United States Office of Water EPA 440/2 -80-085
Environmental Protection Regulations and Standards December 1980
Agency Washington. O.C. 20460
Water
v»EPA Economic Impact Analysis of
Proposed Effluent Limitations
Guidelines, New Source Performance
Standards, and Pretreatment
Standards for the Coal Mining
Point Source Category
Volume III- Appendix B,
Methodology Description
-------
ECONOMIC IMPACT ANALYSIS
OF PROPOSED EFFLUENT LIMITATIONS GUIDELINES,
NEW SOURCE PERFORMANCE STANDARDS,
AND PRETREATMENT STANDARDS
FOR THE COAL MINING POINT SOURCE CATEGORY
VOLUME III — APPENDIX B, METHODOLOGY DESCRIPTION
prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER REGULATIONS AND STANDARDS
WASHINGTON, D.C. 20460
by
CONTRACT NO. 68-01-4466
DECEMBER 1980
-------
This document is available through the U.S. Environmental Protection
Agency, Economic Analysis Staff WH-586, 401 M Street, S.W.,
Washington, D.C. 20460, 202-755-2484.
This report has been reviewed by the Office of Water Planning and
Standards, EPA, and approved for publication. Approval does not
signify that the contents necessarily reflect the views and policies
of the Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recommendations
for use.
U;S. Em/iror.r.s.'-.^i! ppfrfcr/'on
-------
PREFACE
This volume is an Appendix to a contractor's study prepared for the
Office of Water Planning and Standards of the Environmental Protection
Agency (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 Coal Mining industry.
The study supplements the technical study (EPA Development Document)
supporting 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 effect 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 Planning and Standards of EPA. This Appendix was submitted
in fulfillment of Contract No. 68-01-4466 by Arthur D. Little, Inc.,
and was completed in July, 1980. The work was performed from June, 1977
through July, 1980; the data sources referred to in the report were
current at the time the work was performed.
This report is being released and circulated at approximately the same
time as publication in the Federal Register of a notice of proposed
rule making. The study is not an official EPA publication. It will
be considered along with the information contained in the Development
Document and any comments received by EPA on either document before
or during final rule-making proceedings necessary to establish final
regulations. Prior to final promulgation of regulations, the accom-
panying study shall have standing in any EPA proceeding or court
proceeding only to the extent that it represents the views of the
contractor who studies the subject industry. It cannot be cited,
referenced, or represented in any respect in any such proceeding as
a statement of EPA's views regarding the Coal Mining industry.
-------
APPENDIX B
TABLE OF CONTENTS
SECTION PAGE NUMBER
I. INTRODUCTION B-l
A. Main Sources of Data Used in The Imapct Analysis B-l
B. Four Main Premises Used for the Analysis B-2
II. IMPACT ANALYSIS METHODOLOGY B-4
III. REGIONALIZATION OK THE MINE FILE (MODULE 1) B-8
IV. MINIMUM REQUIRED PRICES (MODULE 2) B-8
A. Number of Different Sets of Model Mine Parameters
Required B-10
B. Model Mine Production Cost Parameters B-10
C. Mine Flow Volumes and Water Treatment Costs B-15
D. Projection of Changes in Mining Conditions Between
1976 and 1984 B-20
E. Minimum Required Price Calculation and MRP Function
Estimation B-29
F. Details of the Minimum Required Price Calculation B-31
G. Sample Output B-33
H. Estimated Function Parameter Values for MRP, Wages,
Cashflow and Required Investment in Mining Equipment B-33
V. REGIONAL SUPPLY CURVES (MODULE 3) B-35
A. Separation into Major Coal Markets B-35
B. Changes in the Mine Population B-38
C. Detailed and Linearized Supply Curves B-41
VI. COAL MARKET SIMULATION (MODULE 4) B-52
A. Sulfur Content Distributions, Air Quality Control
Standards and Coal Utilization Costs B-52
B. Transportation Costs B- 56
C. Demand for Coal B- 58
-------
Table of Contents
(continued)
Section Page No.
VII. LINEAR PROGRAMMING MODEL OF THE STEAM COAL MARKET B-58
A. Market Clearing Algorithms B-64
B. Linear Programming Formulation of Contract
Market (No Producer's Surplus) B-64
1. Specification of Constraints for
the LP Model B-66
2. Specification of Objective Function B-70
C. Mixed Integer Programming Formulation of
Spot Market (Producer's Surplus) B-70
VIII. OUTPUT OF THE COAL MARKET SIMULATION MODEL (MODULE 6) B-71
IX. LIMITS OF THE ANALYSIS B-75
A. Summary B-75
B. Statistical Significance of the Impact
Estimates B-77
C. Sensitivity of the Impact Estimates to
Systematic Errors B-87
D. Bias Resulting from Aggregation Errors B-93
n
-------
APPENDIX B
LIST OF TABLES
No. Title Page No.
1 Supply and Demand Regions Used for the Impact Analysis B-9
2 Number of Different Combinations of Mining Cost, Water B-ll
Treatment Cost and Fiscal Parameters Required for the
Analysis of One Control Level
3 Example of Input Data for Model Mine Cost Analysis B-13
4 BPT Investment Cost for New Mines B-19
5 BAT Costs for Small Mines (0.025 MMT/Yr) with Different Flow B-21
Volumes (Without Correction for Difference in Mining Region or
Mine Type)
6 BAT Costs for Large Underground Mines (0.6 MMT/YR) With B-22
Different Flow Volumes (Without Correction for Differences in
Mining Region or Mine Type)
7 BAT Costs for Large Strip Mines (2.0 MMT/Yr) With Different B-23
Flow Volumes (Without Correction for Differences in Mining
Region or Mine Type)
8 BAT Costs for Large Strip Mines (4.0 MMT/Yr) With Different B-24
Flow Volumes (Without Correction for Differences in Mining
Region or Mine Type)
9 Multipliers to Correct for Regional Differences in BAT Invest- B-25
ment Costs
10 Projected Escalation in Costs and Labor Productivity Used in B-27
the Model Mine Cost Analysis
11 Results of Model Mine Analysis B-34
12 Estimated Function Parameters for Large Underground Mines in B-36
Northern Appalachia
13 Contract and Spot Market Coal Supplied to Electric Utilities B- 39
in 1976
14 Attrition Rate Parameters B-40
15 Mine Productivity Distributions B-42
16 Existing Production of Contract Mines in 1976 in Different B-44
Size Ranges
17 Estimated Potential Increases in Mining Capacitv Until 1984 B-47
18 Mine Water Flow Distributions B-48
iii
-------
List of Tables
(continued)
No. Title Page No.
19 Detailed Supply Function for Pennsylvania, 1984 (BAT-4) B- 49
20 Linearized Supply Curve for Pennsylvania B-51
21 Ratios of Coal Available in Different Sulfur Content Ranges B-53
22 Ratios of Coal Demand Subject to Different Air Quality
Control Standards B- 54
23 Coal Utilization Costs B- 55
24 Parameter Values for Rail Transportation Cost Equations and
Applicable Origins and Destinations for Contract Market B- 57
25 Transportation Costs Not Based on Rail Rate Equations B- 59
26 Transportation Costs for Links Between 27 Supply
(Horizontal) and 35 Demand (Vertical) Regions (Contract
Market) B- 60
27 Rail Transport Cost Equations and Applicable Origins and
Destinations for the Spot Market B- 61
28 Bridge from Contract to Spot Market Regions B- 62
29 Coal Demand in 1984 (Billions of Btu's) B- 63
30 Nomenclature Used in the LP Model of the Steam Coal
Market B- 67
31 Coal Supply Summary - 1984 Level 4 B-72
32 Coal Burned Cost Summary B-73
33 Example of Calculation of Solution Supply for a Sampled
Curve for Pennsylvania (under BPT) B-82
34 Results of Thirty Supply Samples for Pennsylvania B-84
35 Estimated Supply Impact of BAT-4 on the Contract Market
Compared with the Possible Variation in that Estimate
Due to Uncertain Information on Mine Water Flows B-85
IV
-------
List of Tables
(continued)
No. Title Page No.
36 Possible Variation in Estimated BAT-4 Supply Impact Due B-88
to Uncertain Information on Mine Water Flows
37 Estimated Supply Impact of BAT-4 on the Spot Market B-89
Compared with the Possible Variation in that Estimate
Due to Uncertain Information on Mine Water Flows
38 Possible Variation in Estimated BAT-4 Supply Impact
Due to Uncertain Information on Mine Water Flows B-90
39 Results of Sensitivity Tests in Terms of Changes B-92
in Estimated Supply Impacts
-------
APPENDIX B
LIST OF FIGURES
No. Title Page No.
1 Schematic of Impact Analysis Methodology B-5
2 Coal Supply-Demand Model B-6
3 Coal Mine Water Flows - Appalachia B-16
4 Coal Mine Water Flows - Midwest and Central West B-17
5 Coal Mine Water Flows - Great Plains and West B-18
6 Historical Mining Labor Cost and Mining Equipment
Cost Indices B-26
7 Historical Productivity Indices for Underground
and Surface Mines B-28
8 Flow Diagram of Model Mine Cost Analysis Program B-30
9 Main Function of Module 3: Estimation of Regional
Coal Supply by Price in 1984 B-37
10 Typical Supply Curve Illustrating Alternate Market
Clearing Mechanisms B-65
11 Changes in the Regional Supplies, Marginal Prices B-74
and User Costs
12 Illustrative Examples of Supply Curves Based on B-78
Average versus Sampled Values for Mine Productivity
and Mine Water Flows
13 Illustrative Example of Indeterminate Range of the B-79
BAT Supply Curve Because of Uncertainty About Mine
Water Acidity and Flows
14 Methodology to Derive Maximum Possible Error
in Impacted Regions B-80
15 Calculation of the Interaction Error B-86
vi
-------
APPENDIX B
METHODOLOGY DESCRIPTION
I. INTRODUCTION
This Appendix describes the coal supply-demand model that was deve1-
oped to analyze the economic impact of proposed, more stringent water-
quality control standards ^or point sources in coal production. These
control standards will require coal producers to spend more for water
treatment at the mine site, thereby increasing their overall cos*s of
producing coal.
The relative increase in production costs for individual mines is
influenced by numerous factors that are related to the different
environments in which an operator has to deal, such as:
f Meteorological: relative rainfall;
• Topographical: hilly versus flat terrain;
• Locational: proximity to other mines;
• Geological: coal quality ind depth of coal seam,
presence of acquifers, chemical composition of
overburdens;
• Financial: cost of money, equipment and labor;
• Microeconomic: a mine's market position in terms
of location and producer/user relationships;
• Macroeconomic: energy and feedstock supply demand
balances and the costs o* other fuels and
feedstocks.
This analysis is to estimate the relative dependence of coal
production costs on each of these factors in sufficient detail to
allow reliable estimates of the direction and relative magnitude of
inter-regional shifts of coal production that can be expected to take
place in 1984 if more stringent mine water treatment requirements are
imposed.
A. Main Sources of Data Used in the Impact Analysis
The 1976 MESA Mine File described in Chapter II of Appendix A of this
report is used as a starting point for estimating mine production
costs of individual mines. This file contains detailed production
information for individual users. Some of this detail is lost or
B-l
-------
smoothed out through averaging, when regional supply curves are lin-
earized.
Transportation costs are estimated by statistical analysis of rail
rates and water transport costs, as described in Chapter V of Appendix
A.
Coal sulfur content distributions are derived from an analysis of coal
quality data for coal sales to electric utilities in 1976, as obtained
from the FPC.
Coal utilization costs are developed as described in Chapter VI of
Appendix A of this report. These costs account, for the different Btu
content and sulfur content of different coals, and for the varying air
quality control standards in the different demand regions.
As will be shown in later sections, the limited number of data points
on water flow volumes for coal treated at individual mine sites is the
main limiting factor in the level of regional detail achievable in the
impact estimates.
B. Four Main Premises Used for the Analysis
The estimated economic impact of the regulations is obtained through
simulation of supply and demand in steam coal markets in 1984. The
simulation is structured around the following four premises:
(1) Coal, as a fuel, competes with coal rather than
oil, gas or nuclear energy. This premise is based
on the current trend of increasing oil and gas
costs - as discussed in Section VI.1.3 of Appendix
A - the restrictions on the use of gas as utility
boiler fuel, and the growing resistance to further
rapid growth in the use of nuclear energy. The
analysis used a fixed demand estimate for 1984 and
concentrated on the estimation of inter-regional
shifts of coal supplies which can be expected to
take place because of relatively higher estimated
pollution control costs in some regions (as result-
ing from differences in the relative wetness of
mines in the various coal-producing regions).
(2) Coal producers, in order to continue to mine coal,
will have to recover their operating expenses and
investment costs (but not sunk costs), including a
return on those investment costs over the remaining
life of the mine. In this analysis this
requirement establishes the minimum price at which
coal will be sold. If mine operators cannot get
that minimum price, then they will close the mine
and cease to produce.
D-2
-------
(3) It Is assumed that the steam coal contract and spot
markets operate independently of each other. Only
two interactions are allowed in the analysis:
Large mines producing more than
50,000 tons per year 'i.e., contract
market mines) are assumed to sell
some of their output in the spot
market at variable costs.
Contract market mines which are
found to be non-competitive in
contract markets are assumed to sell
to the spot market (i.e., a market
dominated by small mines producing
less than 50,000 tons of coal per
year).
(4) The nature of transactions between producers and
users in contract and spot markets is assumed to be
completely different. In the contract market all
producers are assumed to sell coal on a cost-plus
basis; the large, virtually "infinite" resource
base creates a very competitive situation on the
supply side, always allowing large buyers to select
among a large number of potential suppliers,
thereby driving prices down.
In the spot market all coal from a region is
assumed to sell at the marginal price. Buyers in
this market are more concerned with timely supply
than with price. Suppliers in this market know
what the cost of coal on the margin is and are able
to negotiate prices up to that margin. In other
words, in the contract market all producers are
assumed to obtain the same return on their invest-
ment; in the spot market producers with low-cost
coal will have a higher return than producers with
high-cost coal.
Given these four main premises, the analysis concentrates primarily on
specifying the differences between costs of supply at different output
levels in the various supply regions, allowing for:
• Differences in coal mining conditions 'including
mine wetness);
• Different transportation costs because of different
distances between supply and demand regions;
B-3
-------
0 Differences in user costs because of different roal
quality;
t Differences in user costs because of the differ-
ences between air quality control regulations in
different demand regions.
The remainder of this chapter explains the above in more detail.
II. IMPACT ANALYSIS METHODOLOGY
Figure 1 shows a schematic of the impact analysis methodology.
Regional supply volumes and costs are projected for 1984 for thp steam
coal markets (spot and contract^ and for the met coal market, allowing
for differentials in coal prices because of differences in production
costs, transportation costs and utilization costs.
As shown, ADL has made estimates of potential supply and related costs
for 27 different supply regions. These supply estimates are combined
with estimates of demand in 35 demand regions 'as obtained from EPA
which used these demand projections recently for an analysis of the
impact of air quality control regulations).
These two sets of estimates, together with estimates of coal quality
differentials, and transportation and utilization costs, are used in
the coal market simulation model to obtain estimates both beforp and
after the proposed regulations. This allows us to derive the dif-
ferential economic impact resulting from production cost increases
caused by more stringent water clean-up requirements at the mine site.
As shown in Figure 1, the primary impact measures are:
t Increased user costs both as an average and on the
margin in different demand regions;
• Increases in marginal prices in supply regions;
• Lost production in supply regions;
• Number of closed-mines;
• Number of jobs lost;
• Resulting decreases in regional mine workers'
wages;
• Mine cashflows versus increased investment
requirements for pollution control equipment.
As shown in Figure 2, the model itself consists of six modules, each
composed of one or more data files and computer programs, which use
those data as input. These modules perform the following main opera-
tions:
B-4
-------
FIGURE 1
SCHEMATIC OF IMPACT ANALYSIS METHODOLOGY
ADL
Coal Supply
Model
1984 Supply
Estimates in 27
Supply Regions
AOL:
Coal Quality
Differentials
National
Coal Supply
Model
ADL:
Transportation Costs
ADL:
Utilization Costs
Coal Simulation
Market Model
1984 Demand
Estimates in 35
Demand Regions
BEFORE REGULATION IN 1984 (by region):
Number of producing mines
Total cost of coal to the user
Tons of coal produced
Coal prices
Number of mine workers employed
Wages earned after tax cashflows
Required investment in mining equipment
AFTER REGULATION IN 1984 (by region):
Number of producing mines
Total cost of coal to the user
Tons of coal produced
Coal prices
Number of mine workers employed
Wages earned after tax cashflows
Required investment in mining equipment
PRIMARY IMPACT OF THE REGULATION (by region):
Increased cost to the user
Lost production
Number of closed mines
Number of jobs lost
Decrease in wages
Decrease in cashflows
Increased investment required
B-5
-------
FIGURE 2
COAL SUPPLY-DEMAND MODEL
MODULI 1
Hfllllll I V
ta
MODULE 3
MODULE 4
MODULE 5
/Model Mine
/Costs by Re-
(i|ion and Mine I
V Type
Mater
Treatment
Cost Data
rPROGRAM 2:
(Price Analysis
\suppi
/Spot Market /^
/and Met Coal /
I Market Deei- I
\£lon Rules V
/rapacity
/Growth and
(Mine Replace* |
Went Para-
\meters
Cost and
fPrice Functv
Parameters
i
XlMne f
( Hater Flow I
ypistributionsV
/Hal
/Tre<
Mater
Treatment
( Cost
V Parameters
I
i
/PROGRAM 3:\
I
Detailed
(nd Linearized/
(Supply Curves
by Major
Market
Coal Sulfur/^
Content /
Distribution
S Coal
/Utilization
Cost Data
V
I Cost
/Air
I Stai
Regional
Control
Standards
c
Transporta-
tion Costs
Demand
Forecast
/PROGRAM 4:
/ Market
\Fomulation
\-
Estimated
Supply and
Cost of
Supply
lUDULC 6
-------
MODULE 1: Separates the mines contained in the MESA file
into 27 supply region files which serve as
inputs to Module 3.
MODULE 2: Uses engineering estimates of production costs
for different types of model mines in dif-
ferent supply regions. It calculates minimum
required prices (MRP^ for these mines for
different labor productivities, as well as
different investment costs in pollution
control equipment resulting from different.
mine water flow volumes. These price esti-
mates are used to calculate a functional
relationship between minimum required price
(the dependent variable^, mine productivity
and required investment in pollution control
equipment per annual ton of coal produced 'the
independent variables).
MODULE 3: Uses the MRP function resulting from Module 2,
plus the regional mine files, to calculate
three supply curves for each of the supply
regions for the steam coal spot and contract
markets. The supply functions, specifying
cumulative potential supply at increasing
prices at the mine, are first calculated on a
mine-by-mine basis 'i.e., detailed) and then
linearized for use as an input to the linear
program used in Module 5 to balance supply and
demand in coal markets.
MODULE 4: Organizes all the information used in the
market simulation in a linear program which is
solved in Module 5. In addition to the lin-
earized supply curves for the 27 supply re-
gions, information on sulfur content distri-
bution and transportation and utilization
costs of different types of coal with
different sulfur content is provided.
MODULE 5: Provides a solution to the linear program set
up by Module 4.
MODULE 6: Takes the output from Module 5 and organizes
it into several reports to allow more rapid
evaluation of the results of the analysis.
The nature of the data and the computations performed by the programs
used in these modules are described in the following sections.
B-7
-------
III. REGIONALIZATION OF THE MINE FILE (MODULE 1)
The MESA file, which contains 1976 information on mine location, mine
type, status (i.e., open, temporarily or permanently closed^, number
of hours worked, and tons of coal produced, is organized into the 27
different supply regions. These supply regions, as shown in Tablp 1,
coincide largely with the supply regions used in the National Toal
Model.
IV. MINIMUM REQUIRED PRICES (MODULE 2)
The 1984 minemouth prices used in the analysis are based on estimated
production costs. Since actual production costs for individual mines
are not available, engineering estimates are made of unit investment
and operating costs for mine models, categorized by:
• Location (region);
• Mine type (underground or surface);
• Mine size (large or small);
• Remaining producing life (new or existing mine);
• Required water treatment costs for compliance with
EPA regulations.
Within each mine category, costs are specified to changp with:
• Mine productivity (tons of coal produced per mine
shift};
0 Level of water flow to be treated.
As shown in Figure 2, the input to the computer programs of Module ?
consists of a set of model mine parameters and water treatment costs;
the output consists of a set of parameter values, relating minimum
required price (MRP) to productivity and dollars of required invest-
ment in water treatment equipment.
The calculation of the parameter values which relate the MRP for a
mine to productivity and treated water flows is performed in two
stages:
• Calculate MRP's for each mine category for
different mine productivities and for different
flow levels;
• Estimate function parameters relating MRP's,
remaining required investment in mining equipment,
B-3
-------
TABLE 1
SUPPLY AND DEMAND REGIONS USED
FDR THE IMPACT ANALYSIS
MCfl REGIONAL CtNTROIDS WITH
rHEIGHT STATION ACCOUNTING CODES
Supply Region
Pennsylvania (PA)
Ohio (OH)
Maryland (HO)
West Virginia. north (NV)
Nest Virginia, Much (SV)
Virginia (VA)
Kentucky, cast (CK)
Tennessee (TM)
•Alabama (AL)
Illinois (It)
Indiana (IN)
Kentucky, vest (MCI
lo»a
ourl (HO)
•s (KS)
Arkansas IAR)
Oklahom (OK)
Toaas (TX)
North Dakota |ND)
Montana, east (EH)
Wyoming (wr) *^
Colorado (CS)
Utah (irn
Arizona (AZ)
Hew Mexico (NH)
Washington (WA)
Centroid
Johnstown
Cambridge
Lonaconing
Clar keburg
Blurfleld
Charleston
Appalechia
Ha sard
Clinton
Cordova
Ceniralia
Hunt Ingburg
Central City
OttiMwa
Clinton
Pittsbwrg
Russollviiio
TUlua
Con. icana
Hilton
Sidney
Casper
Carbondalc
Sunnyside
Nlm.low
Gallup
Cent ralia
Freight Station
Accounting Code
05001151
05001549
83909238
05000469
SS0039S5
OS003642
72401230
44402631
7240S420
72407457
07624120
72300344
44404124
07620241
69300162
40000130
49406080
69300839
69401202
0765769'
076S9225
07632236
19702416
19709106
02210286
022101 SB
14004514
Deasnd Region
Hew England
Centcoid
freight Station
Accounting Code
South Atlantic
MV Concord, K.H. 0690012!
MC Sprxnqfield, Mass. C2219122
Middle Atlantic NU Osweoo, N.Y. 62210667
UP Fittsbur?, Pa. 62204727
PJ Trenton, H.J. 62200203
VM Baltlnore, Md. 05000129
Norfolk. Va. 55001001
W Wheeling, H.Va. 05000161
Huntlnaton. H.Va. 05000768
CA Charlotte, H.C. 72402535
T Atlanta, Ca. 71213699
ria. (Barge only)
Palm Beach, fla. 71226590
Eaat north
Central
Bast South
Central
Hast North
Central
West South
Central
Mountain
Pacific
SP w.
ON Cleveland. O. 05002168
OH Coluefeus. O. 05001640
OS Marietta, O. 05001648
MI Detroit. Mich. 62206874
XL Peoria, 111. 62208884
IN Indianapolis, Ind. 62208463
Ml Milwaukee, Hie. 140C!<>«9
EX Winchester, Ky. 44402033
NX LouUville, Ky. 44401000
ET Kiorville. Tena. 44405683
NT Memphis. Term. 44406380
AH Biminahaa, Ala. 44407206
DM Minneapolis. Minn. 14005898
KM Topeka. Kan. 02202571
XA Das Moinea, la. 14500759
MO St. Louis. No. 69300752
AO Little Rock. Ark. 14507420
TX Dallas, Tex. 02209354
MM Billings. Mont. 07630841
UN Salt Lake City, Utah 80200081
CO Denver. Col. 07620977
AN Phoenix, Aris. 02214194
NO Seattle. Hash. 07602199
CM Oakland, Calif. 72100013
CS Los Angeles. Calif. 07612141
Wyoming has been separated into Wyoming Powder River Basin and Wyoming, Other Regions.
B-9
-------
annual wages, and cashflow per annual ton of coal
produced (the dependent variables) with productiv-
ity and mine water flow volumes 'the independent
variables).
A. Number of Different Sets of Model Mine Parameters Required
As shown in Table 2, model mine specifications for both old and new
mines required estimates for the following parameters:
• Investment and operating costs;
0 Fiscal parameters, such as royalty payment,
federal, state and local taxes;
• Water treatment costs.
A total of 50 different combinations of these three sets of parameters
resulted after careful analysis of available engineering estimates of
mine operating and investment costs, state taxes and royalties and
available information on mine water flows in conjunction with cost
estimates for different water treatment levels provided to us by the
EPA (see Table 2). Each pollution control level required evaluation
of such a set of 50 mines. The impact analysis considered the effect
of two different proposed BAT treatment levels relative to the exist-
ing BPT treatment level. Therefore, a total of 150 different mine
models had to be evaluated.
B. Model Mine Production Cost Parameters
Table 3 shows an example of the model mine investment and operating
costs used for the calculations described in this section.
Preliminary analysis showed that within the mine size ranges, judged
to be representative for the two major coal markets analyzed (contract
and spot), minimum required prices were only very slightly related to
mine size. Therefore, one representative mine size was used for each
mine category in the different regions.
As shown in Table 3, the cost parameters are divided into three
groups:
• Investment cost data for the mine facilities and
mining equipment;
• Operating cost data, together with the average
labor productivity for which they are specified;
• Operating cost ratios, specifying by how many
percentage points certain categories of operating
B-10
-------
NUMBER OF DIFFERENT COMBINATIONS OF MINING COST.
WATER TREATMENT COST AND FISCAL PARAMETERS REQUIRED
FOR THE ANALYSIS OF ONE CONTROL LEVEL'
TYPE OF PARAMETERS
Coal Mining Cost
REGIONS
1. NA Ug
Su
2M. CA & SA Ug
Su
co
i
^j
~* 4. MW Ug
Su
5. CM Ug
Su
6. Tx Ug
Su
L
S
L
S
L
S
L
S
L
S
L
S
L
S
L
S
L
S
L
S
OLD
Ao
Bo
Co
Do
Ao
Bo
Co
Do
Ao
Bo
Eo
Fo
Ao
Bo
Eo
Fo
Ao
Bo
Go
Fo
NEW
Al
Bl
Cl
01
Al
Bl
Cl
Dl
Al
Bl
El
Fl
Al
Bl
El
Fl
Al
Bl
Gl
Fl
Uater Treatment Cost
OLD
ao
bo
bo
bo
CO
CO
bo
bo
ao
ao
do
do
CO
CO
do
do
CO
CO
bo
bo
NEW
al
al
bo
bo
cl
cl
bo
bo
a2
a2
do
do
c?
c2
do
do
CO
CO
bo
bo
Royalties/
Taxes
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Z
2
2
?
Parameter Combination
Old Mines
Aoaol
Bobol
Cobol
Dobol
Aocol
Bocol
Cobol
Dobol
Aoaol
Boaol
Eodol
Fodol
Aocol
Bocol
Eodol
Fodol
Aoco2
Boco?
fiobo2
Fobo2
New Mines
Alall
Blall
Clbol
Dlbol
Alcll
Blcll
Clbol
Olbol
AlaZl
Bla?l
Eldol
Fldol
Alc21
Blc21
Eldol
Fldol
Alco2
Blco2
Glbo2
Flbo?
Model Mine #
Old
1
3
5
7
9
11
5
13
1
16
18
20
22
11
18
20
25
27
29
31
New
2
4
6
8
10
12
6
14
15
17
19
21
23
24
19
21
26
28
30
32
The control level could be the existing BPT regulation or any of the proposed RAT regulations considered by the EPA
-------
TABLE 2 - Cont'd
REGIONS
7. NGP
MA
ND/SD
Wy
^ Rocky M.
849. Nev/NMex.
10. WA
Coal Mining Cost
Old" New
Su
Su
Ug
Su
Ug
Su
Su
L
S
L
S
L
S
L
S
L
S
L
S
L
S
Ho
Fo
Io
Fo
Ao
Bo
Jo
Fo
Ao
Bo
Jo
Fo
Jo
Fo
Hi
Fl
11
Fl
Al
Bl
Jl
Fl
Al
Bl
Jl
Fl
Jl
Fl
Mater Treatment Cost
Old" ' '
bo
bo
bo
bo
CO
CO
bo
bo
CO
CO
bo
bo
bo
bo
New
bo
bo
bo
bo
CO
CO
bo
bo
cl
cl
bo
bo
bo
bo
Royalties/
Taxes
3
3
4
4
4
4
4
4
1
1
1
1
1
1
1'iiriiiiieter Combination
Old Mines
Hobo3
Fobo3
Iobo4
Fobo4
Aoco4
Boco4
Jobo4
Fobo4
Aocol
Bocol
Jobol
Fobol
Jobol
Fobol
New Mines
Hlbo3
Flbo3
Ilbo4
Flbol
Alco4
Blco4
Jlbo4
Flbo4
Alcll
Bocll
Jlbol
Flbol
Jlbol
.Flbol
Model Mine '
Old
33
35
37
39
41
43
45
39
9
11
47
49
47
49
New
34
36
38
40
42
44
46
40
10
12
48
50
48
50
-------
TABLE 3
1 ..1 I.I 1.
iiti.it : ••
L
1
I
J
I
1
1
1
1
I
1
1
1
1
1
J
I
I
1
1
\
J
J
I
I
1
•L
\
I
1
PL A MM IN
....
EXAMPLE OF
.I'l.l t i.i ii inn 1. 1 f'l II
1 n|' K'l (M. I'd -1 INl I'1!
"Ui*iH PilU, it r.li'il
U U t< -J l . U 0 U 0
. 140'i i . r?.m
ir./n i (i'ioo
I'. ,'0
. J £. C f '
. .. ' •'.'
'"•••U
. :-5 -o
."7>n
. VJ v!j
. •"! J i
.-. 'lv>
.i "J J
.. J l,ll
. .VA
.'•'«.-.'!
1 1 "' *, V
r ,
rle. • "*
.r. ..,.5
. vfu
• .•
. i. •. .j
•t; n
.vi. •<
/•..-: 7
V.ir /
. 1 ""'i'j
•-• j ' r
• -yc •
•-'::;,:'
G_vfARS
1 . U A 1 0
1 . Uti^O
i . J in; 0
j . ; .-60
i . j»t' n
1 . I 7.70
i . i?:.u
l .nl.T?
1 . .'ll"^3
i , ''-..31
1 .::•/'.. 5
1. '.IVl
j . ^u: ,
i . ". f.' f
1 .'ii'Ol
1 .•''•Ml
j . l''..67
1 . 'v!4S->3
i . ".il"»ci
1. . !.''! t •*'
J . ti /i'ȣl
1 .t>([Z J
1 . dll OP
1 . v . ".-3
• J .7iW.fi
1 .7't09
1. 77i,7
INPUT DATA
VI 1 '1 ''I i.l
A'-l '•>
l 01 II PHI NT
1.0000
1 .032P
1 .t'l-'.O
l.p-cMO
1 . ii >?0
\. I'vjn
1 . liV t$
\.OT- .50
i . on,:u
l . HV.JI*
l . Ii i!V
1 . i n t1:
J . J HVi
i. t IVi
l .i.'il
l .1:",;
1 . r .-.s
l . i .••::n
- . i '•"•:>
1 . I1!'"1!-
l.ii-i-i
l.U.-T'
1 l..f.V
i . \ '.':>:,
1 .IVi.iJ
i . i P-I •>
\ . i -I..-
1 . 19bl
i.:io:>i
FOR MODEL MINE COST ANALYSIS
ml l
" " "1
I
J
1
1
j
1
J
1
J
I
1
1
1
I
1
1
1
1
j
1
1
1
1
)
1
1
I
1
1
o-Ti rr
TiWt)
.1710
. 37.-.0
.1900
.20 50
. '.MC'O
. '.-'3 *0
pil70
.'.!£ f 0
. :J". *o
. 2''"i'!
. -M'i<#
' 1M .0
. •I'm:;
. V .'/.~<
. 3V "3
. ^.i ' j
.'I'jfiO
.'i '.:»:
.'!•.•> 7
.•137.'
•r. •>.'
. M ' .: »
. iiin'j
. 'j"1 .'>
.ii'l/.V
. w">;o
. HI'1*!
.6031
. A.-:'3
Sr.'"..I VIVj TV PARAMETERS
Tie K iT " I'LoNNINCrVL'ARS
NUMl'tR OF HCNE SIZES 1
HUH!."! fir MINE PROUUCTTVITIEIS
M'.I'LW Ol" MljlLJ LlVbS 3
. ".'n.. <: or r ois i
NUnpuR CF n.i]U3 5
f3.iC'il. Milt F INAMUAl. F'ARAMKTiTRS
f e'll K-;i" fn/' r--:.lV" "
S'V:-:KAKCL TAX RAT:
\ii-i Ml l KEDIT
l.;",:
-------
.uitiMji nirJi u v\ I. 'i MM UN I
MKJ: irniiiK 1 1 vi i u '.
lji'ii|ii)i ! IVMV v Mil '... MA
TABLE 3 - (cont'd.)
•JO
CO
I
•rii.'f " :.'!/[ uir.l/Yt.MO
:?."•: IT viMjj: - EXPLORATION cosrs
w.i.i-J.if.;-, r:r.
Pi;.! iif .-'.Llir EIJfLDtNGS
i-'ji1!;-'!! ,'Ji h'tir \.n~\~
r -J".' :•! fl! - Jn Yl A PS
Fi-j11. ' !• .1: I., 'it Ik1-.-
Hi' I r • • '•-• '.I1'", I OM'i
fi*'. ' ,.! . r'f'i-i .- I j VI IY
,'.'ll , 'Mil/ Yl iV-1 '
U I i,l'! H.i'S Hl.nCH LUNG
I • i • --jii.i! (r;N » : L
li'ln Kl :. I I ••.'•: 1>
I'm, '." I •'.'r.ki.Z
,,14:- J IjlJI ii i'CF
Pi" ..:if '-lil LlJiiTIJ'ii. rOS]_S
ji"vt '.-fi 1 •*! <:ii:"' I ;".", ~i«HT
or1: -. i. '.•: :J>, |:
.OiUlU
. ii :i P u
.211-8
. ''i?.'
. (•:.>> o
.0000
.0300
.0000
.ODOU
.0000
.0000
.iHiftO
.(Id no
. Dl'OO
. !} <} fl 0
. niuuj
illljO
. -K' I"1
. ll U fl '1
.0010
.00(111
.cnsio
, is ij n i.i
. fi fj U U
. f. o n ii
. prmo
.00f»0
• 7P'tJ
. 'L 7?:.'
,U.'99
.0000
.0090
. 0000
.0000
.0000
.coca
. orj'i.j
.0!>0i>
.OH II j
. 0 0 fl 0
. I'JIlllO
.00(10
.Ullfl'l
. (lilf;!1
• Ol'Uti
.0009
. I) II II (1
.0-00
.0000
.Oi-i'O
.00(10
. 0 ii 'J 0
-1.7-1
. iri.,1
.fi:.'r3
.0000
.0000
.0000
.0000
.0000
.onon
.once
.!>OuO
. (lUCO
.1)000
. ufiuO
.(IOI..O
.uoou
.PUOO
.Ofi li>)
. uaon
. r, ii i) o
.0"L-!l
. GDO'l
.O'SU'J
.0090
. CfiO't
~n. .- 'MI
. i it:.'
.o:1.,-!
-------
costs will be higher because of add-ons 'such as
indirect costs).
Working capital requirements are estimated to be equivalent to one
month of wage payments (i.e., 3% of annual wage payments).
As mentioned previously, the model mine costs are estimated for a
specific average labor productivity. To allow for analysis of minimum
required prices as a function of mine productivity, four other values
for mine productivity are specified for each model mine in the range
over which productivities for actual mines had been found to vary 'see
Chapter II.4 of Appendix A of this report for a discussion of the mine
productivity analysis).
C. Mine Flow Volumes and Water Treatment Costs
The analysis of available data on coal mine water flows demonstrates
that, when expressed in gallons per ton of coal produced, they are
generally larger for smaller mines. This is shown in Figures 3, 4,
and 5.
Average flow levels obtained from the data are highest in the three
Appalachian regions, somewhat less in the Midwest and Central West,
and much less (by a factor of about 10 ) in the rest of the coal
mining regions (Northern Great Plains, Texas, Rocky Mountains, Nevada,
New Mexico, Washington). Also, the range of possible flows is much
narrower in these "dry areas" than in the "wetter" Appalachian
regions, the Midwest and the Central West.
The increase in the MRP as a function of water treatment costs for a
given model mine is calculated for five different volumes, ranqing
from 1 to 10,000 gallons per ton for the wet Appalachian regions and
the regions in the Midwest and Central West and ranging from 0.1 to
1,000 gallons per ton for the other drier areas.
It is assumed that by 1984 all existing mines will have water
treatment installations in compliance with BPT standards. The
increase in required price for these mines will be through costs to
upgrade treatment levels from BPT standards to BAT standards.
Investment costs in BPT equipment are treated as sunk costs and are
not considered in the calculation of the minimum price per ton of coal
required by the mine to stay in production. However, investment costs
for BPT equipment are considered in the MRP calculations for new mines
since the investment has to be incurred to open up the new mine. The
estimated BPT investment costs for different model mines used in the
analysis are shown in Table 4. Operating costs are judged to be
negligible for BPT.
The investment and operating costs for the two different possible BAT
treatment standards, for which the potential economic impact is
estimated, are derived from data provided by the EPA.
B-15
-------
o
3
I
10
Mine Production, Tons/Year
FIGURE 3 COAL MINE WATER FLOWS - APPALACHIA
B-16
-------
10
,5 -
tnzr
'
r*
__ _
ET==
1 'V^ •
—
kJ
L_.
=
i —
^x
-
f
;
I
1
' M—
1 i — i—
—
r^:
_t-
-
\ 1
_^—_
1
^
— m — i
— i- — — —
i — ,_ — , — |
I
3
I
I
10°
10 t
" "" ' *~ ' ' •-*"• ••*••'- - V? — *—.__.- - ^ •
v-£^:^4^4.t±==
I
10"
10=
10C
10'
Mine Production, Tons/Year
FIGURE 4 COAL MINE WATER FLOWS-MIDWEST AND CENTRAL WEST
B-17
-------
10=
I I I T I i
10"
0
i
I
m
10"
10° 10C
Mine Production, Tons/Year
10'
FIGURE 5 COAL MINE WATER FLOWS - GREAT PLAINS AND WEST
B-18
-------
TABLE 4
BPT Investment Cost for New Mines (SMM)
(1978 S)
Small Large Large Large Large Small
UG SU UG Texas NGP SU
Mine size MMT/yr 0.025 0.600 0.600 2.00 4.00 0.025
NA.CA.SA
$MM 0.015 0.090 0.366 0.00375
S/TPY 0.610 0.150 0.610 0.150
MW,CW,TX
$MM 0.036 0.120
S/TPY 0.060 0.060
NGP,ROCKY,NEV.N.MEX
SMM 0.0015 0.054 0.360
S/TPY 0.060 0.090 0.090
B-19
-------
The BAT investment and operating costs for different flow levels as
calculated for the different representative mine sizes in the
different regions are shown in Tables 5, 6, 7, and 8. The costs
shown in these tables are not corrected for construction cost
variations between different regions. The multipliers used for this
correction are shown in Table 9.
D. Projection of Changes in Mining Conditions Between 1976 and 1984
Mining conditions are expected to change significantly between 1976
and 1984 because of:
• Changes in constant dollar unit costs of labor,
power and equipment;
t Changes in labor productivity.
As shown in Figure 6, constant dollar labor and equipment costs have
increased significantly since the early 1970s. Mining equipment costs
have increased at a very high average rate of 11% per year between
1973 and 1976, slowing to 1% per year between 1977 and 1979. The
labor cost index has increased relatively steadily at an average rate
of about 2% per year between 1968 and 1979.
Given the required continued growth in coal supply, labor costs can be
expected to continue to increase in real terms. As shown in Table 10
a continued annual growth rate of 2% is used for the period from 1978
to 1984. The actual average increase of labor costs between 1977 and
1978, resulting from the end-1977 settlement of wage negotiations
between the mine workers' union and coal mine management, is
calculated to be 14% over and above inflation.
Real increases in equipment costs (including pollution control
equipment) are judged to gradually decrease from about 11% per year
between 1975 and 1976 to about 0.5% per year between 1978 and 19R4.
The high increase in equipment costs - 11% per year - between 1973 and
1976 are judged to have been created by temporary bottlenecks due to
the sudden renewed interest in coal as a fuel. This renewed interest
was brought about by increases in imported crude oil prices at the end
of 1973; these bottlenecks are expected to gradually disappear.
Power costs are projected to increase at an average of 2% per year,
reflecting continued real increases in the costs of all fuels (see
Table 10).
As shown in Figure 7, labor productivity in both underground and
surface mines has decreased dramatically during the last decade.
These decreases are mainly attributable to regulations protecting the
mining environment and mine workers' health and safety. This decline
is expected to continue until 1981, especially for surface mines in
B-20
-------
TABLE 5
to
i
to
BAT COSTS
(Without
Flow G/T
Flow MMG/D
Level 2
Log of Flow
INV $MM
OPC $MM/Y
Level 4
Log of Flow
INV $MM
OPC $MM/Y
FOR SMALL MINES (0.025 MMT/yr) WITH
Correction for Differences in Mining
104
0.685
-.1643
.0350
.0115
-.1643
.2575
.0421
103
0.0685 0.
-1.1643 -2.
.0350
.0063
-1.1643 -2.
.1494
.0271
DIFFERENT
Region or
102
00685
1643
0350
0060
1643
1179
0270
FLOW VOLUMES
in Mine Type
10
0.000685
-3.1643
.0350
.0060
-3.1643
.1100
.0270
1.0
0.0000685
-4.1643
.0350
.0060
-4.1643
.1100
.0270
-------
TABLE 6
BAT COSTS FOR LARGE UNDERGROUND MINES (0.6
(Without Correction for Differences in
Flow G/T
Flow MMG/D
Level 2
7 Log of Flow
to
INV $MM
OPC $MM/Y
Level 4
Log of Flow
INV $MM
OPC $MM/Y
104
16.438
1.2158
.0350
.0485
1.2158
.9039
.1979
103
1.6438
.2158
.0350
.0162
.2158
.3433
.0580
MMT/yr) WITH DIFFERENT FLOW VOLUMES
Mining Region or in Mine Type
102
.16438
-.7841
.0350
.0075
-.7841
.1772
.0299
10
.016438
-1.7841
.0350
.0060
-1.7841
.1236
.0273
1.0
.0016438
-2.7841
.0350
.0059
-2.7841
.1187
.0250
-------
TABLE 7
bo
t
BAT COSTS FOR LARGE STRIP MINES (2.0 MMT/yr) WITH DIFFERENT FLOW VOLUMES
(Without Correction for Differences in Mining Region or Mine Type)
Flow G/T
Flow MMG/D
Level 2
Log of Flow
INV $MM
OPC $MM/Y
Level 4
Log of Flow
INV $MM
OPC $MM/Y
103
5.479
.7387
.0350
.0278
.7387
.5481
.1030
102
.5479
-.2613
.0350
.0107
-.2613
.2410
.0393
10
.05479
-1.2613
.0350
.0061
-1.2613
.1441
.0267
1.0
.005479
-2.2613
.0350
.0060
-2.2613
.1171
.0260
0.1
.0005479
-3.2613
.0350
.0050
-3.2613
.1100
.0250
-------
TABLE 8
w
to
*.
BAT COSTS FOR LARGE STRIP MINES (4.0 MMT/yr) WITH DIFFERENT FLOW VOLUMES
(Without Correction for Differences in Mining Region or Mine Type
Flow G/T
Flow MMG/D
Level 2
Log of Flow
INV $MM
OPC $MM/Y
Level 4
Log of Flow
INV $MM
OPC $MM/Y
103
10.96
1.0398
.0350
.0392
1.0398
.7455
.1535
102
1.096
.0398
.0350
.0138
.0398
.2988
.0495
10
0.1096
-.9602
.0350
.0068
-.9602
.1629
.0283
1.0
0.01096
-1.9602
.0350
.0062
-1.9602
.1207
.0282
0.1
0.001096
-2.9602
.0350
.0060
-2.9602
.1200
.0280
-------
TABLE 9
Region
NA
CA
SA
MW
CW
Texas
NGP
Rocky Mtns.
Nev., N. Mexico
MULTIPLIERS TO CORRECT FOR REGIONAL
DIFFERENCES IN BAT INVESTMENT COSTS
Multipliers
All Surface
and New Underground
1.32
1.32
1.28
1.12
1.08
1.00
1.00
1.36
1.26
Existing
Underground
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
B-25
-------
FIGURE 6
:i
•i
• • 1 .• 6 •-
• • • I ; l-.
r
-r.2-
s
T
1
1
N
It
£ D'8
X
' ' D'.i
Oil
i
^r-
i
4
50 51
1
-
y*-*^
*^~-
™ S3
HIS'
*/
|**>S
•*——"*"
5* c-.-
55
FORICAL
z~*~
^
~-* —
^•"
lj u ,1
5A 57
MINING LABOR C
^..i..
J 1
56 ay
& £,
V""----V
*-— -•*_. J
.!„
60 61
OST AND
~ a... ...j
* ~---~
i . - r
62 63
YEAR
MINING EQUIPMENT COST
! :
t^-r. -ja~.'...T3
-H
'"""
* M1N1N
A .TOTAL
V UNlifc.h
64 65
^•-"~i-^
^— __^"
INDICE
jS_— «•
", LABOR COST 1NȣX
.MIMING -EQUIPHEW-T .CU51 .1
iKQIIND ri;IN]NG fciUUlPMEMT
66 67 ^ 69
70 71
B (1972=
.~*-^
~^i
iJ!>tX . . .
COS! JUT
1 .1. ._
72 73
=100)
"?/
../•./..
. /
i/ ^
ft _-"<
(i--"
L"X
1 —
74 75
•
>7
. V
r-"""*'
_...>——-*.
*• -*'"
:
•
.. .-J_ .L L. .J
76 77 i78 79
60
-------
TABLE 10
PROJECTED ESCALATION IN COSTS AND LABOR PRODUCTIVITY
USED IN THE MODEL MINE COST ANALYSIS
(1977 is 1.00)
Year
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986 and
following
year
Wages
1.000
1.140
1.157
1.157
1.180
1.203
1.228
1.253
1.277
1.02x
Power
Costs
1.000
1.020
1.040
1.061
1.081
1.104
1.126
1.149
1.172
(1.02x)
Mining
Equipment
Cost
1.000
1.032
1.048
1.056
1.061
1.067
1.072
1.077
1.083
(1.005x)
Underground
1.000
0.990
0.985
0.980
0.990
1.000
1.010
1.020
1.030
(xl.Ol)
Productivity
Surface
East. Regions
1.00
0.90
0.85
0.80
0.808
0.816
0.824
0.832
0.841
(xl.015)
Surface
Other Regions
1.00
0.95
0.92
0.90
0.914
0.927
0.941
0.955
0.970
(xl.015)
(1) Surface other includes strip mines in the West, Midwest and Texas
-------
FIGURE 7
HISTORICAL PRODUCTIVITY INDICES FOR UNDERGROUND AND SURFACE MINES (1967=100)
ea
N5
00
* .UNDERiROUND.HINE.PROI JC1IVITY
-------
the Eastern regions (the relatively small surface mines in these
regions are expected to suffer most from the Surface Mine and
Reclamation Act; the difficult terrain will make it costly to restore
mined areas to conditions required by the Act). Beyond 1981, produc-
tivity is expected to rise again at an average rate of 1.5*. per year,
mostly due to a maturing work force.
E. Minimum Required Price Calculation and MRP Function Estimation
Given the input parameter values for mining, productivity and water
treatment costs discussed in the previous sections, the minimum
required prices could be calculated for different mine categories.
The general algorithm of the computer program developed for this
purpose is shown in Figure 3. Given input parameter values for a
specific type of mine model, the program can calculate minimum
required prices for:
• Different planning years in the future;
t Different treatment cost levels;
• Different remaining lives for the mine;
• Different mine labor productivities.
Originally the computer program allowed investment costs and operating
costs to change with mine size. However, analysis of the sensitivity
of MRPs to changes in mine size and productivity revealed that MRP
variations are dominated by productivity variations. Mine size was
subsequently dropped as an independent varirable in the calculations.
This significantly reduced the number of computations required for the
analysis. As shown in Figure 8, the results are ranges of values for:
• Minimum required price (in dollars per ton);
• Annual wage payments Mn dollars per annual ton^;
• Cashflow generated in the planning year (in dollars
per annual ton);
t Required investment in mining equipment in the
planning year (in dollars per annual ton>.
These calculated values (i.e., the dependent variables* are retained
in relation to the pre-specified values of the independent variables,
which are:
• Labor productivity (in tons per miner shiftV,
• Mine water flow volumes (in million gallons per
year).
B-29
-------
FIGURE 8
FLOW DIAGRAM OF MODEL MINE COST ANALYSIS PROGRAM
Read data
Next water flow
I
Next planning
I
Next remaining
I
Calculate investment requirements
Project after- tax operating costs
for different productivities
Calculate minimum required price,
wages, cashflow, required remain-
ing investment, treatment opera-
ting and investment costs
Store results
f
Next remaining life?
|
Next planning year?
I
Next water flow level?
*
NO
*
STOP
B-30
-------
These data are used to estimate parameter values for MRP functions as
described in Section H below.
F. Details of the Minimum Required Price Calculation
The minimum required prices are calculated subject to the following
two conditions:
CONDITION 1: After tax revenues should cover all costs plus a return
1) Sum (PV (INV. (l-credit}+ WC + OC . (1-t) - DEPP . t^ =
Sum (PV (PROD . MRP . (1-roy - sevtax^ . 'l-'l-depP . t ^
CONDITION 2: In each year of the planning period
2) [PROD . MRP . (1-roy - sevtax) . (1-fl-depl) . O - OC . 'l-t> +
D'EPR . t"1 = T(PROD . MRP . (1-roy - sevtax)] . CONSTANT
where,
MRP the minimum required price in a given year
of the mine's life.
INV the investment in a given year, including
investment in water treatment equipment.
OC the • operating costs in a given year,
including water treatment costs.
DEPR the depreciation allowance in a given year.
depl the depletion allowance rate.
PROD the production in a given year.
credit the investment credit rate.
t the federal corporate income tax rate.
roy the royalty rate.
sevtax the state and local severance tax rate.
PV the present value operator for year t.
Sum summation over the remaining mine life.
WC the working capital requirement (treated as
an investment occurring in the first year
of the mine's life).
B-31
-------
The actual value of the "CONSTANT" will result from the calculations.
For example, if the remaining mine life is 30 years, then we will have
30 equations of type (2) and one equation of type '1^, or ?l equa-
tions, which in order to have a unique solution will need to have 31
unknowns: 30 MRPs (one MRP for each year) and the "CONSTANT."
The calculation of the MRPs and the "CONSTANT" is broken down into
four steps:
STEP 1: Calculate: P*t I*, d*, where:
P* = PROD . (1 - roy - sevtax^
I* = Sum . fPV (INV . '1 - credit) + WO)
d* = (i - (1 - depl) . t )
STEP 2: Calculate: A* = Sum 'PV (OC '1 - O -
DEPR . t ))
STEP 3: Calculate: CONSTANT = d* - d* . A*/ (A*
+ I*)
STEP 4: Calculate: MRP = 'OC . '1 - t.) - DEPR .
t)/ (P* . (d* - CONSTANT) =
A/ (P* . {D* - CONSTANT, WHERE:
A = OC . (1 - t) - DEPR . t
Substitute: (d* - CONSTANT) . = A* . d*/
(A* + I*)
MRP = A . (A* + I*) / P* . A* . d*
To relate changes in costs with changes in productivity, a standard
production function is used:
C/C* = (PRTY*/PRTY)F, where:
C = the unit operating or investment costs 'in
dollars per ton) related with a produc-
tivity PRTY (in tons per miner shifO.
C* = the unit operating or investment costs
related with the benchmark productivity
PRTY*.
F = a factor with a value between zero and one.
B-32
-------
Higher labor productivity within the same mine implies lower unit
labor costs and higher mining equipment costs. Therefore, a positive
value of F relates labor costs with productivity and a negative valup
of F relates equipment costs with productivity. In the analysis a
value of +0.8 was used to relate unit equipment costs with labor
productivity; a value of -0.8 was used to relate unit labor costs with
labor productivity.
G. Sample Output
A detailed example of the results of the calculations described in the
two previous sections is shown in Table 11. The results are for a
large underground mine in Northern Appalachia, where mine water flows
are mainly acid (see Table 3 for input parameter values). The
treatment costs are for level 4.
Table 11 shows the results of the calculations for control cost levels
corresponding with a high, medium and low flow for an existing mine
and for a new mine. The price difference, because of different flows,
can be as high as $.50/ton. The price difference because of different
productivities can be as high as $18.9/ton.
H. Estimated Function Parameter Values for MRP, Wages. Cashflow and
Required Investment in Mining Equipment
The last step in the calculations performed by Module 2 consists of
the estimation of function parameters relating the dependent variables
- minimum required price, wages, cashflow and required investment in
mining equipment (all in dollars per annual ton produced) - with the
independent variables - productivity (in tons per miner shift) and
water flow level (in gallons per annual ton of coal produced^.
The following functional formula provided a good correlation between
the values of dependent variables calculated by the program described
in the previous section, and the pre-specified values of the indepen-
dent variables:
DV = LogL) + b . Log(FLOW) + c . Log'PRTY)
a
DV = the dependent variable (MRP, wages,
cashflow, investment in mining equipment,
all in dollars per ton)
PRTY = productivity (in tons per miner shift^
FLOW = mine water flow volume (in million gallons
per year)
a,b,c = estimated parameter values
The parameter values a, b, and c are estimated using a standard,
ordinary least-squares regression program.
B-33
-------
TABLE 11
RESULTS OF MODEL MINE ANALYSIS
LARGE UNDERGROUND MINES IN NORTHERN APPALACHIA: m
REQUIRED PRICE. WAGES, CASHFLOW. INVESTMENT. TREATMENT COSTSu;
HIGH FLOW (10,000 G/T)
Price
Wages
Cashflow
(2)
Investment* '
Treatment Cost:
Operating
Investment
Price
Wages
Cashflow
(2)
Investment '
Treatment Cost:
Operating
Investment
NEW
Labor
Low
7
43.1
13.4
6.7
35.2
0.38
2.1
42.6
13.4
6.6
35.2
0.06
0.42
MINE
Productivity
Medium
10
35.0
10.1
6.7
43.6
0.38
2.1
Medium Flow
34.4
10.1
6.6
43.6
0.06
0.42
OLD MINE
(T/MSH)
High
24
24.2
5.0
8.6
77.7
0.38
2.1
(100 G/T)
23.8
5.0
8.5
77.7
0.06
0.42
Labor
Low
7
41.5
13.4
6.0
15.1
0.38
1.6
41.2
13.4
5.9
15.1
0.06
0.32
Productivity
Medium
10
33.3
10.1
5.9
20.1
0.38
1.6
33.0
10.1
5.9
20.1
0.06
0.32
Low Flow (1 G/T)
Price
Wages
Cashflow
(2)
Investment* '
Treatment Cost:
Operating
Investment
42.6
13.4
6.6
35.2
0.05
0.28
34.4
10.1
6.6
43.6
0.05
0.28
23.8
5.0
8.5
77.7
0.05
0.28
41.2
13.4
5.9
15.1
0.05
0.21
33.0
10.1
5.9
20.1
0.05
0.21
High
24
22.8
5.0
7.9
40.4
0.38
1.6
22.^
5.0
7.8
40.4
0.06
0.32
22.4
5.0
7.8
40.1
0.05
0.21
1
All costs are in dollars per annual ton.
"Required investment in mine equipment (and facilities) to open the mine or
to keep the old mine producing.
B-34
-------
Table 12 shows the estimated values for a, b, and c for the same mine
model used as an example in the previous section. As shown by the
values of the correlation coefficients (see Table 12\ the values of
MRP, wages, and required investment in mining equipment, estimated by
the functions, deviated less than !<& from the original values; the
estimated cashflow values deviated less than 7% from the orgininal
values.
V. REGIONAL SUPPLY CURVES (MODULE 3^
Module 3 operates on the regional mine files with production and pro-
ductivity information for individual mines in 1975 obtained from the
MESA file. As shown in Figure 9, Module 3 performs six main func-
tions:
• Organizes the mines in each region into three major
categories (met coal, contract and spot market
steam coal);
• Changes the population on these regional files by
retiring old mines and opening new mines;
• Estimates mine waterflows for the individual mines
and calculates the required investment and opera-
ting costs to treat these flows;
• Calculates estimates for minimum required price,
wages, cashflow and required investment for these
mines;
• Rank orders the resulting regional mine population
by increasing price;
• Fits a linearized function to the resulting list of
mines, relating increasing price with increasing
production from the region.
A. Separation into Major Coal Markets
As explained in the introduction it is assumed that the steam coal and
the metallurgical coal markets operate completely independently of
each other. The steam coal market consists of the spot and contract
markets; these markets are assumed to have only limited interaction
with each other.
In order to create three different mine files for each region, two
decision rules are used:
• All underground mines with productivities of less
than nine tons per miner shift are assigned to the
met coal file;
B-35
-------
TABLE 12
ESTIMATED FUNCTION PARAMETERS FOR
LARGE UNDERGROUND MINES ON
NORTHERN APPALACHIA
New Mines
Dependent Variable
Minimum Required Price
Wages
Cashflow
Investment In
Mining Equipment
Existing Mines
Dependent Variable
Minimum Required Price
Wages
Cashflow
Investment In
Mining Equipment
Function
a
104.9542
63.6449
4.2516
10.0138
Function
a
105.2204
63.6449
3.603
3.1785
Parameters
b
0.001538
0.0
0.001842
0.000
Parameters
b
0.001104
0.000
0.000431
0.000
c
-0.4731
-0.800
0.2086
0.6419
c
-0.4934
-0.800
0.2316
0.800
Correlation Coefficient
0.99474
1.000
0.9327
0.99966
Correlation Coefficient (R )
0.99409
1.000
0.93239
1.000
1
The function parameters a,b, and c relate the dependent variables (MRP, wages,
cashflow or investment) with the independent variables, mine labor productivity
(PRTY) and log of flow.
DV = log x log of flowb x PRTYC
a
B-36
-------
FIGURE 9
MAIN FUNCTION OF MODULE 3:
ESTIMATION OF REGIONAL COAL SUPPLY BY PRICE in 1984
Met Coal Mines
in 1976
Production Data for Existing
Mines in 1976
Spot Marker Mines
in 1976
Old & New Miies in 1984
O
Old & New Mir
Separation Rule
•Contract Market Mines
in 1976
Mine Attrition Rate
New Productive Capacity
New Mine Parameters:
Size, Productivity
Flow Distribution (T
es with Flows
.Price and Cost Function (4)
Parameters
Old & New Mines with Price
and Cost Data
-Rank Order by Ascending Price
File with Mines in 1984 by
Ascending Price
•Linearize
Linearized Supply Curves
B-37
-------
• All small mines - mines with less than 51,000 tons
of production in 1976 - are assigned to the spot
market file.
These decision rules are admittedly crude. However, the 1976 supplies
from these met coal mines obtained in this manner for the different
supply regions corresponded remarkably well with the available infor-
mation on actual supplies from the same regions.
The hypothesis that underground mines with low productivities are most
probably metallurgical coal mines is based on two considerations.
Firstly, metallurgical coal is a relatively scarce resource, occurring
mostly in deeper and thinner seams that are generally more difficult
to mine. Generally, therefore, met coal mines have lower labor pro-
ductivities and correspondingly higher production costs. Secondly,
underground mines with productivities 'as measured in this analysis^
below nine tons per man-day are not competitive at today's prices.
As shown in Table 13, the total amount of coal ourchased in spot
markets by utilities in 1976 is about double the amount produced by
small mines assigned to the spot market file, using the decision rule
mentioned above. The differential amount of coal supplied to the spot
market in the analysis is taken from contract mines, using the
percentages shown in Table 14 to estimate the relative volumes of spot
market coal from the different supply regions. These additional
volumes are assumed to be available at variable costs because fixed
costs are already paid for by contract sales.
B. Changes in the Mine Population
Steam coal supply is expected to grow at an annual rate of 4-5*,
between 1979 and 1984. In the analysis it is assumed that all of the
increased production will come from large mines. Small mines 'i.e.,
those with less than 50,000 tons per year production) will be more
affected by the various environmental regulations 'including the water
control standards analyzed here). Also, the majority of coal use in
the future is expected to be by electric utilities, which will be more
interested in longer term contracts because of their emphasis on
security of supply.
The estimates of the number of large mines closing between 1976 and
1984 are calculated using annual attrition rates obtained from the
MESA Mine File analysis. The resulting attrition rates, compounded
for the total period of nine years between 1976 and 1984, are shown in
Table 14.
In the simulation of changes in regional mine populations, the
smaller, least productive mines from the 1976 mine file are retired
first. For mine sizes smaller than 200,000 tons, the size of the
replacement mine is assumed to be twice the size of the old mine. The
B-38
-------
TABLE 13
Origin
Pennsylvania
Ohio
Maryland
Northern WV
Southern WV
Virginia
E. Kentucky
W. Kentucky
Tennessee
Alabama
Illinois
Indiana
Iowa
Missouri
Kansas
Oklahoma
Texas
North Dakota
Montana
Wyoming
Utah
N. Colorado
S. Colorado
Ari zona
New Mexico
Washington
TOTAL
CONTRACT AND SPOT MARKET COAL SUPPLIED TO ELECTRIC UTILITIES IN 1976
(000 Tons)
Total
ia 40,940
40,135
2,412
IV 26,981
IV 17,558
13,642
:y 52,722
,y 49,440
7,225
13,940
49,140
23,817
601
4,056
2,317
2,503
11,867
ita 10,031
24,958
25,781
4,632
lo 1 ,487
lo 4,174
10,258
i 8,465
i 3,600
452,682
Contract
29,058
31,265
902
23,402
15,883
11,589
39,225
44,644
6,154
9,967
45,897
20,078
444
4,019
1,898
1,832
11,855
10,011
24,858
24,956
4,391
1,041
4,053
10,258
8,465
3,600
389,745
Spot
11,882
8,870
1,510
3,579
1,675
2,053
13,497
4,796
1,071
3,973
3,243
3,739
157
37
419
671
12
20
100
825
241
446
121
0
0
0
62,937
Percent
Spot
29.0%
22.1%
62.6%
13.3%
9.5%
15.0%
25.6%
9.7%
14.8%
28.5%
6.6%
15.7%
26.1%
0.9%
18.1%
26.8%
0.1%
0.2%
0.4%
3.2%
5.2%
30.0%
2.9%
0.0%
0.0%
0.0%
13.9%
Spot as Percent
of
Total Spot
18.9%
14.1%
2.4%
5.7%
2.7%
3.3%
21.4%
7.6%
1.7%
6.3%
5.2%
5.9%
0.2%
0.1%
0.7%
1.1%
0.02%
0.03%
0.20%
1.3%
0.40%
0.70%
0.20%
0.00%
0.00%
0.00%
100.00%
: FPC Form 423
B-39
-------
TABLE 14
ATTRITION RATE PARAMETERS
The following tables specify the attrition rates used for mine retire-
ment and replacement simulation when projecting changes in the contract mine
population between 1976 and 1984.
The data are organized in three sets of four records (a record is a
line in the table). The first set specifies attrition rate parameters for
underground mines and the last set of records is for surface mines.
The first line in each set of records has the attrition rate for the
corresponding size range; the second and third records give, respectively,
the lower and upper limits of the mine size range; and the fourth record gives
the maximum mine size for replacement mines. The last or thirteenth record
in the table specifies the increase in mine size of replacement mines: mines
with sizes of up to half a million tons, which were closed between 1976 and
1984, were assumed to be replaced with mines twice the size of the closed
mines.
Annual attrition rate:
Mine size: lower limits
Mine size: upper limits
Maximum Mine Sizp:
Annual attrition rate:
Mine size: lower limits:
Mine size: upper limits:
Maximum Mine Size:
Replacement mine size
Multiplier:
0.52 0.35
0.05 0.10
0.10 0.20
in.o in.o
0.19 0.13 0.131
0.20 0.50 1.00
0.50 1.00 10.00
10.0 10.0 10.0)
underground mines
0./5
0.05
0.10
10.0
2.00
0.59
0.10
0.20
10.0
2.00
0.35
0.20
0.50
10.0
2.00
0.25
0.50
1.00
10.0
1.00
0.19"
1.00
10.00
10.0
1.00
.surface mines
Notes mine size in millions of tons per year.
B-40
-------
productivity of the replacement mine is obtained by the sampling of a
productivity distribution derived from analysis of the MESA tape.
These productivity distributions are shown in Table 15.
The simulation of the closing of old mines is continued until all are
retired. Simultaneously, the opening of replacement, mines is
simulated, insuring that the sum total of productive capacity for thp
individual mines does not exceed total projected productive rapacity.
If, after the retirement of old mines and the opening of replacement
mines, additional productive capacity is still needed, the simulation
continues to add new mines. The simulation of new mines starts at thp
high end of the mine size range and gradually works toward the low end
until all projected new capacity for the region is realized. The
percentage distribution of mines over different mine size categories,
determined for the 1976 mine population (see Table 16^, is used to
allocate new mines to different size classes. The result of the
simulation is a file containing old, replacement, and new mines.
Table 17 shows the projected increases in capacity for the differpnt
regions.
For each mine added to the mine file the annual water flow to be
treated is estimated. This is done by calculating an average flow as
a function of the size of the mine and by sampling a distribution
which specifies the spread around that average. The parameter values
for, and functional relationships between, average flow and mine size,
and the distributions of the spreads around the average are shown in
Table 18.
Given the size and productivity of the mine and the watpr flow
estimates, the minimum required price, wage payments, cashflow,
required investment in mining equipment, and additional required
investment in water treatment equipment can be calculated. This is
done with the functions estimated in Module 2 by modelling the mine
cashflow analysis, as discussed in Section TV.
C. Detailed and Linearized Supply Curves
The result of this sequence of calculations is a list of old and new
mines in the region under analysis. This list, after being ordered by
ascending price, is stored for future use in the impact analysis.
Table 19 shows an example of this mine file for Pennsylvania with
estimated water treatment costs specified for BAT level 4. This list
of mines, specifying cumulative potential coal supply in 1984 by in-
creasing price, is essentially the projected "detailed" supply curve
for that region in 1984.
As shown in Table 20, linear approximation of the detailed supply
curve is used in the coal market simulation model. The linear
'including replacement mines
B-41
-------
TABLE 15
MINE PRODUCTIVITY DISTRIBUTIONS
The following tables (15.1, 15.2, 15.3, 15.4) contain data on mine
productivity distributions, used to obtain replacement and new mine pro-
ductivity estimates as discussed in Chapter I.
The first three lines in the tables contain parameter values to cal-
culate average productivity (in tons per miner shift) as a function of
mine size for underground mines, a dummy mine type and surface mines,
respectively. The next three lines contain, .nine fractile distributions
of the ratio of the actual versus average calculated productivity. The
fractiles for these distributions are-0.0, 0.01, 0.05, 0.25, 0.50,
0.75, 0.99, and 1.0.
The first parameter in each of the first three lines of the tables
specify the average tons per mineshift of mines with an annual production
(in million tons per year) as specified by the second parameter value.
The third and fourth parameter values (P3 and P4) allow calculation of the
average productivity (PRODTY) for a mine with a size of S million tons per
year by
PRODTY = P3 + P4 . In (In S)
B-42
-------
TABLE 15.1: LARGF UNDERGROUND AND SURFACE MINES III IMF EAST
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
(!)
2)
(3)
1
(2)
(3)
0)
(2)
(3)
(1)
(2)
(3)
0.60
0.00
0.35
0.65
0.00
0.39
0.65
0.00
0.85
0.65
0.00
0.85
12.0
0.00
25.0
0.60
0.00
0.40
15.
0.
30.
0.71
0.00
0.41
15.
0.
100.
0.71
0.00
0.85
15.
0.
60.
0.7]
0.00
0.85
200. -2.72 9.50 6.0
0. 0.00 0.00 0.0 average produr. i« i LV para'-flrr-.
200. -1.00 15.79 10.0
0.650.75 0.95 1.201.90 1.95 2.00 Drodllcll », ty <"s - ,-,!,.-Li0-s ( ,< ,
o.ooo.oo ooo o.ooo.oo o.oo o.oo „? «r .,L , '. «H, Z7i, , •
0.460.67 0.87 1.041.30 1.50 1.50 ratio of «,vei ,•-••„, oducliviL-, ,
TABLE 15.2: LARGE UNDERC-WiD W2 SUT/'CE MINES
IN THE MI QUEST AND CENTRAL WEST
200. -2.72 11.12 6.0
0. 0.00 0.00 0.0 average product i>-i 'y parameters
200. -1.00 18.50 10.
0.84 0.94 1.00 1.C3 1.16 1.29 1.35
0.00 0.00 0.00 0.00 0.00 0.00 0 00 productivity diM t ibutions (as a
0.47 0.67 0.88 1.03 1.33 1.48 1.50 ratio of average productivity)
TABLE 15.3: LARGE UNDERGROUND AND SURFACE MINES
IN THE NORTHERN GREAT PLAINS
200. -2.72 11.12 6.0
0. 0.00 0.00 0.0 averanc productivity parameters
5000. .-226 .060 152.25 30.
°:« S-S J:§S i:!S 0:Jo i:3 J:™ ^^^ distributions {u .
0.870:920.99 K03 1.06 1.17 1.20 ratio of average productivity)
TABLE 15.4. LARGE UNDERGROUND AND SURFACE MINES
IN THE ROCKY MOUNTAINS, SCUTHl.'EST AND TEXAS
200. - -2.72 11.12 6.0
0. 0.00 0.00 0.0 average productivity parameters
5000. -24.76 39.94 20.
0.840941.00 1.031.16 1.29 1.35 _ ., _t ,„, tu ., tf.hllt.nn, / „ .
n nn n nn n nn n 11 n n« n nn n nn productivity distributions Us a
0.00 0.00 0.00 0.10 O.Ow 0.00 0.00 ratio of averagc productivity)
0.86 0 93 0.99 1.03 1.06 1.17 1.20
(1) Underground mines
(3) Surface mines
B-43
-------
TABLE 16
1. ALA
3. ARI
4. ARK
5. COLO
6. ILL
8. IOWA
7. INO
9. KAN
14. EKY
15. WKY
12. MD
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
EXISTING PRODUCTION OF CONTRACT MINES IN 1976
IN DIFFERENT SIZE RANGES
(Size
0.05-0.1
nd 0.0
2.706
nd
nd 0.
0.371
nd 0.
0.0
-------
Table 16 cont'd
0.05-0.1 0.1-0.2 0.2-0.5 0.5-1.0 1.0-10.0 TOTAL
• j. MISS
16. NMEX
19. OKLA
18. OHIO
20. PENN
28. MONT-E
'" ND
SD
22. TENN
23. TEX
24. UTAH
25. VA
10. WVA-N
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
Underground
Surface
0.
0.
0.
0.48B
0.087
3.476
0.835
5.956
0.422
0.540
0.
0.
3.213
3.076
.100
0.500
0.
0.101
1.022
0.305
4.646
2.461
8.551
0.
0.128
0.936
1.743
0.960
2.330
3.109
.100
0.
0.
0.
0.222
0.320
6.329
6.602
11.120
0.288
0.747
0.516
1.419
1.289
4.2C7
2.097
.300
0.
3.058
0.830
0.865
0.
5.649
5.507
6.111
8.129
0.
0.
(0.844)
0.282
4.364
0.692
0.514
.80
.45
0.
2.636 5.797
0.830
7.979 8.947
0.
1.618 3.292
4.211 10.4
7.701 27.1
9.613 25.05
2.183 34.67
0.
25.792 25.795
0.
11.389 12.269
1.881
3.349
0. 0.
19.216 14.2
1.038 7.6R7
0.
9.989
8.463
15.00 16.715
6.350 7.026
B-45
-------
Table 16 conc'd
11. WVA-S
26. WASH
29. WYO
Total
Underground
Surface
Underground
Surface
Underground
Surface
0.05-0.1 0.1-0.2 0.2-0.5
.150 .050 .300
.100 .100 .200
0.
0.
0.
0.
34.465 50.877 69.953
0.5-1.0
0.500
.550
2.417
83.577
1.0-10.0
25.40
9.90
4.084
26.479
301.289
TOTAL
26.779
10.991
0.
4.084
0.203
28.985
540.151
B-46
-------
TABLE_J_7
ESTIMATED POTENTIAL INCREASES IN MINING CAPACITY UNTIL 1984
1-PENN
2-OHIO
3-MARYLAND
4-rf. VA. -N
5-iJ. VA. -S
6-VIRGINIA
7-E. KY.
8-TENNESSEE
9-ALABAMA
10-ILLINOIS
11-INDIANA
12-W. KY.
' OWA
It-HlSSOURl
15-KANSAS
16-ARKANSAS
17-OKLAHOMA
18-TEXAS
19-N. DAK
2CI-S. DAK
21-MONTANA E.
22-MONTANA U.
23-UYOMING
24-COLO - N
25-COLO - S
26-UTAH
27-ARIZONA
28-N. MEXICO
29-WASHINGrON
NEW
UG
15.9
16.6
1.8
1.5
13.8
3.1
10.4
0
4.0
22.6
2.0
3.5
2.0
MINES
SURFACE
0.4
2.6
1.0
8.3
1.6
0
21.5
0
4.3'
9.6
10.6
4.5
3.1
MHT Per
Year
REPLACEMENT MINES f
UG
4.6
1.5
0.05
2.2
3.5
3.4
6.8
1.0
0.41
4.2
0.13
3.1
SURFACE
15.3
10.4
0.26
1.7
2.3
5.0
14.3
0.50
7.3
6.6
6.1
7.1
0.13
1.3
0.13
0.28
1.4
0.39
1.3
1.4
NEW + REPLACEMENT MINES
UG
20.5
18.5
2.0
4.0
17.5
6.5
17.5
1.0
4.5
27.0
2.5
6.5
2.0
7.0
4.5
18.0
SURFACE
16.5
13.5
1.5
10.5
4.0
5.0
35.0
0.5
11.5
16.5
17.0
11.5
0.5
1.5
0.5
0.5
4.5
40.0
70.0
0.0
75.0
0.0
220.0
11.0
0.0
7.5
2.5
35.5
,U1AL
97.2
67.5
32.58 80.7
15.95 611.5
B-47
-------
TABLE 18
MINE WATER FLOW DISTRIBUTIONS
Parameter Values to Calculate Average Flow (Flow in
MMG/day) as a Function of Mine Size (Size in MTons/year)
FLOW = A . SIZE8
REGIONS; A B
NA, SA, CA 0.00078 0.4754
MW, CW, TX 0.01951 0.1651
NGP, RM, NEV,
NMEX 0.06364 0.0
Frequency Distributions of Multipliers to Allow
for the Variation Around Average Flow Frequency
NGP.RM,
Multiplier NA.CA.SA MW.CW.TX NEV. NMEX
45.2548 0.0 0.018 0.0
22.6274 Q.027 0.070 0.0
11.3137 0.072 0.053 0.0
5.6569 0.162 0.140 0.222
2.8284 0.072 0.140 0.111
1.4142 0.144 0.158 0.223
0.7071 0.135 0.088 0.111
0.3536 0.126 0.088 0.222
0.1768 0.135 0.070 0.111
0.0884 Q.054 0.070 0.0
0.0442 0.018 0.070 0.0
0.0221 0.018 0.035 0.0
0.0110 o.009 0.0 0.0
0.0055 o.028 0.0 0.0
B-48
-------
TABLE 19
NUMBER PRODUCTION
COST
DbTAILEU SUPPLY FUNCTION FOR PENNSYLVANIA, 1984
Supply Function for Mine Type: 1
Pennsylvania. 1984
PRICE NO. MINES EMPLOYMENT WAGES FLOW CASH FLOW MINE EOPT CONTROL EQPT
442
443
w
445
446
44 T
44B
449
450
45?
453
454
455
4S6
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
4T4
475
476
477
478
479
480
481
(1)
44. ?n
44.30
44.40
44.50
44.60
44.70
44.80
44.90
45. no
45.10
45.20
45.30
45.40
45.50
45. f>0
45.70
45. "0
45.90
46.00
46.10
46.20
46.30
4ft. 40
46.50
46.60
46.70
46. HO
46.90
47.00
47.10
47.20
47.30
47.4fl
47.50
47. hO
47.70
47. BO
47.90
46.00
48.10
976.4$
979.42
98?. 30
985.35
988.32
991. £9
994.77
997. <>5
1000.24
1001.23
100*. ?2
1004.22
1012. ?4
1015. 76
lOlfl.?*
1021.10
1024.32
1027.34
1030.36
1033. JB
103A.40
1031.4?
104?*44
1045.46
1040.48
1051.30
1054.'i2
10«7.S5
\Of-O.Sfi
1063.60
1066.63
10W. '.5
107?. 68
1075.71
I07S.74
1081.77
1084.30
1087.64
1090.88
1093.91
29.66
29.06
29. '.6
29. ^6
29.72
<>9.72
29. MS
29.^5
29. *9
29.i?9
21.79
29. *9
so. n
30.21
30.21
30. ^1
30.^1
30.21
30.21
30.21
30. ?1
30.21
30.21
30.21
30.21
30. •'I
30.r?6
30. -:6
30. £*>
30.26
30. ?6
30.29
30. £9
30.i?9
30.29
30.29
30.37
30.37
30.37
30. J7
180
180
180
180
181
181
18?
18?
183
183
184
184
IBS
185
185
185
185
185
185
185
185
185
18'j
185
185
185
186
186
186
186
186
187
187
187
187
188
189
189
189
189
25425.48
25425.48
25425.48
7S4P5.48
25719.48
2VM9.48
25877.48
2587 f. 48
2590B.-J3
25908.93
26117.93
261 17. 93
26149.51
26149.51
26149.51
26)49.51
26149. SI
26149.51
26149.5]
26149.51
2614Q.51
26149.51
2614«J.51
26149.51
26149.51
?614<).S|
272SB.SI
2725/1.51
27258. SI
27258.51
27258.51
27280.29
2f».nn.?9
27280.^9
27280.29
27339. ?9
27411.07
27411.07
27411.07
27411.07
301.42
301.42
301.42
301.'i(!
303. Cl
303.1} 1
305.72
30S.72
307.22
307. ?Z
308. S4
308.^4
321.13
321.13
321.13
321. U
321.13
321.13
321.13
321.13
3?1.13
321.13
321.13
321.13
321.13
321.13
3?4.H5
3?4.'J5
3?4.A5
3P4.85
324. i'j5
3?7.f,0
32r.>-0
327.^0
327. f>0
329. 06
335.15
335.15
335.15
335.15
40.54
40.54
40.54
40.54
40.75
40.75
40.«(S
40.86
40. H9
40. «9
41.01
41.01
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41.19
41. 19
41.'
24-J.06
243.06
241. Ofi
?45.67
245.67
245.67
2'«5.67
246.25
251.97
251.97
251.97
251.97
1209.81
1209.P1
1209.81
1209.81
1221. 51
1221.51
1226.89
1226.89
1231,28
1231.28
1233.89
1233.89
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
1269.57
12^5.43
1295.41
1295.43
1295.43
1295.43
1314. <>0
1314.60
1314. hO
1314.60
1317.40
1359.33
1359.33
1359.33
1359.33
44.33
44.33
44.33
44.33
44.33
44.33
44.62
44.62
44. »6
44. A6
44.86
44.86
45.08
45.08
45.08
45.08
45.08
45.08
45.08
45.08
45.08
45.08
45.06
45.08
45.08
45.01
45. 0"
45. OH
45.08
45.08
45.08
45.55
45.55
45.55
45.55
45.55
45.55
45.55
45.55
45.55
Number of increments of 0.1 million tons of production capacity
-------
W
I
NUMBER PRODUCTION
482
483
484
485
486
4fl7
488
489
490
491
492
493
494
495
496
497
498
499
SOO
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
COST
PRICE
NO.
48. ?0
48.30
48.40
48.50
48.60
48.70
48.80
48.90
49.00
49.10
49.20
49.30
49.40
49.50
49.60
49.70
49.80
49.90
50.00
50.10
50.20
50.30
50.40
50.50
bO.60
50.70
50.80
50.90
51.00
51.10
51.20
51.30
51.40
51.50
51.60
51.70
54.80
51.90
52.00
52.10
1006.95
1099.99
1103.02
1106.06
1109.10
1112.14
1115.18
1118.21
1121.75
1124.29
1127.33
1130. J7
1133.40
1136.44
1139.V8
1142.32
1145.56
1140.60
1151.65
1154.69
1157.74
1160.80
1161.85
116«-.90
1169.96
1171.01
117'-. 07
117Q.12
118?. 18
11«A.23
lin«.29
1191.34
1194.39
1197.45
1200.50
1201.56
1206.62
1209.69
1212.76
1215.83
30. 1/
30.37
30.17
30.37
30.18
30. JB
30. 38
30.. IB
30.18
30.18
30.- 18
30. IB
10. 18
30. JB
30. 38
30.- 18
30.36
30.46
30.<«6
30.46
30.55
30.55
30.55
30.55
30. S5
30. SS
30. 5b
3P.55
30.55
30.55
30.55
30.55
30.55
30.55
30.55
30. b5
30.67
30.67
30.69
30.69
TflRIF 1Q front 'H }
SUPPLY
MINES
18«>
189
189
189
190
190
190
190
190
190
190
190
190
190
190
190
190
192
192
193
193
193
193
193
193
193
193
193
193
193
193
193
193
193
193
194
195
195
FUNCTION FOR
Pennsylvania
EMPLOYMFNT
27411.07
27411.07
27411.07
27411.07
27442.98
27442.98
27442.98
27442.98
27442.98
27442.98
27442.98
27442.98
27442.98
274«2.9B
27442.98
27442.98
27442.98
27631.98
27811.98
27811.98
27831.76
27833.76
27833.76
27833.76
27833.76
27831.76
27833.76
27833.76
2783 ». 76
27833.76
2783.1.76
27833.76
27833.76
27833.76
27833.76
27833.76
28198.76
28190.76
28231.03
28231.03
MINE TYPE: 1
, 1984
WAGES FLOW CASH FLOW MINE EOPT
335.15
335.15
335.15
335.15
3*6.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
346.12
347.60
350.02
350.02
362.15
362.15
362.15
361.15
36?,15
36?. lb
362.15
36;-. 15
36r'.lb
36?. 15
36/7.15
362.15
365
43. ?5
43.8?
4.1.8?
43.82
43. H?
43.0?
43. H?
4.1.8?
43.82
43.8?
43.82
43. H?
43.82
43.8?
43.96
44.17
44.17
45.65
45.65
45.65
45.65
45.65
45.65
45.65
4?,. 65
4f,.6S
45.65
45.65
45.65
45.65
45.65
45.65
4S.65
45.66
45.66
45.70
45.70
251.17
251.97
251.97
251.97
260.7.9
260.29
260.29
260. ?9
260. ?9
260. ?9
260. ?9
260.29
260.29
260.29
260. ?9
260.29
260.29
260.87
261.81
261. «3
273.08
273.08.
273.08
271.08
273.08
271.0H
273.08
27.1. P8
P73.08
27.1.08
273.08
273.08
273.08
273.08
273.08
273.08
274.88
274.88
277.96
277.96
1359.33
1359.33
1359.33
1J59.33
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1390.00
1392.79
1397.41
1397.41
1479.46
1479.46
1479.46
1479.46
1479.46
1479.46
1479.46
1479.46
1470.46
1479.46
1470.46
1479.46
1479.46
1479.46
1479.46
1479.46
149?. 52
1492.52
1503.71
1503.71
CONTROL EOPT
45.55
45.55
45.55
45.55
45. «5
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
45.85
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.23
46.42
46.42
46.62
46.62
(1)
Number of increments of 0.1 million tons production capacity
-------
TABLE 20
LINEARIZED SUPPLY. CURVE FOR PENNSYLVANIA
1
2
3
4
5
6
7
8
9
Supply
(1012Btu's)
2.28
15.96
31.92
79.80
134.52
335.16
679.41
971.22
1424.90
(BAT-4)
Cumulative
Costs
($MN)
0.98
7.82
17.08
50.11
94.98
289.61
688.94
109 .63
1289.13
Unit Price
($/MMBtu)
.43
.50
.58
.69
.82
.97
1.16
1.38
1.61
Number
Of Hines
1
3
7
13
23
59
131
170
243
B-51
-------
approximation is obtained by ordinary least-squares regression of a
series of linear functions between cumulative cost (i.e., production
times price) and cumulative potential production; the length of the
linear segments aTid tfienumber 67segments follow from a minimum
accuracy requirement.
For the impact analysis, minimum required price values specifying the
beginning and the end of linear segments are not allowed to deviate by
moffe than 10% from the estimated highest or lowest values of the
previous segment.
Ml. COAL MARKET SIMULATION (MODULE 4)
Module 4 organizes the results of the 1984 projections (i.e., regional
supply, demand, transportation and utilization costs) in a linear pro-
gram, which balances supply and demand in coal markets.
A. Sulfur Content Distributions, Air Quality Control Standards and
Coal Utilization Costs
Sulfur content distribution for coals from the different supply re-
gions is derived through analysis of information on 1976 coal pur-
chases by large utilities, obtained from the Federal Power Commission.
The results of this analysis are shown in Table 21.
The cost of coal use for a specific plant is a function of the air
emission control standard with which the plant must comply. Two
different sets of standards are required for the analysis. The first
set (the so-called SIP), issued by the individual states, establishes
permissible emission limits for plants constructed prior to 1975. The
second set (the so-called NSPS), which applies to all plants cons-
tructed between 1976 and 1984, is issued by the federal government.
The estimated ratios of demand by old plants subject to different
sulfur emission limits set by SIP standards, (in terms of pounds of
sulfur found in stack gas per million Btu of coal burned) are shown in
Table 22. These ratios are essentially the samp as those used in the
National Coal Model. The sulfur emission limit set by the NSPS stan-
dards is 0.6 pounds per million Btu of coal burned.
The derivation of utilization costs for different coals is extensively
discussed is Chapter VI of Appendix A of this report. Table 23 shows
the utilization costs which are used in the analysis for new and old
plants subject to different sulfur emission limit standards and
burning coals with different sulfur contents. The utilization cost
differentials shown in Table 23 are required because of significant
differences in other coal quality characteristics, Such as Btu per
pound and ash and moisture content.
B-52
-------
TABLE 21
RATIOS OF COAL AVAILABLE IN DIFFERENT SULFUR CONTENT RANGES
ESCALAIUR
1.062
L8/HM8Tu
COAL
F COAL
LB/MMflHj
3 MU
0.03 0.6b
0.14 O.S4
0000.0.10
0.
.0
.12
<5 *A
0000.0.05
OOOO.OOOU
oooo.oooo
OOOO.OCOU
0000.0000
onoo.oooo
OOOO.OOOU
onoo.oooo
0000.0000
cooo.oooo
).01 0.02
>.01 0.02
§0
0
.00
.00
,00
.00
oo
0.60
0.2P
.0000.
.0.00
.0000.
.0000.
,0100.
,0000.
,0000.
oooo.
,0.4S
0.39
0.39
0.00
0.00
0.00
0.00
o.oc
o.oo
S.oo
.00
O.?o 0.03 OOUO.OOOO.
O.i?7 n.05 00.03 0.02
0.16 0.0A 0.04 _
0.23 0.16 0.14 0
,00
0.40 0
..Ob
0.00 0.00
0.00 n.QO 0.00 0.00
0000.0000.0.00 0.00
oooo.oooo.oooo.oooo
0000.0000.0000.0000,
OOOO.noOO.OOOO.0.00
0000.0000.0000.0.00
OOOO 0000
-------
TABLE 22
RATIOS OF COAL DEMAND SUBJECT TO DIFFERENT
AIR QUALITY CONTROL STANDARDS
••::.,? L^r^'oT ;%sKbR u;TuTbL7.J *?."**? r.sNDAPU
1 !. ."7 *5J
2 •••- • -' '"3 ,, .37
3 '"' b6
5 VJ *16 .'J1- *''b
o V- -SB •« .28
7 *V !j .69 '16
9 Ot- «33 >ei7 |.
10 ;^
11 ^ »;B
13 L'S ».
!«, Ml .19 .f*1 ,,,
Ib IL «hl
1», IN .96 .0<»
IB tl 1%P l'°
19 -r. .l'c .f"' ^
20 tT ."fc '7,.
21 »T .*" ^^ B;B .11
23 OM .lf< •*! •*l
/"• N.« lt 1.0
.) / .\l) I."
1.0
.ov .•»:»
.b? .-3
31 (-0 1.
3? AN 1.
33 t>0 J«
3*. CN 1.
34 CS 1.
B-54
-------
TABLE 23
COAL UTILIZATION COSTS
Cost of Burning as a Function of SIP Limit and Coal Sulfur Content
C'JbT OF Bi-i'M1..
COAL S Ltl/fMliTU
SIl <0.3
SI- <0.h
SIP <1.S
blr- M.'S
b' ' ••<•.. 0
iJ C*!rjTI.1G PLANTS N3/^.-dTU)»COAL S LU/MMBTU HOR.t SIP LIMIT
0.0 0.6 0.
.V)
.-P
.'-P
I.1-"
.58
.59
.SH
.5*
.54
.25
.b3
. /4
.67
.b9
.S4
• ^2
.b2
.52
«b2
.^2
.SO
.h4
.74
• <0
.56
.b6
.56
.56
3.00 '
.90 i
.66
.79
.74
.70
.67
.64
.64
.64
.64
*.00 5.00
?.00 2.11
.96 2.G9
.92 2.05
.86 2.00
.82 1.95
.79 1.91
.tS 1.78
.7b 1.78
.75 1.78
.7S 1.78
cSCALATCP 1.062
COST Of ritiKMMG F0'4 NFW PLANTS AND ULO PLANTS SUBJECT TO SIP'S < 0.3 LB/MMRTU
COAL 1> LlVKMbTU (1.3 n.6 0.9 1.3 l.b I.7S 2.0 3.0 4.0 5.0
CO" l.SR l.i". 1.67 1.74 1.74 1.7S 1.76 1.86 1.96 2.09
Cost of.Burning Differentials Due to Differences in Moisture,
Ash and Btu Content of Different Coals
I
3
<•
1
S
6
r
8
4
] 1
On
M!i
^'J
PA
vS
VA
CK
TN
AL
J\L
I'.OO
•).0(l
O.CO
0.01.
O.JO
0.00
0.00
r.oo
O.PO
.0 «2
12**
lilA
1—«0
le-TX
iobl)
PbCS
2buT
03<>
.a-.-
.ic-
.011
.074
.074
.011
.Jll
.011
.053
.053
.051
B-55
-------
B. Transportation Costs
The transportation costs for appropriate origin-destination links are
estimated from statistical analysis of rail rates and water transport-
ation costs developed in Chapter V of Appendix A.
The basic determinant of transportation costs for all links is dis-
tance. The distances between coal-producing and coal-using regions
developed for the National Coal Model are used. These distances are
between centroids of production and consumption in the producing and
demand regions. The actual locations of mines and major use centers
are taken into account. The resulting distance is a weighted average
among all points of production within a producing region and all
points of consumption within a demand region. The weighting factor is
the quantity of coal shipped.
The second factor in determining transportation costs is the number of
rail line changes required over a specific link. The line changes
have been developed as part of the rate-qathering exercise. For an
individual point-to-point movement, the number of line changes must be
an integer. The transportation cost estimated for the model is an
average of all the movements expected to take place. The number of
line changes required for the average movement between a producing and
consuming region may be a non-integer value because a different number
of line changes may be required to move between different parts of the
origin and destination regions. The line change values used for the
model are rough estimates and are applied to those links where line
change is a factor in the transportation cost equation.
The statistical analysis of rail rates shows that there are two
separate cost estimation equations depending on the area of the
country in which the coal moved. Western rates are statistically
different from Eastern rates; therefore, the estimation equation used
must apply to specific origin-destination pairs.
The data base for the statistical analysis of rates excluded rates for
coal moving from the Western plains to the East Coast because there is
little current movement on these links. However, it is felt that
higher costs of Eastern rail operations should be taken into account.
Therefore, the cost for a combined Eastern and Western railroad move-
ment is estimated by an equation whose intercept and distance coef-
ficients are the average of the Eastern and Western cost equations.
These cost equations and the range of origin-destination pairs to
which they are applied are shown in Table 24.
The data and equations described above are used to calculate rail
transportation costs for all origin-destination links for which a
distance is specified. The program sets the transportation cost to
zero for any link where the distance is specified as zero. A zero
transportation cost prevents any coal from being shipped over that
B-56
-------
TABLE 24
PARAMETER VALUES FOR RAIL TRANSPORTATION COST EQUATIONS
AND APPLICABLE ORIGINS AND DESTINATIONS FOR CONTRACT MARKET
Equation Coefficients
Intercept Distance Line Change Origins Destinations
3.288 13.568 - 1-12 1-35
2.379 11.262 1.072 21-26 1-22
1.529 8.956 1.072 M3-20,27-30 1 - 35
II3-20,27-
l 21-26
23 - 35
B-57
-------
link. The transportation costs for coal movements within production
regions are specified separately. These transport, costs are generally
smal 1.
There are specific links in thp system which arp not servpd by rail-
roads. For these links externally determined costs are rpad in by thp
computer program. Table 25 shows the origin-destination links for
which the rail rate estimated costs are replaced.
The final transportation costs used in the linear program simulation
of the contract coal market are shown in Table 26. The origins
(supply regions) are columns, the destinations are rows (thres rows
per demand region), and the values are in dollars per ton.
The transportation costs used for the simulation of the spot coal
market are higher than the costs used for the contract market because
of the smaller volumes shipped. These higher transport costs are
estimated by using greater distance coefficients for the estimation
equations as shown in Table 27.
Certain regions, which have significant contract coal poroduction
and/or demand, produce and/or consume negligible quantities of spot
market coal. These regions are therefore omitted from the spot markPt
simulation. The computer program which generates the transport file
estimates only those transportation costs required for the spot markPt.
origins and destinations. The specific spot and contract market sup-
ply and demand regions used in the anlysis are shown in Table 28. Thp
final transportation costs for spot market coal are shown in Appendix
IX.
C. Demand for Coal
The EPA Office of Air Quality Standards recently published a pro-
jection of the demand for coal as part of an anlysis of the impact of
air quality standards. The demand for coal within the 35 demand
regions used in this study is taken directly from this EPA air quality
study. The projected demands are shown in Table 29.
The use of these demands implies that the water effluent control
standards do not significantly alter the total demand for coal in any
of the 35 demand regions. The demand projections would be altered if
increased costs for water treatment are so large as to render other
fuels cost competitive with coal in certain demand regions. As shown
by the impact estimates, this is not the case.
VIL. LINEAR PROGRAMMING MODEL OF THE STEAM COAL MARKET
A fundamental assumption in the ADL coal model is that the market is
cleared on the basis of minimum total cost. Two distinct formulations
are used depending on whether or not a producer's surplus of coal is
B-58
-------
TABLE 25
TRANSPORTATION COSTS
NOT BASED ON RAIL RATL EQUATIONS
Destination Cost
2 3 $ 3.50
4 3 3.10
5 3 5.10
5 7 3.10
5 13 3.70
5 15 7.35
2 19 7.70
4 19 7.75
5 19 5.40
7 21 10.95
7 23 14.75
4 26 10.85
5 26 6.93
5 27 8.00
7 27 12.50
12 23 7.53
11 17 17.48
21 13 19.43
22 13 20.43
23 13 18.43
22 14 16.13
21 20 19.15
22 20 20.15
23 20 18.15
27 32 2.55
28 32 2.34
B-59
-------
TABLE 26
TRANSPORTATION COSTS FOR LINKS BETWEEN 27 SUPPLY (HORIZONTAL)
AND 35 DEMAND (VERTICAL) REGIONS (CONTRACT MARKET)
10. Hb912.9.1?ll. 34512.-
o.n ._ o.o n.o o.o
_ _ _
1 0.0
910.7
9.<>0<> 0.0 9.57910.7(^12.3
C.O 0.0 0.0 0.0 0.0
24.767U/ 4. |9o h
/.Slrt 0
0.0
6:5(i>
'.»*0
9
0
U.O
0.0
*}•!)•*' *• • •**f *
9.92610.694
0.0
0.0
0.0
o.n
u.o
Q.Q
u.o
\* tf«U UV1' HVV
059 6.039 9.054 8.926
.24011.734
.0
..
t.140 3l«44
b./22 H.7U8
b.57? S.JfaO 7.?H/ /.3bl
9.081 9.H4«10.?9310.77^
tf "» f
.?93
U.O 19.4/0??. 133 0.0
4.31B 3.491 7.^4J H
10. 32M1.09J11. 34912
4.31B 3.491 7.S4J H.017
»).32M1.09Jil.349l:
(>
U.
<>.
IS.
u.
0.
u.
0.
u.
16.
0.
U.
Ib.
6,{911.<,H1 9.629 0,
/03 9.^87 0.0 10,
lOJJ 6.097 0.0 b,
/4bl4.2b311./45 0,
n 0.0 0.0 0,
0 0.0 0.0 /,
^t/21b.p4713.262 U,
/Ib U.O 19.80618,
n o.n o.o /,
J/«Mb.95713.372 0,
0 0.0 0.0 0,
0 U.O 0.0 0,
b04 3.349 6.08010,
0 U.O 0.0 0,
0 0.0 0.0 U.
S69 9.1HO 9.99'. 7
- o.n
o.o
. 0.0
040 H.Q61 6.046
" ' ~ " '
U.O
U.O
0.0
0.0
0
. .
41/17. ;,/9]>>.*3f o
o u.o o.o o
0 O.n 0.0 0.0
n u.o
330 U.b4?
229 0.0
0 U.O
1?9 6.417
397 O.U
0 0.0
91515.070
764 0.0
0 0.0
7131J.V
H05 0.0
0 0.0 I
137U.48J I
8M2liU219 Olo Olo I
51810.001 9.361 9.617 I
960 6.7/9 /.Jl'J U.O I
76815.105 0.0 0.0 I
54711.85711./92IO.244 I
394 4.125 b.900 0.0 I
0 0.0 0.0 0.0 I
?4616.10716.68314.4941;
760 6.48U 5.857 3.954 I
0.0 I
0.0 I
0.0
.735 <5
.0 0
..in no
. 8;.n o
.0 n
.?S9 fl
.O'j5 7
.0 0
.1401?
.01611
.0 0
.27614
.732 7
.910 0
91913
. 00') \ 1 .
.0 0.
. IVilO.
.0 /.
.0 0.
.njoio.
.834 V.
.0 0.
.45411.
.018 9.
.0 0.
.3b1U.
.1U7 b.
.0 0.
.06713.
0 0.0 U.O
0 0.0 0.0
o n.o o.o
76310. 745 0.0
0 0.0 0.0
----- --
0.0
0.0
- -
0 16.41514.216 0.0
PHI 4.62913.04/13.679
??7?|.40620.76621.0?2
0 11.426 9.652 0.0
954 7.0/810.207 9.115
n o.o o.o u.o
o o.o o.o o.o
03712.698 2.406 2.208
?«?j:*b:iz?:i56 8:8
0/211.245 0.0 0.0
o n.o o.o o.o
SSSfWIH:!118:
o n.o o.o
?7119.24?17.400
91811.666. H.It
_.
o.o
523
-------
TABLE 27
RAIL TRANSPORTATION COST EQUATIONS
AND APPLICABLE ORIGINS AND DESTINATIONS FOR THE SPOT MARKET
Equation Coefficients
Intercept
Distance
Line Change
Origins
Destinations
3.288
2.379
1.529
21.030
17.456
13.882
1.072
1.072
1 - 12
21 - 26
13 - 20
21 - 26
1 - 35
1 - 22
1 - 35
23 - 35
B-61
-------
TABLE 28
BRIDGE FROM CONTRACT TO SPOT MARKET REGIONS
SUPPLY REGIONS
DEMAND REGIONS
Contract
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Spo
1
2
3
4
5
6
7
8
9
10
11
12
_
_
_
-
13
-
-
-
14
-
-
-
15
_
_
-
-
_
Contract
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Spot
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
B-62
-------
TABLE 29
COAL DEMAND IN 1984
(Billions of Btu's)
Region
Number Region Name Contract Market Spot Market
1 Vermont 31.4 0.0
2 Massachusetts, Connecticut 206.0 0.0
3 Pennsylvania West 514.3 238.1
4 Pennsylvania East, N.Y., N.J. 420.7 177.4
5 New York Upstate 190.3 106.0
6 Virginia, Maryland, Delaware 330.0 155.0
7 Virginia West 675.4 150.8
8 Carolina, North & South 720.3. 154.5
9 Georgia, Florida North 972.0 125.6
10 Florida South 18.0 0.0
11 Ohio North 224.0 125.8
12 Ohio Central 173.7 81.1
13 Ohio South 568.1 231.9
14 Michigan 545.1 150.3
15 Illinois 717.8 164.2
16 Indiana 919.0 251.7
17 Wisconsin 381.1 41.3
18 Kentucky East 123.2 22.4
19 Kentucky West 567.7 98.7
20 Tennessee East 181.3 41.4
21 Tennessee West 273.5 53.2
22 Alabama, Mississippi 699.2 267.3
23 Dakota, North & South 412.9 32.9
24 Kansas, Nebraska 271.2 113.0
25 Iowa 265.0 29.7
26 Missouri J87.7 63.5
27 Arkansas, Oklahoma 548.6 50.1
28 Texas 1,260.2 125.0
29 Montana, Wyoming, Idaho 445.5 19.0
30 Utah, Nevada 303.7 27.9
31 Colorado 266.7 21.6
32 Arizona 391.7 20.4
33 Washington, Oregon 123.6 22.8
34 California North 16.5 0.0
35 California South 27.0 0.0
14,272.4 3,162.6
B-63
-------
assumed. In either case, the market clearing mechanism is described
as minimization of a linear function subject to a series of linpar
inequality constraints. Such a formulation is called a "linear
program" and well-developed mathematical techniques are available for
its solution. This section describes the details of thp two market
clearing algorithms and their formulation as a linear program.
A. Market Clearing Algorithms
Figure 10 illustrates the distinction between the two assumed clearing
mechanisms. The solid line represents the estimated supply curve
based on minimum required price as discussed in Sections TIT and TV.
Essentially, two interpretations of the curve are possible.
First, if most of the coal is sold via long-term contracts, it is
reasonable to expect that each mine will bid near to its minimum
required price in order to be as competitive as possible. Tn this
case, the cost of coal supplied is represented by the cross-hatched
area in Figure 10.
On the other hand, if the market is dominated by short-term purchases,
mines are likely to charge whatever price the market will bear. In
this case, the total cost is represented by the cross-hatched area
plus the shaded area since all coal is sold at the highest price
realized. Producers in this case will reap an economic rent, i.e.,
the producers' surplus S represented by the shaded area in Figure in.
The ADL coal model has been developed to accomortate both formulations
in order to simulate the behavior of both the spot and contract mar-
kets.
B. Linear Programming Formulation of Contract Market (No Producer's
Surplus)
In order to develop the model, it is necessary to express the market
clearing algorithm as a minimization of a linear objective function
subject to a series of linear constraints. The constraints, which
represent the supply-demand balance required of the steam coal market,
are of the following four types:
t Constraints representing the total available supply
of coal from the different supply regions;
• Constraints insuring the balance between coal sup-
plied and coal which is shipped;
• Constraints describing the coal demand in different
demand regions subject to applicable SIP standards;
and
B-64
-------
FIGURE 10
TYPICAL SUPPLY CURVE ILLUSTRATING ALTERNATE MARKET
CLEARING MECHANISMS
MRP
MRP'
MRP = F IQ)
Q mln
B-65
-------
0 Constraints describing the sulfur balance between
coal available from the supply regions and the coal
which is burned in the different demand regions.
The objective function assembles the total burned cost of coal and
consists of three components.
t The cost of coal supplied from a specific supply
region;
• The cost of transporting coal from its source to
its point of consumption; and
• The cost of burning coal based on its sulfur con-
tent, type of consumer and applicable pollution
control standards.
1. Specification of Constraints for the IP Model
The methodology used to estimate the cost and constraint, parameters
has been described in the previous sections, so this discussion will
be confined to mathematical formulation of the problem.
nomenclature used in the following sections which
the LP model. It is assumed in
with the methodology of
Table 30 shows the
describe in detail each component of
the discussion that the reader is familiar
the discussion that the reader is familiar with the methodology
supply curve development, specification of costs and specification
demand as discussed in Sections TV and V.
a. Constraints Representing Total Available Coal Supply
of
These constraints are included to ensure that the total coal supplied
from any region (sum of coal supplied at each price level) does not
exceed the total available resources. Mathematically, this is
expressed for each region as follows:
(VI.1)
Additionally, since only a fixed amount of coal is available at any
given price level, the following constraints are included for each
region and each price level:
(VI.2)
B-66
-------
Symbol
TABLE 30
NOMENCLATURE USED IN THE LP MODEL OF THE STEAM COAL MARKET
Definition
'i L
Cost of transporting coal from supply
region i to demand region j (see section
V.2).
Cost of supplying coal from region i at
the Ji'th price level (see section IV.3).
C8jkn
Supply Parameters
SL
ik
Cost of burning coal with sulfur content
k in demand region j subject to SIP
standard n (see section V.I).
Total amount of coal available in supply
regioi i.
Amount of coal available in region i at
price level a.
Amount of coal supplied by region i at
price level SL.
Amount of coal shipped from supply region
i to demand reg-on j.
Fraction of coal mined in region i at
sulfur level k.
B-67
-------
TABLE 30 (cont'ri.)
NOMENCLATURE USED IN THE LP MODEL OF THE STEAM COAL MARKET (cont.)
Demand Parameters
Definition
jkn
Amount of coal subject to air quality
control standard n which is required by
region j.
Amount of coal of sulfur level k which
is burned in region j subject to standard
n.
Indices
J
k
Supply region number
Demand region number
Sulfur level number
Air quality control standard index number
B-68
-------
b. Constraints Ensuring Balance Between Coal Supplied and Coal Which
is Shipped
In order to accurately represent coal pricing, it is necessary to have
two descriptions for each unit of coal. First, the coal must be des-
cribed according to its region of origin and its price 'Y.£>. Second-
ly, in order to model the cost of transportation, it mus^be described
by its origin and destination (X..). This dual representation neces-
sitates the introduction of constraints to ensure that the amount of
coal supplied from each region is the same in each representation. To
accomplish this, the following constraints are included in the model
for each supply region:
This simply states that the total coal leaving region i 'to all
destinations j) is equal to the total coal supplied in region i (at
all price levels a ).
c. Constraints Describing Coal Demand as a Function of Demand Region
and Applicable Air Quality Control Standards
The model is designed to accomodate up to four different cost burning
conditions per region. These costs of burning conditions allow for
consideration of different air quality control standards in different
demand regions in combination with the different burning costs for
different vintages of coal burning plants.' This requires additional
constraints to ensure satisfaction of demand for each set of condi-
tions. The following constraints express this requirement:
k
k'U.ikn ' Djn
This states that the total coal supplied at all sulfur levels k to
region j, under cost of burning conditions n, must equal the total
demand for the region j under the cost of burning conditions n.
d. Constraints Describing the Sulfur Balance
It is necessary to ensure that the volumes of coal with different
sulfur content that are burned do not exceed the total volumes
Consisting of the State's "SIP" standards and the "NSPS" and "New
NSPS" standards, promulgated by the EPA.
2
Consisting of estimated handling, burning and clean-up costs for
plants subject to, respectively, "SIP", "NSPS" and "New NSPS"
standards.
B-69
-------
of coal with different sulfur contents that are availablp. This
condition is imposed by the following constraints:
SLik - r Ujkn
This set of constraints specifies that the volumes of coal U with
sulfur level k burned under all the different cost conditions n in
demand region j has to be equal to the volumes of coal with sulfur
content k shipped to demand region j from all supply regions i.
2. Specification of Objective Function
The objective function used in the coal model represents the total
burned cost of coal. This cost contains three components: a trans-
portation cost, which is proportional to the coal flow between any two
regions; the cost of supply, which is a function of the volume of coal
supplied by each region; and finally the cost of burning, which
depends on the demand region, air quality control standard, sulfur
content and plant type. This cost function has the following
mathematical formulation:
«• • 1.J U MUU j.n CBjknUjkn
The total cost C, represented by this objective function, is minimized
subject to the constraints given by expressions VII.1 to VII.5. *\s
previously mentioned, this is a standard mathematical problem that can
be solved on the computer. For the impact analysis the IBM MPSX
mathematical programming system is used.
C. Mixed Integer Programming Formulation of Spot Market (Producer's
Surplus)
In order to clear the market with a producer's surplus, 'see Section
VILA), it is necessary to modify the steam coal market formulation
described in the previous sections to ensure that all coal supplied
from a given region is sold at the highest price paid for coal from
that region. This requires that the model is formulated as a "mixed
integer" program in which some variables can assume only integer
values.
The above is accomplished by modifying Yij, to be the total volume of
coal supplied up to the price level «., instead of the total volume
B-70
-------
supplied ^t level j, . This ensures that the price coefficient Hi 2. will
apply to all volumes of coal up to and including the specified level.
Given this formulation, it is necessary to ensure that only one Y^fc is
non-zero for each region i. This is accomplished by introducing an
integer variable 6^ corresponding with each Y^. Equation VI.? is
then replaced by the following:
where: S..* = the total amount of coal available in region i at or
below price level .
Since Y^ >_ 0, this ensures that only one Y^ ^ can be greater than
zero when the following constraints are added to the model:
(VI. 8)
.0
1 J6
integer
Thus, constraint VI 1. 8
variables 6^ is equal
zero. This constraint
desired conditions: only
ensures that exactly one of the integer
to one and that all the others are equal to
combined with constraint VII. 7 ensures the
one Y^ represents the total coal flow out of
the supply region. The remainder of the model is the same as for the
contract market clearing algorithm.
VIII. OUTPUT OF THE COAL MARKET SIMULATION MODEL (MODULE 61
Module 6 organizes the results of the coal market simulation into two
reports, which are shown in Tables 31 and 32. The first report shows
the amount of coal supplied by the different supply regions and the
marginal cost of these regional supplies (see Table 31 1. The second
report shows the average user cost of burning coal in the different
demand regions. These average user costs are specified separately by
plant type - each plant type being subject to different sets of air
quality control standard: - and as a weighted average using the
relative amounts of coal consumed by those plant types as weights.
As shown in Figure 11, changes in the regional supplies, marginal
prices, and user costs are derived by subtracting the results of Coal
Market Model runs made with and without the cost increases resulting
from EPA regulations.
B-71
-------
TABLE 31
Region
PA
OH
PL)
WVAN
WVAS
VA
EKY
TEN
ALA
ILL
IND
<*> WKY
*** lOitf A
f^
MlbS
KAW
ARK
UKLA
TEX
NDAK
SDAK
MYW0
WMON
WY-P
NCOL
SCOL
UTAH
ARIZ
NHEX
MASH
Total Coal
Supplied
(MTon)
47299.93
332GU.01
1^.99
9.3-3
9.38
1 7 t^'^
17.20
ifO«97
15.61
14.72
15.18
-------
TABLE 32
Region
VERMONT
MASS.CO
FA te
PE.NY,N
NY UPST
W VA
N, S CA
GAtFL N
FLA STH
OHIO N
OHIO C
0«IO S
v.ICH
ILL
wise
KY-w
TENS E
TFNN W
ALA,MIS
DAKNtSt
KAN,NEB
IOWA
MISSOUR
ARK.OKL
TEXAS
MA.WYtl
UT* NEV
COLO
ARI.NME
MASH*OR
CAL N
CAL S
COAL
BURNED COST
Burned Cost (MMBtu)
Average
J.OU
-'.at1
2, jb
3,66
3.66
3.64
J.39
3.60
3.57
o.o
3.49
3.56
3,bl
J. 37
3s2
-------
FIGURE 11
CHANGES IN THE REGIONAL SUPPLIES. MARGINAL PRICES
AND USER COSTS
RUN;
Coal Market Model
without Regulations
RUN:
Coal Market Model
with Regulations
Regional Supplies, Prices,
User Costs
Regional Supplies, Prices,
User Costs^
Changes in Regional Supplies,
Prices, User Costs
Detailed regiona1_
supply curves
Regional economic
baseline
Regional changes in number of
producing mines, mine workers
employed, wages paid, invest-
ment required
I
Changes in regional earnings,
employment, economic growth;
energy prices to users; balance
of payments effect; effect on
growth of the coal industry
3-74
-------
The changes in other primary impact measures follow from the
comparison of the detailed regional supply curves that are used to
derive the linearized supply curves used in the Coal Market Model 'see
Figure 11).
Specifically, with the decrease or increase in a regional coal sup-
ply - obtained by comparison of the Coal Market Model runs with and
without additional pollution control costs (see Table 19) - the
changes in the number of producing mines, mine workers employed, wages
paid, and investment required to keep mines producing is derived.
IX. LIMITS OF THE ANALYSIS
A. Summary
The impact, as measured by the decrease in the consumption of coal
from an impacted supply region, will have been underestimated or
overestimated if the "demand elasticity" of coal from the impacted
supply region(s) was respectively under- or overestimated relative to
the "demand elasticity" of coal from the other regions.
The "demand elasticity" specifies the decrease in demand for coal from
a supply region in response to an increase in the cost, to the user of
an incremental unit of coal from that supply region. This "demand
elasticity" is increased by?the incremental compliance cost estimated
to result from regulations." The increase in the total user cost of
coal will cause a large decrease in the use of that coal if the
"demand elasticity" is high and it will cause a small decrease if the
elasticity is relatively low.
lun j n 4.- •*. .• A(Demand) / A(Cost per Ton)
"Demand Elasticity" = - Demand / Cost per Ton
2
The lowest end of the coal supply curve of an impacted supply region
is made up by mines with low production costs and negligible mine
flows and, therefore, negligible compliance costs (see Figure 12). *\s
a result, the lowest part of the supply curve of a given supply region
will not change when compliance costs resulting from stricter
standards in mine water treatment are added. However, the higher end
of the supply curve is shifted upward when compliance costs are added,
resulting in a higher cost for the same supply from the region. This
higher cost per incremental unit of supply will cause a relatively
larger decrease in the demand of that coal per unit increase in the
cost of that coal: the "demand elasticity" of demand for that coal has
increased.
B-75
-------
The total user cost of an incremental unit of coal from a supply
region will consist of the sum of:
• Production costs;
• Compliance costs;
• Transportation costs; and
• Utilization costs (handling, burning and clean-up
costs).
An under- or overestimation of the "demand elasticity" can occur
because of a systematic under- or overestimation in any of these four
different costs. This systematic error in the different types of
costs can be caused by aggregation errors in the 'non-sampled) data
used in the analysis.
The use of sampled data for labor productivities of new mines, mine
water flows and mine water acidity establish a range within which the
impact estimate cannot be determined: the impact estimate is
statistically insignificant within this range. The "demand
elasticity" is indeterminate within that range because the underlying
sampled data for mine water acidity, mine water flows and new mine
productivities are indeterminate within a corresponding range.
The impact estimates for BAT-4 are generally significant in a statis-
tical sense: the estimated impact exceeded the range within which
impact estimates are indeterminate because of the use of sampled data.
The extent of systematic errors possibly existing in the data cannot
be estimated. Sensitivity tests demonstrate that the supply impact
estimate is relatively insensitive to systematic errors in the user
cost of the coals from the different supply regions. However, the
impact estimate for the impacted supply regions - regions where supply
decreases because of relatively high compliance costs - is highly
sensitive to an underestimate of the compliance costs (but relatively
insensitive to an overestimate of the compliance costs^.
Because data on mine water flows are only available for highly
aggregated supply regions - the Appalachians, the Midwest plus Central
West, and the rest of the-U.S. - the water treatment cost estimates
are the limiting factor in the impact analysis.
The use of average cost data for mine production costs, transportation
costs, and utilization costs in the impact analysis has most likely
resulted in an overestimate of the decrease in the use of coal from
impacted regions (caused by increased compliance costsK
'Including replacement mines
B-76
-------
B. Statistical Significance of the Impact Estimates
The model mine analysis shows that the minimum price required for new
mine openings is highly sensitive to the value of average annual labor
productivity. Since the labor productivities for new mines is not
known, sampled values obtained from mine labor productivity distri-
butions for existing mines are used.
The compliance cost is highly dependent on the volume of acid water to
be treated. To allow for the wide range over which flow volumes for
different mines can vary, a sample of flow values, obtained from
regional mine flow distributions, is assigned to the different mines
in each supply region. The acidity of those mine flows is determined
through sampling of a distribution that indicates what percentage of
mines can be expected to have acid mine water in that region.
The use of these sampled values for productivity and acid mine water
flow results in a better representation of the "true" supply curves
than if, for example, average productivity in mine flows were used.
This is illustrated by Figure 12.
Using an average productivity to estimate the minimum required price
for new mines would result in too low an estimate of the elasticity of
supply because new mines with higher and lower than average produc-
tivities would have been mis-specified, resulting in a flatter supply
curve than the "true" supply curve. Use of these flatter supply
curves for the impacted regions results in a systematic underestima-
tion of the true "demand elasticity" and this 'other things being
equal) would cause a systematic overestimation of the impact.
Given that sampled values are used for the flow, acidity and produc-
tivity data, it is impossible to specify the "true" supply curve
exactly: the probabilistic or uncertain nature of these data result in
an indeterminate range for the "true" supply curve.
As illustrated by Figure 13, the width of the range, caused by the
uncertainty about acidity and flows, determines whether the shift in
the supply curve resulting from increased compliance costs is
statistically significant; the impact estimate is statistically
insignificant if the BPT supply curve lies within the indeterminate
range of the BAT supply curve. Figure 13 shows an example of a
statistically significant shift of the BPT supply curve.
The possible variation due to the use of sampled values for new mine
labor productivity will be the same in both the BPT and BAT supply
curve and, therefore, cancel when the impact is calculated.
In Figure 14, it is shown how an estimate of the maximum possible
variation resulting from the use of sampled values for water quality
and water flow is obtained.
B-77
-------
FIGURE 12
ILLUSTRATIVE EXAMPLES OF SUPPLY CURVES BASED
ON AVERAGE VERSUS SAMPLED VALUES FOR MINE
PRODUCTIVITY AND MINE WATER FLOWS
t
Minimum
Required
Price
(S/T)
/•
/
/
"Sampled"
FiAf Sup,.iy i in ve
, "Average'1
BAT Supply Curve
/"Sampled"
/ f;PI Supply Curve
"Averaqc"
BPT Supply Curve
Supply (MMT)
B-78
-------
FIGURE 13
ILLUSTRATIVE EXAMPLE OF INDETERMINATE RANGE
OF THE BAT SUPPLY CURVE BECAUSE
OF UNCERTAINTY ABOUT MINE
"WATER ACIDITY AND FLOWS '
Minimum
Required
Price
($/Ton)
"Sampled" BAT
Supply Curve
Range over which BAT
Supply Curve is
indeterminate due to
uncertainty about
mine water acidity
and flows.
BPT
»' Supply Curve
Supply (Millions of Tons)
B-79
-------
FIGURE 14
METHODOLOGY TO DERIVE MAXIMUM POSSIBLE
ERROR IN IMPACTED REGIONS
30 Samples
of Pennsylvania
BPT
Supply Curve
Calculate Mean Supply
and Maximum
Deviation from Mean
Due to Productivity
Sampling
30 Samples
of Pennsylvania
BAT4
Supply Curve
Calculate Mean Supply
and Maximum
Deviation from Mean
Due to Productivity, Water Quality
and Flow Sampling
Derive Maximum Error
Due to Flow Sampling
for Pennsylvania
Derive Maximum Variation
Due Directly to Flow Sampling
for Other Impacted Regions
Total Possible Variation
Due to Interaction for
All Impacted Regions
1
Total Possible Variation
in Impact Results
Due to Flow Sampling
B-30
-------
Thirty samples of the Pennsylvania BPT and BAT-4 supply curves are
obtained. These samples are taken with different starting seeds for
the random number generator to obtain a different sequence of
productivity and flow samples for each supply curve. (Thirty samples
are considered to be adequate to obtain a statistically significant
sample of the underlying distribution.)
The mean supply and the maximum deviation from this mean due to
productivity sampling is calculated for the BPT case. In a similar
manner, the mean supply and the maximum deviation from this mean due
to both productivity and flow sampling is calculated from the results
of the thirty samples in the BAT-4 case.
The supply taken from the thirty different supply curves represents
the volumes that would have resulted in the linear program solution
under the assumption that no changes had occurred in all other supply
regions. (This last assumption implies that the total cost of coal
supplied from Pennsylvania would have remained the same in all the
thirty different solutions.)
Table 33 gives an example of how the "sampled" supply is calculated.
The first supply curve shown in Table 33 is the one which is obtained
in the original solution used for the impact analysis. The amount of
coal supplied by Pennsylvania in that solution is 1,093.89 million
Btu's at a price of $1.61 per million Btu. The volume which would
have resulted - using the second supply curve shown in Table 33 in the
solution - is obtained by calculating the incremental supply from the
last linear segment - at that linear segment's minimum required
price - that equates the cumulative costs of coal supplied in thp
original solution (i.e., $1,071.80 million).
Three different types of variation result from the sampling of produc-
tivity, water quality, and flow distributions. The first type of
variation is caused by productivity sampling. The two other types of
variation result from sampling of flow distributions for the impacted
supply region considered. They are the "direct" and the "interaction
variation."
The procedure, previously described, allows the derivation of an
estimate of the "direct variation." The other type of variation, the
"interaction variation," results from variation in estimated volumes
supplied in any of the impacted regions because of variation in the
supply curves of the other impacted regions.
estimated impact of BAT-2 compliance costs was found to be
negligibly small and therefore statistically insignificant.
B-81
-------
TABLE 33
EXAMPLE OF CALCULATION OF SOLUTION SUPPLY
FOR A SAMPTED SUPPLY CURVE FOR PENNSYLVANIA (UNDER BPT)
Pennsylvania Supply Curve Used In Contract Coal Market Simulation
MMBtu's
2.28
15.96
31.92
79.80
134.52
335.16
679.41
971.22,
1093.89^
1424.90
MM$
0.98
7.82
17.08
50.11
94.98
289.61
688.9f.
1091.63
1289.13
1822.06
$/MMBtu
0.43
0.50
0.58
0.69
0.82
0.97
16
38
61
1.61
Sample of Pennsylvania Supply Curve Obtained With a Different
Starting Seed for the Random Number Generator
2.28
15.96
31.92
79.80
102.60
321.48
636.09
994. 02, -^
1071. 80u;
1454.54
0.98
7.82
17.08
50.11
68.81
283.31
654.55
1159. 23, 9x
1289. 13u;
1928.30
0.43
0.50
0.58
0.69
0.82
0.98
1.18
1.41
1.67
1.67
Supply from Pennsylvania in contract coal market solution is 1093.89
^Supply calculated for this case is 1071.80:
1071.80 = (1289.13 - 1159.23)/!.67 + 994.02
B-82
-------
The variation caused by sampling of the productivity distributions can
be ignored because of its systematic nature. Using the same starting
seed for the random number generator ensures that the same produc-
tivity numbers are used for the same mines in the three different cost
scenarios (i.e., the BPT, the BAT-2 and BAT-4 case^. Therefore, the
variation in the BAT-4 and BPT supply estimates - caused by
productivity sampling - are systematic and cancel in the calculation
of the impact estimate, when BAT-4 and BPT supplies are subtracted.
Having obtained mean and standard deviation estimates for the BPT and
BAT-4 cases, the maximum variation due to water acidity and flow
sampling is derived for the Pennsylvania supply curve. From this an
estimate of the maximum variation due directly to water quality and
flow sampling for other impacted regions and due to interaction
between impacted regions is estimated. Comparing the sum total of
these two types of variation for individually impacted regions with
the estimated supply impact shows whether or not the change in supply,
attributed to the regulations, is statistically significant.
The results from the thirty supply curve samples obtained for Pennsyl-
vania are shown in Table 34. Pennsylvania is used in this case
because it is shown to have the largest impact in the BAT-4 case. The
mean supply under the BPT and BAT-4 case are, 47.3 million tons and
46.8 million tons. The difference in maximum deviation due to water
quality and flow sampling under BAT-4 is found to be approximately
0.44 million tons or about 1% of the mean.
A region with a smaller number of mines than Pennsylvania, such as
Ohio, will have a larger possible variation. A region with a larger
number of mines than Pennsylvania will have a smaller possible
variation. The ratio of the maximum possible variation of the two
different regions will be the inverse of the square root of the number
of mines producing in the different regions.
Table 35 shows the derived values for these possible direct variations
in supplies for what has been called the "impacted regions" and the
"balancing regions." The "impacted regions" in this particular case
are the regions which have a decrease in supply volumes because of
increases in treatment costs. The "balancing regions" are the regions
which have an increase in supply.
The second type of variation, the interaction variation, is calculated
as shown in Figure 15. The total possible variation in the supply of
balancing regions that resulted from the sampling of flows is
allocated to the impacted regions and vice versa.
In the case of the "impacted regions," the total possible variation in
the supply of the "balancing regions" supply is calculated and
assigned to "impacted regions" using the decrease in supply volumes
B-83
-------
TABLE 34
RESULTS OF THIRTY SUPPLY CURVE SAMPLES
FOR PENNSYLVANIA?!)
Compliance
Costs
Solution
Supply
Mean Supply
(MMT)
Difference
in Max
Deviation
(MMT)
(2)
(Max. Dev
T Mean)
BPT
BAT-4
48.0
45.1
47.3
46.8
1 0.44
1 0.01
(1)
(2)
Pennsylvania has 189 contract nines.
I.e., the maximum deviation due to water quality and flow
sampling under BAT-4.
B-84
-------
TABLE 35
ESTIMATED SUPPLY IMPACT OF BAT-4 ON THE CONTRACT MARKET COMPARED WITH
THE POSSIBLE VARIATION IN THAT ESTIMATE DUE TO UNCERTAIN INFORMATION ON MINE WATER FLOWS
oo
THE IMPACTED REGIONS
POSSIBLE VARIATION
IN IMPACT
ESTIMATES
(in MMT)
POSSIBLE VARIATION
IN IMPACT
ESTIMATES
THE BALANCING REGIONS (in MMT)
123 BY 4
:GION MINES SUPPLY LOSS DIRECT INTER- TOTAL
MMT MMT ACTION
PA 189 48.00 2.90 0.5 0.60 1.10
UT 1 0.33 0.33 0.0 0.07 0.07
REGION
OH
WV(S)
VA
AL
MT
1
MINES
103
117
90
88
15
2
SUPPLY
MMT
31
26
14
22
75
.2
.2
.2
.6
.9
3
GAIN
MMT
0
0
0
0
0
.4
.2
.3
.1
.5
BY
DIRECT INTER-
ACTION
0
0
0
0
0
.4
.3
.2
.2
.0
0
0
0
0
0
.07
.03
.05
.02
.06
4
TOTAL
0.47
0.33
0.25
0.22
0.06
WY(P) 26 201.4 2.8 0.0 0.04 0.40
-------
FIGURE 15
CALCULATION OF THE INTERACTION ERROR
TNTFRAfTTON FRRnR - fLoss or Gain *n SuPPly* ( Total Possible Error in x
INTERACTION ERROR - ( Total LQSS Qr ^ ') ("Balancing Regions" Supply)
INTERACTION ERROR = Maximum possible deviation in supply due to interaction.
Loss or Gain in Supply = Loss or Gain in Impacted Region
Total Loss or Gain
Total Loss or Gain of All Impacted
Regions with a Loss or a
Gain
Balancing Regions
All Regions with a Gain
or a Loss Offsetting the Loss or Gain
in Impacted Regions
B-06
-------
due to impact as weights. In the case of the "balancing regions," the
total possible variation due to flow sampling in "impacted regions" is
calculated and assigned to individual balancing regions using the
increase in supply volumes due to impact as weights. As shown in
Table 35, for the heaviest impacted regions 'e.g., Pennsylvania^, the
maximum possible variation due to interaction is larger than the
maximum possible "direct" variation.
As shown in the right hand column of Table 36, for a number of regions
with relatively small impact estimates, the sum of the direct and
interaction variation is larger than the estimated impact. Therefore,
the impact estimate for these regions is inconclusive. However, the
contribution by these regions to the total impact estimate calculated
for larger regions (e.g., Northern Appalachia) is relatively small.
The same analysis is done for the impact estimates obtained for th»
spot market, as shown in Tables 37 and 38. The direct variation
resulting from the flow sampling are estimated to be about twice as
large for the small spot market jnines as the variation obtained for
the larger contract market mines.
The results of this error analysis for the spot market show again that
the impact estimate for region(s) with relatively small estimated
impacts is inconclusive.
C. Sensitivity of the Impact Estimates to Systematic Errors
To test the sensitivity of the impact estimates to systematic errors,
it is estimated how the supply impact changes if large systematic
errors occur in the different costs (i.e., production, transportation,
burning and clean-up costs) used in the analysis. As discussed in the
introduction, systematic errors in these cost estimates for the
impacted regions result in an over- or underestimation of the "demand
elasticity" for coal from those regions and this results in an over-
or underestimation of the supply impact attributable to increased
compliance costs.
It is not possible to determine what actual systematic error may be
present in the different cost estimates used in the analysis. There-
fore, two extreme cases are constructed by:
The analysis of mine water flows shows water flows 'in gallons per
ton produced) to be significantly higher for smaller mines than for
larger mines. As a result, errors in estimates of minimum required
prices, including an estimate for water treatment costs, for smaller
mines will be larger, resulting in large possible errors in total
supply available at a given total cost.
B-87
-------
TABLE 36
POSSIBLE VARIATION IN ESTIMATED BAT-4 SUPPLY IMPACT
DUE TO UNCERTAIN INFORMATION ON MINE UATER FLONS U)
(Contract Market)
Pennsylvania
Ohio
W. Virginia South
Virginia
Alabama
Montana
Wyoming,
Powder River
Utah
Supply
With
BPT
48.0
31.2
26.2
14.2
22.6
75.9
201.4
Impact
With
BAT-4
MMT/Y
-2.9
+0.4
+0.2
+0.3
+0.1
+0.5
+2.8
Maximum
Possible
Variation
MMT/Y
± 1-1
+ 0.47
+ 0.33
+ 0.25
+ 0.22
+ 0.66
+ 0.4
Inconclusive
Impact Estimate
Yes
Yes
Yes
0.33
-0.33
+ 0.07
(1)
Concerning acidity and alkalinity and volume of flow
D-88
-------
TABLE 37
REGION
at
00
* PA
OH
ESTIMATED SUPPLY IMPACT OK BAV-4 ON
THE SPOT MARKET COMPARED WITH
THE POSSIBLE VARIATION IN THAT ESTIMATE DUE 0) THE UNCERTAIN INFORMATION ON MINE WATER FLOWS
THE
IMPACTED
MINES SUPPLY
401
89
MMT
7.25
1.84
REGIONS
LOSS
MMT
0.44
0.43
POSSIBLE VARIATION
IN IMPACT
ESTIMATES
(in MMT)
BY
DIRECT INTER- TOTAL
ACTION
0.07 0.03 0.10
0.15 0.04 0.19
POSSIBLE VARIATION
THE
1
REGION
MD
KY(E)
OK
BALANCING REGIONS
2
MINES
14
366
3
SUPPLY
MMT
0.27
6.15
0.33
GAIN
MMr
0.04
0.43
0.08
IN IMPACT
ESTIMATES
(in MMT)
BY
DIRECT INTER-
ACTION
0.02 0.02
0.08 0.20
0.00
TOTAL
0.04
0.28
0.03
-------
TABLE 38
POSSIBLE VARIATION IN ESTIMATED BAT-4 SUPPLY IMPACT
DUE TO UNCERTAIN INFORMATION ON~MINE WATER FLOWS (0
(Spot Market)
Pennsylvania
Ohio
Maryland
Kentucky East
Oklahoma
Supply
With
BPT
MMT
7.25
1.84
0.27
6.15
0.33
Impact
With ,.i
BAT=4u;
MMT/Y
-0.44
-0.43
+0.04
+0.43
0.08
Possible
Error
MMT/Y
+_ 0.10
+ 0.19
+ 0.04
+ 0.28
+ 0.03
Inconclusive
Impact
Estimate?
Yes
(1)
(2)
Concerning acidity versus alkalinity and volume of flow
Net losses in total coal supply are expected to be made up
by supplies from large contract mines in the Wyoming Powder
River Basin.
B-90
-------
t Increasing or decreasing the production and utili-
zation costs of coal from the most impacted region
(i.e., Pennsylvania) by approximately 10% for both
the BPT and BAT-4 case;
• Increasing or decreasing the transportation costs
from all regions by 10% for both the BPT and BAT-4
case;
• Increasing or decreasing the estimated BAT-4 com-
pliance costs by approximately 30%.
The changes in production, transportation, and utilization costs
result in a lower and a higher estimate of the "user cost elasticity"
of demand relative to the cost elasticity used in the impact analysis.
The impact estimate obtained with the lower estimate of the "demand
elasticity" for coal from Pennsylvania in combination with the higher
estimate of compliance costs demonstrates what the impart estimates
had been if user cost increases at the margin 'per incremental unit of
coal from Pennsylvania) would have been systematically overestimated
in the analysis while compliance costs were underestimated. Likewise,
the impact estimate obtained with the higher "demand elasticity" for
Pennsylvania coal in combination with the lower compliance costs
demonstrates what the impact estimate had been if the impact analysis
had underestimated the marginal user cost increase for Pennsylvania
coal while overestimating the compliance costs.
As shown in 'Table 39, if the marginal user cost for coal from
Pennsylvania in 1984 is 10% higher than as estimated for the impact
analysis, then the projected demand for coal from this region in 1984
will be approximately 6.7 million tons per year (or 14%) lower. If,
in addition, the compliance costs for mines in Pennsylvania in 1984
are 30% lower than as estimated in the impact analysis, then the
impact of the BAT-4 mine water treatment standards will be an increase
in coal supply from Pennsylvania of 2.7 million tons to offset the
decline in supply of coal from Ohio (by 4.1 million tons per yearN and
Alabama (by 3.2 million tons per year).
If the marginal user cost of coal from Pennsylvania in 1984 is
lower than as estimated for the impact analysis, then the projected
demand for coal from this region will be approximately 13.5 million
tons per year (or 27%) higher than as estimated for the impact
analysis. If the BAT-4 compliance costs for mines in Pennsylvania is
30% higher than as estimated in the impact analysis, then the impact
in terms of reduced coal supply for Pennsylvania in 1984 will be 2.5
million tons per year, or approximately 13% less than as estimated in
the impact analysis. Apparently, the compliance costs are so small that
a 30% increase of these costs for one region (PA) will allow the model
to find a computer- initiative solution: the answers given by the model
become nonsensical.
B-91
-------
TABLE 39
03
O
ro
RESULTS OF SENSITIVITY TESTS IN
TERMS
OF CHANGES IN ESTIMATED SUPPLY IMPACTS
Marginal User Cost Marginal User Costs
for PA Coal 10% for PA Coal 10% lower;
ORIGINAL
•^ ESTIMATES ^
PA
OH
WV (S)
VA
AL
WY (P)
UT
ILL
TX
BPT
Supply
48.0
31.0
26.2
14.2
22.6
201.4
0.3
41.4
17.7
BAT-4
(MMT)
45.1
31.6
26.2
14.5
22.7
204.1
0.0
41.4
17.7
Impact
(MMT)
-2.9
+0.4 •
+0.2
+0.3
+0.1
+2.9
-0.3
0.0
0.0
Higher; Compliance Compliance Cost
*- Cost 30% Lower — *• *«-— 30% Higher *
BPT
Supply
42.3
31.2
22.6 .
14.2
22.6
228.9
0.0
41.4
11.9
BAT-4
(MMT)
45.1
27.1
23.9
14.5
19.4
228.9
0.0
41.4
17.7
Impact
(MMT)
+2.7
4.1
+1.3
+0.3
-3.2
0.0
0.0
0.0
+6.8
BPT
Supply
62.5
31.2 -
26.2
19.8
22.6
166.8
0.3
45.6
17.7
BAT-4
(MMT)
60.0
31.6
26.4
19.8
22.7
170.5
0.0
44.6
17.7
Impact
(MMT)
-2.5
+0.4
+0.2
0.0
+0.1
+3.7
-0.3
-1.0
0.0
-------
These results demonstrate that the impact estimate for specific
regions is highly sensitive to any systematic underestimation that may
have occurred in compliance costs resulting from the misspecifications
of the treatment costs and/or the mine water quality and flow volumes
for individual regions. ^ systematic underestimation by 30^ of the
compliance costs of one region relative to the compliance costs of th*>
other impacted supply regions will result in a complete missper.ifi-
cation of the supply impact for that region; i.e., the supply in that
region will go up rather than down as a result of the BAT-4 standards.
However, a systematic overestimation of the compliance costs for one
supply region relative to the estimated compliance costs for other
supply regions will result in a relatively small error in the supply
impact for that region.
D. BIAS RESULTING FROM AGGREGATION ERRORS
Use of average values for the different costs used for the impact
estimate probably results in an overestimation of the supply impact
attributable to an increase in compliance costs.
As discussed earlier, the impact will have been overestimated if the
"demand elasticity" for coal supply from the impacted regions is
overestimated. The impact analysis shows that the main supply impact
is expected to occur in Pennsylvania, resulting in a decrease in coal
supplies from that region. The overestimation of the "demand elastic-
ity" for coal supply from Pennsylvania results from:
• The assumption that all mines use a discount rate
of 10% per year to estimate their minimum required
price;
• The use of average mine investment and operating
costs to calculate minimum required prices;
• The use of average transportation costs;
• The use of average coal quality characteristics for
coal produced in Pennsylvania; and
• The use of average burning and cleaning costs for
Pennylvania coal.
In the impact analysis, it is assumed that mines will close if the
compliance costs render it impossible to make an average (DCF) rate of
return on investment of 10% per year. In reality, the more heavily
impacted mines, i.e., mines incurring higher compliance costs, will
most probably continue to operate even if their rate of return drops
below 10% per year. Therefore, the supply impact is probably
overestimated.
B-9:
-------
The use of average mine investment and operating costs to calculate
minimum required prices (even if appropriate ffTlowance"" is made for
mine type and mine labor productivity^ and the^use- of. average trans-
portation costs, coal quality characteristics and average burning and
clean-up costs will result in an overestimation o^ the "demand elas-
ticity." In reality, these costs are different for the mines that, in
the analysis, are assumed to have the same production, transportation
and burning costs. As a result of these differences in costs, the
actual "demand elasticity" is lower than the "demand elasticity" used
in the analysis; i.e., a coal production cost increase in a supply
region because of increased compliance costs in reality will cause
less of a change in demand for coal from that region than shown by the
impact analysis. Therefore, the aggregation errors resulting from the
use of average costs probably result in a conservative (i.e., too
high) estimate of the supply impact.
<»U. S. GOVERNMENT PRINTING OF F I CE I I 98 1 -3« I -085 /46 4 1
B-94
------- |