysis.  Employment estimates
for the production facilities projected  to  close  are available from the EPA
308 Survey.  Meanwhile, employment  in  the remaining plants  is  not expected  to
decrease because there will be no  price  increase  as the  result of the
            2/
regulations.

     Community impacts result primarily from employment  impacts.  The critical
variable is the ratio of Electronic  Crystals industry  unemployment to total
employment in the community.  Data  on  community employment  are available
through the Bureau of the Census and the Bureau of Labor Statistics.  Some-
times county-wide data will have to  be  relied upon in  the  absence of
community-specific data, although  this  was  not  necessary in this  study.
II
  Some increase in employment might  occur  at  the  remaining  facilities  as
  substitutes for the eliminated  capacity  from  the  closed plants.
  Quantitative evaluation of  this  effect was  not  developed  in  this  study.
                                      1-13

-------
     The assessment of industry structure changes  is based on  examination  of
the following before and after compliance with  the  regulation:

     •  Numbers of firms and plants;
     •  Industry concentration ratios; and
     •  Effects of plant closures on specialty markets.

     Imports and exports are important determinants of  pricing behavior  in the
Electronic Crystals industry.  The role of these variables is  qualitatively
evaluated in Chapter 3 of this report.  Basically,  the  threat  of  imports
appears to deter domestic producers from excessive  non-competitive  pricing
behavior.

2.10  STEP 9:  SMALL BUSINESS ANALYSIS
     The Regulatory Flexibility Act (RFA) of 1980,  (P.L. 96-354)  requires
Federal regulatory agencies to consider "small entities" throughout  the
regulatory process.  The RFA requires an initial screening analysis  to be
performed to determine if a substantial number of  small entities  will be
significantly impacted.  If so, regulatory alternatives that eliminate or
mitigate the impacts must be considered.  This step in  the study  addresses
these objectives by identifying the economic impacts which are likely to
result from the promulgation of BPT, BCT, BATEA, NSPS,  PSES, and  PSNS
regulations on small businesses in the electronic  crystals manufacturing
industry.  The primary economic variables covered  are those analyzed  in  the
general economic impact analysis such as plant  financial performance, plant
closures, and unemployment and community impacts.   Most of the information and
analytical techniques in the small business analysis are drawn from  the
general economic impact analysis which is described above and  in  the  remainder
of this report.  The specific conditions of small  firms are evaluated against
the background of general condition in the electronic crystal  markets.

     A specific problem in the methodology was developing an acceptable
definition of small entities.  The Small Business Administration  (SBA) defines
small entities in SIC 3674 (Semiconductors and Related  Devices  and SIC 3679
                                     1-14

-------
(Electronic Components, n.e.c.)  as  firms of  fewer  than  500  employees.   The  SBA
definition was found to be  inappropriate as  a  basis  for defining  small
entities in the electronic  crystal  manufacturing  industry  for  purposes  of
developing water pollution  regulations.  Instead,  a  definition was  sought
which would account for firm  size  in  comparison to total industry size  and  in
comparison to unit compliance  costs (unit  compliance costs  increase signif-
icantly in reverse proportion  to plant  size).  Moreover, since the  available
data on compliance cost and production  was on  an  individual  plant basis,  the
individual production  facility rather than the firm  was used as the basis for
the analysis.

     In the absence of an appropriate specific definition  for  a small business
in the electronic crystals  manufacturing industry, three size  definitions for
electronic crystal manufacturing plants based  on  plant  revenues were selected
for examination.  Plant revenues were selected to  distinguish  plant size
because plant level employment data were considered  less reliable and plant
production volume data did  not allow  consistent comparisons  across  the  product
groups.  The three size categories  are  plants  with revenues  less  than $1
million, $3 million and $5  million,  respectively.  The  use  of  several
different size definitions  provides EPA with alternatives  in defining small
electronic crystal operations.

     The impacts on small plants under  each  definition  were  assessed by
examining the distribution  by  plant size of  the number  of  electronic crystal
plants, plant revenues, wastewater  volumes,  compliance  costs and  potential
closures from the regulations.
                                      1-15

-------
                           3.  INDUSTRY DESCRIPTION

     This section describes the manufacturing  processes  and  the  plant  and  firm
characteristics of the electronic crystals  industry.   In  addition,  the
characteristics and end uses of crystals and the extent of foreign  competition
are also examined.

3.1  MANUFACTURING PROCESSES
     As indicated earlier, the scope of this study includes  the  plants  that
grow and fabricate piezoelectric, semiconducting and  liquid  crystals.   The
manufacturing processes for these electronic crystals  are generally very
similar.  First, single crystals are grown  from polycrystalline  raw materials.
 The most common technique for growing single  crystals is the Czochralski
method in which the raw material is melted; a  single  crystal bar  (seed) is
then dipped into the melt and raised slowly with rotation to allow  the  molten
material to freeze onto the seed to form a  single crystal rod.   Other methods
based on some variations of the Czochralski method are also  used  for growing
electronic crystals and are described in the Final Development Document.

     After the single crystal rod has been  grown, it  must be fabricated into a
suitable form, which for most applications  is  a wafer.  Generally,  the  crystal
rod is ground to a preferred shape.  It is  then sliced into  thin  pieces using
diamond blade saws or slurry saws.  Once sliced, the  crystal wafers  are then
lapped to obtain a desired thickness and a  smoother finish.  Finally,  the
slices are polished to a mirror finish, and are then  ready for further
processing to produce various electronic components.

3.2  FIRM AND PLANT CHARACTERISTICS
     According to EPA estimates, there are  approximately  70  producers  of  elec-
                                    2/
tronic crystals in the United States  , with total employment of the industry
  EPA, Final Development Document for Effluent Limitation Guidelines
  for the Electrical and Electronic Components Point Source
  Category. March, 1983.
2/Ibid.
                                     1-16

-------
estimated at around 10,000 employees.  Table 3-1 lists the number of plants by
type of crystals.  As indicated in the table, with the exception of quartz
crystal the production of electronic crystals is highly concentrated with a
small number of firms accounting for most of the production of each type of
crystal.

     The geographical distribution of the plants is shown in Table 3-2.  Four
states account for 57 percent of the number of plants identified (California,
Ohio, Pennsylvania, and Texas, with 26, 12, 12, and 7 percent of the number of
plants, respectively), with the remainder widely dispersed throughout the
country.

     Based on the survey data obtained on 52 plants, about 60 percent of the
plants have less than $3 million in product shipments or employ less than 100
employees.  Table 3-3 presents a distribution of the electronic crystals
plants by plant size.  Most of the plants appear to be highly specialized in
the fabrication of electronic crystals.  Only 12 plants report integration to
the manufacture of electronic devices at the same production facilities.  The
companies that own the crystal plants, on the other hand, are generally more
diversified and often produce other electronic devices in addition to
crystals, usually at other production facilities.

3.3  PRODUCT CHARACTERISTICS AND USES
     As indicated in Section 1, the Electronic Crystals subcategory as defined
in this study includes three types of crystals:

     •  Piezoelectric crystals which include quartz, ceramic,
        yttrium-iron-garnet, and lithium niobate crystals;
     •  Semiconducting crystals, which include silicon, gallium
        arsenide, gallium phosphide, indium arsenide, indium anti-
        monide, bismuth telluride, and sapphire; and
     •  Liquid crystals.
                                     1-17

-------
     TABLE 3-1.   NUMBER OF ELECTRONIC CRYSTAL PLANTS IN THE UNITED STATES
                                ESTIMATED
                                TOTAL NUMBER
TYPES OF CRYSTALS               OF PLANTS

Piezoelectric Crystals

     Quartz                          40
     Ceramics                         8
     Yttrium Iron                     3
       Garnet (YIG)
     YAG                              2
     Lithium Niobate                  3

Semiconducting Crystals

     Silicon                          8
     Gallium Arsenide/Phosphide       6
     Indium Arsenide/                 1
       Indium Antimonide/
       Bismuth Telluride
     Gallium Gadolinium               3
       Garnet (GGG)
     Sapphire                         1

Liquid Crystal                        2

         TOTAL                       70 a/b/
                                                  NUMBER OF PLANTS
                                                  WITH RETURNED
                                                  QUESTIONNAIRES
                                                         27
                                                          8
                                                          3

                                                          2
                                                          3
                                                          8
                                                          6
                                                          1
                                                          3

                                                          1

                                                         _2

                                                         57
a'Columns do not add up because some plants make more than one type of
  product.
b/
  Projected by JRB assuming that non-respondent plants are single product
  plants.
Source - EPA, Final Development Document, March 1983
                                     1-18

-------
      TABLE  3-2.   GEOGRAPHICAL  DISTRIBUTION  OF  ELECTRONIC  CRYSTAL PLANTS






                   STATE                         NUMBER OF PLANTS






                   Arizona                              1




                   California                          15




                   Colorado                             1




                   Connecticut                           2




                   Florida                              1




                   Illinois                             1




                   Kansas                               1




                   Massachusetts                        1




                   Missouri                             1




                   Montana                              1




                   New Jersey                           3




                   New York                             2




                   North Carolina                       1




                   Ohio                                 7




                   Oklahoma                             2




                   Oregon                               1




                   Pennsylvania                         7




                   South Dakota                         1




                   Texas                                4




                   Utah                                 2




                   Not Reported                         2




                        TOTAL                           57







Source:  EPA 308 Survey
                                     1-19

-------
               TABLE  3-3.  DISTRIBUTION  OF  PLANTS  BY  PLANT  SIZE
a/
Plant Size '
Number of
Sample, /
Plantsb/
Estimated Number of
Percent of Plants for
Total Total Industry
Gallium Arsenide/Indium
Arsenide Crystals
Extra Small
Small
Medium
Large


2
3
-
3
8

25
38
-
38
100

2
3
-
3
8
Other Electronic Crystals

     Extra Small
     Small
     Medium
     Large
         10
         16
         10
                             44
 23
 36
 23
 18

100
14
23
14
                                          62
  Extra Small Plant = Plants with less than $1 million in value of shipments
                      or with less than 20 employees.
  Small Plant
  Medium Plant
  Large Plant
b/
= Plant with $1 million to $3 million in value of
  shipments or with 20 to 100 employees.

= Plant with $3 million to $5 million in value of
  shipments or with 100 to 250 employees.

= Plant with over $5 million in value of shipments or with
  more than 250 employees.
  Data on plant size are available only for 52 of 57 plants with survey
  responses.

Source:   EPA Survey
                                     1-20

-------
Quartz and silicon crystals  are  by  far  the most widely  used  crystals.   The
principal end-uses of various electronic  crystals  are summarized  in  Table 3-4.

     The potential for substitution of  electronic  crystals in  their  current
applications is limited.  For instance, silicon's  exceptional  physical  and
chemical properties make this crystal the best and most  important material  for
semiconductor applications.  Similarly, the  special  characteristics  of  the
other crystals make them hard to be replaced  in most of  their  applications.

     Potential substitution  among various crystals  is also limited.  Except
for gallium arsenide, gallium phosphide,  and  liquid  crystals,  which  are all
used in the manufacturing of display devices,  each electronic  crystal has
quite distinctive applications for which  it  cannot be easily replaced (see
Table 3-4).  Gallium arsenide and gallium phosphide  can  substitute for  each
other in the fabrication of  light emitting diodes  (LEDs)—gallium arsenide  for
red displays and gallium phosphide  for  green  displays—however, each manufac-
turer can easily produce either  or both materials.   Therefore,  this  substi-
tution does not present any  problem to  the industry.  Meanwhile, LEDs and
liquid crystal displays (LCDs) compete  against each  other; however,  the
potential for substitution is limited due to  the special  characteristics of
each product.  Because it requires minimum use of  power,  LCD has replaced LED
in many battery-operated devices such as  electronic  watches  and calculators;
however, LCD is not as visible as LED,  and thus its  substitution  for LED is
limited.

3.4  TRENDS
     Data are not available  on U.S. total production of  fabricated (i.e.,
sliced, polished, lapped) electronic crystals.  However,  the U.S. Department
of Commerce reported that shipments of  quartz  and  other  crystal filters and
frequency control devices (which account  for  most  of the  demand for  piezo-
electric crystals) totaled $148.6 million in  1978.   It was estimated that  the
value of the crystal accounted for about  75  percent  of  the value of  the
finished devices,   thus, U.S. shipments  of  piezoelectric crystals in 1978
were valued around $110 million.  According  to the U.S.  Department of Commerce,
37
  Interview with Jack Clifford  of U.S.  Department  of-Commerce.
                                      1-21

-------
            TABLE 3-4.  MAJOR APPLICATIONS FOR ELECTRONIC CRYSTALS

TYPE OF CRYSTAL                     MAJOR APPLICATIONS
Piezoelectric Crystals

     •  Quartz



     •  Ceramic
        Yttrium-Iron-Garnet
          (YIG)
Semiconducting Crystals
        Silicon
        Gallium Arsenide,
        Gallium Phosphide
     •  Indium Arsenide, Indium
        Antimonide

     •  Bismuth Telluride
        Sapphire
        Gallium-Gadolinium-
        Garnet (GGG)
Liquid Crystals
Timing devices in watches; frequency
control, modulation, and demodulation in
oscillators and filters

Used in transducers, oscillators, ultrasonic
cleaners, gas igniters, audible alarms,
keyboard switches, medical electronics

High-frequency control in microwave circuits
for electronic devices such as sonar
equipment.
Wafers for semiconductor devices such as
integrated circuits, diodes, rectifiers,
transistors and other circuit elements

Light emitting diodes (LEDs) used as display
devices in calculators, digital watches and
other electronic equipment; high-performance
transistors such as field effect transistors
(FET)

Components of power measuring devices
Devices to cool small components of
electrical circuits

Substrate for silicon-on-sapphire (SOS)
semiconductor devices; infrared detector
cell windows; ultraviolet windows and
optics; high power laser optics;
ultracentrifuge cell windows

Substrate for magnetic garnet films used in
the manufacture of magnetic bubble memories

Liquid crystal displays (LCDs)  for
calculators, digital watches, etc.
Source:   EPA, Final Development Document
                                     1-22

-------
U.S. shipments of piezoelectric  crystals  had  been  growing  slowly.   The  reason
for this modest growth was  the  inability  of U.S. companies  to  compete with  the
Japanese in several  fast-growing markets  such as those  for  electronic watch
and CB radio crystals (for  which  the  average  annual  growth  rates  between  1973
and 1978 were 26 and 24  percent  respectively  ').   As  a  result,  U.S.  firms are
generally more specialized  in the market  for  high  performance  crystals  where
foreign competition  is not  as strong.  Although  the  number  of  devices that  use
these high-quality crystals has  been  increasing  significantly,  the  demand for
the crystals has been growing slowly  because  technological  advances  have
reduced the number of crystals  required  in each  device.  For  instance,  the
introduction of phase lock  loop  system has reduced the  number  of  crystals used
in communications equipment.     U.S.  Department  of Commerce data  show that
U.S. shipments of crystal filters and  frequency  control  devices fluctuated
widely between 1973  and  1978, and averaged about 5 percent  annual growth  rate
during that period (see  Table 3-5).   However,  the  validity  of  the data
reported is very questionable,  especially for earlier years.

     It was estimated that  the  cost of a  semiconductor wafer represented  about
5 percent of the value of a semiconductor discrete device,  and  about
7 to 8 percent of the value of  an integrated  circuit.    Based  on the value of
shipments of semiconductor  devices  reported in Table  3-6,  the  U.S.  shipments
of semiconducting and liquid crystals  in  1979 were valued  around  $450 million.
Between 1969 and 1979, the  demand for  semiconducting  and liquid crystals had
been growing rapidly as  indicated by  the  17 percent  average annual  growth rate
of U.S. shipments of semiconductor  devices (see Table 3-6).
  U.S. Department of Commerce, U.S. Industrial Outlook 1980, p. 400.
  Interview with Jack Clifford of U.S. Department of Commerce.
6/Ibid.
  U.S. Department of Commerce, A Report on  the U.S. Semiconductor  Industry,
  September, 1979.
                                     1-23

-------
              TABLE 3-5.   U.S.  SHIPMENTS OF PIEZOELECTRIC CRYSTAL
                       DEVICES, 1973 - 1978 ($ MILLIONS)
         Year
   Frequency
Control Devices
Filters
          Estimated
          Value of
Total      Crystal
1973
1974
1975
1976
1977
1978
98.2
90.6
94.3
143.3
111.1
107.1
22.1
29.9
114.3
76.9
40.3
41.5
120.3
120.5
208.6
220.2
151.4
148.6 110. Oa/
AVERAGE ANNUAL
COMPOUNDED
GROWTH RATE
           b/
       4.3%
  10.1%
   5.3%
a/  About 75 percent of value of finished devices (interview with Jack
    Clifford of U.S. Department of Commerce).
b/
    Calculated using semi-log regression technique.
Source:   U.S.  Department of Commerce, Current Industrial Reports - Selected
         Electronic and Associated Products,  Including Telephone and Telegraph
         Apparatus, selected issues.
                                     1-24

-------
              TABLE 3-6.  U.S. SHIPMENTS OF SEMICONDUCTOR DEVICES
                           1969 - 1979 ($ MILLIONS)
Year
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
Discrete
Devices
1013
987
870
1093
1469
1519
1320
1694
1835
1880
1957
Integrated
Circuits
480
594
696
1268
1709
2056
1718
2598
2697
3950
4671
Total
1493
1581
1566
2361
3179
3575
3038
4292
4532
5830
6628
AVERAGE ANNUAL
COMPOUNDED
GROWTH RATE
           a/
                                             25%
                                                             17%
a/
t
Calculated  using  semi-log  regression  technique.
Sources:  U.S. Department of Commerce
          Semiconductor Industry Association
                                     1-25

-------
3.5  FOREIGN TRADE
     Data on U.S. imports  and  exports  of  electronic  crystals  are  not  available
directly.  However,  the  available  evidence  indicates  that  the U.S.  manufac-
turers of semiconducting crystals  seem to be  able  to  successfully compete with
                                                                        8/
their foreign producers.   According  to a Department of  Commerce Report,    many
of the U.S. semiconductor  device manufacturers  have  transferred  their assembly
operations overseas  to take advantage  of  lower  wage rates.  However,  these
same companies still maintain  the  production  of semiconductor wafers  in  the
country.  The primary reason for this  is the  capital  intensive nature of the
growing and fabricating process for  wafers.   However, this  does not mean that
there is no potential for  competition  from  abroad.  In  fact,  a number of firms
in Japan and Europe have the necessary technology  and capacity and,
consequently, may serve to limit the ability  of  the domestic  producers of
                                                                             9/
semiconductor wafers to pass through the costs  of  environmental regulations.

     For the manufacture of piezoelectric crystals, foreign competition  is  not
expected to be too significant.  As  indicated earlier,  the  piezoelectric
crystals currently produced by U.S. manufacturers  are generally high-perfor-
mance crystals and competition from  foreign producers in this market  does  not
seem to be very important.  In the past, U.S. firms had tried to  penetrate  the
booming market for quartz  crystals for  electronic  watches and CB  radios.
However, aggressive competition from the Japanese  producers,  who  now  dominate
the world market for these lower quality crystals, has  driven most of these
U.S. companies out of the market and forced them to specialize in the  low-
volume, high-quality crystal market  instead.
8/
  U.S. Department of Commerce, A Report on the U.S. Semiconductor Industry,
  September 1979.                                         "         ~~
97
  Interview with Jack Clifford of U.S. Department of Commerce.
10/Ibid.
                                     1-26

-------
                 4.   BASELINE PROJECTIONS OF INDUSTRY CONDITIONS

     This  section provides projections of conditions in the U.S. electronic
 crystals  industry to 1990 in the absence of further water pollution control
 requirements  resulting from the Clean Water Act.   The baseline projections in
 this report provide  a general point of reference  for the analysis and are not
 intended  to be  a comprehensive, authoritative forecast of future industry
 conditions.   These projections provide a plausible picture of future develop-
 ments,  and  thus can  be used as a benchmark for comparison.  The primary vari-
 ables of  interest are the industry's projected sales, capital requirements,
 and employment.

 4.1  U.S. ELECTRONIC CRYSTAL SALES PROJECTIONS
     As indicated in Table 3-5, shipments of piezoelectric crystals grew at a
 modest  5 percent average  annual rate between 1973  and 1978.   The U.S.  Bureau
 of Mines projected that U.S.  demand for  lasca,  a  feedstock for growing
 cultured  (i.e.,  artificial) quartz, will increase  at an annual rate of
 approximately 3 percent through 1990.  Since quartz  crystal  is the  most  common
 piezoelectric crystal and cultured quartz represents 95 percent  of  all
 electronic grade quartz,  it is  projected that  U.S.  annual  production of  piezo-
 electric crystals will grow about  3 percent  a  year  and  will  total $130 million
 in 1983 and $160 million  in 1990.     These  projections  are  shown in Table  4-1.

     U.S.  production  of semiconducting and  liquid crystals  is  expected to  grow
more rapidly in  the coming  years.   As  summarized in  Table  4-1, shipments of
semiconductors  by U.S. manufacturers  were projected  to  be  $11.3  billion  for
1983 and $27.0  billion for  1990, up from $6.6  billion in  1979.   Assuming that
the crystals used in  the  manufacturing of semiconductor components  would
continue to be  supplied mainly  by  domestic producers, U.S.  production  of
semiconducting and liquid crystals  is projected to grow from  an  estimated
$450 million in 1979   to $770 million  in  1983 and $1.84  billion in 1990.
  U.S.  Bureau of Mines, Mineral Commodity Summaries 1981, pp. 120-121.
                                     1-27

-------
   TABLE 4-1.  PROJECTIONS OF U.S. ELECTRONIC CRYSTAL SHIPMENTS ($ MILLIONS)



                      Piezoelectric        Semiconductor
     Year               Crystals  	      and_ Liquid Crystals     Semiconductors

1978 (Estimated)

1979 (Estimated)

1983 (Projected)

1990 (Projected)

     (NA) = Not available

     a/     JRB estimates (see Section 3.4).

     b/     U.S. Department of Commerce forecasts  (1983 U.S. Industrial
            Outlook).

     c/     Assume 3  percent annual growth  rate.

     d/     Assume same growth rate as semiconductors.

     Source:   JRB Associates estimates
iioa/
(NA)
130C/
160C/
390a/
450a/
770d/
l,840d/
5,830
6,628
ll,300b/
27,000b/
                                    1-28

-------
 4.2  CAPITAL EXPENDITURES PROJECTIONS
      Plants that produce semiconducting and piezoelectric crystals are
 classified under SIC 3674 (Semiconductors and Related Services) and SIC 3679
 (Electronic Components, Not Elsewhere Classified); respectively.  Between 1972
 and 1977,  annual capital expenditures of the plants under SIC 3674 ranged
 between 6  percent and 11 percent of the industry's value of shipments.  A
 large share of the capital expenditures was for plant and equipment used in
 the manufacture of semiconductor devices which have experienced rapid
 obsolescence due to rapid changes in production technology.  Since the manu-
 facturing  processes of crystals are not expected to change as rapidly, the
 annual capital expenditure rates are estimated to be lower and similar to
 those of the more stable segment of the industry, SIC 3679, which ranged from
 3 percent  to 5 percent between 1972 and 1977.  For this reason, the annual
 capital expenditures rates were projected to average about 4 percent of the
 industry's value of shipments in the 1980s.

      Table 4-2 below presents the projections of the capital expenditures
 required to maintain the projected levels of sales for the 1980s.   It shows
 that new capital expenditure requirements in 1990 for the electronic crystal
 industry would be around $80 million.

           TABLE 4-2.  PROJECTIONS OF U.S. ELECTRONIC CRYSTAL INDUSTRY
                      NEW CAPITAL EXPENDITURES ($MILLIONS)

                         Industry Value of        Industry New Capital'
Year                    	Shipments           	Expenditures
 1983                            900                       40
 1990                          2,000                       80

Source:  JRB  Associates estimates.

4.3   EMPLOYMENT PROJECTIONS
      Based  on EPA survey data,  total employment  in the industry was estimated
at  10,000  in  1978.   Since the manufacturing processes  of electronic crystals
are  already very capital-intensive,  it appears  likely that the value of ship-
ment  per employee  ratio would remain fairly stable in  the 1980s.
                                      1-29

-------
     Table 4-3 presents the projections of employment for the electronic
crystal industry.  The employment projections are based upon the ratio of
value of shipments per employee and show that total employment in 1990 would
be about 40,000.
                 TABLE 4-3.  PROJECTIONS OF EMPLOYMENT IN THE
                       U.S. ELECTRONIC CRYSTAL INDUSTRY
                        Industry Value of
Year                    	Shipments           Industry Employment
                           ($Millions)
1978                           500                      10,000
1983                           900                      18,000
1990                         2,000                      40,000

Source:  JRB Associates estimates.

4.4  SUMMARY
     Projections of conditions in the U.S. electronic crystal industry have
been estimated in this section as a means of looking at future expectations.
Our estimates show that by 1990, U.S. electronic crystal industry value of
shipments will be $2 billion (an increase of $1.5 billion from 1978), industry
new capital expenditures will be $80 million, and industry employment will be
40,000 (an increase of 30,000 from 1978).
                                     1-30

-------
                            5.  COST OF COMPLIANCE

5.1  OVERVIEW
     The recommended water treatment control systems, costs, and effluent
limitations for the Electronic Crystals subcategory are enumerated in the
Final Development Document for Effluent Limitation Guidelines for the
Electrical and Electronic Components Point Source Category.  This document
identifies various characteristics of the industry, including manufacturing
processes, products manufactured, raw waste characteristics, supply, volume,
and discharge destination of water used in the production processes, sources
of waste and wastewaters, and the constituents of wastewaters.  Using that
data, pollutant parameters requiring effluent limitations or standards of
performance were selected by EPA.

     The EPA Final Development Document also identifies and assesses the range
of control and treatment technologies which apply to the Electronic Crystals
subcategory.  This assessment involved an evaluation of both in-plant and
end-of-pipe technologies that could be designed for this subcategory.  This
information was then evaluated for existing direct dischargers to determine
the effluent limitations achievable based on the best practical control
technology currently available (BPT), and the best available technology
economically achievable (BAT).  Similar evaluations were performed for new
direct dischargers to develop new source performance standards (NSPS).
Finally, pretreatment standards for existing sources (PSES) and pretreatment
standards for new sources (PSNS) were developed for dischargers to publicly
owned treatment works (POTWs).  The technologies identified were analyzed to
calculate cost and performance of each.  Cost data were expressed in terms of
investment, operating and maintenance costs, depreciation, and interest
expense.

5.2  POLLUTANT PARAMETERS
     The selection of pollution parameters to be considered for regulation was
based primarily on laboratory analyses of wastewaters sampled from several
electronic crystals plants and responses to a mail survey of electronic
        ft.
crystals manufacturers.  The specific approach to selecting the pollutant
                                     1-31

-------
parameters was presented in Sections 5 and 6 of  the Final Development
Document.  The specific pollutants selected  for  regulations  are  pH,  total
suspended solids (TSS), fluoride, total toxic organics  (TTO),  and  arsenic.
Arsenic is to be regulated only at facilities that produce gallium arsenide or
indium arsenide crystals.  The pollutants comprising TTO are  listed  in
in the Final Development Document.

5.3  RECOMMENDED TREATMENT TECHNOLOGIES
     Based on the analysis of the potential  pollutant parameters and  treatment-
in-place in the electronic crystals industry, EPA identified  six treatment
technologies that are most applicable for the reduction of the selected  pol-
lutants.  These treatment technologies are described in detail in  Section 7 of
the Final Development Document and are listed below:

     •  Option 1:  Segregation and collection of spent  solvents  containing
                   toxic organics for reuse, resale or  contract  hauling
                   (referred to as solvent management)  plus end-of-pipe
                   treatment for pH control
     •  Option 2:  Option 1 plus end-of-pipe precipitation/clarification for
                   control of fluoride, arsenic, and suspended solids
     •  Option 3:  Option 1 plus precipitation/clarification  of  concentrated
                   fluoride wastes for control of fluoride
     •  Option 4:  Option 2 plus recycle of  treated effluent  to  further  reduce
                   pollutant discharges
     •  Option 5:  Option 2 plus filtration  to further  reduce  pollutant
                   discharges
     •  Option 6:  Option 5 plus activated carbon to further  reduce  toxic
                   organics.

     The Agency's evaluation of treatment options 4 and 6 concluded  that these
technologies would not be technically feasible for all  crystal plants  in the
case of option 4, and would remove little, if any, pollutants  in the  case of
option 6.  For this reason, the economic impact  analysis concentrated  on
treatment options 1,  2, 3, and 5 only.
                                     1-32

-------
5.4  TREATMENT COST ESTIMATES
     Costs of compliance were estimated by EPA for treatment options  1, 2, 3,
and 5.  Treatment costs of option 1 are for monitoring costs only.  Informa-
tion available indicates that most electronic crystal plants are practicing
some degree of solvent management to control toxic organics (53 percent of the
plants are estimated to be already meeting the TTO limit), and that the
incremental costs of disposal tend to be offset by resale of recovered
solvent.  Moreover, all these plants are controlling the pH of discharges by
end-of-pipe neutralization.

     The toxic organics monitoring costs are expected to apply to no  more than
approximately 50 percent of the plants in this subcategory because the
regulation provides for a certification alternative to monitoring.  EPA is
unable to predict precisely which plants will incur monitoring costs;
therefore, the economic impact was assessed for each plant.  However, for
purposes of estimating total compliance costs for this regulation, only half
of these costs are included in the total since only half of the plants at most
are expected to incur costs.  Appendix I-A contains a sensitivity analysis on
option 1 costs and economic impacts.

     In developing the compliance cost estimates for treatment options 1, 2,
3, and 5 the following major assumptions were made by EPA:

     •  All costs are expressed in end-of-year 1979 dollars.
     •  The treatment facilities were assumed to operate 8 hours per  day,
        260 days per year for plants with discharge less than 60,000  gpd;
        24 hours per day, 260 days per year for plants with 60,000 gpd to
        200,000 gpd; and 24 hours per day, 350 days per year for plants with
        more than 200,000 gpd.
     •  Labor costs are based on an hourly rate of $20, including fringe
        benefits and plant overhead.
     •  The cost of land is valued at $12,000 per acre.
     •  Energy costs are based on $306 per horsepower.
     •  Sludge disposal costs are included.  Available information indicates
        that the sludge to be disposed from the removal of fluoride is not
        defined as hazardous wastes by RCRA.
                                     1-33

-------
     •  The capital investment and annual operating and maintenance costs  for
        monitoring the wastewater of a plant are $2,000 and $3,500 (based  on
        an average sampling frequency of once every three months),
        respectively.
     •  Capital costs are amortized at 5 years and 13 percent  interest.

5.4.1  Existing Sources
     Based on the above assumptions, capital investment costs  and annual
compliance costs at several selected flow rates were estimated  for treatment
options 1, 2, 3, and 5.  Chapter 9 in the Final Development Document present
the cost curves for these options.  These cost curves were used to estimate
compliance costs for the seven model non-arsenide crystal plants and the eight
actual arsenide crystal plants.  The key operating characteristics of the
seven model plants are summarized in Table 5—1.  The seven model plants
(non—arsenide crystal) were based on the plant sizes defined in Table 3—3,
wastewater flow rates, and other data developed from the EPA 308 survey.

     Information obtained by EPA indicates that all the identified non-
arsenide direct dischargers are already in compliance with the  alternative
effluent standards.  Tables 5-2 and 5-3 present the estimated  investment and
annual cost of compliance for the 53 non—arsenide indirect dischargers.  Table
5-4 summarizes the compliance costs for the eight arsenide crystal plants.

5.4.2  New Sources
     For the Electronic Crystal subcategory, the final NSPS and PSNS discharge
limitations are the same as those for BAT and PSES.  Since new source costs
are defined as incremental costs from BAT and PSES, costs of NSPS and PSNS are
zero.
                                     1-34

-------
                          TAILS 5-1,  PRQPILB 39 N0N-AR81NIBB MBTAL 6RVITAL, M8BBI PLANTS
                                  Numbi? if
Value if
Book Value el
flanfe Alien
XS1.
XS2.
SI.
v "•
w
Ul
S3.
M.
L.
ttw«* * cBttav gniBcwjeeei nae_e
Extra Small Plant with
Low Flow Rate 19 19
Extra Small Plant with
High Flow Rate 29 159
Small Plant with Lew
Flow Rate 89 15
Small Plant with Medium
Flow Rate 89 959
Small Plant with High
Flow Rate 89 5,999
Medium Plant 115 1,999
Large Plant 399 22,599
^^^^^psa^^i^^^
(iQQQi
299
199
2,999
1,799
1,799
4,999
11,999
^nr
25
29
25
15
15
15
29
(9QOPJ
149
229
1,299
799
799
2,199
1,199
Source:   JRB Aiaoeiatei  estimate!

-------
                           TABLE 5-2.  COMPLIANCE COSTS OF  INDIRECT DISCHARGER NON-ARSENIDE METAL CRYSTAL MODEL PLANTS
                                                      (Thousand Dollars)








M
1
OJ
cr.

Plant
Types
XS1
XS2
SI
S2
S3
M
L
Flow
Rate
(GPD)
10
150
65
550
5,000
6,500
22,500
Compliance Capital Investment Per Plant

Option 1
2.0
2.0
2.0
2.0
2.0
2.0
2.0

Option 2
20.0
20.0
20.0
20.0
102.6
114.1
189.0

Option 3
2.0
2.0
2.0
63.5
63.5
63.5
103.7

Option 5
20.0
20.0
20.0
20.0
116.0
127.9
207.6
Annual Compliance Costs Per Plant

Option 1
4.1
4.1
4.1
4.1
4.1
4.1
4.1

Option 2
15.2
15.5
15.2
17.9
72.3
80.6
135.9

Option 3
4.1
4.1
4.1
33.8
52.5
52.5
85.2

Option 5
15.2
15.5
15.2
17.9
80.2
88.7
146.8
Source:  JRB Associates estimates

-------
                            TABLE 5-3.  TOTAL COMPLIANCE COSTS FOR 53  INDIRECT DISCHARGER  NON-ARSENIDE CRYSTAL PLANTS
                                                                 (Thousand DoIlars)
                                                                                                                                                  a/
Plant
Types
XS1
XS2
SI
S2
S3
M
1 M
u>
L
Number of
Sample
Plants
4
3
5
5
4
10
7
Estimated
Tota 1 Number
Plants
5
4
7
7
6
14
10
Option
5
4
7
7
6
14
10
Compl lance Capital
1* Option 2
95
76
133
133
610
1.583
1,880
Investment
OjJtlon 3
5
4
7
438
375
875
1,027
Option 5
95
76
133
133
690
1,777
2,066
Option 1*
10.3
8.2
14.4
14.4
12.3
28.7
20.5
Annual Compl
Option 2
65.8
53.8
92.0
110.9
421.5
1,099.7
1,338.5
lance Costs
Option 3
10.3
8.2
14.4
222.2
302.7
706.3
831.5
Option 5
65.8
53.8
92.0
110.9
468.9
1,213.1
1,447.5
Total
a/
                  38
                                  53
                                                  53
4,510
2,731
  Toxic organ Ics monitoring costs are  Incurred by 50 percent of the plants only.

   'Selected option

Source:  JRI3 Associates estimates
4,970
                                                                                                             108.8
3,182.2
2,095.6     3,452.0

-------
                        TABLE M.  COMPLIANCE COSTS  Of  EIGHT ARSENIDE  CRYSTAL PLANT!  (Thousand Dollars)
Plant ip Status

Tl 1
Tl 1
n \
T4 1
Tl 1
Tl 1
T7 1
TS g
Tefal
Nust&eF @f
IBB levees PI ex Rate
(OPO)
12 9
71 9
§ 1,119
41 499
41 7,199
41,199
893 111,989
1,179 189,999

Value ef gamp Mange gap If al invest
iflJMents gBflen 1 3Bfl@n 1* Q
(iooo) '
na 3 9
na 9 9
§99 98' 8a/
1,999 1,9 1,9
1,184 1,3 118,1
11,339 1,3 246,7
41,399 1,9 111,1
78,9896/ 88' 88/
8,9 760,9
8,9 760,9
3 9
menf Annual gawp 1 lanee
atl§n I apt 1 en, \ eetlef) i»
99 9
9 9 9
9a' 88' 98/
1,9&/ 4,1^ 4,16/
111,1 4,1 84,1
171, § 4,1 171,4
441,7 4,1 191,8
9" 9" f
856,1 16,4 §§9,4
856,1 16,4 §§9,4
99 9
Sssts
eat leg §
9
9
98'
4Jb/
91,1
194,1
111,9
98'
§11,§
613,6
9
na = N@f available,

       plant  is
                      in
This plant is already in gawp I lanes with flusHde and  arsenie gHluent
          eg  §@lv@nf manai@fn@nt ta8hni|u@§ t§e§ Final  §evelapfflsnt g@gyn)int),
                                                                                                        need t@
               en
                         ef

-------
                              6.  ECONOMIC  IMPACT

     This section provides an assessment of the economic  impacts which are
likely to occur as a result of the costs of the proposed  effluent treatment
technologies described in Chapter 5.  It is based upon an examination of the
estimated compliance costs and other economic, technical, and financial
characteristics of both model and actual plants of various sizes and the
analytical methodology described in Chapter 2.  The primary economic impact
areas discussed include changes in industry profitability, plant closures,
substitution effects, changes in employment, shifts in imports and exports,
etc.

6.1  PRICE AND QUANTITY CHANGES
     As described in Chapters 2 and 3,  the Electronic Crystals industry
exhibits characteristics associated with non-competitive  markets.  For
example, the industry has unusually high profit rates and most of the product
groups have high concentration ratios and low demand elasticities.  These
three characteristics are often indicative of non—competitive markets.  All
other things being equal firms in this  type of market would be able to recover
part or all of the compliance costs in  the  form of higher prices.

     However, as previously mentioned,  the existence of the threat of foreign
competition serves to mitigate the ability of firms in the industry to raise
prices.  Such price action would cause  the domestic industry to  lose market
share.  Consequently, it is assumed that there is an equilibrium trade-off
between domestic and foreign market share which is determined largely by the
price differentials; and that domestic  producers would like to maintain the
baseline equilibrium by maintaining the price spread between domestic and
foreign producers.  Since the domestic  producers have high profit rates, they
can afford to maintain the spread by absorbing the increased costs.  For this
reason, it is assumed that for small and moderate levels  of compliance cost,
domestic electronic crystals producers will absorb the costs and not adjust
prices.  Consequently, the price changes due to the regulations would be zero.
It follows, also, that quantities demanded are not expected to change because
of the regulations.
                                      1-39

-------
6.2  RESULTS OF SCREENING ANALYSIS
     As described in Chapter 2, a ratio of annual compliance cost  to  plant
revenues of one percent was used as a threshold value below which  plants  are
considered to have low impacts from the regulations.  The calculated  ratios
are shown in Table 6-1.  Most non-arsenide model plants have compliance costs
above this value for option 5.  Therefore, the profitability analysis was
conducted for all non-arsenide model plants.  For the arsenide metal  crystal
plants only one plant (plant number T5) has compliance costs greater  than one
percent.

     The one arsenide metal crystal plant and all of the non-arsenide plants
were subjected to a more detailed analysis of plant profitability.

6.3  PROFIT IMPACT ANALYSIS
     As indicated in Chapter 2 above, the assessment of the impact  of com-
pliance on plant profitability is based on the plants' after-compliance return
on investment (ROl) ratios.

     Table 6-2 summarizes the estimated profit impact for the seven non-
arsenide model plants.  As indicated in the table, there are significant
reductions in ROI at some plants under options 2 and 5, smaller profit
reductions under options 1 and 3.  However, all non-arsenide model  plants
remain above the threshold level for profitability.  Their after-compliance
ROIs range between 18.5 percent and 39.7 percent at treatment option  5  as
compared to 11.5 percent average yield of U.S. Treasury bonds for  1980   ,
therefore are not considered to be potential plant closures.
     Table 6-3 presents the estimated profit impact on  the  potential  "high
impact" arsenide metal crystal plant.  This table  indicates  that  after-
compliance ROI under treatment options 2 and 5 are very low for plant T5.
However, this plant is already operating at an extremely  low ROI  without  the
regulation, according to the responses to the EPA  308 survey.  It  is  suspected
that this low baseline ROI results from the internal transfer  pricing of  the
respondent.  Therefore, the baseline ROI is believed to be  understated.
  Standard and Poors Corp., The Outlook.
                                     1-40

-------
                        TABLE 6-1.  SCREENING ANALYSIS
                          ANNUAL COMPLIANCE COSTS TO VALUE OF SHIPMENTS (%)
PLANT
Arsenide Crystals
Tl
T2
T3
T4
T5
T6
T7
T8
Non-Arsenide Crystals
XS1
XS2
SI
S2
S3
M
L
Option 1 Option 2

0 0
0 0
0 0
0.2 0
0.2 3
* 0
* 0
0 0

2.1 7
1.4 5
0.2 0
0.2 1
0.2 4
0.1 2
* 1




.2
.5
.8
.7


.6
.2
.8
.1
.3
.0
.2
Option 3

NA
NA
NA
NA
NA
NA
NA
NA

2.1
1.4
0.2
2.0
3.1
1.3
0.8
Option 5

0
0
0
0.2
3.9
0.8
0.7
0

7.6
5.2
0.8
1.1
4.7
2.2
1.3
NA = Not applicable.  Treatment option not selected by EPA for this product
     group.

*    Less than 0.05 percent.

Source:  JRB Associates estimates.
                                     1-41

-------
TAiLl 6=2,
                                       OF  PROFIT  IMPACT  AlilliMINT  FOR  NQN-ARS1NIB1  61VSTAL MOBIL PLANT!
Value of
Model Plants Shipments
XS1
XS2
SI
S2
S3
M
£ M
L
($000)
200
300
2,000
1,700
1,700
4,000
11,000
Assets
Value
($000)
140
220
1,200
700
700
2,100
6,300
Pre-Cotnplianee
Return on Sales 801
(3)
23,0
20,0
25,0
13,0
15,0
15,0
20,0
(*)
35,7
27,3
41,7
36,4
36,4
28.6
34,9
Option 1
After-eomplianee 801
Option 2
Option 3
Option 5

32,3
25,2
41,3
35,7
35,7
28,3
34,8
21,8 32,3
18,5
39,7
32,9
22,8
23,5
31,8
25,2
41,3
29,0
26.5
25.3
33,0
21,8
18,3
39,7
32,9
21,4
22,9
31,6
Source:   JRB Aseoeiates Estimates

-------
        TABU 6-1,  PRQfi? IMPA6? ASSgllMINT 0F POTENTIAL  "HIQH  IM?A6TM ARISNtDI 0RYSTAL PLANTS
Plane ID
               Value of
           A§§el§
            ill
               TfOOO!
leluffl ea
                                 Al6§f=eetBBlUflie  RDt  (I)
2,184      4,100
                                                          2,0
                                              0,2         0,1
Source i  JRi A§s§§iaiee

-------
6.4  CAPITAL REQUIREMENTS ANALYSIS
     As described in Chapter 2, the  assessment  of  the  firms'  ability to raise
the required capital is based upon ratios of  compliance  capital  investment to
revenues (CCI/R) for specific plants compared with  the difference  between the
industry's highest and lowest ratios of average  annual capital expenditures to
average value of shipments between 1970 and 1977.   The highest and lowest
ratios for SIC 3679 (Electronic Components, n.e.c.) were  5.1  percent and 2.8
                                                     II
percent over the 1970-1977 time period respectively;     therefore,  2.3  percent
was selected as the threshold level  for the capital availability analysis.
Plants whose CCI/R ratios are lower  than 2.3  percent are  considered  to  have
low levels of impacts.  Those plants with higher CCI/R ratios  are  subjected to
the cash flow test described in Chapter 2.

     Table 6-4 presents the compliance capital  investment  to  plant  revenue
ratios for the non-arsenide crystal  plants.  As  indicated  in  the table,  com-
pliance capital requirements for the proposed treatment options  appear  to be
high for model plants XS1, XS2, S2,  S3, and M.  The cash  flow  test  for  these
five plants is shown in Table 6-5.   The plant annual cash  flows  (i.e.,
depreciation plus net profit after taxes) seem  to be sufficient  to meet  total
capital requirements (compliance capital investment plus minimum annual
capital expenditures for plant and equipment) of all plants at each  treatment
option.

     Table-6-6 summarizes the capital requirements  analysis for  the  eight
arsenide crystal plants, where plant T5 is predicted to have  some  difficulties
financing treatment options 2 or 5.  However, plant T5 is  owned  by  a large
integrated electronics company and the required compliance capital  investments
for both treatment options represent less than  0.1  percent of  this  firm's
total annual capital expenditures in 1978.    Since the capital  requirements
is so small relative to total capital expenditures  for this company, it  is
assumed that the company can readily finance  the investment.
2/
  U.S. Department of Commerce, 1977 Census of Manufactures
 'Standard and Poor's Corp., Stock Market Encyclopedia.
                                     1-44

-------
                TABLE 6-4.  COMPLIANCE CAPITAL REQUIREMENTS ANALYSIS - NON-ARSENIDE CRYSTAL MODEL PLANTS
Value of
Model Plants Shipments

XS1
XS2
SI
S2
S3
M
L
($000)
200
300
2,000
1,700
1,700
4,000
11,000
Assets
Pre-Compliance
Compl
Value Profit Margin(%) ROI(%)
($000)
140
220
1,200
700
700
2,100
7,200

25.0
20.0
30.0
15.0
15.0
15.0
20.0

35.7
27.3
50.0
36.4
36.4
28.6
34.7
Option 1

1.0
0.7
0.1
0.1
0.1
0.1
**
iance Capital Investment to
Value of Shipments (%)
Option 2

10.0*
6.7*
1.0
1.2
6.0*
2.9*
1.7
Option 3

1.0
0.7
0.1
3.7*
3.7*
1.6
0.9
Option 5

10.0*
6.7*
1.0
1.2
6.8*
3.2*
1.9
 * Greater than threshold value.
** Less than 0.05 percent.
Source:  JRB Associates estimates

-------
                       TABLE 6-5.   C0WPAKIS4HI OIF CASH FLOW
                          AUTO TOTAL, CAPITAL, REQDTIREBIEH1TS
                         (as a Percent of Plant Revenues)
                            Internally            Total Capital
                             (generated
     Model Plants           (percent)         Optima 2     CTptiom 3    gtptifflm 5


          XS1                   18.5                12.8          3.8        12.8



                                15.1                 9.5          3.5         9.5



                                11.1                 4.0)          6.5         4.0)



                                11.1                 8L8B          6.5         9_fe



                                11.6                 5.1          4.4         6.0)
              tto> melt pircDfritt maarg;iimK ((assmniiimjg am aweorag^ cxiiirpaiiraitte:  imconiE
  irate ct£ 401 pemnemlt)) plus dtepureciLatiLcntL ((10) peirceimfi:
  assets x assuiuffli ten toe 50) percemC &E ttcuttal
  Qampliance capital imvesttnueiffitt plms minminiinni anmial capital expsmfiittviiires fenr
  plant ami eqpxipfflient to maimtain a targeted jnnaduictticnii  level Cassxamieai ten be
  2.8  percent of plant rewentuDstsJ).

Source:  JIB Associates estimates
                                         1-46

-------
           TABLE 6-6.  COMPLIANCE CAPITAL REQUIREMENTI ANALYSIS - ARSENIDE CRYSTAL PLANTS
Plant ID

Tl
T2
T3
T4
7 T5
T6
T7
T8
Value of Shipments 6em§lian@i
(8000) £j

na
na
600
2,000
2,384
23,000
45,000
78,000b'
Compliance capital investment represents Uss than
total annual capital expenditures in 1971,
'e><.hJ«AfeA V .

GasiUl Investment To
itun i Option 2

0 0
0 0
0 0
0,1 0,1
0,1 5,0a/
* 1,1
* ,9
0 0
0,1 peteeni §f firm's

Value of Shipments (X)
Option 5

0
0
0
0.1
5.6*'
1.2
1.0
0


* Leas than 0.05 percent,
Source:  JRB Asioeiates esfcimalee,

-------
6.5  POTENTIAL PLANT CLOSURES
     Based on the profit impact assessment and  capital  requirements  analysis
discussed in Sections 6.2 and 6.3, only one arsenide  crystal  plant,  model
plant T5, might be considered to be a potential closure at  treatment options  2
and 5.  However, plant T5 is owned by a large integrated  electronic  company
and supplies gallium arsenide and gallium phosphide crystals  to  be used  in  the
production of LEDs at other production facilities of  the  firm.   The  plant's
low reported baseline profitability may be due  to the firm's  intercompany
transfer pricing policy and actual profit at market price may be very satis-
factory.  Therefore, it is expected that the firm will  keep this plant
operating to maintain control on the source of  supply for its raw material.

6.6  EMPLOYMENT EFFECTS
     The preceding section projects no closure  potential  for  the proposed
treatment technologies.  Thus, no employment effects  are expected from the
promulgated regulations.

6.7  SUBSTITUTION EFFECTS
     As described in Section 2.2, because of their special properties, the
potential for substitution of electronic crystals in  their current applica-
tions is limited in the base case.  In addition, the  preceding impact analysis
assumes that there will be no price changes resulting from the proposed
regulations.  For these reasons, no substitution effects are expected  to
result from the promulgated regulations.

6.8  FOREIGN TRADE IMPACTS
     As described in chapters 2 and 3, the United States is currently very
competitive with foreign producers on the world market  for electronic
crystals.  However, there exists considerable technical expertise among
foreign producers, which is perceived as a potential source of competition.
If there was a significant price effect from the regulations, the U.S.
producers could lose market share.  However,  as indicated in the profit  impact
analysis described in Section 6.4, U.S. producers appear to be able  and  likely
                                     1-48

-------
to absorb the additional pollution control costs.  Consequently, no foreign
trade impacts are expected to result from the regulations.

6.9  NEW SOURCE IMPACTS
     As indicated in Section 5.4.2, the incremental costs of NSPS and PSNS are
zero.  Because there is no difference in compliance costs between existing and
new sources, the regulations would not foster any competitive advantage or
disadvantage between new and existing sources as long as treatment require-
ments for each are the same.

6.10  SUMMARY OF IMPACTS OF PROMULGATED REGULATIONS
     The estimated economic impacts for the  treatment options selected  for
promulgation in the Electronic Crystals subcategory are summarized in Table
6.7.

     Toxic Organics
     The control of toxic organics and pH under  option  1 is not  expected  to
cause any incremental compliance costs for non-arsenide crystal  direct
dischargers under BPT and BAT.  These plants already  control  their discharges
to the selected treatment level for regulation.

     Non-arsenide crystal indirect dischargers are expected to  incur
monitoring costs to comply with the total toxic  organics limitation under
PSES.  These costs, totaling $53,000 in capital  investment and  $109,000  in
total annual costs, assume that at least the 50  percent of the  plants already
in compliance will choose to certify that they do not dump their solvents into
the effluent instead of monitoring.  The remaining plants are expected,  on
average, to monitor quarterly.  Since the incremental costs of  solvent
disposal tend to be offset by resale of recovered solvent, EPA  did not
estimate costs for solvent management.

     These monitoring costs are expected to  cause moderate reductions  in
profitability for small non-arsenide crystal plants, but no plant closures  are
expected as a result of this level of control.
                                      1-49

-------
                TABLE 6-7.  SUMMARY OF ESTIMATED ECONOMIC IMPACTS
ECONOMIC IMPACT VARIABLES
Gallium/Indium Arsenide Crystals
                                     OPTION 1 OPTION 2 OPTION 3
         SELECTED
OPTION 5  OPTION
  Annual Compliance Costs/Revenues (%)
    Indirect Dischargers
    Direct Dischargers
  Change in Price (%)
  Change in Quantity (%)
  Change in Profitability
    Indirect Dischargers
    Direct Dischargers
  Capital Requirements
    Indirect Dischargers
    Direct Dischargers
  Plant Closures due to Regulations
  Employment at Closed Plants
  Balance of Trade Changes
  Industry Structure Changes

Non-Arsenide Crystals
  Annual Compliance Costs/Revenues
    Indirect Dischargers
    Direct Dischargers
  Change in Price (%)
  Change in Quantity (%)
  Change in Profitability
    Indirect Dischargers
    Direct Dischargers
  Capital Requirements
    Indirect Dischargers
    Direct Dischargers
  Plant Closures due to Regulations
  Employment at Closed Plants
  Balance of Trade Changes
  Industry Structure Changes
0-0.2
0
0
0
0-3.5
0
0
0
N/A
N/A
N/A
N/A
0-3.9
0
0
0
0-3.5
0
0
0
                                        Low    Moderate   N/A    Moderate  Moderate
                                         0       0        N/A      0         0
Low
0
0
0
None
None
Moderate
0
0
0
None
None
N/A
N/A
N/A
N/A
N/A
N/A
Moderate
0
0
0
None
None
Moderate
0
0
0
None
None
                                        0-2.1  0.8-7.6 0.2-3.1  0.8-7.6    0-2.1
                                         00000
                                         00000
                                         00000
                                      Moderate  Signif. Moderate Signif.  Moderate
                                         ooooo
Low
0
0
0
None
None
Moderate
0
0
0
None
None
Moderate
0
0
0
None
None
Moderate
0
0
0
None
None
Low
0
0
0
None
None
  N/A = Not applicable.
        group.
                         Treatment option not selected by EPA for this product
                                       1-50

-------
     In addition to the above  analysis,  EPA conducted  a sensitivity analysis
for option 1 costs consisting  of  two  parts.   First,  the impact  of monthly
monitoring was estimated  for all  facilities  (some  facilities may monitor as
frequently as once per month,  although EPA cannot  identify  precisely which
ones.)  In addition to this analysis, costs  and  impacts were also determined
for facilities which may  incur "worst case"  solvent  disposal costs.   These
costs are developed for facilities  that  could fall under the requirements of
the Resource Conservation and  Recovery Act (RCRA)  for  disposal.   The analysis
of these costs shows that while the effects  on profitability are higher using
these costs, the effects  are not  expected  to significantly  increase the impact
on these facilities and are not expected to  cause  any  plant closures.

     Arsenic
     Three arsenide crystal plants  are expected  to incur investment costs of
$761,000 and $560,000 annually (including  monitoring)  as a  result of the
control of arsenic under  option 2 for PSES.   One arsenide crystal plant shows
relatively high impacts under  this  control.   However,  this  plant is owned by a
large electronics company and  supplies crystals  to be  used  in  the production
of LEDs at other production facilities of  the firm.   It is  expected that the
firm will keep this plant operating to maintain  control of  the  source of
supply for its raw material.   Consequently,  no plant closure  is  expected.

     New Sources
     The effluent standards and associated technologies for new sources are
identical to those for existing sources.  Consequently, the economic impacts
for new sources will, mirror those of  existing sources  and the  promulgated
regulations are not expected to foster competitive advantages  or disadvantages
between new and existing  sources.
                                      1-51

-------
                          7-  SMALL BUSINESS ANALYSIS

     The Regulatory Flexibility Act (RFA) of 1980 (P.L. 96-354) requires
Federal regulatory agencies to consider "small entities" throughout  the
regulatory process.  The RFA requires an initial screening analysis  to be
performed to determine if a substantial number of small entities will be
significantly impacted.  If so, regulatory alternatives that eliminate or
mitigate the impacts must be considered.  This analysis addresses  these
objectives by identifying and evaluating the economic  impacts of the
aforementioned regulations on small electronic crystal manufacturers.  As
described in Chapter 2, the small business analysis is developed as  an
integral part of the general economic impact analysis  and is based on the
examination of the distribution by plant size of the number of electronic
crystal plants, plant revenues, wastewater volumes, compliance costs and
potential closures from the regulations.

     As explained  in Chapter 2, in the absence of a specific definition for a
small electronic crystal business, three plant size definitions based on plant
revenues were used to provide the EPA possible alternative definitions of
small electronic crystal manufacturing operations.  These are:

     •  Plants with less than $1 million in value of shipments
     •  Plants with less than $3 million in value of shipments
     •  Plants with less than $5 million in value of shipments.

     The number of electronic crystals plants falling  into each size category
is shown in Tables 7-1 and 7-2 for the arsenide and non-arsenide crystal
plants, respectively.  The tables also show the plant revenues of  the
different sized plants, along with the percentages of  the industry total for
each.  As the tables show, the industry is characterized by a few  large plants
which account for most of the production and many smaller plants producing a
smaller portion of industry output.

     The information available for this study precluded a forecast to the
1985-1990 period of change in the distribution of industry output by plant
                                     1-52

-------
                 Table 7-1.  SUMMARY OF SMALL BUSINESS ANALYSIS
                             FOR ARSENIDE CRYSTAL PLANTS


No. of Plants - Number
- % of Total
Flow rate - 000 gpd
- % of Total
Plant Revenues - %
- $ million
- % of Total
Treatment Option 1 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
Treatment Option 2 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
Treatment Option 5 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
All
Plants3/
4

185.8


72.4


8

16

*

761

560

0.8

856

614

0.8
Plants with Annual Revenues
<1 1-3 3-5
0 20
50.0
0 7.6 0
4.1

0 4.4 0
6.1

0 40
50.0
0 80
50.0
0.2

0 121 0
15.9
0 88 0
15.7
2.0

0 135 0
15.8
0 97 0
15.8
2.2
(in millions)
> 5
2
50.0
178.2
95.9

68.0
93.9

4
50.0
8
50.0
*

640
84.1
472
84.3
0.7

721
84.2
517
84.2
0.8
  Four arsenide crystal plants are already in compliance and are excluded from
  this analysis.

 *Less than .05 percent.
Source:  JRB Associates estimates.
                                     1-53

-------
                 TABLE 7-2.  SUMMARY OF SMALL BUSINESS ANALYSIS
                            FOR NON-ARSENIDE CRYSTAL PLANTS


No. of Plants - Number
- % of Total
Flow Rate - 000 gpd
- % of Total
Plant Revenues -
- $ Million
- % of Total
Treatment Option 1 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
Treatment Option 2 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
Treatment Option 3 Costs
Investment - $000
- % of Total
Annual - $000
- % of Total
- % of Revenues
Treatment Option 5 Costs
Investment - $000
Annual - $000
- % of Total
- % of Revenues
All
Plants
53

351.0


204.3


53

109

0.5

4,510

3,182

1.6

2,731

2,096

1.0
4,970

3,452

1.7
Plants with Annual Revenues (
<1
9
17.0
0.7
0.2

2.2
1.1

9
17.0
19
17.4
0.9

171
3.8
120
3.8
5.5

9
0.3
19
0.9
0.9
171
3.4
120
3.5
5.5
1 - 3
20
37.7
34.3
9.8

36.1
17.7

20
37.7
41
37.6
0.1

876
19.4
624
19.6
1.7

820
30.0
539
25.7
1.5
956
19.2
672
19.5
1.9
3-5
14
26.4
91.0
25.9

56.0
27.4

14
26.4
29
26.6
0.1

1,583
35.1
1,100
34.6
2.1

875
32.0
706
33.7
1.3
1,777
35.8
1,213
35.1
2.2
in millions)
> 5
10
18.9
225.0
64.1

110.0
53.8

10
18.9
20
18.4
*

1,880
41.7
1,338
42.1
1.2

1,027
37.6
832
39.7
0.8
2,066
41.6
1,447
41.9
1.3
*Less than .05 percent.
Source:  JRB Associates estimates
                                     1-54

-------
size.  However, no reason was found to expect the proportion of  industry
output produced by these small plants to change.  For these reasons, the
assessment of impacts on small business were developed from the  data on
existing plants.

     The assessment of economic impacts on small plants begins with an
evaluation of compliance costs incurred at the plant level.  Tables 7-1 and
7-2 also show the estimated investment and total annual compliance cost for
electronic crystal manufacturing by plant size.  These tables also show the
proportion of electronic crystal output and industry compliance  costs
attributed to both small and large plants.  For example, at treatment option 2
the nine non-arsenide crystal plants with less than $1 million in revenues
discharge only 0.2 percent of the total wastewater discharged while they
account for 4.2 percent of total annual compliance costs.  In contrast to
these small plants, the ten plants with sales greater than $5 million
discharge 64.1 percent of the wastewater and incur only 41.3 percent of total
annual compliance costs.  For all regulatory options, compliance cost as a
percent of revenues is significantly larger for the smaller than for the
larger plants.

     Because the compliance costs of small plants are greater than that of
larger plants and because they compete for the same markets as larger plants,
their profits as measured by return on assets would decrease more than those
of larger plants.  Likewise, their capital requirements for compliance rela-
tive to their revenues will be greater than that of larger plants.  However,
as shown in Chapter 6, the magnitude of these impacts for the selected options
are not enough to cause any closures or any other significant impact among
these smaller plants.  Consequently, there will be no employment, community
impacts, or industry structure changes resulting from the regulations.

     As described in Chapters 2 and 6, the regulations will not  cause changes
to the prices of electronic crystals.  As a result, manufacturers will absorb
the full amount of the compliance costs and, therefore, will experience lower
profits from electronic crystal manufacturing.
                                     1-55

-------
                        8.  LIMITATIONS OF THE ANALYSIS

     This section discusses the major Limitations of the economic  impact  anal-
ysis.  It focuses on the limitations of the data, methodology, assumptions,
and estimations made in this report.  Information pertaining to the estimation
of the compliance costs for specific plants and its limitations are outlined
in Section 9 of the EPA Development Document.

     The economic impacts assessed are the result of this regulation only.
The assessment does not include the economic impacts associated with other
regulations for air pollution control, OSHA requirements, and solid waste
requirements.

8.1  DATA LIMITATIONS
     The accuracy of the conclusions of this report depends largely on  the
accuracy of the data used in the analyses, especially that of the  estimated
compliance costs, and plant financial and economic characteristics.

     A critical data input to this study is the compliance cost estimates.
The compliance costs used were developed from compliance cost curves provided
by EPA.  The assumptions relating to the estimation of compliance  costs are
outlined in Section 9 of the EPA Development Document and summarized in
Section 5.4 of this report.

     In addition to the compliance cost assumptions, plant financial profiles
used in the analysis are subject to the following major assumptions and
limitations:
     •  Baseline financial data of the eight existing arsenide crystal plants
        are obtained from EPA 308 survey and follow-up telephone  interviews
        with plant representatives;
     •  Plant characteristics such as size and flow rate of the 53 non-
        arsenide crystal plants are inferred from survey data collected by EPA
        on 38 existing plants;
     •  Baseline financial data of the 53 non-arsenide crystal plants are
        inferred from survey data on 26 plants;
                                     1-56

-------
     •  The impact analysis for non-arsenide crystal plants was based on  seven
        model plants which were developed from the above survey data; and

     •  Plant financial data are based on 1978 economic conditions and will
        remain essentially the same over the period of the study.


     Because individual plant financial characteristics vary from the model
plant profiles, the model plants developed for this study should be considered
order of magnitude estimates of the typical operating characteristics of  the
non-arsenide crystals manufacturing industry.  However, it is believed that
the sample of plants was sufficient to reflect considerable intra-industry
variations in conditions.  Therefore, although the confidence interval of the
conclusions are indeterminate, there is no reason to suspect that the
conclusions are biased toward overestimating or underestimating the degree of
potential impact.


8.2  METHODOLOGY LIMITATIONS

     In addition to the data limitations described above, this report is also
subject to the following limitations of the methodology used:


     •  It is assumed that firms will not attempt to pass through the
        compliance costs to their customers by raising their product prices;

     •  The required rate of return on investment (i.e., profit impact
        threshold value) is estimated to be 11.5 percent, the 1980 average
        yield of U.S. Treasury bonds;

     •  The threshold values of a firm's capital availability are assumed to
        be the plant's estimated annual cash flow and/or the industry
        historical annual rates of capital expenditures as reported by the
        Census of Manufactures; and

     •  Plant closure criteria do not include the financial structure of  the
        plants, and the time value of money, or special plant-specific
        conditions.


8.2.1  Price Increases Assumption

     The analysis assumes that the electronic crystals industry will not
increase the price of their products and pass through any portion of the
compliance costs to their customers.  While this is a reasonable assumption
for some types of crystals it may not be true for the entire industry.  It  is
                                     1-57

-------
possible that the price pass-through percentage will vary  among  the various
types of crystals.  Adequate information on  these  segments  is not  available.
Therefore, no further analysis is carried out to measure the extent of  a  price
increase which may occur as the production costs of  the plants  increase with
the installation of pollution control.

     The estimation error which might result  from  the  price change assumptions
depends primarily on the cross elasticity of  demand with respect to the ratio
of domestic to foreign prices.  That is, if  increased  domestic  prices shifts
demand production overseas, there would be a  drop  in aggregate  quantities
produced by domestic producers and, consequently,  unemployment,  and possibly
other impacts.  Thus, the remaining domestic  crystal producers  would have
healthy profits, but account for a smaller portion of  world market  share.  On
the other hand, if such shifts were smaller,  then  domestic producers could
raise prices to cover compliance expenditures without  losing market share.

     If the latter situation prevailed in the real world,  the economic  impacts
estimated in this report are overstated.  If  the former situation  prevailed,
the impacts would take a different form, but would probably not exceed  the
levels estimated in Chapter 6 of this report.

8.2.2  Profit Impact Threshold Assumptions
     The profit impact assessment is determined by comparing the plant's  after
compliance return on investment (ROI) to the  average yield of U.S. Treasury
bonds which is 11.5 percent in 1980.  Plants with  after-compliance ROI  below
11.5 percent are assumed to be potential plant closures.

     Although investment decisions based on discounted cash flow techniques
generally provide more consistent and theoretically correct investment
decisions than those based on ROI analysis, ROI analysis is widely used by the
investment community and does provide acceptable results in comparative
analyses of industry-wide conditions, such as that presented in  this report.
It is reasonable to assume that plants cannot continue to  operate  as viable
concerns if they are unable to generate a ROI that is  at least  equal to the
opportunity cost of other lower risk investment alternatives (e.g., U.S.
Treasury bonds).
                                     1-58

-------
8.2.3  Capital Availability Threshold Assumptions
     A firm's ability to make the required pollution control investment  is
evaluated by comparing the required compliance capital investment to the
industry historical annual rates of capital expenditures and the plant's
estimated internally generated annual cash flow.  Although the above analyses
do not precisely indicate whether or not firms can afford to make the
investments, they do provide a good indication of the relative burden created
by the compliance requirement.

8.2.4  Plant Closure Assessment
     The criteria for plant closure are not all  inclusive.  They do not
address the financial structure of the plants, the time value of money,  or
other factors which affect closure decisions.  A more detailed closure
analysis would require more precise data on the  capital structure, and cash
flow of plants in the electronic crystals  industry.  However, the costs  of the
recommended treatment technology are small, and  a more detailed analysis will
not necessarily yield different results.

8.3  SUMMARY OF LIMITATIONS
     Although a number of assumptions and  data limitations may significantly
bias the economic impact conclusions, the  potential changes to the conclusions
resulting from the elimination of  these biases are probably small.   Sensitiv-
ity analyses of plus and minus 20  percent  were performed for each of the above
assumptions.  These analyses concluded that if each of the  aforementioned
variables changed by plus or minus 20 percent, there would  be no plant
closures as a result of the regulations and the  other results of  the study
would not be significantly impacted.  Consequently, the conclusions  of  this
study appear to represent a reasonable industry-wide assessment of potential
impacts.
                                      1-59

-------
                                 APPENDIX I-A
                    SENSITIVITY ANALYSIS ON OPTION I COSTS

I.   INTRODUCTION

     The costs used to analyze the economic impact of option 1 in the crystals
subcategory analysis are based on two assumptions:  (1) that the only costs
associated with compliance consist of toxic organics monitoring costs, and
(2) that monitoring will occur, on average, once per quarter for the affected
plants.  This analysis consists of two parts.  The first part estimates
the economic impact of monthly monitoring on individual plants in place of
the quarterly monitoring.  The second part estimates the economic impacts
of "worst case" incremental solvent disposal costs.

II.  MONITORING COST SENSITIVITY ANALYSIS

     EPA estimates that facilities covered under this regulation will, on
average, monitor their effluent for toxic organics once per quarter.  However,
some facilities may monitor as frequently as once per month.  Because EPA
cannot determine precisely which facilities will be required to monitor on a
monthly basis, this analysis is performed for all plants to determine the
impact of the costs associated with more frequent monitoring.

     The capital investment cost required for monitoring monthly is  the same
as that for monitoring quarterly, $2,000.  The annual operating and  maintenance
(O&M) costs for quarterly monitoring are $3,500 (as reported in Chapter 5);
therefore, the annual O&M costs for monthly monitoring are $10,500 (Quarterly
costs times 3).

     The methodology for analyzing the impacts of monthly monitoring is the
same as described in Chapter 2 of this report.  Tables A-l through A-3 sum-
marize the results of this analysis.  For the non-arsenide plants, the costs
of monthly monitoring have some effect on profitability, with reductions
in the return on investment (ROI) ranging from 0.2 to 8.3 percentage points
                                  1-60

-------
               TABLE A-l.  ESTIMATED COMPLIANCE COSTS BASED ON
                 ANNUAL MONITORING COSTS OF $11,000 PER PLANT3/
        PLANT
  Arsenide Crystals

         Tl
         T2
         T3
         T4
         T5
         T6
         T7
         T8
                           ANNUAL COMPLIANCE COSTS PER PLANT (SOOO)
                            OPTION 1
                          0

                          Ob/
                         11.
                         11.1
                         11.1
                         11.1
                          Ob/
                                    OPTION 2
  0
  0
  Ob/
 11.
 91.1
186.4
299.8
  Ob/
           OPTION 3
NA
NA
NA
NA
NA
NA
NA
NA
           OPTION 5
  0

  Ob/
 11.
 99.3
201.2
330.0
  Ob/
  Non-Arsenide Crystals
     (Model Plants)

        XS1                   11.1
        XS2                   11.1
        SI                    11.1
        S2                    11.1
        S3                    11.1
        M                     11.1
        L                     11.1
                                          .2
                                          .5
 22.
 22.
 22.2
 24.9
 79.3
 87.6
142.9
11.1
11.1
11.1
40.8
59.5
59.5
92.2
 22.2
 22.5
 22.2
 24.9
 87.2
 95.7
153.8
NA


a/

b/

c/
Not applicable.  Treatment option not selected by EPA for this
product group.

Include $10,500 annual O&M costs and $600 annual capital cost.

This plant is already in compliance.

This plant is already in compliance with fluoride and arsenic effluent
standards.  Only toxic organics need to be reduced through effective
solvent management techniques  (see Final Development Document).
SOURCE:  JRB Associates estimates.
                                      1-61

-------
               TABLE A-2.  SCREENING ANALYSIS BASED ON ANNUAL
                    MONITORING COSTS OF $11,000 PER PLANT3/
        PLANT
ANNUAL COMPLIANCE COSTS TO VALUE OF SHIPMENTS (%)
  Arsenide Crystals

        Tl
        T2
        T3
        T4
        T5
        T6
        T7
        T8
  Non-Arsenide Crystals

        XS1
        XS2
        SI
        S2
        S3
        M
        L
OPTION 1
0
0
0
0.6
0.5
*
*
0
5.6
3.7
0.6
0.7
0.7
0.3
0.1
OPTION 2
0
0
0
0.6
3.8
0.8
0.7
0
11.1
7.5
1.1
1.5
4.7
2.2
1.3
OPTION 3
NA
NA
NA
NA
NA
NA
NA
NA
5.6
3.7
0.6
2.4
3.5
1.5
0.8
OPTION 5
0
0
0
0.6
3.9
0.9
0.7
0
11.1
7.5
1.1
1.5
5.1
2.4
1.4
NA = Not applicable.  Treatment option not selected by EPA for this
     product group.

a'    Include $10,500 annual O&M costs and $600 annual capital costs.

*    Less than 0.05 percent.


SOURCE:  JRB Associates estimates.
                                   1-62

-------
                                  TABLE A-3.  SUMMARY OF PROFIT IMPACT ASSESSMENT BASED ON ANNUAL
                                               MONITORING COSTS OF $11,000 PER PLANT3/
i
&
OJ

PLANT

Arsenide Crystals
T5
Non-Arsenide Crystals
XS1
XS2
SI
S2
S3
M
L
VALUE OF
SHIPMENTS
($000)

2,384

200
300
2,000
1,700
1,700
4,000
11,000
                                             ASSETS
                                             VALUE

                                             ($000)
                                             4,800
  140
  220
1,200
  700
  700
2,100
6,300
             PRE-COMPLIANCE
           RETURN ON SALES  ROI
                  AFTER-COMPLIANCE ROI
                 4.0
2.0
       OPTION 1   OPTION 2   OPTION 3   OPTION 5
1.8
0.1
NA
-0.1
25.0
20.0
25.0
15.0
15.0
15.0
20.0
35.7
27.3
41.7
36.4
36.4
28.6
34.9
27.4
22.0
40.7
34.7
34.7
28.0
34.7
17.4
15.6
39.2
32.0
21.9
23.1
31.7
27.4
22.0
40.7
28.1
25.6
25.0
32.9
17.4
15.6
39.2
32.0
20.6
22.6
31.4
         NA = Not applicable.  Treatment option not selected by EPA for this product group.

         a/   Include $10,500 annual O&M costs and $600 annual capital costs.

         SOURCE:  JRB Associates estimates.

-------
(Table A-3).  However, these reductions are not expected to cause reductions
in production or plant closures because the ROIs remain above the ROI threshold
level (see Chapter 2).  One arsenide plant, T5, appears to have a relatively
large reduction in ROI, from 2.0 percent to 0.1 percent for the selected
option 2 (which includes the monthly monitoring costs).  However, as explained
in Chapter 5, this plant is owned by a large, integrated electronics firm.
The low pre-compliance ROI is likely the result of transfer-pricing within
the company that understates the profits at this plant.  We believe that
this plant is likely to continue to operate despite this low profitability,
to maintain control of supplies to the downstream plants in the company.

     Thus, while the impact of monthly monitoring is expected to be greater
than for quarterly monitoring, the impacts are not expected to cause plant
closures.

III.  SOLVENT DISPOSAL SENSITIVITY ANALYSIS

     EPA has determined that the incremental costs associated with improved
solvent management techniques tend to be offset by resale of the solvents.
Consequently, EPA did not cost out any small costs for the purposes of the
impact analysis.  However, some facilities may have to haul away their solvents
under the requirements of the Resource Conservation and Recovery Act (RCRA).
This analysis examined "worst case" costs for all plants under RCRA hauling
because EPA cannot determine which plants, if any, will fall under these
requirements.  These costs are not considered to be ordinary or average costs
of compliance for facilities in this industry.

     EPA estimated individual plant costs for each facility where possible.
For remaining plants, and for model plants, the costs were extrapolated from
plants with similar volumes of spent solvents.  The costs used in the impact
analysis appear in Table A-4.
                                     1-64

-------
               TABLE A-4.  ESTIMATED COMPLIANCE COSTS ASSUMING
                     INCURRENCE OF SOLVENT DISPOSAL COSTS
        PLANT
  Arsenide Crystals

        Tl
        T2
        T3
        T4
        T5
        T6
        T7
        T8
                                ANNUAL COMPLIANCE COSTS PER PLANT ($000)

                            OPTION 1     OPTION 2     OPTION 3     OPTION 5
                              0

                              Oa/
                             12.Ob/
                             12.0
                             12.0
                             12.0
                              Oa/
  0

  Oa/
 12.Ob/
 92.0
187.3
300.7
NA
NA
NA
NA
NA
NA
NA
NA
  0

  Qa/
 12.Ob/
100.2
202.1
330.9
  Non-Arsenide Crystals

        XS1                  12.0
        XS2                  12.0
        SI                   12.0
        S2                   12.0
        S3                   12.0
        M                    12.0
        L                    12.0
                                           23,
                                           23,
                                           23,
                                            80
                                            88
                                           25.8
                                           144.8
              12.0
              12.0
              12.0
              41,
              60.
              60.
  .7
  .4
  .4
              94.1
              23,
              23,
              23,
 25.8
 88.
 96,
             155.7
NA = Not applicable.
     product group.
                      Treatment option not selected by EPA for this
     This plant is already in compliance.

     This plant is already in compliance with fluoride and arsenic effluent
     standards.  Only toxic organics need to be reduced through effective
     solvent  management techniques (see Final Development Document).
SOURCE:  JRB Associates estimates,
                                     1-65

-------
     The impact analysis was performed including the costs of monthly
monitoring.  This was done to determine the maximum possible impact on any
one plant assuming that (1) the plant is required to monitor toxic organics
monthly; and (2) the plant must dispose its spent solvents under RCRA.

     The analysis followed the methodology described in Chapter 2 of this
report.  Table A-5 shows the results of the screening analysis, and Table A-6
summarizes the results of the sensitivity analysis.  The profit reductions
and levels are only slightly different than with monthly monitoring costs
alone.  For the non-arsenide plants, the costs of monthly monitoring plus
RCRA solvent disposal have some effect on profitability, with reductions
in ROI ranging from 0.2 to 8.9 percentage points.  These reductions are not
expected to cause reductions in production or plant closures because the ROIs
remain above the threshold level.  The one arsenide plant with a relatively
high impact, T5, is not expected to close because it is owned by a large,
integrated electronics firm which uses the products for its "downstream1'
process facilities.

     Thus, while the impact of monthly monitoring plus RCRA solvent disposal
costs are greater than quarterly monitoring costs alone, the impacts are
not expected to cause plant closures.
                                     1-66

-------
                   TABLE A-5.   SCREENING ANALYSIS ASSUMING
                     INCURRENCE OF SOLVENT DISPOSAL COSTS
        PLANT
  Arsenide Crystals

        Tl
        T2
        T3
        T4
        T5
        T6
        T7
        T8
  Non-Arsenide Crystals

        XS1
        XS2
        SI
        S2
        S3
        M
        L
ANNUAL COMPLIANCE COSTS TO VALUE OF SHIPMENTS (%)
OPTION 1
0
0
0
0.6
0.5
0.1
*
*
6.0
4.0
0.6
0.7
0.7
0.3
0.1
OPTION 2
0
0
0
0.6
3.9
0.8
0.7
0
11.6
7.8
1.2
1.5
4.7
2.2
1.3
OPTION 3
NA
NA
NA
NA
NA
NA
NA
NA
6.0
4.0
0.6
2.5
3.6
1.5
0.9
OPTION 5
0
0
0
0.6
4.2
0.9
0.7
0
11.6
7.8
1.2
1.5
5.2
2.4
1.4
NA = Not applicable.  Treatment option not selected by EPA for this
     product group.

*    Less than 0.05 percent.
SOURCE:  JRB Associates estimates.
                                   1-67

-------
                            TABLE A-6.   SUMMARY  OF  PROFIT  IMPACT ASSESSMENT ASSUMING
                                       INCURRENCE OF SOLVENT DISPOSAL COSTS

PLANT

Arsenide Crystals
T5
Non-Arsenide Crystals
XS1
XS2
SI
S2
S3
M
L
VALUE OF
SHIPMENTS
($000)

2,384

200
300
2,000
1,700
1,700
4,000
11,000
                                    ASSETS
                                    VALUE

                                    ($000)
                                    4,800
                                       140
                                       220
                                     1,200
                                       700
                                       700
                                     2,100
                                     6,300
  PRE-COMPLIANCE
RETURN ON SALES  ROI
                   AFTER-COMPLIANCE ROI
        OPTION 1   OPTION 2   OPTION 3   OPTION  5
      4.0
     25.0
     20.0
     25.0
     15.0
     15.0
     15.0
     20.0
 2.0
35.
27.
41,
36.
36.
28.6
34.9
.7
.3
.7
.4
.4
        1.7
26.8
21.6
40.
34.
34.
28.0
34.7
.6
.6
.6
             0.1
16.8
15.3
39.1
31.8
21.8
23.1
31.7
                     NA
26.8
21.6
40.6
27.9
25.5
24.9
32.9
                      -0.1
16.8
15.3
39.1
31.8
20.5
22.6
31.4
NA = Not applicable.  Treatment option not  selected  by EPA for this product group.
SOURCE:  JRB Associates estimates.

-------
         PART II
SEMICONDUCTOR SUBCATEGORY

-------
                               1.  INTRODUCTION

1.1  PURPOSE

     The purpose of this report is to analyze the economic impacts that may
result from the promulgation of EPA's effluent regulations on the semiconductor
industry subcategory of the Electrical and Electronic Products point source
category.  The effluent standards and limitations which are issued under
Section 301(b) of PL 92-500, the Clean Water Act of 1977, require the Admin-
istrator (EPA) to establish the following:

     1.  Effluent limitations based on the Best Practicable
         Technology Currently Available (BPT), Best Conventional
         Pollutant Control Technology (BCT), or Best Available
         Technology Economically Achievable (BATEA or BAT) to
         be met by industrial dischargers
     2.  Pretreatment standards for existing and new dis-
         chargers (PSES and PSNS) to publicly owned treat-
         ment works (POTWs)
     3.  New Source Performance Standards (NSPS) to be met
         by new source industrial dischargers.

In order to assist in the development of these limitations this study explic-
itly assesses the effects of various pollution abatement regulatory options
upon the costs of production, capacity expansion and replacement, profita-
bility, and the potential for plant closures in the semiconductor industry.
In addition, the likely effects on employment, communities, foreign trade,
and small businesses are estimated.
                                     II-l

-------
1.2  SCOPE

     The semiconductor subcategory as defined for this study includes the
manufacture of electronic components made from elements such as silicon and
germanium that contain small amounts of impurities which make them neither
good electrical conductors nor good insulators.  Among other functions,
semiconductors can amplify, switch, and rectify electronic current.  They
are increasingly used in place of vacuum tubes or electron receiving tubes,
mechanical relays, switches, and ferromagnetic memory coils (in computers).
Semiconductors are used in many electrical and electronic products, such as
radio and television receivers, and in transmitters, amplifiers, calculators,
computers, auto ignitions, telephone switching systems, cameras, and digital
watches.

     The industry can be segmented into four separate product classes.  These
product classes and their associated standard industrial classification (SIC)
codes are shown below.

                  SIC               INDUSTRY PRODUCT CLASS

                 3674         Semiconductor and Related Devices
                 36741        Integrated Microcircuits
                 36742        Transistors
                 36745        Diodes and Rectifiers
                 36749        Semiconductor Devices, n.e.c., and parts

     The semiconductor industry includes both discrete or individual function
devices such as diodes, rectifiers, and transistors; and integrated circuits,
including both monolithic (which may contain hundreds or thousands of tran-
sistors and diodes on a single chip 1/4 inch square) and hybrid integrated
circuits (on which discrete semiconductors have been attached).  Also included
in this industry are such specialized devices as light-emitting diodes (used
                                     II-2

-------
as the display device in many calculators and digital watches) and light
sensitive devices (such as photo diodes).  Moreover, there are thousands of
different kinds of diodes, transistors, and integrated circuits within each
of these groups.

1.3  ORGANIZATION OF PART II OF THIS REPORT

     The remainder of Part II of this report is organized in the following
manner.  Chapter 2 provides the methodology used to analyze the economic
impacts of the regulations.  Chapter 3 describes the semiconductor industry
characteristics and performance in the domestic and international markets.
Chapter 4 provides projections of the growth and performance of the semicon-
ductor industry over the regulatory impact period assuming there are no
additional water pollution control requirements.  In Chapter 5 the costs of
pollution control for the treatment options are presented.  Chapter 6 pre-
sents the results of the economic impact assessment.  Finally, Chapter 7
discusses the limitations of the economic impact analysis.
                                      II-3

-------
                             2.  STUDY METHODOLOGY
2.1  OVERVIEW
     Figure 2-1 shows an overview of the analytical approach used  to  assess
the economic impacts likely to occur as a result of the costs of each  proposed
regulatory option.  For the Semiconductor subcategory, four regulatory options
are considered in the economic impact study.  The basic approach used  in  this
study is to (1) develop an operational description of the  price and output
behavior of the industry, and (2) assess the likely plant-specific responses
to the incurrence of the compliance costs enumerated in Chapter 5.

     The operational description of the price and output behavior  is used, in
conjunction with compliance cost estimates supplied by EPA, to determine  new
post-compliance industry price and production levels for each regulatory
option.  Individual plant data is then analyzed under conditions of the
post-compliance industry price levels, for each regulatory option, to  isolate
those plants whose production costs are projected to change significantly more
than the estimated change in their revenues.  These identified plants  are
subjected to a more detailed financial analysis to determine likely plant
closures.  The industry description is then revised, for each regulatory
option, to incorporate the reduced supply into the analysis.  Finally,  other
effects which flow from the basic price, production, and industry  structure
changes are determined.  These include employment, community, and  foreign
trade impacts.  Specifically, the study proceeds in the following  steps:

      1.  Description of industry characteristics
      2.  Industry supply and demand analysis
      3.  Analysis of cost of compliance estimates
      4.  Plant level profitability analysis
      5.  Plant level capital requirements analysis
      6.  Assessment of plant closure potential
      7.  Assessment of other impacts
      8.  Small business analysis.
                                     II-4

-------
                   EPA POLLUTION
                   CONTROL COSTS
  INDUSTRY
SEGMENTATION

  INDUSTRY
  STRUCTURE

   MARKET
  STRUCTURE

  FINANCIAL
    DATA
   INDUSTRY


MICROECONOMIC
   ANALYSIS
PRICE INCREASE
   ANALYSIS
                                                                    COMMUNITY
                                                                   EMPLOYMENT
                                                                     EFFECTS
IDENTIFICATION
   OF  HIGH-
    IMPACT
   SEGMENTS
                        MODEL
                      FINANCIAL
                      ANALYSIS
                        PLANT
                      CLOSURES
                     FIGURE 2-1.   ECONOMIC ANALYSIS STUDY OVERVIEW

-------
Although each of these steps is described  separately  in  this  section,  it  is
important to realize that  there are  significant  interactions  between them,  as
shown in Figure 2-1.

2.2  STEP 1:  DESCRIPTION  OF INDUSTRY CHARACTERISTICS
     The first step in the analysis  is  to  describe  the basic  industry
characteristics.  These characteristics, which include the  determinants of
demand, market structure,  the degree of  intra-industry competition,  and
financial performance, are described in  Chapter  3 of  this report.

     The sources for this  information include government reports,  trade
association data, discussions with various  trade association  representatives
and individuals associated with the  industry, and an EPA industry  survey  under
Section 308 of the Clean Water Act.

2.3  STEP 2:  SUPPLY-DEMAND ANALYSIS
     The purpose of the supply-demand analysis is to determine  the likely
changes in market prices and industry production levels resulting  from each
regulatory option.  The estimates of post-compliance price  and  output  levels
are used in the plant-level analysis to  determine post-compliance  revenue and
profit levels for specific plants in each  product group.  If  prices  are
successfully raised without significantly  reducing  product  demand  and
companies are able to maintain their current financial status,  the potential
for plant closings will be minimal.  If  prices cannot be raised  to fully
recover compliance costs because of  the  potential for a significant  decline in
product demand or because  of significant infra-industry competition,  the  firms
may attempt to maintain their financial  status by closing higher cost/less
efficient plants.  The supply-demand analysis was divided into  four  basic
components:  description of industry structure, determination of industry
pricing mechanism, projection of possible  changes in industry structure during
the 1980s (when the primary economic impacts of  the regulations  will be felt),
and determination of plant- and firm-specific operational parameters (e.g.,
production costs, profit rates, etc.).
                                  II-6

-------
     Short-run pricing behavior depends  upon  the market  structure  of the
industry, which can range  from competitive,  to  oligopoly,  and  to monopoly
situations.  Many economic  impact  studies  begin by  assuming  perfect
competition.  However, the  product groups  covered  in  this  study exhibit some
characteristics that are indicative of non-competitive pricing mechanisms.

     The perfectly competitive market structure is  one in  which there  are many
buyers and sellers and the  actions of any  one of these do  not  significantly
affect the market.  Firms  in a competitive market generally  earn a "normal"
rate of return on their assets.  If it is  assumed  that (1) the market  for a
competitive good is currently at equilibrium, or will be when  the  regulations
become effective, and (2)  firms will attempt  to maintain their current
financial status by passing through industry-wide cost increases in  the form
of higher prices, the post-compliance equilibrium price and  quantity level can
be derived from the interaction of the elasticities of supply  and  demand.

     Because of the existence of speciality markets,  non-competitive pricing
behavior might be expected.  In this situation, a number of  pricing  schemes
may be used to characterize the behavior of  oligopolistic  firms, depending on
specific industry characteristics.  The  oligopolistic pricing  scheme is
applicable for those industry groups which exhibit  characteristics of
oligopoly markets, such as  the following:

     •  Few firms in the product group
     •  High industry concentration
     •  Low degree of foreign competition
     •  Abnormally high profitability
     •  Low demand elasticities
     •  Highly capital intensive
     •  Large degree of integration of production,  marketing,  and  distribution
     •  Large degree of specialized knowledge.

     Industries which exhibit the  first  three of  these  characteristics are
those in which the pricing  and output actions of  one  firm  will directly affect
those of other firms in the industry.  While  these  conditions  do  not guarantee
                                      II-7

-------
oligopolistic behavior, they are necessary conditions  for  an oligopoly  and
good indicators that one exists.  Abnormally high  profits  in an  industry
would, in time, normally attract new entrants  to the  industry, thereby
increasing price competition.  However, very high  profits  over long  periods  of
time which are not explained by such factors as excess risk, unusual  amounts
of technological innovation, or firm size may  be an indicator that an
imperfect market structure exists.  Such conditions may occur when entry  into
an industry is difficult.  The last three of the above points are indicators
of difficulty of entry into the market.

     Although the domestic semiconductor industry  exhibits  some  of these
characteristics of non-competitive markets, the specific nature  of the  pricing
mechanism is not clear from the available evidence.  That  is, the existence  of
specialty markets, low-demand elasticities, and high levels of specialized
knowledge characteristic of this industry are  indicative of non-competitive
markets.  On the other hand, there are a number of characteristics which  are
indicative of competitive markets, such as the existence of many firms  in the
industry, a significant amount of competition  from foreign producers, and the
fact that barriers to entry into the industry  are not  particularly severe.
Little can be concluded from the four-firm and eight-firm concentration
ratios, because they are in a middle (moderate) range.  For these reasons and
because of the dynamic nature of this new industry, the pricing mechanisms
that dominate the industry cannot be specified with certainty.   Instead,  it
is assumed that semiconductor plants will attempt to absorb all  of the costs
of compliance and that in the short to intermediate period there will be  no
change in semiconductor prices resulting from  the regulations.   This  assump-
tion derives from two observations:
     •  The international competitiveness in the industry has been increasing
        in recent years and domestic manufacturers are quite sensitive to  this
        fact.
     •  If, at any point in time, it were possible to determine profit rates
        attributable to different semiconductor product lines within a given
        plant, there would probably be a wide variation in the profit rates.
                                     II-8

-------
     The U.S. semiconductor manufacturers  operate  in a very competitive world
market.  There is significant  trade  and  competition among the  western
developed nations for market share.   Consequently,  U.S.  producers cannot
unilaterally attempt to pass additional  regulatory  costs through to their
customers without losing  significant  portions  of  the domestic  and inter-
national markets.  If they do  atttempt  to  pass through the  pollution control
costs, the U.S. products  would  become less  competitive,  forcing domestic and
foreign purchasers to buy semiconductor  products  in foreign markets that do
not have these added costs.  As U.S.  products  became more expensive, imports
of semiconductor products into  the U.S.  would  increase,  exports would
decrease, and sales and profitability of U.S.  plants would  significantly
decrease as U.S. producer prices rose.   Since  U.S.  semiconductor producers are
quite sensitive to these  conditions,  their  most likely option  would be to
absorb the regulatory costs  in  the short and  intermediate period and continue
operations if possible.

     The existence of differences in profits  from one product  line to another
within a firm implies the potential  for  price  flexibility.   That is, the most
mature product lines operating  with  narrow profit margins may  have some of
their compliance cost burden borne by the most profitable newer products.  For
this reason, there generally exists  within  the industry the capability for
absorbing the cost of the pollution  control equipment.

     In summary, although the market  structure is not precisely determined, it
has been shown that the industry has  both  the  incentive and the capability to
not raise prices in response to small and  moderate  levels of mandated
pollution control costs.  For  these  reasons,  it is  assumed  that there will be
no price effect resulting from  the proposed regulations in the short and
intermediate time period.

     The remainder of the economic impact  analysis, therefore, focuses on the
extent that plants in the semiconductor  industry  can remain financially viable
and competitive if they are required  to  install pollution control equipment.
The measures of financial viability  that are  considered relate to the plants'
profitability and their ability to raise the  initial investment funds for
                                      II-9

-------
pollution control equipment in addition  to other  capital  equipment  needed  for
expansion, growth, and competitiveness.

     Having described the current industry structure,  it  was  necessary  to
determine if the key parameters would change significantly during  the 1980s.
Projections of industry conditions begin with a demand  forecast.   The demand
during the 1980s is estimated via trend  analysis  and market research analysis.
The projections of industry conditions indicated  that  only minor changes  in
market structure would occur in the base case.  For this  reason, it was
concluded that the market structure previously described  can  be used to
determine price changes due to the regulation.

     The post-compliance market price levels (i.e., zero  price increases due
to the regulation) are used in a later step to assess  the financial condition
of individual semiconductor manufacturing facilities.

2.4  STEP 3:  COST OF COMPLIANCE ESTIMATES
     Investment and annual compliance costs for the recommended treatment
options  1, 2, 3, and 5 were estimated by EPA's Effluent Guidelines  Division
for treatment systems of various selected sizes .   Based on these cost
estimates, compliance cost curves were developed  and then used to  estimate
plant-specific compliance costs.  A description of the  control and  treatment
technologies and the rationale behind these compliance  cost estimates appear
in Chapter 5.

2.5  STEP 4:  PLANT LEVEL PROFITABILITY  ANALYSIS
     The basic measure of financial performance used to assess the  impact  of
the proposed regulations on the profitabilities of individual plants  is  return
on sales (ROS).  The use of this technique involves a  comparison of the  after
compliance ROS with a critical return.
                                      11-10

-------
     The return on sales  is defined  as  the  ratio of annual  profits before
taxes to the total revenues of  a  plant.   The  principal  virtue of this
technique is its simplicity and its  common  use  in comparative analyses of
operating efficiencies of  financial  entities.   Its  principal  shortcomings are
that it is based on accounting  income rather  than cash  flow and  that  it fails
to account for the amount  of  invested capital and the timing  of  cash  flows,
thereby ignoring the  time  value of money.

     The profit impact assessment is determined by  calculating the after-
compliance ROS for each plant and comparing them to a threshold  value which
represents the lowest ROS  which the  industry  typically  considers to be
acceptable to remain  in operation.   Plants  with after-compliance ROS  below the
threshold value are considered  potential  plant  closures.  The underlying
assumption is that plants  cannot  continue to  operate as viable concerns if
they are unable to generate a return on sales that  is at  least equal  to the
threshold value.  Equation (1)  summarizes this  technique:

           ACC
              P
(1)   PM - —— < PM..   Significant  Profit  Impact
           VS   ~—   ij
             P
where
     PM     = the pre-compliance  average  annual profit  margin (ROS) for the
              industry
     PK.     - the lowest ROS at which plants will remain  open
     ACC
        E.   = the ratio of the  total annual compliance  costs  (ACC ) to annual
      "p      value of shipments  (VS )  for  each plant.
     The lowest acceptable PM was determined  to be  2.7  percent.   This figure
was derived from the observation  that it  is the lowest  annual profit  margin
for the semiconductor industry  over  the 1971-1977 period.     The precompliance
profit margin is curved to be 4.5 percent,  which is the average  for the 7-year
period.   Since 2.7 percent is the lowest ROS  recorded by  operating plants, it
is assumed to be a rather  conservative  threshold.
  U.S.  Department of Commerce, U.S. Semiconductor  Industry,  September 1979.
                                     11-11

-------
     The values of shipments  for  the  21  sample  plants  were estimated by
multiplying the industry average  value of  shipments  per production employee
by the number of production employees working  at  the specific sample plant.
The industry average value of  shipments  per  production employee was derived
from the 1977 Census of Manufactures.  The value  of  shipments data along with
the total annual compliance cost  estimates are  used  to analyze the performance
of the plants after compliance.

2.6  STEP 5:  CAPITAL REQUIREMENTS ANALYSIS
     This step  in the analysis assesses  the  ability  of firms to make the
initial capital investment needed to  construct  and  install the required
treatment systems.  Some plants which are  not  initially identified as
potential closures in the profitability  analysis  may encounter problems
raising the amount of capital  required to  install the  necessary treatment
equipment.  The limit on a given  firm's  ability to  raise  capital to finance
investment expenditures is quite  variable, depending upon factors such as the
firm's capital  structure, profitability, future business  prospects, the
industry's business climate,  the  characteristics  of  the financial markets and
the aggregate economy, and the firm management's  relationships with the
financial community.  The precise limit, considering all  these factors, is
ultimately judgmental.  Thus,  even given firm-specific data, a limit on a
firm's ability  (or willingness) to raise funds  for  capital investment would be
difficult to estimate.  Because firm-specific  data  for this study is scarce
the problem becomes even more  difficult.   For  these  reasons, the analysis of
capital availability  focused  upon general  indicators of the relative burden
created by  the  compliance expenditures  for each semiconductor firm.  That is,
to assess the firms'  ability  to commit the capital  necessary to install the
specified pollution control systems,  the following  ratio is calculated for
each sample plant:

        	Compliance  Capital  Investment
           Annual New Plant  and  Equipment  Expenditures
                                      11-12

-------
     This ratio reflects the typical  level  of  capital  expenditures  of the
industry, and is used as a budget  standard  of  the  capital  expenditure level
that firms in the industry generally  hold.   Although a specific  threshold
value for this parameter is difficult to  specify,  a value  that  is  useful for
comparison is the average annual value  of this ratio for  the  entire industry.

     Although this ratio provides  a good  indication of the  relative burden
created by the compliance requirement,  it does not precisely  indicate whether
or not firms can afford to make the investments.   If,  for  example,  the same
investment requirements were placed on  a  firm  which is already highly
leveraged (as indicated by a high  debt/equity  ratio) and a  firm  which is not
leveraged (as indicated by a debt/equity  ratio of  zero), the  highly leveraged
firm is likely to experience the most significant  impact.   In addition,  the
capital requirements must be evaluated  together with other  factors,  such as
profitability.  For example, a plant  that is extremely profitable would
consider the risk of more leverage or increased cost of capital  resulting from
investment more worthwhile than would a less profitable plant.

     Because of the absence of plant-specific  financial data, the annual
capital expenditures of the plants are  inferred from the  industry average new
                                                    2 /
capital expenditures per production employee ratios.    These annual capital
expenditure estimates, along with  capital investment cost  for the  pollution
control options, are used to indicate the significance of  the capital impacts.

2.7  STEP 6:  PLANT CLOSURE ANALYSIS
     The plant-level analysis examined  the  individual  production units in each
product group to determine the potential  for plant closures and  profitability
changes.  The decision to close a  plant,  like  most major  investment decisions,
is ultimately judgmental.  This is because  the decision involves a  wide
21
  These ratios are derived  from  the  1977  Census  of Manufactures.
                                      11-13

-------
variety of considerations, many of which  cannot  be  quantified  or  even
identified.  Some of the most  important  factors  are:

     •  Profitability before and after compliance
     •  Ability to raise capital
     •  Market and technological integration
     •  Market growth rate
     •  Other pending Federal, state, and  local  regulations
     •  Ease of entry into market
     •  Market share
     •  Foreign competition
     •  Substitutability of the product
     •  Existence of specialty markets.

     Many of these factors are highly uncertain, even for the  owners  of  the
plants.  However,  this analysis was structured to make quantitative estimates
of the first two factors, as described above, and to qualitatively consider
the importance of the others.   In this analysis, the first two factors are
given the greatest amount of weight and the importance of the  other factors
varies from plant to plant.

2.8  STEP 7:  OTHER IMPACTS
     "Other impacts" include economic impacts which flow from  the basic  price,
production, and plant-level profitability changes.  These include impacts on
employment, communities, industry structure, and balance of trade.

     The estimate  of employment effects flows directly from the outputs  of the
industry-level analysis  and the plant closure analysis.  The algorithms  used
are:
       direct    =  employment at  +  ( S )/(S/employee)
     employment   closed facilities
      S  = change in revenues at the remaining plants, which is derived  from
           the tnicroeconomic and plant level analyses (in this case S  =0).
     S/employee = baseline revenue produced per employee.
                                     11-14

-------
Employment  estimates  for the production facilities projected to close are
available  from  the  EPA 308 Survey.

     Community  impacts result primarily from employment impacts.   The critical
variable is  the  ratio  of semiconductor industry unemployment to total employ-
ment in the  community.   Data on  community employment  are available through the
Bureau of  the Census  and the Bureau  of Labor Statistics.

     The assessment of industry  structure changes  is  based  on examination of
the  following before  and after compliance with the regulation:

     •  Numbers  of  firms and plants
     •  Industry concentration ratios
     •  Effects  of  plant closures on specialty markets.

     Imports and  exports are important determinants of  pricing  behavior  in the
semiconductors  industry.   The role of  these  variables is  qualitatively
evaluated  in Chapter 3  of this report.   Basically,  the  threat of  imports
appears to deter  domestic producers  from passing through  compliance  costs.

2.9  STEP  8:  SMALL BUSINESS ANALYSIS
     The Regulatory Flexibility  Act  (RFA)  requires  Federal  regulatory agencies
to consider small entities  throughout  the  regulatory  process.   This  analysis
addresses thess  objectives  by identifying  and  evaluating  the economic impacts
that are likely  to result  from the promulgation of BPT, BCT,  BATEA,  NSPS,  PSES
and PSNS regulations on  small business in  the  semiconductor manufacturing
industry.   The primarv economic  variables  covered  are those analyzed in  the
general economic  impact  analysis such  as  plant  financial  performance, plant
closures,  and unemployment  and community  impacts.  Most of  the  information and
analytical  techniques  in  the  small business  analysis  are  drawn  from  the
general economic  impact  analysis which is  described above and in  the remainder
of this report.   The specific conditions  of  small  firms  are evaluated against
the background of general conditions in  semiconductor markets.
                                     11-15

-------
      A specific problem in the methodology was developing an acceptable
 definition of small entities.   The Small Business Administration (SBA) defines
 small entities in SIC 3674 (Semiconductors and Related Devices)  and SIC 3679
 (Electronic Components,  n.e.c.)  as firms of fewer than 500 employees.  The SBA
 definition was found to  be inappropriate as a basis  for defining small
 entities  in the semiconductor  manufacturing industry for purposes of
 developing water pollution regulations.   Instead,  a  definition was  sought
 which would account  for  firm size  in  comparison  to total industry size and in
 comparison to unit  compliance  costs (unit  compliance costs increase signif-
 icantly in reverse  proportion  to plant size).  Moreover, since the  available
 data  on compliance  cost  and production were on a plant  basis,  the individual
 production facility,  rather than the  firm,  was used  as  the basis  for the
 analysis.

      Alternative  definitions for "small"  semiconductor  manufacturing plants
were  selected  for examination.  Number of  production  employees was  the primary
variable used  to  distinguish plant size.   This is  because  plant  level
employment  data were  considered most  reliable.

      The impacts  on  small  plants under each definition  were  assessed by
examining  the  distribution of  semiconductor plants by plant  employee size,
wastewater volumes, compliance costs  and potential closures  from  regulations.
                                     11-16

-------
                          3.   INDUSTRY  DESCRIPTION

     This chapter provides an overview of  the  semiconductor  industry.  It
 describes the economic  and market  characteristics of  the industry and the
 factors that contribute to the growth  of the industry.  Particular attention
 is given to the demand  and end-use markets  for semiconductor products.  Addi-
 tionally, information is provided  on the industry structure, foreign competi-
 tion, and price trends  because these factors determine  the future outlook and
 competitiveness of  the  industry, and producers' ability to afford additional
 capital outlays for pollution control  equipment.

 3.1  MARKET CHARACTERISTICS

 3.1.1  Major Product Groups  and Trends

   In 1982, the value of shipments for all  semiconductor devices in the U.S.
 domestic industry was $9.5 billion (Table  3-1).  This represented an increase
 of 5 percent over the 1981 level of shipments.  The economic downturn over
 1981-1982 was responsible for the  slow growth  over this period.  Historically,
 this industry lias experienced very high rates  of growth.  During the 1970s,
 the semiconductor industry grew at a compound  annual  rate of about 19 percent,
 which was substantially higher than the 12  percent annual growth experienced
 during the 1960s.   Moreover,  for some  segments  such as  integrated circuits,
 shipments have grown by about 18.3 percent  annually,  over the  1972-1982 period.
 This industry segment's  rapid growth,  however,  has been at the expense of
 other semiconductor devices,  such  as discrete  resistors and  capacitators,
 which it has displaced  in many types of equipment.

     The recent penetration  of  integrated circuits into the  semiconductor
market is reflected in  the fact that while  integrated circuits accounted for
 54 percent of the domestic market  share in  1972, they represented 72 percent
by 1982.   In contrast,  the discrete devices segment of  the market has been
                                     11-17

-------
      TABLE 3-1.  VALUE OF SHIPMENTS OF THE U.S. SEMICONDUCTOR INDUSTRY
                                 1972 - 1982
                                 ($ Millions)
YEAR
1972
1977
1979
1980
1981a/
1982a/
SEMICONDUCTOR PRODUCTS
DISCRETE DEVICES
1,093
1,835
2,696
2,687
2,576
2,719
INTEGRATED CIRCUITS
1,268
2,697
4,660
6,768
6,500
6,820
TOTAL SEMICONDUCTORS
2,361
4,532
7,356
9,455
9,076
9,539
a/ Estimated.
SOURCE:  U.S. Department of Commerce,  U.S.  Industrial Outlook 1983.

-------
increasing at a much slower rate and has been losing its market share over the
years.  In 1972 the market share of this industry segment was 46 percent, but
by 1982 it had dropped to 28 percent of total industry shipments.  If new appli-
cations for discrete devices, such as those used in automotive engines, had not
been introduced, the decreases in market share for this industry segment could
have been even greater.

3.1.2  End-Use Markets
     Since the invention of semiconductor devices in the late 1950s, their
uses have pervaded almost every sector of our society.  The principal end-uses
of semiconductor devices can be grouped into four major market categories:

     •  Computer
     •  Industrial
     •  Consumer
     •  Government, including military/space.

The growth of the semiconductor industry depends on the performance of each of
these end-use markets.  The semiconductor industry end-use markets, with few
exceptions, have been outperforming the general economy in recent years.  The
relative strength of these end-use markets is derived from the integrated and
pervasive role which electronic components (specifically semiconductors) play
in all sectors of the economy.  These end-use markets are expected to continue
to grow in the future.

     The relative growth of these four markets is shown in Table 3-2.  Each of
these markets has experienced different historical development patterns due to
the relative trade-offs between performance, reliability, and cost of semicon-
ductors in product applications.^'  For example, component costs are far more
important in highly competitive consumer product industries than in other mar-
kets.  Computer and industrial equipment manufacturers require high-performance
   U.S. Department of Commerce, A Report of the U.S. Semiconductor Industry,
   September 1979.

                                     11-19

-------
       TABLE 3-2.   DISTRIBUTION OF SEMICONDUCTOR PRODUCTS BY MAJOR END-USERS
                                   (1973 - 1975)
End-Users
Computer
Industrial
Consumer
Government, inc.
military/ space
TOTAL
1973
% of
Total
1974
% of
Total
1975
% of
Total
($mil.) ($mil.) ($mil.)
$590
605
510
378
$2,083
28
29
25
18
100
$855
775
575
420
$2,625
32
30
22
16
100
$1,005
905
715
450
$3,075
33
29
23
15
100
Percent
72/73
Change
73/74 74/75

66
54
72
26
55
45
28
13
11
26
18
17
24
7
17
SOURCE:   Fairchild Camera and Instrument Corp.,  cited in "Coleman and Company
         Electronics Letter," B.  M.  Rosen,  March 8,  1974.
                                       11-20

-------
and reliable components, and price is of lesser importance than in consumer
products.  The military/space market has very high performance and reliability
requirements and is willing to pay for the development of premium products.
Semiconductor devices used in military and space applications must be capable
of operating under extreme weather and other adverse conditions.  This market
also requires specialized, high-performance semiconductors, such as infrared
detectors, radiation detection devices, and light-sensing devices, which have
very limited civilian applications.  Because they are spread over a small pro-
duction volume, unit development and production costs are high.

3.2  INDUSTRY STRUCTURE

3.2.1  Plant Characteristics

     The 1977 Census of Manufacturers reported 545 establishments whose primary
activity is the manufacture of semiconductors.^'  In that year, these plants
shipped over $5.3 billion worth of products.  Table 3-3 provides information
on the distribution of plants in the industries by their employee size classi-
fications and value of shipments.  This table indicates that the industry is
composed of a disproportionately large number of very small and medium-sized
plants, and the number of these plants has been increasing at a faster rate
than large plants.  However, the larger plants are responsible for most of the
semiconductor products shipped.  For example, plants that had over 500 employees
shipped 79 percent of the products in 1977, but they represented only 7.5 per-
cent of the plants in the total industry.  The information shown in the table
also indicates that the plants in the semiconductor industry are highly
specialized.  That is, approximately 94 percent of their product shipments
were semiconductor devices in 1977.  Furthermore, the average coverage ratio
for the industry is high.  In 1977, 545 plants in the industry manufactured 93
percent of the products classified under SIC 3674.
   A significant number of these plants, especially the smaller ones, may be
   R&D and other facilities that are not necessarily involved in wafer manufac-
   turing and are not covered by the proposed regulation.
                                     11-21

-------
      TABLE 3-3.  DISTRIBUTION OF SEMICONDUCTOR PLANTS.AND VALUE OF SHIPMENTS
                                  BY SIZE OF PLANT
                                  (Selected Years)


PLANT SIZE
Establishments with:
1-99 employees13/
100 - 499 employees
500 - 999 employees
Over 1,000 employees
Total
Coverage Ratioc'
Specialization Ratio^/
SEMICONDUCTOR INDUSTRY (SIC 3674)

PLANTS
(No.)
1972 1977
225 414
65 90
17 19
18 22
325 545
a/
PLANTS
(Percentage)
1972 1977
69.0 76.0
20.0 16.5
5.2 3.5
5.5 4.0
100.0 100.0
VALUE OF
SHIPMENTS
(Millions)
1972 1977
263.3
860.0
529.5
3696.8
5349.6
VALUE OF a/
SHIPMENTS
(Percentage)
1972 1977
4.9 4.4
14.4 16.2
10.4 10.0
70.3 69.4
100.0 100.0
92.0 94.0
89.0 93.0
a'  Percentage of total sampled plants within each employment category of the total
   plants.


"I  A number of these plants, especially the smaller ones, may be R&D and other
   facilities that are not necessarily involved in wafer manufacturing.


c/  Coverage ratio is the ratio of a given industry's primary product shipments to
   total shipments of this product by all industries.


"'  Specialization ratio is the ratio of primary product shipments to total
   primary and secondary product shipments.
SOURCE:  U.S. Department of Commerce, Bureau of Census, Census of Manufactures,
         1977, Industry Series: Electronic Components and Accessories, MC77-1-36E
         (Washington, D.C. 1980).
                                      11-22

-------
     As shown in Table 3-4, the plants of the U.S. domestic semiconductor indus-
try are concentrated in a few states.  In 1977, approximately a third of the
plants were located in California, and over two-thirds were located in six
states: California, New York, Massachusetts, New Jersey, Pennsylvania, and
Texas.  Most of the largest semiconductor companies have several domestic
plants (usually in the same region of the country) and several foreign plants.
In many cases, increases in the number of plants in a particular firm have been
the result of growth in the industry's market and limitations in adjacent land
needed to enlarge the size of existing plants.  This phenomenon is especially
true in the northern California area where the scarcity of large industrial
tracts has influenced companies to build many plants on different sites.  In
some cases semiconductor companies have also constructed plants that specialize
in the production of specific products.  For example, a semiconductor company
may have an integrated circuit plant and a separate transistor and diode plant,
or separate MOS and bipolar integrated circuit plants.

3.2.2  Industry Concentration

     The semiconductor industry has shown a consistent level of market share
concentration over time.  In 1972, the four largest manufacturers of semi-
conductors provided half of the industry's total value of shipments, only 1
percent lower than in 1957.  Table 3-5 indicates that firms that produce
transistors are more highly concentrated than firms in other semiconductor
product groups.  In 1972, the concentration ratio for the largest four com-
panies producing transistor products was 68 percent, as opposed to 57 percent
for firms in the integrated circuits, and 45 percent for those in the diodes
and rectifiers product groups.

3.3  FOREIGN TRADE

     The market for semiconductors is worldwide.  The United States is a
significant exporter and importer of semiconductor devices.  It is important
to note that most large and many of the smaller U.S. semiconductor companies
conduct the design and wafer fabrication processes in the United States and
                                    11-23

-------
TABLE 3-4.  GEOGRAPHICAL DISTRIBUTION OF DOMESTIC SEMICONDUCTOR PLANTS IN 1977a/
STATE
New England Division:
Maine
Vermont
Massachusetts
Rhode Island
Connecticut
Middle Atlantic Division:
New York
New Jersey
Pennsylvania
East North Central Division:
Ohio
Indiana
West North Central Division:
Minnesota
Missouri
South Atlantic Division:
Florida
East South Central Division:
Tennessee
West South Central Division:
Oklahoma
Texas
Mountain Division:
Idaho
Colorado
Arizona
Utah
Pacific Divison:
Oregon
California
Total U.S.
NUMBER OF
ESTABLISHMENTS

3
2
46
4
13

59
34
31

13
5

3
4

19

1

3
36

1
8
16
3

4
180
545
NUMBER OF
ESTABLISHMENTS
WITH 20 OR MORE
EMPLOYEES

1
1
23
2
6

22
11
16

5
2

1
1

9

1

1
16

1
3
9
1

2
80
219
a' The totals do not add up because of lack of state-specific information.

SOURCE:  U.S. Bureau of the Census, Census of Manufactures, 1977, Industry Series:
         Electronic Components and Accessories, MC77-1-36E (Washington, D.C., 1980).
                                    11-24

-------
  TABLE 3-5.  CONCENTRATION RATIOS OF U.S. DOMESTIC SEMICONDUCTOR SHIPMENTS
                             1957, 1965, AND 1972
                 (PERCENT OF TOTAL U.S. DOMESTIC SHIPMENTS)*/
                                           PERCENT OF TOTAL U.S. SHIPMENTS
 NUMBER OF COMPANIES                       1957          1965         1972

 All semiconductors
   4 largest companies                       51            50           50
   8 largest companies                       71            77           66
   20 largest companies                      97            90           81
   50 largest companies                     100            96           96

 Integrated circuits
   4 largest companies                                     69           57
   8 largest companies                                     91           73
   20 largest companies                                    99           91
   50 largest companies                                   100          100

 Transistors
   4 largest companies                                     51           68
   8 largest companies                                     83           84
   20 largest companies                                    95           98
   50 largest companies                                   100          100

 Diodes and Rectifiers
   4 largest companies                                     38           45
   8 largest companies                                     56           61
   20 largest companies                                    85           88
   50 largest companies                                    97          100
a'  Data include value of production in captive facilities.

SOURCE:  Compiled from data collected in the Department of Commerce, Quarterly
         Survey of Production Capabilities for Electronic Parts.
                                    11-25

-------
then ship the chips abroad to Southeast Asia or Mexico for wire bonding and
assembly.  After assembly, the devices are imported back into the United
States subject to special U.S. tariff provisions.  These tariff provisions
specify that goods shipped abroad for assembly and then shipped back to the
United States shall be subject to duty only on the value added abroad.

     Japan is one of our most important trading partners for finished semicon-
ductors, particularly for integrated circuits.  Its growing equipment and com-
ponents industries have become a major market for U.S. semiconductors and also
a major supplier of semiconductor containing products to this country.

     According to the Department of Commerce, the U.S. exports of semiconductor
devices rose 5.1 percent in 1982, to $3.65 billion.  Imports increased 7.3
percent, to $3.81 billion.  During the 1972-1982 period, exports grew at a
compound annual rate of 22.8 percent and imports at a rate of 27.7 percent.
Exports as a proportion of product shipments rose from 19.9 percent to 38.8
percent, while imports rose from 12.3 percent of new supply to 28.6 percent.
The import growth reflects an increase in U.S. offshore operations and compe-
tition from Japan, particularly in the random access memory (RAM) segment
of the semiconductor industry.

     In recent times, the Japanese have made significant inroads into the high-
quality and technologically advanced markets of the semiconductor industry.
They have already captured 40 percent of the market for the 16K chip, and are
considered the leaders in the 64K RAM market.5/  Key factors that contributed
to the Japanese technological advancements in these areas, which were tradi-
tionally dominated by U.S. firms, are the Japanese's high level of manufac-
turing discipline, team effort, and long-range planning capabilities.  If
these trends continue, the balance of trade position for semiconductor products
will deteriorate as U.S. electronic manufacturers use increasingly more of the
Japanese low-priced, high-quality semiconductor products.
5/
   Fortune,  Japan's Ominous Chip Victory, December 14, 1981, pp. 52-57-
                                    11-26

-------
3.4  PRICING BEHAVIOR AND TRENDS


     Since their introduction in the marketplace, there has been a downward

trend in the prices of semiconductor devices.  This downward trend is illus-
trated in Table 3-6, which shows price indexes for major groups of semicon-

ductor products over the 1975-1980 period.  This table indicates that the

prices of three of the four product groups declined over this period.  The
only price increases are in the discrete segment of the industry (diodes and
rectifiers) which represents a much slower growing and mature segment of the
semiconductor industry.


     There are several reasons for the decline in prices of semiconductor
devices over the years.  Some of the major factors contributing to this
phenomenon are the following:


     •  The life cycle of most semiconductor products is short.
        The products pass from sales of prototypes through market
        expansion to a technologically mature product in only a
        few years.  Price reductions are generally very rapid
        during the market expansion phase.  The prices tend to
        fall most rapidly on those new devices for which there has
        been new entry by several firms or for new devices which
        can be easily imitated.

     •  Competition and trade in both the domestic and international
        markets are strong.  As a result, high profits on new products
        generally lead to new entry by firms that, in turn, cause
        prices to fall.  Semiconductor firms also recognize that large
        price drops may slow or prevent new firms from entering the
        industry, and they use price cuts to retain or increase their
        market share.

     •  In addition, the learning or experience curve appears to
        have a strong influence on the downward trend in prices
        of this industry.  The experience curve theory implies
        that as the cumulative output increases by a fixed per-
        centage, prices also tend to fall by some fixed percentage.
        An example of the learning curve for integrated circuits
        from 1964 through 1975 is depicted in Figure 3-1.  This
        figure indicates that for each doubling of cumulative unit
        output, the selling price in constant dollars has decreased
        by about 27 percent.
                                    11-27

-------
             TABLE 3-6.   PRICE INDEXES FOR SEMICONDUCTOR DEVICES
                                (1975 - 1980)


YEAR
1975
1976
1977
1978
1979
1980

INTEGRATED
MICRO CIRCUITS
(36741)a/
100.0
93.3
82.5
69.3
65.8
71.1


TRANSISTORS
(36742)a/
100.0
97.8
94.0
91.7
90.8
98.2

DIODES AND
RECTIFIERS
(36743)a/
100.0
102.3
103.4
101.6
101.4
102.2
OTHER SEMI-
CONDUCTOR
DEVICES
(36749)b/
-
100.0
90.9
85.8
85.8
86.1
a/ 1975 = base year

b/ 1976 = base year
SOURCE:  Bureau of Labor Statistics, Producer Prices and Price Indexes
         (selected years).
                                    11-28

-------
      5100.00
        10.00 -
                                                     (rrom 1254 '..".roujn 19?:, eac.n
                                                      doubling of cumulative ^
                                                      results in i 27.5 * dsciine in
                                                      constant dollar average ;ric:.)
                                             100
                                      Cumulative Unit Volume
                                       (millions of circuits)
1,000
10,000
                    Integrated  Circuit  Learning  Curve Data
Year
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
Annual
2
10
29
69
133
253
299
362
602
1,060
1,477
1,375'
Cumulative
2
12
41
110
243
496
795
1,156
1,759
2,319
4,296
5,571
                                                         Averase Sallir.z  Price
        Millions of  I.C. Units    Annual  Value
                                    (Saillior.s)
                                          41
                                          7?
                                         149
                                         22B
                                         303
                                         413
                                         433
                                         443
                                         608
                                      1,049
                                      1,359
                                      1,100
Scurca:   "Morgan Stanley Zlectrcr.ics latter"  October  15,  1976,
Current
Dollars
513.50
8.33
5.05
3.32
2.28
1.63
1.45
1.23
1.01
0.99
0.92
0.30
Constant '72
Dollars
524.34
11.21
6.58
4.20
2.76
1.88
1.59
1.2S
1.01
0.?4
0.7?
0.52
                  FIGURE  3-1.   INTEGRATED  CIRCUIT LEARNING CURVE
                                        11-29

-------
     It should be noted that these factors may impede the industry's ability or
inclination to pass through the costs of pollution control to their customers
by raising prices.  The extent to which some of the costs of pollution control
could be passed through to customers may be observed in the form of slower
declines in the prices of semiconductor products.
                                    11-30

-------
                          4.  BASE CASE PROJECTIONS

     This section provides projections of the growth and performance of the U.S.
semiconductor industry over the regulatory impact analysis period in the absence
of water pollution control requirements resulting from the Clean Water Act.  The
baseline projections in this report provide a general point of reference for the
analysis and are not intended to be a comprehensive, authoritative forecast of
future industry conditions.  When the impact estimates (based on the estimated
compliance costs) are compared against the baseline, the difference represents
the estimated impact resulting from these regulations.  The primary variables of
interest are the projected sales, employment, and capital requirements.

4.1  U.S. SEMICONDUCTOR SALES PROJECTIONS

     Growth in the U.S. and foreign markets for semiconductor products has
been slow during the 1981-1982 period.  The recession depressed the end-use
markets causing shipments of semiconductor products to remain low over this
period.  In 1981 the value of shipments fell by about 4 percent from $9.4
billion to $9.1 billion.  However, the industry shipments grew by about 5 per-
cent in 1982 as lower prices stimulated the demand for semiconductor products.

     The demand for semiconductor products is expected to increase rapidly as
economic activity in the United States and rest of the world expands.  According
to the Department of Commerce, the value of shipments of semiconductor products
will increase to $11.3 billion in 1983, or by 18.4 percent over the 1982
level.6/  Between 1982 and 1987, industry shipments of semiconductor devices,
are projected to grow at an annual rate of 16 percent.  Shipments by 1987
are expected to reach $27 billion (Table 4-1).7/
6/ U.S. Department of Commerce, U.S. Industrial Outlook 1983.
7/ Ibid.
                                    11-31

-------
        TABLE 4-1  U.S. SEMICONDUCTOR SHIPMENT FORECASTS ($ MILLIONS)
YEAR
1979
1980
1981
1982a/
1983b/
1987b/
VALUE OF
SHIPMENTS
7,356
9,455
9,076
9,539
11,294
27,000
a/ Estimated




b/ Forecast




SOURCE:  U.S. Department of Commerce, U.S.  Industrial Outlook 1983.
                                    11-32

-------
     Growth in the semiconductor industry will be fueled by continued advances
in both product and process technology.  Many of the new devices, at the
leading edge of technology only a few years ago, will find broad application
by equipment producers.  Devices such as microprocessors, high speed memories,
and integrated power semiconductors, will be a part in a wide range of com-
mercial and consumer electronic products.

4.2  CAPITAL EXPENDITURES PROJECTIONS

     Capital investment requirements are generally high in the U.S. semicon-
ductor industry compared to other industries.  During the 1973-1977 historical
period, the capital expenditures in this industry increased approximately
eightfold from $52.4 to $409 million, and ranged from 6 percent to 12 percent
of the industry's value of shipments.  The high level of capital expenditures
over this period is due mainly to rapid changes in both product and production
technology which results in a high rate of equipment obsolescence.  In order
to maintain market share and growth semiconductor firms have had to invest
continually in new equipment and technology.  This trend is expected to con-
tinue over the regulatory impact period.

     Capital expenditures are projected based on the average 1963-1977 his-
torical capital expenditures/value of shipment ratio which equals  .083.  Using
this historical relationship the estimate for new capital expenditures require-
ments would be $0.9 billion in 1983, and $2.2 billion, in 1987 (Table 4-2).

4.3  EMPLOYMENT PROJECTIONS

     In the past, employment in the semiconductor industry has been increasing
at a slower rate than the value of shipments.  Table 4-3 shows that between
1972 and 1977, employment increased by 17 percent, while value of  shipments
nearly doubled.  The lower employment growth rate was due to improved produc-
tivity which was a result of a combination of factors, including technological
advancement and increasing transfer of assembly and other production activities
                                    11-33

-------
        TABLE 4-2.  NEW CAPITAL EXPENDITURES FOR THE SEMICONDUCTOR
                 INDUSTRY, HISTORICAL3/ AND FORECASTb/ VALUES
YEAR
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1980
1983
1987
EXPENDITURES
($ MILLIONS)
52.4
43.4
61.3
123.6
131.1
118.4
143.6
98.4
94.3
170.3
334.6
479.2
282.9
362.3
409.0
784.1
931.4
2,241.0
PERCENT OF VALUE OF
INDUSTRY SHIPMENTS
7.6
6.0
6.7
11.0
11.5
9.0
9.1
6.3
5.9
6.3
9.2
11.1
8.6
8.1
8.3
8.3
8.3
8.3
a'  Historical data on capital expenditures obtained from the U.S. Department
   of Commerce, Annual Survey of Manufactures.

"'  Projection based on the average historical capital expenditure to value of
   shipments relationship.
                                    11-34

-------
           TABLE 4-3.  SEMICONDUCTOR INDUSTRY EMPLOYMENT FORECASTS
YEAR
1972
1973
1974
1975
1976
1977
1980P
1983P
1987P
EMPLOYMENT
(000)
97.6
120.0
133.1
96.7
102.7
114.0
149.2
197.4
280.1
VALUE OF
SHIPMENTS
($ Millions)
2,704.8
3,647.7
4,305.1
3,276.9
4,473.8
5,322.6
9,455
11,294
27,000
VALUE OF SHIPMENTS/
EMPLOYMENT
($000/Employee)
27.7
30.4
32.3
33.9
43.6
46.7
57. 2a/
68. 8a/
96.1a/
P = Projected




a/ Projected based on 1972-1977 historical trend




SOURCE:  U.S. Department of Commerce, 1977 Census of Manufactures.
                                    11-35

-------
to U.S.-owned offshore facilities to take advantage of lower labor costs.  It
is expected that these productivity increases, measured in terms of value of
shipment per employee, will continue as demonstrated in the past.  The ratio
of value of shipments to employees was projected over the regulatory impact
period based on the historical trend of these variables.  Using the forecasts
of value of shipments (calculated above) and the time trend projections of
the employee/value of shipment ratios, the level of employment in the semicon-
ductor industry was calculated for this 1980-1987 period.  These projections
are also shown in Table 2-3.  Total employment is projected to be 197,400 in
1983 and as high as 280,100 by 1987.
                                    11-36

-------
                           5.  COST OF COMPLIANCE

5.1  OVERVIEW

     The water treatment control systems, costs, and effluent limitations
recommended for the semiconductor industry are enumerated in the Development
Document for Effluent Limitation Guidelines and Standards of Performance for
the Electrical and Electronic Components Point Source Category.  That document
identifies various characteristics of the industry including manufacturing
processes, products manufactured, volume of output, raw waste characteristics,
supply, volume and discharge destination of water used in the production pro-
cesses, sources of waste and wastewaters, and the constituents of wastewaters.
Using that data, pollutant parameters requiring effluent limitations or stand-
ards of performance were selected by EPA, and the costs of the treatment con-
trol systems to achieve certain standards of performance were estimated.

     The EPA Development Document identifies and assesses the range of con-
trol and treatment technologies which apply to the semiconductor industry's
effluent.  This assessment involved an evaluation of both in-plant and end-
of-pipe technologies that could be designed for this subcategory.  This
information was then evaluated for existing direct industrial dischargers to
determine the effluent limitations achievable based on the "best practicable
control technology currently available" (BPT), and the "best available tech-
nology economically achievable" (BAT).  Similar evaluations were performed
for new direct dischargers to develop new source performance standards (NSPS).
Finally, pretreatment standards for existing sources (PSES) and pretreatment
standards for new sources (PSNS) were developed for dischargers to publicly
owned treatment works (POTW).  Each of the technologies identified was
analyzed to calculate cost and peformance.  Cost data was expressed in terms
of investment, operating and maintenance costs, depreciation, and interest
expense.  Pollution characteristics were expressed in terras of median and
mean concentration levels (per liter of water).
                                     11-37

-------
 5.2  POLLUTANT PARAMETERS

 5.2.1   Pollution Parameters Analyzed

     The selection of pollution parameters to be considered  for  regulation was
 based primarily on laboratory analyses of wastewaters  sampled  from  several semi-
 conductor plants and responses to a mail survey of semiconductor manufacturers.
 The specific approach to selecting the pollutant parameters  is presented in
 Sections 5 and 6 of the Development Document.  The chemical  pollutants analyzed
 fall into three groups:

     •  Conventional
     •  Toxics
     •  Nonconventional.

 Conventional pollutants are BOD, TSS, oil and grease,  and pH.  The  toxic
 pollutants are comprised of 129 chemicals that are identified  as priority
 pollutants.  The nonconventional pollutants are those  which  are neither
 conventional nor are toxic pollutants.

 5.2.2   Pollutants to be Regulated

     The specific pollutants selected for regulation in the  semiconductor sub-
 category are pH, fluoride, and total toxic organics (TTO).   These pollutants
 are commonly generated during the manufacture of semiconductor products.
 See the Final Development Document for a description of the  sources and levels
 of concentration of these pollutants in the wastewater.

 5.3  RECOMMENDED TREATMENT TECHNOLOGIES

     Based on the analysis of the potential pollution  parameters and treatment-
in-place in the semiconductor industry. EPA identified six treatment technolo-
gies that are most applicable for the reduction of the selected pollutants.
                                    11-38

-------
These treatment technologies are described in detail in Section 7 of the
Development Document and are listed below.
     •  Option 1:
     •  Option 2:


     •  Option 3:


     •  Option 4:

     •  Option 5:

     •  Option 6:
Segregation and collection of spent solvents
containing toxic organics for reuse, resale,
or contract hauling (referred to as solvent
management), plus end-of-pipe treatment for
pH control.
Option 1 plus end-of-pipe precipitation/
clarification for control of fluoride, arsenic,
and suspended solids.
Option 1 plus precipitation/clarification of
concentrated fluoride wastes for control of
fluoride.
Option 2 plus recycle of treated effluent
to further reduce pollutant discharges.
Option 2 plus filtration to further reduce
pollutant discharges.
Option 5 plus activated carbon to further
reduce toxic organics.
     The Agency's evaluation of treatment options 4 and 6 concluded that these
technologies would not be technically feasible for all semiconductor plants
in the case of 4, and would remove little, if any, additional pollutants in
the case of 6.  For this reason, the economic impact analysis concentrated
on treatment options 1, 2, 3, and 5 only.

5.4  TREATMENT COST ESTIMATES

     Costs of compliance were estimated by EPA for treatment options 1, 2, 3,
and 5.  Treatment costs of option 1 are only for monitoring.  Information
available indicates that most semiconductor plants are practicing solvent
management to some degree to control toxic organics, and that the incremental
costs of disposal tend to be offset by resale (an estimated 53 percent of the
facilities already meet the TTO limit).  Moreover, all of these plants are
controlling the pH of the discharges by end-of-pipe neutralization.
                                    11-39

-------
     The monitoring costs for Option 1 are expected to apply to no more than
approximately fifty percent of the plants in this subcategory because the

regulation provides for a certification alternative to monitoring.  EPA is

unable to predict precisely which plants will incur monitoring costs; there-
fore, the economic impact was assessed for each plant.  However, for purposes
of estimating total compliance costs for this regulation, only half of
these costs are included in the total since only half of the plants at most

are expected to incur costs.


     Appendix A-l contains a sensitivity analysis on Option 1 costs and
e conomi c impa c t s.


     In developing the compliance cost estimates for the treatment options,
the following major assumptions were made by EPA:


     •  All costs are expressed in end-of-year 1979 dollars.

     •  The treatment facilities were assumed to operate 8 hours
        per day, 260 days per year for plants with discharge less
        than 60,000 gpd; 24 hours per day. 260 days per year for
        plants with 60,000 gpd to 200,000 gpd; and 24 hours per day,
        350 days per year for plants with more than 200,000 gpd.

     •  Labor costs are based on an hourly rate of $20, including
        fringe benefits and plant overhead.

     •  The cost of land is valued at $12,000 per acre.

     •  Energy costs are based on $306 per horsepower.

     •  Sludge disosal costs are included.  Available information
        indicates that the sludge generated from the treatment of
        fluoride is not defined as hazardous waste by RCRA.

     •  Capital costs are amortized at 5 years and 13 percent
        interest.

     •  The capital investment and total annual costs associated
        with quarterly monitoring the wastewater of a plant are
        $2,000 and $4,100, respectively.8/
   Total annual costs of monitoring include $3,500 annual operating and
   maintenance costs, and $600 annual capital costs.

                                    11-40

-------
5.4.1  Existing Sources

     Based on the above assumptions, capital investment costs and annual
compliance costs were estimated for treatment options 2, 3, and 5.  The
Final Development Document presents the cost curves for these three options.
These cost curves were used to estimate the compliance costs for 21 semicon-
ductor plants.  Key operating characteristics for the 21 plants are presented
in Table 5-1, and the average investment and annual costs for pollution con-
trol are presented in Table 5-2.  The five model plant size groupings were
based on the plant sizes defined in Table 3-2 as well as wastewater flow
rates and other data developed from a technical EPA survey of the industry.

     Using the average costs for the various pollution control options by
employee size classification in Table 5-2, the compliance costs for the
total industry were calculated and are shown in Table 5-3.^/

5.4.2  New Sources
     The basis for the new source standards (NSPS and PSNS) is the best avail-
able demonstrated technology (BDT), including in-plant controls and end-of-
pipe treatment technologies that provide the maximum pollution reductions
feasible.  For the semiconductor subcategory, the proposed NSPS and PSNS
discharge limitations are the same as those proposed for BAT and PSES.  Since
new source costs are defined as incremental costs from BAT and PSES, costs of
NSPS and PSNS are zero.
   The average investment and annual compliance costs were multiplied by the
   total number of plants in each size classification to determine the  total
   compliance cost by size classifications.  The results were  then summed  to
   determine the total industry compliance costs.  Note that the number of
   direct plants incurring costs for Option 3 is different because of treatment-
   in-place.
                                    11-41

-------
           TABLE 5-1.  SEMICONDUCTOR PLANT OPERATING STATISTICS
PLANT
ID #
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
NUMBER OF
PRODUCTION
EMPLOYEES
4,650
2,395
40
353
600
3,600
2,100
733
1,800
2,500
800
60
300
150
50
300
1,150
33
12,450
60
595
TOTAL
FLOW RATE (GPD)
(OOO's)
2,494
540
43
360
290
2,013
408
280
432
3,026
663
50
324
2
3
14
199
151
792
35
41
FLOW RATE
CONCENTRATED
ACID WASTE
6,290
1,340
120
912
700
5,090
1,030
696
1,080
7,560
1,660
120
65
24
4
24
504
384
1,990
720
98
FOR
FLUORIDE
(GPD)





















SOURCE:   The EPA Effluent Guidelines Division.
                                  11-42

-------
                                        TABLE 5-2.  MODEL PLANT COMPLIANCE COSTS
                                               (thousands of 1979 dollars)

SIZE OF PLANTS
Small (1-99
Employees)
Medium-Small
(100-499
Employees)
Medium-Large
(500-999
Employees)
Large (Over
1,000
Employees)
NUMBER
OF SAMPLE
PLANTS
5
4
4
8
AVERAGE
FLOW RATE
(GPD)
56,772
174,450
318,325
1,238,224
AVERAGE COMPLIANCE COST PER PLANT ($000)
a/
OPTION 1
CAPITAL ANNUAL
INVESTMENT COSTS
2.0 4.1
2.0 4.1
2.0 4.1
2.0 4.1
OPTION 2
CAPITAL ANNUAL
INVESTMENT COSTS
245 180
405.1 305
504.4 396
891.6 698
OPTION 3
CAPITAL ANNUAL
INVESTMENT COSTS
63.5 38
68.05 49
75.61 59
116.64 94
OPTION 5
CAPITAL ANNUAL
INVESTMENT COSTS
280.2 197
458.7 333
583.8 427
1,002 764
a/ The costs associated with Option 1 include only monitoring costs.  The costs associated with solvent disposal
   are negligible.
SOURCE:  Compiled by JRB Associates,

-------
                                     TABLE 5-3.  TOTAL INDUSTRY COMPLIANCE COSTS
                                           (thousands of 1979 dollars)

SIZE OF PLANTS
Small (1-99
Employees)
Medium- Smal 1
(100-499
Employees)
Medium-Large
(500-999
Employees)
Large (Over
1,000
Employees)
TOTAL
ESTIMATED
NUMBER OF
PLANTS
Direct 41
Indirect 95
Subtotal 136
Direct 25
Indirect 58
Subtotal 83
Direct 5
Indirect 13
Subtotal 18
Direct 6
Indirect 14
Subtotal 20
Direct 77
Indirect 180
Total 257
TOTAL COMPLIANCE COST
OPTION 1*
CAPITAL ANNUAL
INVESTMENT COSTS
41 84
95 195
136 279
25 51
58 119
83 170
5 10
13 27
18 37
6 12
14 29
20 41
77 158
180 369
257 527
OPTION 2
CAPITAL ANNUAL
INVESTMENT COSTS
10,041 7,295
23,270 16,904
33,313 24,199
10,104 7,585
23,441 17,597
33,545 25,182
2,575 1,971
6,700 5,121
9,275 7,092
5,344 4,177
12,468 9,745
17,812 13,922
28,078 21,126
65,870 49,368
93,947 70,394
OPTION 3**
CAPITAL ANNUAL
INVESTMENT COSTS
l,674a/ l,000a/
5,938 3,417
7,612 4,417
1.1313/ 809a/
3,889 2,703
5,020 3,512
222a/ 119a/
970 752
1,192 931
577a/ 465a/
1,619 1,294
2,196 1,759
3.6043/ 2,452a/
12,147 8,167
15,751 10,619
OPTION 5
CAPITAL ANNUAL
INVESTMENT COSTS
11,447 7,991
26,522 18,516
37,969 26,507
11,444 8,285
26,549 19,218
37,993 27,503
2,914 2,124
7,577 5,519
10,491 7,643
6,006 4,570
14,014 10,662
20,020 15,232
31,811 22,968
74,662 53,916
106,473 76,884
 a' Incremental costs for Option 3 from Option 1 apply to the following numbers of plants:  small 27, medium-
    small 17, medium-large 3, large 5 = 52.  See text for explanation of why this distribution is different.

 *  Selected Option for Indirect Dischargers Costs are computed assuming that 50% of all plants will monitor;
    the remaining plants are assumed to certify under the provisions in the regulation.

**  Selected Option for Direct Dischargers.

SOURCE:  Compiled by JRB Associates from data provided by the Effluent Guidelines Division.

-------
                             6.  ECONOMIC IMPACTS

     This chapter provides an assessment of the economic impacts that are
likely to occur as a result of the costs of the effluent control/treatment
technologies described in Chapter 3.  The analysis is based upon the examina-
tion of the estimated compliance costs, other economic and financial char-
acteristics of 21 sample plants, and the analytical methodology described in
Chapter 2.  The primary economic impact areas discussed include the effect
of the pollution control costs on semiconductor prices, the profitability
of semiconductor manufacturing, plant closures, employment, changes in imports
and exports, and industry structure.  The analysis is not intended to be plant-
specific; instead, the sample plants serve only as a basis for deriving
indications of potential industry-wide impacts.

6.1  PRICE AND QUANTITY IMPACTS

     The price and quantity impacts of the compliance costs are determined
through an examination of industry market structure and pricing behavior.
Economic theory predicts different pricing behavior for different market
structures (e.g., competitive, oligopoly, monopoly).  As described in Chapters
2 and 3, the market structure for this industry is not precisely determined
in this study.  Nevertheless, there is evidence that the semiconductor industry
has both the incentive and the capability to completely absorb all small and
moderate levels of mandated pollution control costs.  The incentive not to
raise product prices is in the form of significant international competition
for world market share; and the capability is implied by the variations in
profit rates among various semiconductor products within multi-product firms.
For these reasons it is assumed that there will be no price changes resulting
from the regulation in the short and intermediate time period.
                                    11-45

-------
6.2  PLANT LEVEL PROFIT IMPACT ANALYSIS

     The basic profitability measure used in this study is before-tax return
on sales, or profit margin (PM).  The profitability of a plant is considered
to be significantly impacted if the total annual compliance costs reduce the
return on sales of a plant to an extent that it is equal to or less than the
lowest average industry PM.  As described in Chapter 2, The lowest acceptable
PM was determined to be 2.7 percent.  This is based on the performance of the
semiconductor industry over the 1971-1977 period.10'

     Table 6-1 presents data on the actual number of production employees and
the estimated value of shipments for each sample plant.  The value of shipments
data, along with the total annual compliance cost estimates for each of the
recommended treatment options for the semiconductor plants, are used to
analyze the performance of the plants after compliance.

     Table 6-2 presents the percent of annual revenues that is estimated to
be spent annually to operate, maintain, and finance the proposed pollution
control technologies for each of the options considered for the sample plants.
This table was developed from the value of shipments data in Table 6-1 and
the compliance cost data from EPA.  Compliance costs for option 1 (Selected
Option for PSES) are estimated to be negligible for the reasons discussed
previously.  (A sensitivity analysis for option 1 appears in Appendix II-A.)
Option 3 (Selected Option for BAT) costs are considered relatively insignifi-
cant; there are only three plants that have treatment costs greater than 1
percent of the estimated annual revenues of the plant.  The post-compliance
profit margins for these plants are not below the threshold value for profita-
bility.  The annual compliance costs associated with options 2 and 5 are
much more significant.  Eight of the plants examined may incur costs that
range from 1-12 percent of their annual revenues.  Four of the eight sample
plants have PMs below the threshold value, all of which are small plants.
    U.S. Department of Commerce, U.S. Semiconductor Industry, September 1979.
                                    11-46

-------
      TABLE 6-1.  SALIENT STATISTICS FOR THE SEMICONDUCTOR SAMPLE PLANTS
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
PRODUCTION*/
EMPLOYEES
4,650
2,395
40
353
600
3,600
2,100
733
1,800
2,500
800
60
300
150
50
300
1,150
33
12,450
60
595
VALUE OFb/
SHIPMENTS
389,765,186
200,750,026
3,352,819
29,588,626
50,292,282
301,753,692
176,022,987
61,440,405
150,876,846
209,551,175
67,056,376
5,029,228
25,146,141
12,573,071
4,191,024
25,146,141
96,393,541
2,766,076
1,043,564,852
5,029,228
49,873,180
NEW CAPITALb/
EXPENDITURES
29,950,393.70
15,426,062.99
257,637.80
2,273,653.54
3,864,566.93
23,187,401.57
13,525,984.25
4,721,212.60
11,593,700.79
16,102,362.20
5,152,755.91
386,456.69
1,932,283.46
966,141.73
322,047.24
1,932,283.46
7,407,086.61
212,551.18
80,189,763.78
386,456.69
3,832,362.20
a/ SOURCE:  EPA, Effluent Guidelines Division

b/ Estimates derived from average industry ratios obtained from the Bureau
   of Census, Census of Manufactures, 1977.
                                     11-47

-------
             TABLE 6-2.  SEMICONDUCTORS - ANNUAL COMPLIANCE COSTS3/
                         AS A PERCENTAGE OF REVENUES
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
TREATMENT
OPTION 1
.00
.00
.12
.01
.01
.00
.00
.01
.00
.00
.01
.08
.02
.03
.10
.02
.00
.15
.00
.08
.01
TREATMENT
OPTION 2
.27
.27
5.35
1.51
.81
.31
.27
.65
.32
.54
.90
3.81
1.70
.39
1.40
1.18
.36
11.10
.06
3.24
.35
TREATMENT
OPTION 3
.03
.04
1.02
.23
.12
.04
.04
.10
.05
.07
.12
.68
.24
.27
.81
.13
.06
1.81
.01
.08b/
.07
TREATMENT
OPTION 5
.29
.34
5.65
1.65
.88
.34
.29
.71
.35
.58
.94
4.21
1.85
.44
1.58
1.30
.39
12.17
.07
3.59
.39
a' The annual compliance costs for each of the treatment options include
   monitoring costs.  These costs are estimated at $4,100 per plant annually.

°' Information on concentrated fluoride acid waste was not available for this
   plant.
                                     11-48

-------
6.3  CAPITAL REQUIREMENTS ANALYSIS

     As described in Chapter 2, the assessment of the firm's ability to raise
the required capital is based on ratios of compliance capital investment to
average annual compliance expenditures.  Because of the absence of plant-
specific financial data, the annual capital expenditures of the plants are
calculated from the industry average new capital expenditures per production
employee ratios reported in the 1977 Census of Manufactures.  The estimated
annual capital expenditures from Table 6-1, along with capital investment
costs for each of the pollution control options, are used to indicate the
significance of the capital impacts.

     Table 6-3 presents the compliance capital investment to plant capital
investment ratios for the 21 sample plants.  These ratios reveal that the
proposed pollution control costs would range from zero to 30 percent of the
average annual capital expenditures for options 1 and 3.  Investment costs
of these magnitudes are not by themselves enough to cause plant closures.
The pollution control investment costs are more significant for options 2
and 5.  Four of the plants, all of which are small, would have to invest more
than 50 percent of their annual capital expenditures in pollution control to
comply with the proposed regulation.  These results will be used below,
together with other information, to assess the potential for plant closure.

6.4  POTENTIAL PLANT CLOSURES

     As described in Chapter 2, potential plant closures are estimated by simul-
taneously examining the results of the profit and capital impact assessments
for the sample plants, and qualitatively assessing other nonquantified factors.
Using these results, possible closures are identified, and  then used as  the
basis for extrapolating industry closures.

     It is assumed that a plant would close if the costs of pollution  control
caused its profit margin to fall below 2.7 percent, and/or  if  a plant  had to
                                    11-49

-------
          TABLE 6-3.  SEMICONDUCTORS - POLLUTION CONTROL INVESTMENT
             COSTS AS A PERCENT OF ANNUAL CAPITAL EXPENDITURES3/
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
TREATMENT
OPTION 1
.01
.01
.78
.09
.05
.01
.01
.04
.02
.01
.04
.52
.10
.21
.62
.10
.03
.94
.00
.52
.05
TREATMENT
OPTION 2
4.35
4.51
95.75
25.87
13.92
5.15
4.58
11.24
5.47
8.77
14.67
67.99
29.15
7.23
25.94
20.66
6.22
193.65
1.01
58.26
6.29
TREATMENT
OPTION 3
.57
.59
24.65
3.41
1.82
.67
.60
1.47
.72
1.14
1.92
16.43
3.50
6.57
19.72
3.29
.86
29.88
.13
.52b/
1.66
TREATMENT
OPTION 5
4.88
5.08
109.20
29.24
15.74
5.77
5.17
12.71
6.17
9.81
16.53
77.49
32.95
8.34
29.89
23.44
7.05
219.66
1.14
66.50
7.17
a'  Monitoring costs are included in the pollution control investment costs
   for each of the treatment options.

"'  Information on the concentrate fluoride acid wastes was not available
   for this plant.
                                    11-50

-------
spend more than 100 percent of its annual capital expenditures on pollution
control equipment.

     The pollution control costs associated with the selected options 1 and 3
are not great enough to cause plant closures.  However, the analysis of the
pollution control costs associated with treatment options 2 and 5 indicate
that the annual profit margins for four of the plants examined may be reduced
below 2.7 percent, and that these four plants will have to invest significant
portions of an average year's capital expenditures in pollution control equip-
ment.  Under these circumstances, four, or 19%, of the sample plants are identi-
fied as likely candidates for closure, if they were required to install treat-
ment option 2 or 5.  All four plants are small, with employment ranging from
33 to 60 employees.  The results of the sample plant analysis are used as the
basis for estimating plant closures at the industry level.  These estimates
are shown in Table 6-4.  It is assumed that the percentage of industry plant
closures will be the same as that indicated for the sample plants.  Using this
approach, it is estimated that 19 percent or 48 of the plants in the semicon-
ductor industry are likely to close under options 2 or 5.  Of these 48 plants,
it is estimated that 14 are likely to be direct and 34 indirect dischargers.
The number of direct versus indirect dischargers that are likely to close is
estimated using the pre-compliance ratio of direct to indirect dischargers for
the total industry.11'

6.5  EMPLOYMENT IMPACTS

     In general, a regulation can impact on employment in an industry through
job losses that are associated with closures, and from price increases that
11/ An important qualification to these extrapolations is in order.  First, 136,
    or 53 percent, of the 257 plants in the industry are small (employ fewer
    than 100 production employees).  However, only  5, or 24 percent, of the
    plants in the EPA sample were small plants, indicating that the economic
    impact estimates based on these data may not fully represent the range of
    likely impacts among small plants.  For example, if 80 percent of small
    plants closed (4 out of 5 small plants in the sample), then the costs of
    options 2 or 5 could result in (0.8)(136) = 109 plant closures, instead of
    the 48 estimated above.
                                    11-51

-------
            TABLE 6-4.   POTENTIAL PLANT CLOSURES  AND  EMPLOYMENT  IMPACTS FOR SEMICONDUCTOR PLANTS


SIZE OF
PLANTS
Small


Medium-
Small

Medium-
Large

Large


Total


ESTIMATED
# OF PLANTS
BY DISCHARGE
STATUS
Direct
Indirect
Subtotal
Direct
Indirect
Subtotal
Direct
Indirect
Subtotal
Direct
Indirect
Subtotal
Direct
Indirect
Total
OPTION 1
POTENTIAL
PLANT
CLOSURES
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

CHANGE IN
EMPLOYMENT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
OPTION 2
POTENTIAL
PLANT
CLOSURES
14
34
48
0
0
0
0
0
0
0
0
0
14
34
48

CHANGE IN
EMPLOYMENT
686
1,666
2,352
0
0
0
0
0
0
0
0
0
686
1,666
2,352
OPTION 3
POTENTIAL
PLANT
CLOSURES
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

CHANGE IN
EMPLOYMENT
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
OPTION 5
POTENTIAL


PLANT CHANGE IN
CLOSURES EMPL
14
34 1
48 2
0
0
0
0
0
0
0
0
0
14
34 1
48 2
OYMENT
686
,666
,352
0
0
0
0
0
0
0
0
0
686
,666
,352
SOURCE:   Compiled by JRB Associates.

-------
cause output and employment to fall.  Since no plant closures or price changes
are expected for either options 1 or 3 the regulation is not expected to have
any impact on the level of unemployment for these options.  On the other hand,
48 plants are likely to close if either treatment option 2 or 5 technologies
are required.  Since all of the potential closures are small plants (i.e.,
having between 1 and 99 employees), and the average number of employees in the
small plants is 49, it is estimated that 2,352 employees (i.e., 48 plants mul-
tiplied by 49 employees) may lose their jobs if either of these two pollution
control technologies is required.

6.6  FOREIGN TRADE IMPACTS

     The impact of the regulation on the levels of imports and exports depends
on (1) the extent that semiconductor product prices in the domestic market rise
faster than prices in the rest-of-world market, and (2) the extent to which the
domestic production losses are replaced by imports from foreign countries.
Because the domestic producers are not expected to increase the price of semi-
conductor products as a result of this regulation they should continue to be
competitive with foreign producers.  Therefore, there will be no foreign trade
impact from this factor.  The increase in imports of semiconductor products
because of domestic capacity reductions (i.e., plant closures) is difficult to
quantify precisely.  For this analysis, it is assumed that the production
losses will be made up by U.S. producers only.  Therefore, no balance of pay-
ment impacts are predicted.

6.7  NEW SOURCE IMPACTS

     As indicated in Chapter 5, the incremental costs of NSPS and PSNS are zero.
Because there is no difference in compliance costs between similar existing and
new sources, the regulation would not foster any competitive advantages or dis-
advantages between new and existing sources.
                                    11-53

-------
6.8  SUMMARY OF SEMICONDUCTOR INDUSTRY ECONOMIC IMPACTS

     The estimated economic impacts for the selected options in the semicon-
ductors subcategory are summarized in Table 6-5.

Toxic Organics
     The control of toxic organics and pH under Option 1 for BPT and PSES is
expected to cause compliance costs for monitoring of $257 thousand in capital
investment and $527 thousand annually.  These costs are based on two assump-
tions, as with the crystals subcategory.  First, at least the 50% of the
facilities currently in compliance with toxic organic limitation are expected
to choose to certify that they do not dump solvents into their effluent instead
of monitoring.  Second, these remaining plants that do monitor will, on average,
be required to do so quarterly.  Since the incremental costs of solvent disposal
tend to be offset by resale, EPA did not estimate costs for solvent management.
Monitoring costs are less than 0.2% of annual revenues and are not expected to
cause other than minor profitability impacts.  No plant closures are expected.

     In addition to the above analysis, EPA conducted a sensitivity analysis
(reported in Appendix II-A) for Option 1 costs consisting of two parts.
First, the impact of monthly monitoring was estimated for all facilities
(some facilities may be required to monitor as frequently as once per month,
although EPA cannot identify precisely which ones).  In addition to this
analysis, costs and impacts were also determined for facilities which may
incur "worst case" solvent disposal costs.  These costs are developed for
facilities that could fall under the requirements of the Resource Conservation
and Recovery Act (RCRA) for disposal.

     The analysis of these costs shows that while the effects of these costs are
somewhat higher, they still represent less than 2.3% of annual revenues.  These
would be expected to cause some profit reductions, but not enough to cause any
plant closures.
                                    11-54

-------
              TABLE 6-5.  SUMMARY OF ESTIMATED ECONOMIC IMPACTS
                     FOR THE SEMICONDUCTORS SUBCATEGORY
ECONOMIC IMPACT VARIABLES

  Annual Compliance Costs/Revenues
    Indirect Dischargers
    Direct Dischargers
  Change in Price (%)
  Change in Quantity (%)
  Change in Profitability
    Indirect Dischargers
    Direct Dischargers
  Capital Requirements
    Indirect Dischargers
    Direct Dischargers
  Plant Closures due to Regulations
    Indirect Dischargers
    Direct Dischargers
  Employment at Closed Plants
    Indirect Dischargers
    Direct Dischargers
  Balance of Trade Changes
  Industry Structure Changes
OPTION 1*  OPTION 2  OPTION 3** OPTION 5
                                                                         SELECTED
                                                                          OPTION
  Low
  Low

  Low
  Low

    0
    0

    0
    0
  None
  None
  Signif.
  Signif.

  Signif.
  Signif.

   34
   14

1,666
  6«6
 None
 Signif.
0.01-2
0.01-2
0
0
Low
Low
Low
Low
0
0
0
0
None
None
.0 0.1-12
.0 0.1-12
0
0
Signif.
Signif.
Signif.
Signif .
34
14
1,666
686
None
Signif.
.2
.2












                                                                          Low
                                                                          Low

                                                                          Low
                                                                          Low

                                                                          0
                                                                          0

                                                                          0
                                                                          0
                                                                         None
                                                                         None
 * Selected Option for Indirect Dischargers

** Selected Option for Direct Dischargers
                                     11-55

-------
Fluoride

     EPA's limits on fluoride under Option 3 for BAT will require an estimated
52 of 77 facilities to spend $3.6 million in capital investment and $2.9 million
annually (including monitoring) to comply with the regulations.  These costs are
not expected to cause other than small reductions in profitability.  No plant
closures are expected.

New Sources
     The effluent standards and associated technologies for new sources are
identical to those for existing sources.  Consequently, the economic impacts
for new sources will mirror those of existing sources and the promulgated
regulations are not expected to foster competitive advantages of disadvantages
between new and existing sources.

6.9  LONG-TERM IMPLICATIONS AND OTHER IMPACTS

     In spite of the impressive growth of the semiconductor industry, a pros-
perous future is not guaranteed.  Technological obsolescence and foreign compe-
tition are likely to continue to be of great concern to the industry.  Because
of this, capital expenditures for R&D and new equipment are very high.  Failure
to keep pace on a year-to-year basis may mean the loss of important market
segments to competitors, which in many cases are European or Japanese companies.
The costs associated with EPA regulations could add to these problems if the
added burden were significant.  It may force some firms into a noncompetitive
position and could lead to potential plant closures in the long run, or it
could lead to a firm's relocating their wafer fabrication operations overseas.
Most companies already have moved their labor intensive assembly operations
overseas, and it would be quite possible for them to relocate wafer fabrication
as well.  Such a move would result in a loss of employment, wages, and a sub-
stantial adverse shift in the current balance of trade.
                                     11-56

-------
6.10 SMALL BUSINESS ANALYSIS

     The Regulatory Flexibility Act (RFA) of 1980 (P.L. 96-354) requires Federal
regulatory agencies to consider "small entities" throughout the regulatory pro-
cess.  The RFA requires analyses to be performed which determine if a substan-
tial number of small entities will be significantly impacted.  If so, regulatory
alternatives that eliminate or mitigate the impact must be considered.  This
analysis addresses these objectives by identifying and evaluating the degree of
economic impacts of the aforementioned regulations on small entities (the
primary small entity affected by these regulations is the small semiconductor
manufacturer).  As described in Chapter 2 the small business analysis was
developed as an integral part of the general economic impact analysis described
above in this report.

6.10.1  Definitions of Small Business

     As described in Chapter 2, the definition of small for purposes of regula-
tory analysis is not precise or universal.  The small business defintions used
by other Federal agencies, which were developed for purposes other than regula-
tory analysis was found to be inappropriate for this study.  Instead, informa-
tion on both the amount of effluent discharged and the number of employees
working in the plants was used to determine an appropriate cutoff point that
identifies small entities.  The criterion selected consider plants with fewer
than 100 employees as small entities because at this cutoff point the plants
within this subcategory have some special and distinct properties.  Table 6-6
illustrates the pollution discharge characteristics of plants in the industry
by employment size categories.  The information presented demonstrates that
the small plants (i.e., plants with less than 100 employees) discharge an
average of 57 thousand gallons of wastewater per day.  Furthermore, although
the small plants make up more than 50 percent of the plants in the industry
they discharge only 15 percent of the wastewater.
                                     11-57

-------
   TABLE 6-6.   SEMICONDUCTOR - AVERAGE PLANT AND TOTAL INDUSTRY FLOW RATES

Size of Plant


Small
(1-99 Employees)
Medium- Small
(100-499 Employees)
Medium- Large
(500-999 Employees)
Large
(Over 1,000
employees)
Total
DISCHARGE
STATUS
DIRECT


41

25

5


6
77
INDIRECT


95

58

13


14
180
TOTAL
# OF
PLANTS


136

83

18


20
257
AVERAGE
FLOW RATE
(GPD)


56,772

174,450

318,325


1,238,224
205,038

Total Flow Rates
(GPD)
Total %

7,720,992 15

14,479,350 27

5,729,850 11


24,764,480 47
52,694,672 100
SOURCE:   Compiled by JRB Associates.
                                    11-58

-------
6.10.2  Impacts on Small Entities

     Table 6-7 shows the average ratios of compliance costs to revenues and the
pollution control investment costs to annual new capital expenditures which are
associated with plants of various sizes in the semiconductor industry.  The
information in this table indicates that the smaller plants (i.e., the plants
with less than 100 employees) would be more severely affected than the larger
plants.  In fact, the impact variables examined are at least 3 times greater
for the small plants than for larger plants.  In addition, all the projected
plant closures are small plants if either treatment option 2 or 5 are required.

     However, under Option 1 (selected option for PSES) and Option 3 (selected
option for BAT) the absolute of the economic impacts is not considered large,
with only 3 plants passing the screening threshold level for profitability, and
none of which show significant profitability impacts upon closer examination.
Therefore, the selected options for promulgation are not expected to have a
significant impact on small facilities in this industry.
                                      11-59

-------
                     TABLE 6-7.   SUMMARY PROFIT  AND  CAPITAL  IMPACT  ANALYSES  BY  EMPLOYEE  SIZE  CLASSIFICATION
                                                               (Percent)
Os
O
SIZE CLASSIFICATION
Small
(1-99 employees)
Medium-Small
(100-499)
Medium-Large
(500-999)
Large
1,000 >
AVERAGE ANNUAL COST TO REVENUES (%)
OPTION 1
.09
.02
.01
.003
OPTION 2
4.98
1.19
.68
.30
OPTION 3
.88
.21
.10
.04
OPTION 5
5.44
1.31
.73
.33
AVERAGE INVESTMENT COST TO ANNUAL
CAPITAL EXPENDITURES (%)
OPTION 1
.68
.12
.04
.01
OPTION 2
88.32
20.7
11.53
5.0
OPTION 3
18.24
4.19
1.72
.66
OPTION 5
100.55
23.49
13.04
5.63
                 SOURCE:   Compiled  by  JRB  Associates

-------
                        7.  LIMITATIONS OF THE ANALYSIS

     This section discusses the major limitations of the economic impact
analysis.  It focuses on the limitations of the data, methodology, assumptions,
and estimations made in this report.  Information pertaining to the estimation
of the compliance costs for specific plants and the limitations of the method
are outlined in Section 9 of the EPA Development Document.

     The economic impacts assessed are the result of this regulation.  The
assessment does not include the economic impacts from other regulations for
air pollution control, OSHA requirements, and solid waste requirements.

7.1  DATA LIMITATIONS

     The major assumptions and estimates that have the greatest impact on the
accuracy of the conclusions of this report relate to the data used in the
analysis.  It is important to realize that (1) no economic survey was con-
ducted to collect plant-specific financial and economic data, and (2) these
types of data are not in the public domain for this industry.  As a result,
most of the plant-specific data used are estimates derived from average
industry operating and financial ratios.  The plant-specific financial data
must, therefore, be considered order-of-magnitude estimates, and not  the
actual data for specific plants.

     The value of shipments and new capital expenditures were estimated by
multiplying the number of production employees at each plant by the appro-
priate average industry performance ratio obtained from the Census of Manu-
factures.  This methodology relies on the assumptions that a fixed relation-
ship exists between the key performance variables for all plants, and  that
this relationship remains constant over time.  These assumptions  are  probably
invalid in the real world because of the following factors:
                                    11-61

-------
        Worker productivity varies significantly among plants
        The type, scale, and age of technologies vary widely
        among plants and greatly affect the value added by
        each production employee
        The type of products produced greatly affects the value
        added per employee or unit of raw material used; and
        product mix varies from one plant to another.
     Ignoring these factors could result in overestimating the impacts on
some plants while underestimating those of other plants.  The reason for
these potential errors is that the intra-industry variation in baseline
conditions is underestimated.  That is, baseline conditions for lower-profit
plants are probably overstated and those of high-profit plants understated.

7.2  METHODOLOGY

7.2.1  Price Increases

     The analysis assumes that the semiconductor industry will not increase
the price of their products and pass through any portion of the compliance
costs to their customers.  While this is a reasonable assumption for some
semiconductor products it may not be true for all the product segments in
the semiconductor industry.  It is possible that the price pass-through
percentage will vary among the various semiconductor products.  Adequate
information on these segments is not available.  Therefore, no further
analysis is carried out to measure the extent of a price increase which may
occur as the production costs of the plants increase with the installation
of pollution control.

     The estimation error which might result from the price change assumptions
depends primarily on the cross elasticity of demand with respect to the ratio
of domestic to foreign prices.  That is, if increased domestic prices shift
demand production overseas, there would be a drop in aggregate quantities pro-
duced by domestic producers and, consequently, increased unemployment and
                                    11-62

-------
 possibly other impacts.  Thus,  the  remaining domestic semiconductor producers
 would have healthy profits,  but account  for a  smaller portion of world market
 share.  On the other hand, if  such  shifts were small, then domestic producers
 could raise prices to cover  compliance expenditures without losing market
 share.

     If the latter situation prevailed in the  real world, the economic'impacts
 estimated in this report are overstated.  If the  former situation prevailed,
 the impacts would take a different  form, but would probably not exceed the
 levels estimated in Chapter  6  of  this report.

 7.2.2  Sensitivity Analysis

     In addition to the biases  that may  be generated by the method used to
 estimate revenues and new capital expenditures of individual plants, the
 results of the analysis can  be  influenced by the  pollution control cost data.
 To evaluate some of the problems  associated with  these data, sensitivity
 analyses were performed.  For  this  analysis, the  estimated compliance  costs
 were varied by -20 percent and  +20  percent.  Tables 7-1 and 7-2 present the
 results of the sensitivity analysis on the profitability and capital availa-
 bility assessments for all the  recommended treatment options.  The tables
 show that the changes in the pollution control costs (or equivalently, changes
 in revenue and capital expenditure  estimates)  would not significantly  change
 the post-compliance financial  ratios and the study's results for treatment
 options 1 and 3.  For treatment options  2 and  5,  reductions in the pollution
 control costs by 20 percent  would not cause changes to the previous results.
 Alternatively, increases in  the pollution control costs by 20 percent  may
 cause 2 additional plants to close  (i.e., six  of  the sample plants  instead
 of four plants, as predicted in the impact analysis in Chapter 6, are  likely
 to close).  Therefore, incorrect  costs for these  options  (options 2 and  5)
 could significantly affect the  findings  of the study.  On the other hand,
incorrect costs for options  1 and 3 are  not likely to alter the findings of
 the study.
                                     11-63

-------
              TABLE 7-1.   SEMICONDUCTORS SENSITIVITY ANALYSIS -
             ANNUAL COMPLIANCE COSTS AS A PERCENTAGE OF REVENUES
PLANT
CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
OPTION 1
-20% +20%
0 0
0 0
.096 .144
.008 .012
.008 .012
0 0
0 0
.008 .012
0 0
0 0
.008 .012
.064 .096
.016 .024
.024 .036
.08 .12
.016 .024
0 0
.12 .18
0 0
.064 .096
.008 .012
OPTION 2
-20% +20%
.216 .324
.216 .324
4.28 6.42
1.208 1.812
.648 .972
.248 .372
.216 .324
.52 .78
.256 .384
.432 .648
.72 1.08
5.448 8.172
1.36 2.04
.312 .468
1.12 1.68
.944 1.416
.288 .432
8.88 13.32
.048 .072
2.592 3.888
.28 .42
OPTION 3
-20% +20%
.024 .036
.032 .048
.816 1.224
.184 .276
.096 .144
.032 .048
.032 .048
.08 .12
.04 .06
.056 .084
.096 .144
.544 .816
.192 .288
.216 .324
.648 .972
.104 .156
.048 .072
1.448 2.172
.008 .012
.064 .096
.056 .084
OPTION 4
-20% +20%
.232 .348
.272 .408
4.52 6.78
1.32 1.98
.704 1.056
.272 .408
.232 .348
.568 .852
.28 .42
.464 .696
.752 1.128
3.368 5.052
1.48 2.22
.352 .528
1.264 1.896
1.04 1.56
.312 .468
9.736 14.604
.056 .084
2.872 4.308
.312 .468
SOURCE:   Compiled by JRB Associates
                                    11-64

-------
              TABLE 7-2.  SEMICONDUCTORS SENSITIVITY ANALYSIS -
                      POLLUTION CONTROL INVESTMENT COSTS
                 AS A PERCENTAGE OF NEW CAPITAL EXPENDITURES
PLANT
CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
OPTION 1
-20% +20%
.008 .012
.008 .012
.624 .936
.072 .108
.04 .06
.008 .012
.008 .012
.032 .048
.016 .024
.008 .012
.032 .048
.416 .624
.08 .12
.168 .252
.496 .744
.08 .12
.024 .036
.752 1.128
0 0
.416 .624
.04 .06
OPTION 2
-20% +20%
3.48 5.22
3.608 5.412
76.6 114.9
20.696 31.044
11.136 16.704
4.12 6.18
3.664 5.496
8.992 13.488
4.376 6.564
7.016 10.524
11.736 17.604
54.392 81.588
23.32 34.98
5.784 8.676
20.752 31.128
16.528 24.792
4.976 7.464
154.92 232.38
.808 1.212
46.608 69.912
5.032 7.548
OPTION 3
-20% +20%
.456 .684
.472 .708
19.72 29.58
2.728 4.092
1.456 2.184
.536 .804
.48 .72
1.176 1.764
.576 .864
.912 1.368
1.536 2.304
13.144 19.716
2.8 4.2
5.256 7.884
15.776 23.664
2.632 3.948
.688 1.032
23.904 35.856
.104 .156
.416 .624
1.328 1.992
OPTION 5
-20% +20%
3.904 5.856
4.064 6.096
87.36 131.04
23.392 35.088
12.592 18.888
4.616 6.924
4.136 6.204
10.168 15.252
4.936 7.404
7.848 11.772
13.224 19.836
61.992 92.988
26.36 39.54
6.672 10.008
23.912 35.868
18.752 28.128
5.64 8.46
175.728 263.592
.912 1.368
53.2 79.8
5.736 8.604
SOURCE:  Compiled by JRB Associates
                                    11-65

-------
7.3  PLANT CLOSURES

     The criteria for plant closure is not all-inclusive.  It does not address
the long-term financial condition of the plants, captive operations, the time
value of money, the amount of invested capital, or other factors which affect
closure decisions.  A more detailed closure analysis would require more pre-
cise data on the long-term capital structure, cash flow, and the profitability
of plants in the semiconductor industry.  However, the costs of the recommended
treatment technology are small, and a more detailed analysis will not necessarily
yield different results.

7.4  SAMPLING

     The statistical significance of the 21-plant sample is skewed in such
a way that the large plants are adequately covered.  However, most of the
plants in the semiconductor industry are small, and these plants are repre-
sented by only a few plants in the sample.

7.5  SUMMARY OF LIMITATIONS

     Although a number of assumptions and data limitations may significantly
bias the economic impact conclusions, the potential changes to the conclu-
sions resulting from the elimination of these biases are probably small.
Consequently, the above methodology appears to have provided a reasonable
industry-wide assessment of potential impacts.
                                    11-66

-------
                                APPENDIX II-A
                    SENSITIVITY ANALYSIS ON OPTION I COSTS
1.  INTRODUCTION
     The costs used to analyze the economic impact of option 1 in the semi-
conductors subcategory analysis are based on two assumptions:  (1) that the
only costs associated with compliance consist of monitoring costs, and (2)
that monitoring will occur, on average, once per quarter for the affected
plants.  This analysis consists of two parts.  The first part estimates
the economic impact of monthly monitoring on individual plants in place of
the quarterly monitoring.  The second part estimates the economic impacts
of "worst case" incremental solvent disposal costs.

II.  MONITORING COST SENSITIVITY ANALYSIS

     EPA estimates that facilities covered under this regulation will, on
average, monitor their effluent for toxic organics once per quarter.  However,
some facilities may monitor as frequently as once per month.  This analysis
is performed for all plants to determine the impact of the costs associated
with more frequent monitoring because EPA cannot determine precisely which
facilities will be required to monitor on a monthly basis.

     The capital investment cost required for monitoring monthly is the same
as that for monitoring quarterly, 52,000.  The annual operating and maintenance
(O&M) costs for quarterly monitoring are $3,500 (as reported in Chapter 5);
therefore, the annual O&M costs for monthly monitoring are $10,500 (quarterly
costs times 3).

     The methodology for analyzing impacts of monthly monitoring is the same
as described in Chapter 2 of this report.  Table A-l summarizes the result
                                    11-67

-------
             TABLE A-l.  MONITORING COSTS SENSITIVITY ANALYSIS -
             ANNUAL COMPLIANCE COSTS AS A PERCENTAGE OF REVENUES3/
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
TREATMENT
OPTION 1
.00
.01
.33
.04
.02
.00
.01
.02
.01
.01
.02
.22
.04
.09
.26
.04
.01
.40
.00
.22
.02
TREATMENT
OPTION 2
.27
.27
5.56
1.54
.82
.32
.27
.66
.33
.54
.91
3.95
1.73
.44
1.57
1.21
.37
11.36
.06
3.38
.37
TREATMENT
OPTION 3
.04
.04
1.23
.25
.14
.04
.04
.11
.05
.07
.13
.82
.27
.32
.97
.16
.06
2.06
.01
.22
.08
TREATMENT
OPTION 5
.29
.34
5.86
1.67
.90
.34
.30
.72
.35
.58
.95
4.35
1.88
.50
1.74
1.32
.40
12.42
.07
3.73
.40
a'  The annual compliance costs for each option includes treatment costs,
   plus total annual monitoring costs which are estimated at $11,100.
   An annual operating and maintenance cost for monitoring of $10,500,
   plus an annual capital cost for monitoring of $600 per plant.

SOURCE:  Compiled by JRB Associates.
                                    11-68

-------
of this analysis.  For the sample impact plants, the annual costs for monthly
monitoring is less than 0.4 percent of revenues for option 1, (selected option
for BPT, PSES and PSNS) and less than 2.06 percent when the incremental costs
of option 3 (selected option for BAT and NSPS) are included.  While these costs
have a greater impact than quarterly monitoring, they still remain below the
threshold level for highly impacted plants.

     Thus, while the impact of monthly monitoring is expected to be greater than
for quarterly monitoring, the impacts are not expected to cause plant closures.

III. SOLVENT DISPOSAL SENSITIVITY ANALYSIS

     EPA has determined that the incremental costs associated with improved
solvent management techniques tend to be offset by resale of the solvents.
Consequently, EPA did not cost out any small costs for the purposes of the
impact analysis.  Some facilities may have to haul away their solvents under
the requirements of the Resource Conservation and Recovery Act (RCRA).  There-
fore, this analysis examined these "worst case" costs for RCRA hauling for all
plants because EPA cannot determine which plants, if any, will fall under
these requirements.  These costs are not considered to be ordinary or average
costs of compliance for facilities in this industry.

     EPA estimated individual plant costs for each facility where possible.
For remaining plants, and for model plants, the costs were extrapolated from
plants with similar volumes of spent solvents.  The costs used in the impact
analysis appear in Table A-2.

     The impact analysis was performed including the costs of monthly
monitoring.  This was done to determine the maximum possible impact on any
one plant assuming that (1) the plant is required to monitor monthly; and  (2)
the plant chooses to dispose of its spent solvents under RCRA.

     The analysis followed the methodology described in Chapter  2 of  this
part.  Tables A-3 and A-4 summarize the results of the sensitivity analysis.

                                    11-69

-------
     TABLE A-2.   SEMICONDUCTOR INDUSTRY STATISTICS FOR SOLVENT DISPOSAL
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
TOTAL ANNUAL
COMPLIANCE
COST
$ 1983
1200.*
14400.*
6800.*
1200.*
1600.*
0.*
0.*
4000.
0.*
1600.*
4000.
6800.
4000.
4000.
6800.
4000.
4000.
6800.
4000.
6800.
4000.
TOTAL ANNUAL
COMPLIANCE
COST
$ 1979
966.
11592.
5477.
966.
1288.
0
0
3220.
0
1288.
3220.
5474.
3220.
3220.
5474.
3220.
3220.
5474.
3220.
5474.
3220.
* Plant data provided by the Effluent Guidelines Division of EPA.
                                    11-70

-------
             TABLE A-3.  SOLVENT DISPOSAL SENSITIVITY ANALYSIS -
            ANNUAL COMPLIANCE COSTS AS A PERCENTAGE OF REVENUES'1/
PLANT
ID CODE
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U
TREATMENT
OPTION 1
.00
.01
.49
.04
.02
.00
.01
.02
.01
.01
.02
.33
.06
.11
.40
.06
.01
.60
.00
.33
.03
TREATMENT
OPTION 2
.27
.28
5.72
1.54
.83
.32
.27
.67
.33
.54
.91
4.06
1.74
.47
1.70
1.22
.37
11.56
.06
3.49
.37
TREATMENT
OPTION 3
.04
.05
1.39
.25
.14
.04
.04
.12
.05
.07
.14
.93
.28
.35
1.10
.17
.07
2.26
.Ol^/
.33
.09
TREATMENT
OPTION 5
.29
.35
6.03
1.67
.90
.34
.30
.73
.35
.58
.96
4.46
1.89
.52
1.87
1.34
.40
12.62
.07
3.84
.41
a/ The annual compliance costs for each of the options include monitoring,
   solvent disposal, and treatment costs.

b/ Information on concentrated fluoride acid waste was not available for
   this plant.

SOURCE:  Compiled by JRB Associates
                                    11-71

-------
                      TABLE A-4.  SUMMARY OF SENSITIVITY
                   ANALYSIS BY EMPLOYEE SIZE CLASSIFICATION*/
                                  (Percent)
SIZE CLASSIFICATION
Small
(1-99 employees)
Medium- Small
(100-499 employees)
Medium-Large
(500-999 employees)
Large
(1,000 >
AVERAGE ANNUAL COST TO REVENUES (%)
OPTION 1
.43
.06
.03
.01
OPTION 2
5.30
1.24
.70
.31
OPTION 3
1.20
.26
.12
.05
OPTION 5
5.76
1.35
.75
.37
a'  Profit impacts associated with monitoring, solvent disposal and
   treatment costs by employee size classifications.

SOURCE:  Compiled by JRB Associates
                                    11-72

-------
The compliance cost  to  revenues ratio increases slightly over those for
monthly monitoring only.   For option 1, the compliance costs are less than
0.49 percent  of  revenues,  and less than 2.26 percent when incremental option
3 costs are included.   While these costs have a greater impact than quarterly
monitoring alone,  they  still remain below the threshold level for highly
impacted plants.

     Thus, while the impact of monthly monitoring plus RCRA  solvent disposal
costs is greater than quarterly monitoring costs alone, the  impacts are not
expected to cause plant closures.
                                                     •D.S. GOVEMMBNT PRINTIHO omOE : 1983 O-381-OB2A06
                                      11-73

-------