United States Industrial Environmental Research EPA-600/7-80 009
Environmental Protection Laboratory January 1980
Agency Research Triangle Park NC 27711
Sammis Generating Station:
Meeting S02 and Paniculate
Standards with Cleaned
Ohio Coals
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-009
January 1980
Sammis Generating Station:
Meeting SC>2 and Paniculate Standards
with Cleaned Ohio Coals
by
Gladys Sessler
Teknekron Research, Inc.
Energy and Environmental Systems Division
2118 Milvia Street
Berkeley, California 94704
Contract No. 68-02-3092
Task No. 3B
Program Element No. EHE623A
EPA Project Officer: James D. Kilgroe
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
This report discusses the background and issues related to the control of air
pollutants emitted by a large coal-burning plant in eastern Ohio. This plant has
had a history of severely exceeding the particulate emission limit set forth in
Ohio's State Implementation Plan (SIP). Furthermore, the plant's S02 emissions
have exceeded the limit that Ohio's forthcoming SIP will allow.
One important issue to consider is the extent to which compliance with the SIP
will promote the plant's switching from Ohio coals to southern Appalachian
coals, which produce fewer particulate and S02 emissions, and the consequent
disruption to the Ohio coal mining industry. Addressing this issue, the report
examines the plant's historical coal usage, the production and characteristics of
Ohio and southern Appalachian coals, the relevance of coal-sulfur variability,
and, most important, the feasibility and implications of producing and burning
cleaned Ohio coals as a strategy for complying with Ohio's SIP.
The report discusses the factors that will affect the relative economics of
burning cleaned Ohio coals at the plant in question. The analysis indicates that,
by burning cleaned Ohio coals, the plant's largest and newest units (which
constitute 60 percent of the plant's total capacity) can increase their consump-
tion of Ohio coal by 50 to 100 percent, depending on the characteristics of the
coals and the cleaning processes used.
This report was submitted in fulfillment of the requirements of Work Assign-
ment 3, Task B, of EPA Task Order Contract 68-02-3092 by Teknekron
Research, Inc., under the sponsorship of the U.S. Environmental Protection
Agency. The report covers the period from March 1979 to July 1979.
ii
-------
CONTENTS
Page
ABSTRACT ii
FIGURES , v
TABLES vi
ACKNOWLEDGEMENTS vii
I. INTRODUCTION AND SUMMARY OF RESULTS I
I. I Introduction I
1.2 Summary of Results 2
2. BACKGROUND INFORMATION 5
2.1 The Sammis Generating Station:
Location and Facilities 5
2.2 Legal and Regulatory Issues Affecting Sammis's
Choice of Coals 9
2.2.1 Particulates: The Ohio Implementation Plan and
Interstate Transport 10
2.2.2 Compliance with S02 Emission Standards 14
2.2.3 S02 Compliance and Section 125 16
2.3 Characteristics and Production of Ohio Coals 18
2.3.1 Recent Production 18
2.3.2 Sulfur Content 21
2.3.3 Incombustible (Ash-Producing) Matter 24
2.3.4 Coal-Preparation Practices in Ohio 25
2.4 Coals Historically Used by Sammis and Representative
Compliance Coals 26
2.4.1 Coals Historically Used by Sammis 26
2.4.2 Representative Compliance Coals 30
3. PROSPECTS FOR THE USE OF CLEANED OHIO
COALS AT SAMMIS 33
3.1 Average Coal-Sulfur Values in Relation to S02 Emission
Limits and Coal-Sulfur Variability 34
iii
-------
CONTENTS (Continued)
Page
3.2 One Ohio Plant's Proposal for Using PCC as an
SOj Compliance Strategy 44
3.3 The Washability of Ohio Coals 51
3.4 The Potential Consumption of Cleaned Ohio Coal
at Sammis 59
3.5 The Costs of Coal Cleaning 66
3.5.1 Estimated Costs of Cleaning High-Sulfur
Eastern Coal 66
3.5.2 Cost Advantages of Burning Cleaned Coal 69
3.5.3 Costs of Cleaned Ohio Coal versus Out-of-State
Low-Sulfur Coal for Units 5-7 74
3.6 Institutional Barriers to Implementing PCC in Ohio 78
APPENDIX: SULFUR VARIABILITY AND A COMPARISON OF THE
EFFECTIVE AND MANDATED S02 EMISSION LIMITATIONS 80
NOTES 88
iv
-------
FIGURES
Number Page
I Plant Layout for W.H. Sammis Plant,
Ohio Edison Company 6
2 Location of the W.H. Sammis Plant (Stratton,
Jefferson County, Ohio) 12
3 Sectors of Extremely Persistent Winds in the
Upper Ohio River Basin Area 13
t\ Ohio Coal Production in 1977, by County 19
5 Histograms of Ohio Coal Reserves and Deliveries in 1977 23
6 Illustration of the Effect of Averaging Period on RSD 35
7 A Page of Washability Data from the BOM Rl 8118 52
8 Available Ohio Coal Reserves for Alternative
S02 Standards and Levels of PCC 57
9 Eastern Coal Prices as a Function of Sulfur Content 75
A-1 RSD of Sulfur Content versus Averaging Period
(Lot Size) 82
-------
TABLES
Number Pqge
I Sammis Plant Characteristics 7
2 1977 Ohio Coal Production, by County and Seam 20
3 Reserve Base of Eastern Bituminous Coals 22
4 A Representative Selection of Historic Coal
Deliveries to Sammis in May 1978 27
5 Summary of Sammis's May and November 1 978 Coal
Deliveries by State of Origin 29
6 Representative Compliance Coals for Sammis 31
7 Values of the Relative Standard Deviation (RSD)
of Sulfur Content in Ohio Coals 39
8 Expected Average $©2 Emissions for Sammis Units
under Different Assumptions of Sulfur Variability 42
9 Summary of Results of Bureau of Mines Washability
Tests on Two Samples from the Middle Kittanning
Seam in Coshocton County, Ohio 47
10 Ash and Sulfur Contents of Coal Samples from Middle
Kittanning Seam, Coshocton County, Ohio 49
1 1 Washability Data for Selected Ohio Coals
(Sulfur Content) 53
12 Washability Data for Selected Ohio Coals (Ash Content) 54
13 Homer City PCC Plant: Performance Design Values 60
14 The Allowable Fraction of Cleaned High-Sulfur Coal
at Sammis Units 5-7 65
15 Annual Physical Coal Cleaning Costs (1978 $) for a
High-Sulfur Eastern Coal 68
16 Summary of the Cost of Producing Cleaned Coal 70
1 7 Summary of Costs versus Savings with PCC 77
A- 1 Total Sulfur and Pyritic Sulfur Content:
Compar ison of Var iabi I i ty 85
vi
-------
ACKNOWLEDGEMENTS
Teknekron Research, Inc., wishes to thank members of the U.S. Environmental
Protection Agency for their important contributions to this work. Throughout
the effort, Project Officer James D. Kilgroe of EPA's Office of Energy,
Minerals, and Industry (OEMI) generously provided guidance and relevant infor-
mation reflecting his experience with coal cleaning and related subjects.
Robert Statnick, also of the OEMI, provided useful direction and, indeed, much
of the impetus for the project. F. Richard Kurzynske and Bertram Frey of EPA's
Region V Office provided applicable technical and legal background material.
At Teknekron, Dr. Andrew Van Horn and Dr. David Large carefully reviewed the
technical contents of this report. Barbara Phillips, in her scrupulous editing,
unrelentingly insisted upon clarity in the text, tables, and figures. Production
was carried out by Evelyn Kawahara, Sheryl Klemm, Maureen Ash, and Carol
See; and the various components of the production effort were efficiently
coordinated by Lorraine Gunther.
VI1
-------
I. INTRODUCTION AND SUMMARY OF RESULTS
I.I Introduction
This report deals with the feasibility and implications of burning cleaned Ohio
coals at the W.H. Sammis Generating Station in eastern Ohio. Sammis's choice
of coals is now at issue as a result of state and federal regulations governing the
plant's particulate and SO7 emissions. Because Sammis's particulate emissions
have greatly exceeded the limitations set forth in Ohio's State Implementation
Plan, Ohio Edison, the owner utility, has been involved in litigation with EPA.
And because Sammis will have to comply with S02 emission limits starting in
October 1979, Ohio Edison plans to substitute low-sulfur (and low-ash) out-of-
state coals for the high-sulfur (and high-ash) Ohio coals that have comprised
most of Sammis's coal supply in the past.
The strategy of relying mainly on out-of-state coals is expected to have adverse
repercussions for Ohio's coal mining industry. Another strategy one that
would counter the decrease in Ohio coal use is to burn Ohio coals that have
been physically cleaned, since physical cleaning can remove a significant
fraction of a coal's ash-producing constituents and pyritic sulfur. Whether the
burning of cleaned Ohio coals at Sammis is feasible and what the implications
would be for both Sammis and the Ohio coal industry are the subjects of this
report.
The report contains two main sections. In Section 2, which provides background
information on Sammis, we discuss: (a) the plant's facilities and historic
emissions; (b) environmental, legal, and regulatory issues affecting Sammis's coal
choices; (c) characteristics and sources of reserves and the recent production of
Ohio coals; and (d) the coals Sammis has burned and the compliance coals
currently available.
In Section 3 we discuss physical coal cleaning (PCC), particularly in terms of its
meeting Sammis's coal needs, and more generally in terms of its attenuating the
decline of Ohio coal production. First (in Section 3.1, but also in the Appendix)
-------
we discuss the subject of sulfur variability in Ohio coals in order to relate
Sammis's maximum allowable SC^ emissions to its effective allowable emissions.
This analysis is essential in order to assess the actual coal-sulfur levels that must
be achieved by PCC to render cleaned Ohio coal use feasible at Sammis. Next
(in Section 3.2) we discuss another Ohio plant - C&SOE's Conesville
plant - which plans to meet its $©2 emission standards by cleaning Ohio coals
from nearby sources. We then discuss the available data on the cleaning of Ohio
coals (Section 3.3) and estimate the quantity of cleaned Ohio coal that Sammis
could burn (Section 3.4). In Section 3.5, which deals with the costs and benefits
of cleaning Ohio coals, we discuss estimated production costs, estimated boiler-
related benefits, and estimated differences between the price of non-Ohio coals
that are low in sulfur and ash and the price of high-sulfur, high-ash Ohio coals
that can be washed to meet Sammis's requirements. In the final section, 3.6, we
mention some of the institutional issues that must be addressed in connection
with the production of cleaned Ohio coal.
1.2 Summary of Results
The investigation detailed in Sections 2 and 3 suggests that cleaned Ohio coals
can comprise a sizable fraction of the supplies to be burned at Sammis units 5-7
in compliance with applicable emission limitations. The salient points made in
this study can be summarized as follows:
The EPA, under an interim compliance program, has
ordered Sammis to reduce particulate emissions so that
they do not exceed 0.7 to 0.8 Ib particulates per million
Btu. (Ohio's SIP specifies a limit of O.I Ib particulates per
million Btu.) A major element of the interim compliance
program involves burning coals with less than 10 pounds of
ash-producing material per million Btu, a "quality index"
that represents considerably lower ash content than that
generally found in Ohio coals.
Sammis's SO? emission limitations will be 4.46 Ib per
million Btu for the three largest and newest units (units 5,
6, and 7), which account for almost 70 percent of the
plant's capacity. The SOo limit for the remaining units is
-------
1.61 Ib 862 per million Btu. Units 5-7 have been oper-
ating at very low capacity factors due to operational
difficulties. If they were to operate at 60 percent of
capacity on an annual basis, they would consume almost
4 million tons of coal a year (in 1977 they consumed
2.5 million tons). We show that the fraction of cleaned
Ohio coal that units 5-7 can acceptably burn ranges from
about 50 to 100 percent, depending on the characteristics
of the Ohio coal, the cleaning process used, and the
characteristics of a low-sulfur coal that can be blended
with the cleaned Ohio coal.
Assuming that up to two exceedances of the 50
will be permitted each month, and that more than two
exceedances per month will occur only once every two
years, the allowable mean $©2 emissions from coals
burned in units 5-7 range from aoout 3.2 to about 3.8 Ib
$©2 per million Btu.
The estimated costs of producing cleaned coal are divided
about equally between the PCC plant costs (capital and
operating) and the value of the combustible material
discarded during PCC.
It generally costs more to purchase and clean Ohio coals
than to purchase uncleaned out-of-state, low-sulfur, low-
ash coals. This cost differential between the use of
cleaned Ohio coals and the use of out-of-state coals is
expected to decrease, since the costs of low-sulfur com-
pliance coals are expected to escalate faster than the
prices of Ohio coals. Moreover, when estimated savings
associated with the burning of cleaned coals are con-
sidered, the use of cleaned Ohio coals may be economi-
cally justified. In the case of Sammis these savings
reflect, among other factors, elimination of the need to
build additional barge unloading facilities for increased
out-of-state coal deliveries.
A limited set of sulfur-removal measurements taken at
one Ohio coal-cleaning plant indicates that PCC at that
plant reduces S02 emissions (in Ib S02 per million Btu) by
about 25 to 40 percent. A large, new, relatively sophisti-
cated PCC plant that is coming on line near Cadiz, Ohio
is designed to remove 50 to 70 percent of the total sulfur.
PCC can also reduce the ash content of Ohio coals by
about 25 to 75 percent.
-------
Besides the technological and economic factors relating
to the use of cleaned coals, there are important institu-
tional issues associated with the production of such coals
in Ohio. Especially important is the fact that many Ohio
coal mines are small; they lack the organization and
capital to build a PCC plant on an economically viable
scale. Also important are the interim arrangements the
utility must make with regard to either coal purchases or
emission limitations during the period of approximately
three years between conception of, and production from,
an advanced coal cleaning plant.
-------
2. BACKGROUND ^FORMATION
2.1 The Sommis Generating Station: Location and Facilities
The W.H. Sammis Station is located at Stratton in Jefferson County on the
eastern border of Ohio. The station is bounded on the east by Ohio
Highway 7 which runs along the Ohio River and on the west by a rail spur
(see Figure I). Further, there is relatively little available unused space. The
available area to the north is used largely for coal storage, conveyance from
barge unloading, and ash disposal. The approximately thirteen acres to the south
of the main building contain ash-disposal facilities and underground water ducts.
The rated plant capacity is 2,300 MW(e). In recent years the plant has operated
at about 1,700 MW(e), burning only coal. The Sammis Station comprises seven
units and four stacks. The first six units are owned by the Ohio Edison Company.
The seventh unit is owned by a consortium: Ohio Edison (48.0 percent); Duquesne
Light Company (31.2 percent); and Pennsylvania Power Company (20.8 percent),
of which Ohio Edison owns all the common stock. All seven units comprise dry-
bottom, pulverized-coal boilers equipped with electrostatic precipitators.
As can be seen in Table I:
Boiler units I through 4 which exhaust into stacks I and
2 were built between 1959 and 1962, while units 5
through 7 exhausting into stocks 3 and 4 were built
between 1 967 and 1971
Units 1-4 comprise 32 percent of the total plant capacity
All units together consumed 3.8 million tons of coal in
1977 (Ohio Edison projects 5.8 million tons in 1980 and
5.5 mi II ion tons in 2000)
emissions in 1977 exceeded the scheduled SO, com-
pliance limitations (a 24-hour standard of either 7.91 Ib
per million Btu from each unit or an alternative 24-hour
-------
o\
Figure I
Plant Layout for W.H. Sammis Plant
Ohio Edison Company
/sircrt«iMnE /
Scries .or scrubbers and reio.ed focHHfcs re.oc ,o
-------
Tablet
Sammis Plant Characteristics
1977 Coal Use
Unit
1
2
3
4
5
6
7
Source:
Slack
1
1
2
2
3
3
4
Acurex
Plants,
MW Year of
(Nameplate) Installation
185
185
185
185
317.5
623
623
2,303.5
Corporation,
Final Report
1959
I960
1961
1962
1967
1969
1971
Ohio Tons
364,560
271,150
118,020
4(2,520
397,320
634,700
1,007,240
3,205,510
(84%)
Non-Ohio
Tons
69,440
51,650
22,480
78,580
75,680
120,900
191.860
610,590
(16%)
1977
Capacity
Factor (%)
55.72
40.68
17.18
61.48
35.80
28.67
45.54
JACA Corporation, and Professional Construction Manager
78-311, prepared for U.S. Environmental Protection Ageo
Current SO2
Emissions
(lb/!06Btu)
5.15
5.15
5.15
5.15
5.15
5.15
5.15
Optional SO, Limit
(Ib I0b dlu)
SO, Limit
(Ib/fO15 Btu)
2.91
2.91
2.91
2.91
2.91
2.91
2.91
nent, Inc., Engineering Stud/ for
cy, Division of
For FGD
Design
0.63
0.63
0.63
0.63
6.33
6.33
0.63
Ohio Coal
For Non-
FCO Design
1.61
1.61
1.61
1.61
4.46
4.46
4.46
Burning Power
Stationary Source Enforcement (Mountain
TlewTCalif.: Fort Washington, Pa.; and Cincinnati, Ohio] March 1979), Table I, p. 3.9-2.
-------
standard that Sammis chooses to apply: 1.61 Ib S07 per
million Btu from units I-A, and 4.46 from units 5-7)
Measurements in the vicinity of the Sammis plant have consistently shown
concentrations of particulates in excess of the secondary and primary National
Ambient Air Quality Standards and of the opacity levels set by Ohio's visibility
standards. And particulate emissions have consistently been far in excess of
Ohio's applicable emission limitation of O.I Ib per million Btu. As we indicate
later, EPA is establishing interim measures for reducing Sammis's particulate
emissions.
Despite the existence of a railroad spur adjacent to the plant site, Sammis does
not, and cannot at this time, receive coal by rail. Deliveries are made primarily
by barge and truck. Ohio Edison has stated that no more than 50 percent of
Sammis's coal deliveries can arrive by barge at the harbor north of the plant on
the Ohio River. For this reason, at least half the deliveries at present must
2
come by truck and must therefore consist largely of Ohio coal.
Sammis's tentative plan for 1980 is to blend 0.8 million tons of Ohio coal with
out-of-state, low-sulfur coal (the blending will be done by a bulldozer at the
o
plant's stockpiles). This planned quantity of 0.8 million tons for I960 is
substantially lower than the 3.2 million tons of Ohio coal purchased by Sammis in
1977 (see Table I).
Ohio Edison has had operational problems with units 5, 6, and 7. The boiler-
turbine-generator systems used on these units - sharply scaled-up versions of
similar systems previously built only cs much smaller units - have experienced
an unusual number of unscheduled outages due to failure of generators, turbines,
and boilers. According to an ongoing study by Bechtel Associates, the problems
that have been encountered in the boiler are aggravated by "the poorer quality
coal on the market today, as compared to coal commonly avuilable when the
[i
plant was designed."
8
-------
2.2 Legal and Regulatory Issues Affecting Sammis's Choice of Coals
Four sets of laws and regulations either do or may significantly affect Sammis's
choice of coals. First are the regulations in Ohio's State Implementation Plan
that limit the emissions of particulates from, and levels of opacity in the
vicinity of, Ohio's steam electric power plants. Because particulates and opacity
levels from Sammis have exceeded the limits set by the plan, EPA has served
several legal notices to the utilities that own Sammis. Second, Sections 110 and
126 of the 1977 Clean Air Act Amendments provide EPA and the states with
mechanisms for restricting the interstate transport of pollutants. Several
neighboring states attribute significant degradation of their air quality to the
transport of particulates and S02 from Sammis. Because of this pollutant-
transport effect, the state of West Virginia has joined forces with EPA in legal
action against Sammis. Third, the state of Ohio, after many delays, now has a
plan for limiting emissions of sulfur dioxide from steam electric plants.
Sammis's strategy for compliance involves significantly decreasing its current
rate of purchase of Ohio coals, which are relatively high in sulfur. This strategy
of sharply cutting the use of Ohio coals risks conflict with the fourth regulatory
issue - the "local or regional coal" provision in Section 125 of the Clean Air Act
Amendments.
Before discussing these legal and regulatory issues in somewhat greater detail,
let us look generally at the matter of Sammis's compliance strategies and coal
choices. Ohio Edison has indicated that the earliest feasible time at which it
will be able to comply completely with Ohio's particulate regulations is the fall
of 1986. An EPA consultant, PEDCo, has concluded that compliance will be
possible before 1984. Interim and final plans for compliance with the
particulate regulations in the Ohio Implementation Plan are still to be submitted
by Ohio Edison.
Because of the limited amount of land available at the Sammis plant, Ohio
Edison has determined that the construction of new facilities for reducing
particulate emissions would necessitate the design and construction of a bridge-
like structure over Ohio Highway 7 (which is adjacent to the plant on the east).
-------
Such a structure - which would require approval by the Ohio and U.S. depart-
ments of transportation would of course necessitate adopting safeguards to
preclude interference with the flow of traffic on the highway.
Ohio Edison has been advised by a consultant, Gilbert/Commonwealth, that the
most reliable and cost-effective method of achieving compliance with both SO
and particuiate emission limitations would be to purchase low-sulfur coal from
West Virginia and eastern Kentucky, to retrofit fabric-filter baghouses on
units 1-4, and to install new electrostatic precipitators or baghouses on
o
units 5-7, at an estimated cost of $480 million. PEDCo has estimated the
capital cost for installing new particuiate control facilities at about
Q
$300 million. Ohio Edison has stated that implementing such a strategy (Its
preferred strategy) would be wasteful jf Sammis were subsequently required to
retrofit a flue gas desulfurization (FGD) system on any of its units which
would occur if, for example, the Section 125 proceedings were to result in an
order to bum only Ohio (high-sulfur) coal. In that case, Ohio Edison argues,
(I) some of the particuiate control equipment might be rendered unnecessary,
and (2) the space used for the particuiate control systems might be needed for
FGD systems.10
In the sections that follow we explore more fully the background and implica-
tions of the legal and regulatory issues affecting the Sammis plant.
2.2.1 Particulates: The Ohio Implementation Plan and Interstate Transport
The Ohio Implementation Plan requires that, after June 1975, all large power
plants emit no more than O.I Ib ash per million Btu (AP-3-11). ("Large" power
plants are defined as those which, like Sammis, burn fuel at a rate exceeding
1,000 million Btu per hour.) Furthermore, there are limits to the extent to which
emissions may affect visibility: the opacity of visible emissions is limited to
20 percent with some periodic allowable exceptions (AP-3-07).
10
-------
Almost all Ohio utilities are either in compliance with the Ohio regulations for
particulates or have agreed to a schedule for final compliance. The exceptions
are one unit belonging to Cincinnati Gas and Electric and all forty-seven Ohio
units of Ohio Edison, including Sammis.
EPA has charged that the Ohio limit of O.I Ib ash per million Btu has been
exceeded at Sammis by factors ranging from 10 to 80. EPA has also charged
that Sammis has violated the opacity levels allowed by AP-3-07 of Ohio's plan
and, furthermore, that Sammis has in several instances violated an Emergency
Action Plan. That Emergency Action Plan is triggered during periods of high
ambient concentrations of particulates and certain meteorological conditions to
avoid the buildup of excessive concentrations in vulnerable counties of Ohio,
West Virginia, and Pennsylvania. It calls for having low-ash coals available and
for burning these coals when an alert is issued. We observe that the plan, which
is implemented for relatively short periods (for example, August 23-25, 1978,
and November 4-6, 1978), can be interpreted in effect as an "intermittent
supplemental control" plan superimposed upon the continuous controls that power
plants must apply in order to meet State Implementation Plans or New Source
Performance Standards.
The history of legal actions related to Sammis's excessive emissions of particu-
lates began when EPA issued a Notice of Violation to Ohio Edison on
22 September 1976 and a Notice of Violation to Duquesne Light one year later.
In a recent action (15 January 1979), EPA filed an Amended Motion for a
Preliminary Injunction, in which the State of West Virginia Air Pollution
Commission acted as Intervenor. The Amended Motion is less exigent than a
previously filed motion for a Preliminary Injunction, which it supersedes. The
earlier action, filed in August 1978, sought a final as well as an interim
compliance program. The Amended Motion in effect defers the question of final
compliance to a time when a full trial will be held to decide the merits of a still-
to-be-proposed resolution.
Compliance with the interim terms is expected to reduce Sammis's yearly
particulate emissions from 135,000 tons to 30,000 tons. Even with this 73 per-
cent emission reduction, however, Sammis is expected to emit particulates at
II
-------
about seven or eight times the allowable rate. The coals burned at Sammis
during the interim period may not exceed a "quality index" of 10 pounds of ash-
producing material per million Btu, with the index based on a 30-day weighted,
running average. EPA is currently considering whether the interim plan should
also include an interim mass-emissions regulation that is more lenient than the
state standard of O.I Ib particulates per million Btu (such as a limit in lb
particulates per hour for each unit corresponding to 0.8 Ib particulates per
million Btu when the unit operates at 100 percent of capacity).
Figure 2 illustrates the proximity of the Sammis Station (in Stratton) to the
states of West Virginia and Pennsylvania. The city of New Manchester, West
Virginia is in Hancock County, where over 10.5 percent of the total adult
population of 25,000 signed a petition submitted with the motion for a
13
Preliminary Injunction.
If
^ I.YIRGIXIA ,
? /Nj
HiKr /VIRGIHU
Figure 2
Location of the W.H. Sammis Plant
(Stratton, Jefferson County, Ohio)
Figure 3 shows the sectors of persistent winds in the area. Persistent winds can
be one of the meteorological mechanisms by which pollutant emissions are
transported from their sources to distant locations. The wind directions shown
12
-------
Figure 3
Sectors of Extremely Persistent Winds in the
Upper Ohio River Basin Area
IN. ^z V]
-,\}cu\ |pT <*
3^^Eg^--m73^-
^ **y4<
_J^s-^sJ<
mirf ««».tttNt ,/*
f ;^^"<^ '»~"Vm ' ._ .-
^X?N^V:
/\ fl"5/ X/^-^M:
&^M^₯>55/^
v //* /\/-~^vA««is-4'--/
v f & r % / » T ^~- _ r *r /
WL*^4X A^rs -M.'^7t^7XA».ou
^^/ VIRGINIA />^T:^
v-^Jl^v^
^-4^y7V ^Ae--//i?<^
2k
t^o Kilometers
Note; The arrows indicate schematically the direction in which the wind
persists for six hours' duration within a radius of 96 miles (155 km) from
the origin.
-------
by the sectors in the figure indicate that emissions to the air from Sammis may
degrade the air quality in certain communities in West Virginia and Pennsylvania
as well as In Ohio counties other than Jefferson.
2.2.2 Compliance with SO2 Emission Standards
The state of Ohio has had a stormy history with regard to the development and
implementation of an approvable and enforceable plan for controlling SO.,
emissions from power plants. The governor of Ohio has twice submitted a plan
for S02 and he has twice retracted the plan following challenges by various
parties, including EPA. Because Ohio did not adopt an approvable plan, EPA,
following its mandate under the Clean Air Act, promulgated S02 emission
regulations for the state. These regulations were ruled effective as of 17 June
1977 for all but certain rural power plants. Those plants that plan to comply
with the regulations by burning low-sulfur coals must be in final compliance by
October 1979; those planning to comply by using stack gas scrubbing must meet a
deadline of 13 June I960.14
Prior to EPA's promulgation of these regulations, Ohio was the only major
industrialized state in the nation totally lacking an enforceable implementation
plan. Further, now more than three and one-half years have elapsed since
utilities were to have been in compliance with such a plan, according to the
Clean Air Act of 1970.
Sammis has chosen the low-sulfur-coal compliance strategy and therefore must
comply with the $©2 plan by 19 October 1979. As it applies to Sammis, the plan
calls for limiting emissions to 2.91 Ib per million Btu, or alternatively, for
adopting a formula allowing different levels of emissions from the different units
of Sammis but resulting in an emission level equivalent to 2.91 Ib/IO Btu on a
plantwide basis. Sammis has chosen the alternative, which translates to: 1.61 Ib
per million Btu for units 1-4, which account for 740 MW(e), or 32 percent of
plant capacity; and 4.46 Ib per million Btu for units 5-7, which account for the
remaining 1,600 MW(e) of total capacity. The compliance emissions of 4.46 Ib
14
-------
S07 per million Btu for units 5-7 are not strikingly different from the average
1977 SOj emissions shown in Table I: 5.15 Ib per million Btu. The compliance
emissions for units 1-4, however, are relatively stringent. Sammis's plantwide
maximum of 2.91 Ib SC^ per million Btu is also relatively stringent: it can be
compared, for example, with the allowed maximum of 8.1 Ib SC>2 per million Btu
at Ohio Edison's Toronto Plant also in Jefferson County and near the Ohio
River or with the allowed maximum of 5.66 Ib for the Columbus and Southern
Ohio Electric plant in Conesville, Coshocton County.
When EPA first published the $©2 emission limits for Ohio plants, it did not
specify methods for demonstrating compliance, nor did it specify averaging
periods for sulfur or S02 measurements. Later, in February 1978, the Agency
described "acceptable fuel sampling analysis methods for demonstrating com-
pliance by S02 sources in Ohio." EPA will normally accept a utility's coal-
sulfur analyses if the utility has used EPA-approved sampling and analytical
methods based on 24-hour averaging; thus, S02 stack sampling is not normally
required. EPA does, however, reserve the option to require EPA-approved 502
stack testing, especially as the basis for any enforcement action. Furthermore,
it is expected that Ohio will permit the 50^ emission limit to be exceeded two
.,17 i
days per month.
In its compliance plan, Sammis has rejected the alternative of using flue gas
scrubbers. Major deterrents to the use of scrubbers include the additional space
that would be needed and the costs. Ohio Edison has estimated the cost of using
scrubbers for SO? control: investment costs are estimated at about $837 million;
18
and annual operating costs, at about $100 million. By contrast, Ohio Edison
estimates that the annualized cost to phase in coal in compliance with both 502
and participate standards would be about $181 million for the period from 1979
through I984.19
15
-------
2.23 SO2 Compliance and Section 125
Sammis plans to comply with EPA's S02 limitations by purchasing about
2.4 million tons of low-sulfur coal from Central Appalachian states. This
quantity is equivalent to about 75 percent of Sammis's 1977 consumption of Ohio
coals, which was 3.2 million tons. According to a study prepared for EPA
Sammis's shift away from Ohio coals would reduce the employment of coal
miners in Ohio by about 720 persons. The same study estimated that the shift
from Ohio coal by all the Ohio utilities that plan to comply with the SO plan by
burning out-of-state, low-sulfur coal would decrease purchases of Ohio coal by
about 15.8 million tons per year, and miners' jobs in Ohio by about 5,300. This
loss of jobs represents about 0.3 percent of the state's entire labor force, about
1.9 percent of the workers in the southeastern quarter of the state, an average
of about 8 percent of the working force in the four most important coal-mining
counties, between 25 and 28 percent of the labor force in one county (Harrison
County), and about 39 percent of Ohio's 1977 mining jobs. Additionally,
economic "ripples" resulting from the decline of mining activities would, it was
estimated, cause the loss of 8,000-10,000 nonmining jobs. The associated
unemployment costs would be $36-41 million for 26 weeks, after which time it
might be necessary to replace unemployment payments with welfare
?!
payments.
While these consequences imposed on the state's economy by the switch to out-
of-state coal are considered exaggerated by some (for example, the Council on
??
Wage and Price Stability), the economic and social impacts will undoubtedly be
serious for the Ohio mining communities affected by mine shutdowns or
slowdowns. As a result, Ohio has been urging the application of Section 125 of
the Clean Air Act Amendments. Section 125 provides for corrective action
where it is determined by EPA, the governor of an affected state, or the
president of the United States - that a shift from local or regional coal to an
alternative fuel would cause significant disruption or unemployment in the
community or region. Upon such determination, a utility can be ordered by the
governor or president to enter into contracts for local or regional coal.
16
-------
On 13 July 1978, EPA, in response to petitions by labor groups, Senator Metzen-
baum, and Governor Rhodes, instituted proceedings under Subsection 125(a) to
determine whether "action may be necessary to prevent or minimize significant
local or regional economic disruption and unemployment in Ohio." "Action"
would preclude the planned massive switch to non-Ohio coals by the fourteen
Ohio plants (including Sammis) that are included in the proceedings, and it would
result in the need for some degree of flue gas scrubbing by these plants.
The kinds of questions that the Section 125 proceedings raise are:
How will utility rates compare under the two options?
How reliable will retrofitted scrubbers be?
« How significantly will Ohio's gross annual product be
affected by the unemployment payments and ripple
effects due to the switch away from Ohio coal? (EPA
estimates a loss of $400 million, or 0.4 percent of Ohio's
total gross annual product)
Does "local or regional" denote only Ohio, or does it
denote also some or all of the other Appalachian states?
(A critical question)
If Ohio plants must adhere to a buy-Ohio policy, how can
the benefits to the Ohio mining community be weighed
against the losses to the mining communities in the other
states?
In support of a buy-Ohio interpretation of Section 125, one preliminary study for
EPA concluded that electricity prices would actually be lower in the long run
under the option of scrubbing Ohio coal than under the option of burning out-of-
23
state, low-sulfur coal. Ohio utilities, which would rather switch to low-sulfur
coal than install FGD systems, see things differently. So do non-Ohio coal
producers. A recently formed "Committee to Preserve the Appalachian Coal
Market" consisting of a group of Ohio electric utilities, including Ohio Edison
and coal producers from Kentucky in West Virginia has put forth the following
argument: (Da buy-Ohio policy (and scrubbers) would result in significantly
higher rate increases to utility customers in Ohio; (2) the economic disruption to
17
-------
miners in Kentucky and West Virginia would be serious; (3) if scrubbers were
required, five to seven years would elapse before they could become operative
during which time either out-of-state, low-sulfur coal would be used (with
serious disruptions to Ohio's coal industry) or SC^ standards would remain unmet.
In arguing against a "state" interpretation of the meaning of "regional or local "
U.S. senators from West Virginia and Kentucky insist that the original intent of
the "local or regional coal" amendment was to preclude massive transport of
western coal to the Appalachian region, and not to produce a "monopolistic" buy-
Ohio policy.
Another voice sounding the opinion that "Ohio by itself does not represent a
distinct region for coal" is the Council on Wage and Price Stability. In support of
its opinion, the Council predicts that actions resulting from Section 125 proceed-
ings will have milder employment and economic consequences for Ohio than
those suggested by the EPA study, and that these consequences must be weighed
both against the resulting economic disruptions to Kentucky and West Virginia
and against the higher electricity prices to Ohio consumers. Furthermore, the
Council observes, Ohio is already a major importer of coal, currently purchasing
over one-half of its coal from other Central Appalachian states.
EPA is expected to clarify the definition of "local or regional" soon. Even if the
outcome is contrary to that desired by some or all of the Ohio utilities, the
certainty it will produce vis-a-vis $©2 compliance should be welcome to the
utilities.
2.3 Characteristics and Production of Ohio Coals
2.3.1 Recent Production
In 1977, Ohio produced 47 million tons of coal from 445 reporting mines, all in
the eastern part of the state (see Figure 4). As can be calculated from Table 2
59 percent of Ohio's 1977 tonnage was produced in four of the state's 29 coal-
producing counties - Belmont (which produced 12 of the 47 million tons),
-------
Figure 4
Ohio Coal Production in 1977, by County
100,000 to I Million Tons
Source: Temple, Barker, & Sloan, Inc., Ohio Section 125 Study: Regional
Economic Impact Analysis, report prepared for U.S. Environmental
Protection Agency, EPA Contract No. 68-01-4905 (Wellesley Hills,
Mass., 14 December 1978), Figure 4, p. 111-2, based on data in the
State of Ohio's 1977 Division of Mines Report.
19
-------
Table 2
1977 Ohio Coal Production, by County and Seam
(In short tons)
County
Total
Athens
Belmont
Carroll
Columbiona
Coshocton
Gall la
Guernsey
Harrison
Hocking
Holmes
Jackson
Jefferson
Lawrence
Mahaning
Meigs
Monroe
Morgan
Muskingum
Noble
Perry
Stark
Tuscarawas
Vinton
Washington
Wayne
Total
46,940,131
96,636
11,943,666
310,370
1,173,230
1,829,929
431,599
963,749
5,989,033
1,153,399
680,887
1,045,126
4,052,713
242,181
296,581
1,637,367
1,387,303
264,494
5,78 ,170
357,313
2,304,028
702,562
1,732,068
2,326,289
198,516
33,922
Brookville Clarion
No. 4 No. 4a
1,493,746 2,971,145
_ _
-
-
-
- -
- -
-
_
130,013 88,574
-
193,552 254,122
-
47,340
198,815
1,637,367
-
-
- -
-
-
134,066
63,865
£92,153 991,082
33,922
Lower
Kitlmniog
No. 5
2,375,644
_
-
58,770
5,970
146,583
-
16,793
-
54,170
406,819
347,171
-
200
40,714
-
-
-
50,074
-
877
253,576
830,789
163,138
-
Middle Lower Upper Mohoning-
Kittaming Freepart Freeport Groff
No. 6 No. 6a No. 7 No. 7a
10,529,210 1,210,105 457,427 24,082
64,699 26,338 5,049
_
128,199 40,225 30,603 24,082
953,814 -78,002 133,303
1,683,346 -
_ _ _ _
3,930 - 125,759
922,505 974,940 467
£75,519 90.600
274,068 -
186.320 -
527,137 -
_ _ _ _
57,052 - - -
_
_ _ _ _
_ _ _ _
1.703,305 - 100,443
_ _ _ _
2,212,572 - 13,868
304,411 - 10,489
708,320 - 25,903
124,013 - 11,543
_ _ _
_ _ _
Meigs
Pittsburgh Redstone Creek Waynesburg
No. » No. 8a No. 9 No. 1 1
12,502,575 530,247 8,495,833 5,099,248
550
5,532,454 87,664 2,190,585 3,890,586
_
_ _ _ _
_
39,275 392,324
811,023 - 6,244
2,257,145 - 1,632,766 60,213
_ _ _ _
_
.
2,225,882 50,259 86,070 1,148,449
_ _ _ _
_ _ _ _
- - - -
1,387,303 -
264,494
172,282 - 3,759,845
357,313
76,711 -
_ _ _ _
_
_
198,516
_
Other
1,250,869
.
242,377
28,491
2,141
-
-
-
140,997
114,523
-
63,961
14,916
194,641
-
-
-
-
1,271
-
-
-
103,191
344,360
-
~
Source; State Division of Ohio, Department of Industrial Relations, Division of Mines, 1977 Division of Mines Report (Columbus, Ohio, n.d.), Table 5, p. 7.
-------
Harrison, Muskingum, and Jefferson; and 67 percent of the 1977 tonnage was
produced at three of the 14 minable seams Pittsburgh, Middle Kittanning, and
Meigs Creek. Seventy percent of the 1977 production was strip-mined (Ohio's
mines are relatively shallow - less than 400 feet deep).
2.3.2 Sulfur Content
Ohio coal is not low in sulfur. Table 3 shows that essentially none of Ohio's
estimated reserves are in the low-sulfur category (less than one percent sulfur by
weight), and that 66 percent of the reserves with measured sulfur content are in
the high-sulfur category (more than three percent sulfur by weight). By
contrast, 36 percent of the estimated coal reserves in West Virginia are in the
low-sulfur category. (Fourteen percent of all the estimated eastern bituminous
reserves are low-sulfur; and of those reserves, 53 percent are in West Virginia.)
Figure 5 shows the distribution of Ohio coal by sulfur content for reserves and
for 1977 deliveries to utilities. While 66 percent of the reserves with measured
sulfur content contained more than three percent sulfur (according to Bureau of
Mines data), about 77 percent of the Ohio coal delivered to utilities contained
more than three percent sulfur. (Subintervals of sulfur content were not
specified for the reserves data when sulfur content exceeded three percent.)
Because Ohio coal production generally occurs in mines that are less than
400 feet deep, and because low-sulfur coal is known to have been mined in Ohio
prior to 1910, the Ohio Division of Geological Survey undertook a program of
exploration in the deepest portions of the Ohio coal basin particularly in the
southeastern counties to determine whether the state might have significant
reserves of low-sulfur coal. The results were not encouraging: none of the
samples fell in the low-sulfur range. Only a few of the samples fell in the
medium-sulfur range; and of these, only the Lower Kittanning sample was in a
core of minable thickness. Most of the samples were in Ohio's "normal" (or high-
sulfur) range of 3 to 5 percent.24
21
-------
Toble3
Reserve Base of Eastern Bituminous Coals
(Millions of Tons)
KJ
Reserve Base by Sulfur Content (% S by Weight)
Origin
Total
Eastern
Ohio
West
Virginia
Source:
Production
Method
Deep
Strip
Deep
Strip
Deep
Strip
U.S. Bureau of Mines,
S <_ 1.0%
21,220
5,302
115
19
11,807
3,005
Reserve Base of
1 .0 < S < 3.0%
48,461
6,822
5,450
991
12,583
1,423
U.S. Coals by Sulfur
S > 3.0%
65,992
15,434
10,109
2,525
6,553
270
Content, Eastern
S Content
Unknown
25,811
4,936
1,754
118
4,143
600
States, 1C 8680,
Total
Reserves
161,516
32,511
17,423
3,654
34,378
5,212
PB-243031
(Pittsburgh, Pa., May 1975).
Note; Reserves included are from coal beds east of the Mississippi River that are more than 28 inches thick and
less than 1,000 feet deep. Estimates are for "coal in place"; potential mining losses are not accounted
for.
-------
0.60 '
0.50
0.40
o
5 0.30
g
o
0.20
0.10
Figure 5
Histograms of Ohio Coal Reserves and Deliveries in 1977
Deliveries to Utilities, 1977°
Reserves
:
1.0
2.0
3.0
4.0
i.O
Total deliveries of Ohio coal were 41.6 million tons in 1977. Source;
National Coal Association, 1978 Steam Electric Plant Factors (Washington,
D.C., 1978).
Reserves of Ohio coal in billions of tons were estimated as: total = 21.1;
reserves with > 3% sulfur = 12.6; reserves of unknown sulfur content = 1.87.
If the "unknown" are included in the total, reserves with > 3% sulfur account
for 60% of the total; if the "unknown" are subtacted, they account for 66%.
Source; U.S. Bureau of Mines, Reserve Base of U.S. Coal by Sulfur Content,
1C 8680, PB-2M 031 (Pittsburgh, Pa., May 1975).
23
-------
2.3 J Incombustible (Ash-Producing) Matter
As mentioned earlier, EPA is requiring that Sammis now bum coal with a
"quality index" not exceeding 10 Ib ash per million Btu, computed on the basis of
a 30-day weighted, running average.
Of the 42 million tons of Ohio coal delivered to electric utilities in 1977, the
average value of ash-producing matter was 14.3 Ib per million Btu (about half
of Ohio's coal production is subject to low-level cleaning). That value is
considerably higher than the ash content of Ohio coals indicated in the Bureau of
Mines (BOM) coal reserves data base, as can be seen from the following BOM
average ash content values for Ohio's major coal-producing counties:
Ash Content Btu/lb
County No. of Samples As-Received (%) As-Received Ib Ash/10^ Btu
Belmont 431 10.7 12,500 8.6
Coshocton 83 7.3 12,150 6.0
Jefferson 478 9.9 13,230 7.5
Muskingum 227 10.3 12,670 8.1
Perry 434 9.7 12,670 7.6
Tuscarawas 92 11.0 12,680 8.7
Vinton 59 9.9 11,670 8.5
The BOM coal data base, from which the above values are taken, represents raw-
coal samples taken since the turn of the century, mainly from producing mines.
The considerably higher ash content of recently delivered Ohio coals (some of
which are cleaned) may represent mining practices that yield relatively large
quantities of incombustible material, or, indeed, the decline in the quality of the
Ohio coal mined throughout the century. (A later section - see Table 12 -
depicts the ash content of recently measured samples of Ohio coals the -
according to washability data performed on these samples - are potentially
washable to $©2 compliance levels.)
24
-------
2,3.4 Coal-Preparation Practices in Ohio
In 1976 almost half the Ohio coals produced were prepared with some degree of
"mechanical cleaning," which, on the average, left behind as refuse about
27
30 percent by weight of the feed coal. Although there are no generally
available detailed data on the level and performance of the cleaning processes
used, our conversations with Ohio coal-pre par at ion managers and our review of
published summaries of the types of coal-cleaning equipment used indicate that
the level of cleaning has generally been relatively low. Coarse crushing is
generally performed, the degree of removal of the less coarse incombustible
material is generally relatively low, advanced technology has not been employed,
and sulfur removal has been only incidental. When we asked managers about the
objective of their coal-washing operations, their frequent response was: "Just to
remove stone."
Can a significant number of Ohio's existing coal-cleaning facilities be upgraded
to achieve a higher level of ash and sulfur removal? We observe that eleven of
the seventeen plants listed in the 1977 Keystone Coal Industry Manual clean only
coarse cod with dense media washers or jigs. The fine cod is either discarded
or recombined (uncleaned) with the coarse coal product. It is possible to upgrade
these plants to provide additional ash and sulfur rejection by adding fine-coal
circuits. The decision to upgrade the plants would depend on the cleanability of
the coal being processed and the costs of plant modifications. We observe
further that several companies are now marketing modular coal-cleaning units
that can be placed in operation within six months. These units can be assembled
either to modify existing plants or to serve as independent units.
It appears, therefore, that many of Ohio's older coal-cleaning plants can be
upgraded or replaced, given adequate economic incentive.
Ohio PCC plants, like Ohio cod mines, have for the most part operated on a
smdl scale. Until very recently, the only large PCC Plant in Ohio was the
Consolidated Coal Company plant near Georgetown, in Brown County. Built in
the 1950s, this plant was designed mainly to remove ash from high-ash coals,
mainly Meigs Creek coals. There now appears to be a trend in Ohio to build
large PCC plants incorporating relatively advanced technology. One such plant,
25
-------
soon to come on line, is located near Cadiz (in Harrison County) and is owned by
R6\F Coal Company. This plant will process and blend Ohio cods from the
Pittsburgh, Meigs Creek, and Waynesburg seams.The three types of coal will be
stored in separate silos and blended with variable-speed feeders at rates
determined by automatic measuring systems. The 1,000-ton-per-hour plant will
have different circuits (including heavy-media and water-only cyclones) for
differently sized particles, including fines down to 325 mesh. Located in a
nonattainment area for particulates, the plant will not use thermal dryers. It is
expected that 50 to 70 percent of total sulfur will be removed, and that ash
content will be reduced to about 14.5 percent (in some cases from as much as
25 percent).29
Half the product from the R & F facility will be sold under contract to TVA's
Colbert plant, whose delivered coal must have a heating value of 11,500 Btu per
pound and produce no more than 4.0 Ib SO, per million Btu. PCC will add $6 oer
29
ton to TVA's cost. Negotiations are currently under way for the remaining
output (half of 1.6 mill ion tons per year, if we assume that the plant operates
13 hours per day and 250 days per year).
2.4 Coals Historically Used by Sammis and
Representative Compliance Coals
2.4.1 Coals Historically Used by Sammis
As indicated earlier, Sammis must burn coal that produces: (I) no more than
10 Ib ash per million Btu on a 30-day running average in order to adhere to an
interim particulate-emissian standard; and (2) no more than 4.46 Ib SO7 per
million Btu for about 70 percent of the plant's capacity, and no more than 1.61 Ib
£©2 per million Btu for the remaining capacity, both on a 24-hour basis. The
coals listed in Table 4 represent about 90 percent of the tonnage delivered to
Sammis in May 1978. The entries include the largest deliveries and represent the
total tonnage's range of values for uncontrolled sulfur and ash emissions, heating
value, and delivered cost for that month. A summary of the May and November
1978 coal deliveries to Sammis aggregated by state and showing weighted
26
-------
Table*
A Representative Selection of Historic Coal Deliveries to Sammis in May 1978°
Company
Valley Camp
Cool Co.
-
Consolidation
Coal Co.
Midwestern Region
F & F Mining Carp
Valley Camp
Coal Co.
Black Hawk Mining
Co., Inc.
fO
-~J Youghiogheny and
Ohio Coal Co.
-
Industrial
Mining Co.
Botch Mining Co.
-
North American
Coal Corp.
F & M Coal Co.
-
-
Schiappa Coal
Co., Inc.
C & W Mining Co.
-
Mine, Slate
No. 1, WV
Elkhorn, KY
Georgetown, OH
FAF.WV
Alexander, WV
Black Hawk, KY
Nelms No. 2, OH
Buzzard, OH
Bergholz.OH
Betsy, OH
CC & R, OH
PowhatenNo. 1,
and No. 3. OH
F&M.OH
No. 38, OH
Monwest, PA
No. 43 4
No. 56, OH
No. 3 and
No. 5, OH
Gollatin, PA
a The select lorn, representing the range
County
Ohio
Floyd
Harrison
Boooe A
Fayetle
Marshall
Floyd
Harrison
Jefferson
Jefferson
Jefferson
Columblano
Bebnont
Jefferson
Jefferson
Fayette
Jefferson
Columbiana
Fayette
Seam
Pittsburgh
-
Pittsburgh
-
Pittsburgh
Lower
Freeport
-
Middle
Kittanning
Pittsburgh
-
Pittsburgh
Harlem
-
-
Pittsburgh
Middle
Kittanning
-
of ash and sulfur contents, are from
/i
-------
averages of the coal characteristics and delivered prices - is presented in
Table 5.
Looking at the percentage of ash in the historic coals listed in Table 4, we
observe that these percentages are too high to comply with Sammis's interim
requirements. As for sulfur content, only one of the listed coals comfortably
meets the more stringent SC^ limitation of 1.61 Ib SC^ per million Btu: this is
the coal from the F & F mine in West Virginia, listed at 0.70 Ib S0? per million
Btu. Two of the other cods - those from the Elkhorn (1.44 Ib) and Block Hawk
(1.56 Ib) mines in Kentucky are very close to the 1.61 Ib limit but surely too
close when sulfur variability is taken into account (as it must be for 24-hour
averaging). The average SC^ emissions for five of the listed coals are somewhat
below the 4.46 Ib limit. Taking into account sulfur variability, however, only one
of the coals that from Gallatin, Pennsylvania would probably qualify as a
compliance coal for S02 (see Section 3.1 for a discussion of sulfur variability).
The sample coals listed in the table Illustrate what is already known that
Sammis cannot comply with particulate or SO2 limitations by burning only Ohio
coal, given existing levels of control and preparation.
The weighted average of ash-producing matter in reported Ohio coal deliveries
to Sammis in January 1977 was 16.6 percent. As shown in Table 5, the ash
content of coals delivered to Sammis in May 1978 (in pounds of ash per million
Btu) was 14 for Ohio and Pennsylvania cods and 12 for West Virginia and
Kentucky coals. Ohio Edison has reported that, during the months of December
1978, January 1979, and February 1979, the average ash content of all coals
delivered to Sammis was slightly below the interim limit of 10 Ib per million Btu
on a monthly basis. (The ash reduction was accomplished largely by washing
coals from Gallatin, Pennsylvania; see Table 4). All the ash values cited above
apply to "as-received" rather than "as-burned" coal. In the present context as-
bumed c<" il is coal that has been pulverized and usually also stored for some
time. For reasons that are not understood (but tentatively ascribed to different
measurement techniques), the ash-quality index of "as-burned" coal at Sammis,
measured by Ohio Edison, has been higher than that of the "as-received" coal.
At present it is not yet clear whether the interim requirement of an ash-quality
28
-------
TobleS
Summary of Sammis's May and November 1978 Coal Deliveries by State of Origin
NJ
VO
State of Origin
May 1978
Ohio
Pennsylvania
West Virginia
Kentucky
November 1978
Ohio
Pennsylvania
Maryland
West Virginia
Kentucky
Source: National
Tons
Delivered
(I03)
367 A
103.8
53.6
66.0
291.9
52.7
15.3
20.6
10.7
%of
Total Tons
Delivered
62
18
9
II
75
13
4
5
3
Coal Association, Power
which were reported
Sulfur
(lb/!06Btu)
2.60
1.82
2.05
0.90
2.58
2.12
2.20
2.28
2.08
SO^
(Ib/IO^Btu)
5.2
3.68
4.1
1.8
5.16
4.24
4.40
4.56
4.16
Plant Coal Deliveries (Washington
Ash
(lb/!06Btu)
13.8
14.0
11.8
12.0
NA
NA
NA
NA
NA
, D.C., 1978),
Btu/lb
11,350
11,610
11,780
11,370
11,856
11,740
11,830
11,860
11,490
except for
Delivered
-Price
($/Ton)
25.13
27.36
30.27
34.84
27.11
25.65
25.53
27.32
33.35
ash values,
in Coal Outlook, 28 Auaust 1978.
-------
index of 10 will be applied to the as-received or the as-burned measurements.
We emphasize that the ash-quality index of 10 for Sammis will result in
particulate emissions that exceed by about seven or eight times the emission
limit of the Ohio Implementation Plan (O.I Ib per million Btu). Compliance with
the statewide standard will require a combination of upgrading the particulate
control systems and using coals of lower ash content.
2.42 Representative Compliance Coals
Table 6 presents a set of representative low-sulfur and low-ash coals which, we
determined recently, are available for delivery on contract terms. Listed in the
table are: the sources of the coals, distances by rail and barge from source to
Sammis, sulfur and ash contents, heating values, and estimated availability and
f.oJb. mine prices. The f.oJD. mine prices listed must be considered tentative.
Changing market conditions and actual contract terms may result in negotiated
prices that are different. Although prices of quality coal in the early part of
1979 were depressed (even in the spot market), this situation will probably not
persist. Similarly, the stated availability (tons per year and number of years) is
subject to change.
As can be seen in Table 6, essentially all the compliance coals listed are from
states other than Ohio mainly southern West Virginia but also eastern Ken-
tucky and Pennsylvania. These coals represent the kinds of coals that would
comprise the majority of Sammis's deliveries under Ohio Edison's S02 compliance
strategy. All these coals would be delivered by barge, often after some overland
transport. (Again, Sammis currently can handle only about 50 percent of its coal
deliveries by barge; hence, its planned compliance strategy would appear to
require expansion of the barge unloading and conveyer facilities.) Barge rates,
generally not regulated, are lower than the rates for other modes of transport.
In May 1 979, for example, the rate for the 243-mile barge haul from Charleston,
35
West Virginia, to Pittsburgh, Pennsylvania, was 1.3 cents per ton-mile.
Although not specified as such, some of the compliance coals represented in
Table 6 (and some of the delivered coals listed in Table 4) reflect low levels of
coal cleaning and tnus some degree of sulfur and ash removal.
30
-------
Table 6
Representative SO-, Compliance Coals for Sammis
Source of Ctxil
Company
C<* melton
Industries, W. Va.
Classic Coals, Ky.
Uuckhaniton Sales
Co., W. Va.
C linclitield Cool, Va.
(Kaw/C leaned)
K.C.U. Coals, Ky.
(,loc iul Minerals, Pa.
H&H Mining, W. Vu.
lorsyth Coal
Lxchunge, N. C.
CSR, Inc., W. Va.
Coal Cave, W. Va.
t.
10" Tons/Yr Years
County State Available Available
Kanawha W. Va.
(loading point)
Lawrence Ky.
Upshur W. Va.
Russell Va.
Carroll Ky.
(loading point)
Clarion Pa.
Summers W. Va.
Lawrence Ohio
(loading point)
Upsl>ur, Lewis, Clay W. Va.
-ayette W. Va.
(loading point)
\ 197V, 1980
0.15 I979-?
0.18-0.24 19/9
0.72-0.84 I980-?
0.25
0.24 1979
0.36-0.48 1980-1983?
0.50
1 1979
0.10-0.15 in 3 mo.
0.40
0.14-0.24 I979-?
Transportation
Distance
(miles)
297 Barge
703 Barge
86 Kail
60 Barge
160 Rail
262 Barge
485 Barge
100 Rail
70 Barge
100 Rail
281 Barge
277 Barge
l52(Upsltur),
152 (Lewis),
53 (Cloy) Rail;
281 Barge
303 Barge
<-ri Moisture/
^2 Ash Volatile F/J.B. Mine
(lb/l06Btu)° (lb/IObBtu) Matter Btu/lb (e/l06Blu)D Reference0
1.08 10.55 l.57%/ 12,315 142.1
33.91%
1. 16-1.22 6.15-6.50 12,300- 122-104
13,000
3.84-4.22 7.81 12,800- 96.2-97.7
13,000
1.12 Raw 12.00 Raw 12,500 152 Raw
8.00 Cleaned 1 72 Cleaned
4.16 10.00 12,000 104.2
3.74 7.78 12,850
1.54 7.69 / 34% 13,000 146
4.16 10.00 12,000 105.2
3.34 12,000 117
1.42 10.52 11,900 71.4
1
2
1
2
1
2
2
1
2
1
Assuming all sulfur is emitted us SC^.
May include low-level cleaning. Price may be f.o.b. loading point (see Col. 2) or delivered price (if «).
References:
I. Communications during November 1978 willi Norman Kilpalrick, director of Surface Mining Research Library, Charleston, W. Va., and consultant to Teknekron Research,
Inc.
2. leknekron's ^ inal Report on Work Assignment 3 (R-OI l-LPA-79), EPA Task Order Contract 68-02-3092, 26 January 1979.
3. Uivid Large, as uf fiant for LPA in Civil Action No. C2-78-76, 11 July 1978, p. 7.
-------
Table 6 (Continued)
Source of Coal
Company
Bruce Mining, W. Va.
Vande Linde, W. Va.
Oglebay Norton
Co., Ohio
Oglebay Norton
Co., Ohio
Oglebay Norton
Co., Ohio
Island Creek, Ky.
u>
NJ
Island Creek, Ky.
Island Creek, Ky.
Peabody Coal Co., Ohio
Oglebay Norton
Co., Ohio
County
Barbour
Webster
Wyoming
(loading point)
Greenbriar
Greenbriar
Logan
Upshur
Upshur
Perry
McDowell
I06 Tons/Yr
State Available
W. Va. 0.29
0.50
W. Va. 2
W. Va.
W. Va.
W. Va.
W. Va. 1.6
W. Va.
W. Va. 1.5
3
Ohio 0.12-0.14
W. Va. 2
Years
Available
90 days
180 days
in 6 mo.
1979
stockpiled
by 1980
by 1980
late 1980
by 1981
Jan. I979-?
6.5
Transportation
Distance
(miles)
174 Rail
281 Barge
189 Rail
281 Barge
164 Rail
264 Barge
64 Rail
281 Barge
64 Rail
281 Barge
85 Rail
264 Barge
86 Rail
60 Barge
86 Rail
60 Barge
150 Truck
175 Rail
264 Barge
so2
(lb/!06Btu)a
2.30
1.54-1.60
0.64
1. 14-1. 18
1. 14-1. 18
1.66
3.34
3.34
4.18-4.28
0.64
Moisture/
Ash Volatile
(lb/IO*Btu) Matter
7.69 /34%
7.69-8.00
6.58
5.71-5.93 / 25%
5.71-5.93 / 25%
10.00
8.33
8.33
8.70-10.71
7.2
Btu/lb
13,000
12,500-
13,000
12,500
13,500-
14,000
13,500-
14,000
12,000
12,000
12,000
11,200-
11,500
15,500
F.O.B. Mine
(C/KTBUi)6
107.7
108-112
128-140
161-167
143-148
125-133.3
125
125-133.3
104.4- 107.1*
128
Reference
2
1
2
2
1
2
1
3
3
1
Assuming all sulfur is emitted as SO2-
May include low-level cleaning. Price may be f.o.b. loading point (see Col. 2) or delivered price (if ").
References:
I. Communications during November 1978 with Norman Kilpatrick, director of Surface Mining Research Library, Charleston, W. Va. and consultant lo leknekron Research,
Inc.
2. Teknekron's Final Report on Work Assignment 3 (R-OI l-EPA-79), EPA Task Order Contract 68-02-3092, 26 January 1979.
3. David Large, as affiant for EPA in Civil Action No. C2-78-76, II July 1978, p. 7.
-------
3. PROSPECTS FOR THE USE OF CLEANED OHIO COALS AT SAMMIS
Without physical coal cleaning (PCC), Sammis's proposed strategy of burning
low-sulfur coal will mean that most of the plant's supplies will come, not from
Ohio, but from southern West Virginia and eastern Kentucky. This may have
serious implications for both Sammis and the Ohio coal mining industry. Sammis
will have to augment its barge-unloading facilities, which now receive both out-
of-state coals and some Ohio coals but currently can handle only about
50 percent of the plant's coal deliveries. Furthermore, Sammis will have to
modify some of its existing contracts and purchase coal in a market which, while
weak at this time, is bound to become increasingly competitive. As for the Ohio
coal mining industry, the decision by Sammis (and other big Ohio plants) to
substitute most of the current Ohio coal purchases with out-of-state supplies
could lead to the loss of coal-industry jobs (and to associated economic "ripples")
as well as to the degradation of coal-production facilities and know-how,
particularly in Jefferson and nearby counties. This issue is at the core of the
current Section 125 proceedings.
Ohio coal has a relatively high heating value, it is relatively easy to mine, and it
is easily transported to Sammis. Moreover, several properties of Ohio
coals for example, grindability index, ash fusion temperature, characteristics
of the ash, and moisture content - are generally suitable for the dry-bottom
boilers of the Sammis station. But the ash content and sulfur content of Ohio
coals are generally too high for existing and proposed emission limitations and
control facilities. Since PCC can lower both ash and sulfur content with some
cost in dollars and energy, but with some side benefits as well we examine the
subject of burning cleaned Ohio coal.
In this section we look first at the subject of sulfur variability, including the
question of how PCC may affect values of relative standard deviation (RSD) of
sulfur content. We next discuss the 502 comPl'ance strategy that another Ohio
utility has proposed to EPA in regard to a power plant located in the central part
of the state: the proposed strategy is to burn cleaned coals from current sources
33
-------
near the plant. We then examine the available data on Ohio coal washability
and, from these data, estimate the increase in the use of Ohio coal at Sammis
that may result from PCC. Finally, we discuss PCC in terms of its potential
costs and benefits and compare the use of cleaned Ohio coals with uncleaned
out-of-state, low-sulfur coals at Sammis units 5, 6, and 7.
3.1 Average Coal-Sulfur Values in Relation to SO, Emission
Limits and Coal-Sulfur Variability L
In order to determine the required sulfur content of the mix of coals to be
burned, a plant's fuels manager must know not only the SIP's allowable maximum
$©2 emission level for his plant, but also the applicable effective S07 emission
level. Because of statistical fluctuations, the effective, or mean, S09 emission
limit will be lower than the maximum allowable $©2 emissions. How much lower
will depend upon such factors as: the variability of the SOj emissions, often
described by the relative standard deviation (RSD) of S02 emissions; the
allowable frequency with which the maximum SOo emission level can be
exceeded; the allowable "confidence level," reflecting an acceptable (small)
probability of violating the standard; and the probability distribution (for
example, a normal or lognormal distribution) of measured SO- emission levels.
In the case of Sammis, the manager's choice of coals must be such that the mean
sulfur value of the mix burned in units I through 7 averaged over 24
hours ensures that the probability of meeting the maximum allowable S09
emission level for each unit (with no more than two exceedances each month)
will correspond to a designated confidence level. He will need to know the
difference between the emission limit and the effective, or mean, level of SO,
emissions - and, of course, will prefer that this difference be minimal. In this
section we mention various factors that affect this difference in general and at
Sammis in particular. In the Appendix we present a more detailed discussion of
sulfur variability.
-------
A larger RSD means a larger difference between the allowable maximum and
mean SO? emissions and hence a lower, or more stringent, effective 502
emission limit. One of the factors that increases the RSD of the weight
percentjge of sulfur in a particular coal is a decrease of the lot size of the coal
from which measured samples are drawn, since fluctuations are expected to be
smoothed with larger lot sizes. Since, for a power plant burning coal at a fixed
rate, the lot size to be sampled is related directly to the averaging period, the
variance and RSD also decrease with increased averaging periods. The effect of
smoothing fluctuations with larger lot sizes (or averaging periods) is illustrated
schematically in Figure 6, in which the same coal is sampled at two different
intervals. Compared with the solid - and more fluctuating - curve, the
dotted - less varying - curve represents sampling at less frequent intervals
(that is, larger averaging periods or larger lot sizes). If Ohio power plants were
permitted to determine S02 emissions on the basis of 30-day composite
samples - rather than 24-hour composite samples - the RSD would theoreti-
cally be expected to equal the 2k-hour RSD divided by \/30 . We observe that
the decrease of RSD with increasing lot size implies that the 502 ''mit» ^or a
given averaging period, is effectively more stringent for small boilers than for
large boilers.
Rgure6
Illustration of the Effect of Averaging Period on RSD
Uncontrolled
Emission Level
(Ib S02/l
-------
RSDs of sulfur content vary from coal to coal (for a given lot size). There is no
experimental basis for linking RSDs with coal type or sulfur content. The
assumption (sometimes made for lack of empirical data) that the RSD per unit
weight of a coal is independent of the coal's sulfur content implies a smaller
variance and standard deviation for lower-sulfur coals (since the RSD equals the
ratio of the standard deviation to the mean).
The RSD of the SC^ emissions (in pounds per million Btu) will be determined
largely by the RSD of the sulfur content, but not entirely. The variability in a
coal's heating value affects the RSD of the S02 emissions (Ib per million Btu) to
a small extent; a report on sulfur variability by PEDCo sets the RSDs of SO
emissions (Ib per million Btu) equal to 1.05 times the RSDs of sulfur content.
Two other factors with relatively small effects on the RSD of sulfur emissions
are: (I) the variability of sulfur retention in the ash during combustion (the
fraction of sulfur retained depends largely upon the coal's alkaline content); and
(2) the variability of the small amount of sulfur removal during pulverizing of the
coal at the power plant.
An analysis of a limited number of data sets has shown that the RSD of pounds of
SC>2 emitted per million Btu decreases as a result of physical coal cleaning
(PCC), somewhat more so with somewhat deeper levels of cleaning,38 but that
the RSD of the weight percentage of sulfur in the coal often does not decrease
after PCC. These results indicate the importance of the enhancement of the
cleaned coal's heating value. They also suggest that in the raw coal the
RSD of pounds of S02 emitted per million Btu is greater for the pyritic SO
which is removed by PCC, than for the organic $©2, which is not removed by
PCC (see the Appendix).
An important factor in determining the applicable effective S02 emission levl is
the acceptable confidence level, related to the probability of emissions being
above the established maximum $©2 emission level. The greater the level of
confidence that no exceedances (or an allowable number of exceedances) will
occur in a specified time, the greater will be the difference between the
allowable maximum and mean SO2 emission levels. Thus, the effective SO7 limit
-------
will be mare stringent for a greater confidence level. A confidence level of
95 percent, for example, implies that, for a normal probability distribution (see
below), the probability of exceeding the maximum allowable SCK emission level
is 0.05, or that violations will be tolerated about 18 days per year. A confidence
level of 99.87 percent implies that violations will be tolerated about one day in a
thousand.
The probability distribution of sulfur measurements also affects the allowable
effective level of SC^ emissions, given a maximum SOj emission level and a
value of RSD. For convenience, a normal distribution of sulfur content is often
used. But, in fact, using a distribution skewed toward higher values for
example, a lognormal or inverted gamma distribution has provided a good
empirical fit to a number of sets of coal-sulfur measurements. A lognormal
distribution can be transformed into a normal distribution by setting the mean
equal to the natural logarithm of sulfur content in the lognormal distribution and
setting the standard deviation equal to the RSD of the lognormal distribution.
For a given RSD and confidence level, a coal will have a lower mean sulfur level
if its sulfur content is lognormally distributed than if it is normally distributed.
Given a confidence level for a normal distribution, the difference between the
mean and the maximum SC^ emission limit can be expressed as a specified
multiple of the standard deviation. This multiple is called the normal variate
and can be found in standard tables of "normal curve areas." Here are some
examples of normal variates, their corresponding confidence levels, and their
implications regarding the number of days per year in which the maximum
S02 emission limit can be violated.
Z = Normal Variate
(Number of Standard Deviations Number of Days per Year
Confidence Level (%) between the Mean and Limit) of Tolerated Violations
84.13 1.0 58.0
95.00 1.645 18.0
97.72 2.0 8.0
99.87 3.0 0.5
37
-------
The mean value, m, of a normal distribution is related to the emission limit,
max, by:
or, since RSD = o/m, by:
max-m = Z a.
m = max/( I + Z RSD),
where a is the standard deviation and Z is the normal variate, corresponding to a
given confidence level.
Diagramatically this relationship is illustrated for a confidence level of 95 per-
cent in the following figure:
m
-allowable emissions
max
It is expected that a confidence level of 99.87 percent (calling for three standard
39
deviations between the mean and the maximum) will be required by EPA.
We will show how the factors mentioned above can determine the effective SO7
emission level required at Sammis. Before doing so, however, we discuss
Table 7, which presents values of sulfur variability as RSD, computed from
measurements of sulfur content and heating value in samples of coal from
38
-------
Table?
Values of the Relative Standard Deviation (BSD)
of Sulfur Content in Ohio Coals
County
Tuscarawos
Jefferson
Harrison
Coshocton
Muskingum
QmffV
r^sff
Vinton
Seam Nome
L. Kittanning \
M. Kittamingf
l_ Kittanning
L. Kittanning
N0.7&7A
M. Kittanning
Unknown
Pittsburgh
L. Freeport
Unknown
No. 6
No. 6
No. 6
Waynesburg
No. 5
No. 5
Unknown
Unknown
No. 6
Unknown
L. Kittanning 1
M. Kittanning/
Unknown
No. 8
M. Kittanning
Unknown
M. Kittanning
M. Kittanning
Unknown
No. 6
Unknown
Unknown
Unknown
Clarion
Unknown
Unknown
Clarion \
L. Kittanning/
Unknown
Mining Method
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Underground
Surface
Surface
Underground
Underground
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Underground/
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Underground
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Surface
Preparation
Method
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Row
Raw
Raw
Raw
Raw
Raw
Raw
Washed
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Raw
Tons
1,150
1,426
1,213
1,248
1,190
1,495
1,311
3,361
2,373
947
,294
,388
,044
959
,181
,250
,449
,016
,183
,192
,568
,245
924
,192
,360
992
1,046
2,025
512
989
984
960
349
1,040
Average
SO, Emissions
(IbflO^Btur
7.50
7.26
6.37
6.66
6.59
6.00
6.11
8.26
8.72
9.48
6.75
7.15
7.02
7.13
7.12
8.09
6.84
5.71
6.98
6.68
6.16
6.62
6.05
6.50
7.49
4.54
5.84
1.51
6.25
7.22
6.77
6.52
6.71
5.65
RSD (%)
19.47
21.19
16.65
20.62
22.26
22.13
19.64
21.71
19.15
18.86
15.82
16.99
23.39
24.51
14.77
22.56
7.34
3.15
31.33
17.05
21.07
15.09
8.82
22.34
16.51
12.07
21.04
_
14.72
15.35
18.85
24.66
26.60
"
Number
of
Samples
68
362
337
175
45
116
275
232
454
455
203
531
262
108
33
40
43
8
3
131
103
479
51
6
295
II
262
53
3
3
210
251
176
78
2
Source;
Written Communication from Ray Morrison, U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Durham, North Carolina, April 14, 1979.
a Assuming that all sulfur leaves the stock as S02.
39
-------
various lot sizes and from various counties and coal seams in Ohio. The table
also indicates the number of samples available in each data set, the type of
mining (surface or underground), average values of potential SG>2 emissions, and
whether the data set represents raw or washed coal (only one of the coals listed
is washed). We note that, in general, the data set for a given coal and lot size
can represent either a composite sample of that lot size or the average of
individual measurements made on different samples of the same lot size.
Among the 39 Ohio samples listed in Table 7, the range of RSD values is large:
from 8.82 to 26.60 percent. (This brackets the value of 15 percent that EPA has
used as a typical RSD.) All the SO2 values are high: none falls below the limit
of 4.46 Ib SO2 per million Btu established for Sammis units 5-7, and all are
considerably higher than the 1.16 Ib limit for units 1-4.
To see if there is any obvious correlation among RSD values for samples taken
from the same county, seam, and type of mine, let us examine separately the
pairs of samples shown below, all extracted from Table 7:
County
Seam
Mine Type RSD
Tons
Tuscarawas
Tuscarawas
Coshocton
Coshocton
Coshocton
Coshocton
Coshocton
Coshocton
Muskingum
Muskingum
L. Kittanning
L. Kittanning
No. 6
No. 6
No. 6
No. 6
No. 5
No. 5
M. Kittanning
M. Kittanning
Surface
Surface
Underground
Underground
Surface
Surface
Surface
Surface
Surface
Surface
21.19
16.68
19.15
18.86
21.71
22.56
16.99
23.39
21.07
8.82
22.34
1,425
1,213
2,375
945
3,360
1,250
1,390
1,045
1,570
925
1,190
Difference
between
RSDs
4.5
1.7
0.8
6.4
13.5
There is no obvious correlation between the RSDs in each pair of samples. On
the basis of the listed RSDs, one cannot conclude that the RSD of a sample from
40
-------
a particular county, seam, and type of mine will be close in value to the RSD of
another sample from the same county, seam, and type of mine. There is also no
obvious relationship here between RSD and the number of tons in the population
represented by a sample. We offer two observations, however. First, since the
RSDs in this study seem to vary even among samples of about the same tonnage,
it appears that we are not comparing similar values of RSD per unit weight, and
therefore that tonnage is not the only variable. Second, the range of tonnages in
the above list is relatively small - from about 1,000 to 3,400 tons. (A unit train
typically carries about 10,000 tons, approximately the daily consumption at
Sammis; Sammis units 5-7 currently burn an average of about 6,700 tons per day.)
In Table 8, we present computed values of the required average SOj emissions
from coal to be burned at the Sammis units, given Sammis's maximum allowable
SC>2 emissions. We have used alternative assumptions regarding the value of
the RSD, the confidence level, and the type of probability distribution of sulfur
content. Two values of RSD are compared - 0.15, frequently assumed for raw
coal fed daily to large power plants, and 0.08, the RSD for the corresponding
cleaned coal, according to the best fit computed for nine of the Versar data sets
(see Appendix). Two confidence levels are used: 99.87 percent, corresponding to
three standard deviations above the mean; and 95 percent, corresponding to
1.645 standard deviations above the mean. Finally, two distributions are
considered: a normal distribution and a lognormal distribution.
To the extent that Sammis can bum coal with a higher average sulfur content,
its coal choices will include more Ohio coals and more lower-priced coals. How
much of a difference do the alternative sets of assumptions in Table 8 make for
the allowable mean value of 502 ern'ssions ('n 'b SO2 per million Btu)? From
Table 8 we see that, for the limit of 1.61 Ib and a normal probability distribution,
the mean SO2 emissions at the 99.87 percent confidence level with an RSD of
0.15 must be I.IOIb; with the lower RSD of 0.08, the mean can be
higher - 1.29 Ib (or 17 percent higher). At the 95 percent confidence level, the
highest allowable mean S02 emission level corresponding to the RSD of 0.15 is
1.28 Ib; and here, again, the allowable mean SO2 value is higher with the lower
RSD of 0.08 - 1.42 Ib (or 11 percent higher). Looking at the 4.46 Ib limit, the
41
-------
TobleS
Expected Average SO? Emissions for Sammis Units under
Different Assumptions of Sulfur Variability
Maximum Allowable
Emission
(IbSCyiO^Btu)
1.61
1.61
1.61
1.61
4.46
4.46
4.46
4.46
RSD
.15
.08
.15
.08
.15
.08
.15
.08
Confidenceb
Level
99.87
99.87
95.0
95.0
99.87
99.87
95.0
95.0
Mean Emission
(IbSCWlO^Btu)
Assuming Normal
Distribution
1.10
1.29
1.28
1.42
3.07
3.59
3.58
3.94
Mean Emission,
Assuming Lognocmal
Distribution0
1.02
1.26
1.25
1.38
2.84
3.49
3.42
3.89
For a given maximum SO, emission level, the allowed mean value for a sample of compliance coal will be
determined by the RSD (relative standard deviation = standard deviation/mean), the probability distribution of
the sampled value, and the required confidence level. The RSOs of .15 and .08 have been assumed to apply to
raw and cleaned cools in quantities required daily by a large power plant. No exceedances per month are
assumed beyond those implied by the confidence level.
A 95% confidence level implies that emissions will exceed the emission limit 5% of the time, or one day in
twenty. A 99.87% confidence level implies that emissions will exceed the emission limit 0.13% of the time or
less than one day per /ear. *
Assuming a normal probability distribution, the mean, m, is found by
max ,
m* I * z RSD
where:
z = 3 standard deviations above the mean for a 99.87% confidence level, and
2 = 1.645 standard deviations above the mean for a 95% confidence level; and
max » maximum allowable emission (1.61 or 4.46 Ib S02/I0 Btu for Sammis's generating costs).
Assuming a lognormal distribution, the mean, m, is found by:
m'
m s e , where
m'* In max - z RSD, or
m'3max e'z'RSD
(see note c for symbols)
-------
expected average S02 emissions are 17 percent higher at the higher confidence
level (3.58 !b instead of 3.07 Ib) and 11 percent higher at the lower confidence level
(3.89 Ib instead of 3.49 Ib).
Table 8 also illustrates that, for an assumed value of RSD, the mean sulfur level
to achieve compliance will be lower with a lognormal distribution than a normal
distribution of measured sulfur values.
The reduced RSD (0.08) used here for cleaned coals (and the corresponding
increases in the average coal-inlet sulfur content) was determined by Versar's
best fit of nine coal-data sets representing Levels 2 and 3 of coal cleaning. We
point out that (I) the data used were very limited, and (2) Level 4 (more
intensive) cleaning may yield a somewhat greater reduction in RSD.
Table 8 is useful because it illustrates a method for calculating the maximum
acceplable average sulfur content of the coal burned at a power plant. We
emphasize that the actual values used in the calculation must be determined
empirically for each individual case. If, for example, measurements indicated
that the RSD of the 6,700 tons per day of raw coal delivered to Sammis units 5-7
was 0.20 (rather than O.I 5), then - for a confidence level of 99.87 percent and a
normal probability distribution the highest allowable mean SQ~ emissions from
this coal would be 2.8 Ib per million Btu (rather than 3.1 Ib corresponding to an
RSD of O.I 5).
Implied in the computations of RSD for the cases listed in Table 8 is the
assumption that no exceedances of the 24-hour S02 standard will be permitted,
within a given confidence level. If, in fact, one or more exceedances per month
will be permitted, a higher (more easily attainable) mean S02 value will be
acceptable. Since it is expected that Sammis will be permitted to exceed its
SOo standard up to two times per month, we have also computed the average
sulfur level taking this leeway into account. For the case in which the RSD
equals 0.15, the SO2 standard equals 4.46 Ib SG>2 per million Btu, the confidence
level is 99.87 percent, and a normal probability distribution is assumed, allowing
up to two exceedances per month implies an allowable mean SO2 level of 3.34 Ib
43
-------
S0? per million Btu, which is higher than the mean SO2 level of 3.07 Ib for the
same case when the 24- hour standard can never be exceeded (see Table 8). When
three exceedances per month are allowed, the mean SO2 level for the case
described here is 3.36 Ib SC>2 per million Btu not very different from the mean
of 3.34 Ib found with two exceedances per month (see Appendix for computa-
tional details).
To summarize, for a confidence level of 99.87 percent and up to two allowed
exceedances per month, the allowable mean values for Sammis are as follows:
Allowable Mean Values (Ib SCyiO Btu) for 99.87 Confidence Level
and Two Exceedances per Month
Assumed
RSD
.15
.08
Limit = 1.61
Normal
Distribution
1.20
1.37
lbS02/!06Btu
Lognormal
Distribution
1.15
1.35
Limit = 4.46
Normal
Distribution
3.34
3.78
lbS02/!06Btu
Lognormal
Distribution
3.19
3.73
3.2 One Ohio Plant's Proposal for Using PCC as on S02 Compliance Strategy
One plant in Ohio proposes to meet its new SO2 standard by burning cleaned Ohio
coals from currently used sources in four of seven units, with the cleaning to be
done in new coal-cleaning facilities. The plant - the Conesville plant near
Conesville in Coshocton County (see Figure 4) - is owned by the Columbus end
Southern Ohio Electric Company (C&SOE).
According to Ohio's State Implementation Plan, Conesville units 1-4, which
represent about 70 percent of the plant's total six-unit capacity of !970MW(e),
must meet a limit of 5.66 Ib SO2 per million Btu; and units 5 and 6 (which have
FGD systems) must meet a 1.2 Ib standard. (Compare with Sammis: 4.46 Ib
44
-------
SOo per million Btu for units 5-7, and 1.61 Ib for units 1-4). Conesville's average
S02 emissions in 1977 were (in Ib per million Btu): 6.95 from units 1-3, 7.32
from unit 4, and 1.10 from unit 5. As of this writing, EPA and C&SOE have
agreed that Conesville will comply with the SO-, emission limit for units 1-4 by
41
burning washed Ohio coals.
At the same time, C&SOE is challenging the 5.66 Ib limit on the grounds that it
is unnecessarily stringent for plantwide compliance with the National Ambient
Air Quality Standards. Whether or not this limit is relaxed, however, the
utility prefers the cleaning of nearby Ohio coals as its $©2 compliance strategy.
At present, all the coal burned at Conesville about 3.73 million tons in
1977 originates within about 15 miles from the plant, coming from five seams
and 17 mines. About 25 percent comes from plant-site mining areas. Deliveries
are by truck or conveyor belt. Since there are no rail or barge facilities, it is
infeasible to use distant low-sulfur coal (from, say, eastern Kentucky or southern
West Virginia) and especially desirable to continue using nearby Ohio coals.
Twenty percent of the coals currently used at Conesville come from nearby parts
of the Lower Kittanning (#5) and Pittsburgh (#8) seams. Since these coals are too
high in sulfur, even after washing, the plant would discontinue using them/*2
Eighty percent of the current supply comes from nearby parts of the Middle
Kittanning (//6), Meigs Creek (//9), and Waynesburg (011) seams; C&SOE would
clean these coals. On the basis of washability tests it has conducted, C&SOE
states that the cleaned coals can meet the 5.66 Ib SOo standard for units 1-4.
C&SOE is also considering the burning of these cleaned coals in the two units
with FGDs in order to reduce limestone demand and the generation of scrubber
sludge. Further, C&SOE is attracted by the possibility of improving plant
performance through the use of PCC.
C&SOE is considering coarse crushing (down to 2 x 0 inches) and a 1.6 specific
gravity medium - a relatively low cleaning level often referred to as "Level 2."
The company estimated tentatively that the resulting S02 level of the washed
product will average 5.4 Ib. We observe that, if we use an RSD of 0.08 (the lower
value in Table 8) for the 7,500 tons per day burned in units 1-4, the average
45
-------
502 emissions corresponding to the allowable maximum emissions of 5.66 Ib are
5.0 Ib for a confidence level of 95 percent, and 4.6 Ib for a confidence level of
99.87 percent. Therefore, the average post-PCC value of 5.4 Ib estimated for
the Level 2 cleaning may not allow an adequate design margin to account for
sulfur variability. A more intensive level of coal cleaning, or a somewhat
greater selectivity of raw coals may, however, bring the product into line.
43
The Bureau of Mines washability data include two samples from the Middle
Kittanning seam in Coshocton County. Middle Kittanning is currently
Coshocton's most productive seam, having produced 1.7 million tons in 1977, or
about 45 percent of the county's coaJ output (see Table 2). While it is impossible
to determine how representative the Bureau of Mines samples are of the current
and future coal production from Coshocton's Middle Kittanning seam (the point
values cannot indicate the inevitable variations within seams), it is interesting
nevertheless to compare the PCC results for these samples with the results
expected by Conesville. The two samples potentially emit 10.3 and 6.7 Ib SO
per million Btu before cleaning. Results of washability tests of these samples
are summarized in Table 9 for two of the Bureau of Mines levels of washing. The
first level "Level 2," which involves crushing to Ifc inch top size and a float-
sink medium of specific gravity equal to 1.6 - corresponds to the level of PCC
that C&SOE is considering. The second (more intensive) level, which we refer to
as "Level 4," involves crushing to 3/8 inch top size and a specific gravity of 1.4.
When the relevant statistical factors require that the mean emissions not exceed
4.9 Ib S02 per million Btu in order to meet a limit of 5.66 Ib, Level 4 cleaning
results in enough sulfur reduction from both the coal samples represented in
Table 9. When the mean $©2 emissions cannot exceed 4.5 Ib $©2 per million Btu,
Level 2 cleaning allows compliance with the SO2 standard only for the lower-
sulfur coal sample; the higher-sulfur coal would require the more intensive PCC.
In addition, several other observations can be made about the results in Table 9:
The percentage of coal ash drops from 13.5 to 4.7 and 3.2
when Levels 2 and 4 are applied to the first coal, and
from 10.2 to 4.8 and 3.4 when they are applied to the
second coal.
46
-------
Toble9
Summary of Results off Bureau of Mines Washability Tests on Two Samples from the
Middle Kittanning Seam in Coshocton County, Ohio
Raw Coal After Level 2 Cleaning After Level 4 Cleaning
Sulfur % Sulfur % Sulfur %
Btu IbSO,/ Btu IbJ
Sample Btu/lb Ash % 10° Btu Total Pyritic Loss Ash % Btu/lb 10° Btu Total Pyrltic Loss Ash% Btu/lb 10° Btu Total Pyritic
I 12,300 13.5 10.3 6.4 4.5 8% 4.7 13,590 4.9 3.4 1.5 16% 3.2 13,810 4.0 2.8 1.0
2 12,500 10.2 6.7 4.2 2.3 4% 4.8 13,300 4.2 2.8 I.I 9% 3.4 13,490 3.6 2.5 0.7
Sourcet Joseph A. Covallero et al., Sulfur Reduction Potential of the Cools of the United States. Bureau of Mines Rl 8118 (Pittsburgh, Pas U.S. Department of
Interior, Bureau of Mines, I975T
Note; Level 2 here designates crushing to Ifc Inch top size and a float-sink medium with specific gravity equal to 1.6. Level 4 designates crushing to 3/8 inch
top size and a specific gravity of 1.4.
-------
The Btu losses resulting from PCC are significantly higher
for the more intensive level of cleaning: 16 percent and
9 percent, as compared with 8 percent and 2 percent for
the lower level of PCC. (Monetary costs are necessarily
attached to the energy losses.)
The laboratory procedures used in the Bureau of Mines
washability tests commonly employ heavy organic liquids
to obtain desired specific gravities of separation. Heavy
organic liquids promote a greater degree of sulfur
removal (for the same specific gravity) than does water
(made denser with materials such as magnetite), which is
the basic separating medium normally used in commercial
PCC operations.
The heating values and sulfur and ash contents listed in
Table 9 are presented on a moisture-free basis, so that
they are higher than those for coal on an as-received
basis. For Ib SC^ per million Btu, however, the mostire-
free and as-received values are not very different.
To put Conesville's proposal - "clean nearby Ohio coals" - in perspective, we
list in Table 10 selected characteristics of analytical samples from the Middle
Kittanning seam in Coshocton County, as reported by the Ohio Geological
Society. Measured heating values and percentages of ash and sulfur (total
pyritic, and organic) are listed in columns 1-5. Computed values of potential
emissions of total sulfur and organic sulfur as Ib $©2 per million Btu are shown in
columns 7 and 8. Column 9 lists the organic-sulfur SO2 emissions of column 8
reduced by an assumed value of 10 percent, to reflect an estimated upgrading of
10 percent in the heating value following PCC. Since essentially no organic
sulfur is removed by PCC, emissions of only the organic-sulfur component
represent the theoretically lowest S©2 emissions from a cleaned coal, assuming
no sulfur retention in the ash. While in practice this theoretical limit will not be
achieved, we list it as a guide for understanding Conesville's planning. Certainly
if these theoretically best values were significantly higher than an allowable
mean emission (about 4.5 to 4.9 Ib 502 per m'"'on Btu f01* a clean-coal RSD of
0.08), and if the measured samples were fairly representative of available coal
from the Middle Kittanning seam in Coshocton, the proposed coal-cleaning SO
compliance strategy would not appear worth considering. Since, however, most
of the S02 emissions listed in the last column are below the allowable mean limit
for units 5-7, PCC does appear to merit consideration.
-------
Table 10
Ash and Sulfur Contents of Coal Samples from Middle Kittaming Seam,
Coshocton County, Ohio
(1)
Heating Value
(Btu/lb)
12880
9980
12220
10510
12880
12330
12580
13350
10030
12300
12860
12730
10270
12230
11950
12800
13130
9070
11980
(2)
%Ash
5.1
5.0
4.9
5.2
2.7
3.3
4.4
4.5
7.1
4.8
4.0
3.2
2.9
3.6
5.7
5.5
5.4
5.9
5.6
Cn.iree: G. Botoman, and B.
(3)
Total
3.5
6.7
4.2
8.9
4.3
5.4
3.5
2.3
6.5
3.6
4.0
3.6
4.5
4.0
5.3
3.9
3.7
6.7
4.5
Smith.
(4)
% Sulfur
Pyritic
1.22
3.2
1.67
5.19
2.11
2.86
1.3
0.97
3.84
. 1.59
1.67
1.25
2.17
1.68
2.96
1.76
1.16
4.06
2.15
Analyses of
(5)
Organic
2.0
3.25
2.28
2.72
2.16
2.24
1.67
1.27
1.79
1.57
1.77
1.84
1.98
1.83
2.05
2.13
2.47
2.14
2.22
Ohio Coals,
(6)
IbAsh/
!06Btu
4.0
5.0
4.0
4.9
2.1
2.7
3.5
3.4
7.1
3.9
3.1
2.5
2.8
2.9
4.8
4.3
4.1
6.5
4.7
1C No. 47
(7)
Total S
before PCC
5.4
13.4
8.4
16.9
6.7
8.8
5.6
3.4
13.0
5.9
6.2
5.7
8.8
6.5
8.9
6.1
5.6
14.8
7.5
(8)
lbS02/IO*Btu
Organic S
before PCCa
3.1
6.5
4.5
5.2
3.4
3.6
2.6
1.9
3.6
2.6
2.8
2.9
3.8
2.9
3.5
3.4
3.8
4.7
3.7
(Columbus, Ohio: Ohio Geological Survey,
(9)
Organic Sn
after PCC°
2.8
5.9
4.1
4.7
3.1
3.3
2.4
1.7
3.3
2.7
2.5
2.6
3.4
2.6
3.2
3.1
3.4
4.3
3.4
1978).
Notes Samples were taken in 1976. Values listed are for as-received coals.
a values for Ib SOVI06 Btu are based on the assumption that all sulfur is emitted as SO,. Values for organic sulfur after
ohysical coal cleaning ore based on the assumption that the heating value'of the cleanM coal is 10 percent higher than
that of the raw coal; these values represent a lower bound on S02 emissions following PCC (i.e., all pyritic sulfur removed
by PCC).
-------
The question at this point is "How will the 'theoretically minimum' SO2 emissions
depicted in the last column of Table 10 correspond, in fact, to actual SO,
emissions from the Coshocton County/Middle Kittanning coals?" The answer, of
course, will depend on the particular coal and PCC process. For the data
representing the two samples in Table 9, the "theoretically minimum" values
would be multiplied by factors of 1.4 and 1.3 for Level 4 PCC. These data, then,
suggest that a certain amount of blending of cleaned coal with low-sulfur coal
may be required.
To some extent, excess emissions from units 1-4 could be offset by decreases In
emissions from the two units with FGD systems, which could result if those units
burned some cleaned coal.
While more data and further analysis are needed to determine whether PCC can
be the exclusive S02 control strategy for Conesville's units 1-4, Table 9 does
show that without PCC the use of the nearby Middle Kittanning coals would be
out of the question. Certainly PCC would significantly increase the potential for
using these coals. C&SOE is attracted further by other consequences of PCC:
the removal of incombustible material, the expected decrease in gas flow during
combustion, and the possibility of raising the coal's ash fusion temperature. The
technical aspects of PCC do, therefore, seem attractive. But there remain a
number of insistent and important institutional questions, which relate to the
fact that, although SO2 compliance is required in Ohio by October 1979, the PCC
facilities needed by Conesville do not exist:
Under what institutional arrangement will the needed
PCC facilities be built?
When will they be operational?
What strategy for SO, compliance will Conesville follow
until it can use PCC? Will Conesville be allowed a
variance in SO? emissions during construction of a PCC
facility? And if a variance is not granted, what will
happen vis-a-vis the disuse of local mines and, possibly,
the construction of needed transportation facilities
(assuming no FGD) during development of the PCC
facilities?
50
-------
Accounting for the need to acquire relevant information, to carry out feasibility
and design studies, and to obtain approval from EPA and other regulatory
agencies, Conesville estimates that adequate PCC facilities could not realisti-
cally be expected to come on line before about the spring of 1982. Meanwhile
C&SOE plans to support studies to determine the optimum ownership, construc-
tion, and operational arrangements for PCC. Conesville's preliminary cost
estimates suggest that, if Level 2 PCC is implemented and does meet the
standards, consumers will see about a 2 percent (uninflated) increase in their
o
cost of electricity (reflecting about a 7 percent increase in fuel costs). An
analysis of the net costs will necessarily address the following questions:
What will the levelized capital and operating costs of the
PCC facilities be, and what will the associated incremen-
tal fuel costs be?
What power-plant benefits or problems other than those
related to 502 comPl'ance w'" result from PCC?
3.3 The Washability of Ohio Coals
The Bureau of Mines (BOM) has compiled a computerized data file describing
results of a washability study of 587 U.S. coal samples, 455 of which are
described in the BOM study report. Two examples of the BOM findings were
discussed in the preceding section; here, in Figure 7, we present a sample page of
results.
What do these washability data tell us about the physical cleaning of Ohio coals
as an S02 compliance strategy for Sammis? For units 1-4 (subject to a limit of
1.61 Ib S02 per million Btu), the data indicate clearly that PCC will not result in
S07 compliance, but for units 5-7 (4.46 Ib S02 per million Btu), PCC does seem
promising. In Tables 11 and 12 we list 19 of the 57 Ohio samples described in the
BOM study. These are the samples for which a fairly intensive level of cleaning
Zift
("Level 4") resulted in a product coal with $©2 emissions not exceeding 3.1 Ib
S07 per million Btu, which, as we pointed out in Section 3.1, is comfortably
below the 4.46 Ib S02 standard. Both tables show the county and coal bed and
51
-------
Figure 7
A Page of Washability Data from the BOM Rl 8118
sr»rti o«lu
cauMfrt caiu^*! ANA
CUMJCA M v« «*
tWQOUCT
FiOAr-i.30
'LOAr-i..j
'LOAT-4. .4*1
FLOAT. i. to
torn.
l»A ST»*OA-0
Mooucr
ri.UAr-l.JO
FLOAT- i .<«)
'LOAt-i.a*
flOAT-l .JO
roTAi
(»4 JtAKOAJO
MOOUCT
rLOAr-i.jo
'LuA r- 1 ..o
'LQA r-i .ao
r LUA7-1 .43
TOT 41.
»£COv£»r.4
UtlOHT <(j
7-..S 40.7
96.3 42. 7
>0.. 46.2
»2.2 »7.3
100.0 100.1
4*.a 47.6
»ecov«».»
cisxr aru
44.6 76.2
43. > 40.7
»..! *».»
41.8 97.3
100.0 100.9
SA-m.t
9TU/L*
14217
14019
13*77
137*4
13133
14011
SAOCIE
a ru/L9
14}]4
14134
14044
13441
13242
1340*
5AM»t.£
STUVL*
14341
14217
1.044
13937
13IS3
c»ui>«eo ro
AWt
3.8
S.2
6.1
».7
11.0
4.9
CMUSMCO TO
4 In.
3.0
4.0
4.7
1.4
10.4
7.9
caushco ro
ASM!
2.6
3.4
A. 7
1.7
11.0
coAcacot Mtooit KITTAJ
94> CO At. nOliru^t: 1
SX441LITT QATA
»«SS 1-1/2 [NC
» SUC/UH
»«|tlC
.21
.41
.5.
.»*
1.72
.44
>tSS 3/4 [NCn
4 SULFUR
TdlTIC
.07
.11
.16
.21
l.al
.44
>«SS 14 ^tsH
« su.ruH
»T«IITIC
.96
.J9
.13
.17
l. a;
r*S
1 k
TOTAC
.6*
.4*
1.03
1.16
2.3S
.a*
.4
TOTM.
.51
.57
.42
.4*
2.34>
.az
.4
rotAc
.44
.52
.Si
.39
2.42
Li S02/M
.0
. J
.5
.7
.6
1.20
Li S02/«
.7
.1
.9
.4
3.S
1.2*
L* 50J/»
.7
.7
.4
.a
3.7
10.
COAL
CUMUL-AIIVC ASft4*!LirT
SA>m.i cxusneo ru PASS 1-1/2 incuts
ltU/L8 ASH.4,
1.20
S02/H
vnon.
__
ft n* T* i »A
f t.3* ' * I *
f LQA f | 40
TOTll.
t»4 sr4NOAoe
ooucr
FLOAr-l.30
r AT*! "4*
f i nA r i 40
FLQA 1 1 . Tw
TOH1.
C>4 ST4«OA4)O
MOOUCT
ft ». » 1A
FI.OA I - 1 . ju
FLOA T 1 .40
FLOA r-i . *0
T0t4(.
tiOr
64.^
t.4
»..D
97.1
100.0
»tco
Mi&tir
74. 0
40. j
4».l
ti.S
190. J
4.7
aru
71..
94.J
16.3
9«.g
UK. a
»e»T««
aru
74.4
9».0
97.1
9a.i
ioa.il
aa.9
»ccove»T.»
CIOHT aru
42.7
44j.o
42. J
">...
100. J
««.4
14. S
46. S
47.)
100.0
139*7
13771
13667
13613
11447
5A*1C
»TU/LS
14247
1344,7
13634
13771
13447
1.044
5MX.C
STU/La
14270
i4oai
13446
U*7D
J3377
5.2
6.]
7.1
7. a
4.0
CKU5MCO ra
A3H.4
3 3
1 2
4 1
4 5
9 0
4.S
CHUSHtO TO
AIM.*
3.1
4.4
1.0
]. a
9.2
Prill TIC
.S*
.77
.49
.97
1.44
V4SS 3/4 {MCH
5m.ru*
PT01T1C
.13
.3»
.SO
.57
1.31
.30
»4SS 14 »CSM
sw.ru«i
ruiTie
.17
.22
.27
.32
!.-«
TOTAt.
.90
1.1]
1.2*
1.32
1.71
.4
roTA«.
.49
.91
1.04.
1.11
i.as
.J*
>
TOUH.
.53
.ST
.62
.67
i.rs
1.3
I. a
1.4
1.4
<.*
1 .20
La S02/H ]ru
1.2«
La SI»SM 9ru
.7
,i
.9
1.9
1.6
9*.2
1.20
Source; Joseph A. Covallaro et al., Sulfur Reduction Potential of the CnoU f
thA lJnit*»H States. Rl 81 18 (Pittsburah. Pa li..^. (*)£»,.».,* . I.
the United States. Rl 8118 (Pittsburgh, Pa.:
Interior, Bureau of Mines, 1976).
partment
52
-------
Table 11
Washability Data for Selected Ohio Coals
(Sulfur Content)
% Sulfur (S) in
Row Coal
County
Harrison
Be/mont
Harrison
Betmont
Gallic
Jefferson
J*Herson
Harrison
Harrison
Motioning
Coiumbiana
/» _i, tfnKjma
^QIUTTWIV nf
Coiumbiana
Muskingum
Perry
Perry
Parry
Vinton
Tuscorawas
<^urcet Joseph
Coal Bed
Sewickley
Sewickley
Sewickley
Waynesburg
Pittsburgh
Pittsburgh
Mohan ing
Lower
Freeport
Lower
Freeport
Brookville
Middle
Kittanning
Middle
Kirt arming
Middle
Kittanning
Middle
Kittanning
Middle
Kiftanning
Middle
Kittonning
Middle
Kirt arming
Middle
Kitt arming
Lower
Kltt arming
A. Cavallaro et
Pyritic S
1.3
1.63
1.35
2.06
2,25
2.03
0.89
1.37
1.64
1.65
1.71
1.72
1.40
1.20
3.32
0.10
0.36
0.41
1.72
aU Sulfur
Totals
1.94
3.03
2.22
2.85
3.26
2.98
1.48
2.36
2.45
2.60
2.51
2.35
1.75
2.99
4.49
0.65
1.02
0.99
2.51
Reduction Potential
%Uni«tt_» ;n
flnoiaiure in
Raw Coal
2.1
2.7
2.1
2.9
6.1
1.4
2.3
2.3
1.9
3.2
2.3
1.5
3.9
2.4
5.3
5.5
6.4
7.2
2.1
of the Coals of
LbSOj/IO"
Cleaned
1.8
2.7
2.2
2.5
3.1
2.9
1.0
1.9
1.5
1.5
1.4
0.8
1.3
3.0
2.9
t.O
1.4
1.3
I.I
the United States.
Raw
3.0
4.8
3.4
4.8
5.1
4.5
2.2
3.5
3.7
3.8
3.7
3.6
2.6
1.2
7.7
I.I
1.7
1.5
3.7
RI8II8
%Btu Recovery
90
79
90
72
92
93
92
90
91
90
91
93
94
95
77
88
69
97
92
(Pittsburgh, Pa.: U.S,
Notes' Selected washability indices for all data shown here: specific gravity of 1.4, crushing to 3/8 inch.
This list includes those of the 57 Ohio samples in Rl 81 18 that produced no more than 3.1 Ib SO, per million Btu after
cleaning (assuming no sulfur retention in the boiler). *
Values of sulfur content and heating value are given on g moisture-free basis. Values of Ib SO,/10*
both a moist and moisture-free basis.
> are comparable
53
-------
Table 12
Washability Data for Selected Ohio Coals
(Ash Content)
Heating Value
County
Harrison
Belmont
Harrison
Belmont
Gallic
Jefferson
Jefferson
Harrison
Harrison
Mohan ing
Columbiana
Columbiana
Columbiana
Muskingum
Perry
Perrv
I vi 7
Perry
Vinton
Tuscarawas
Source: Joseph
Coal Bed
Sewickley
Sewickley
Sewickley
Waynesburg
Pittsburgh
Pittsburgh
Mahoning
Lower Freeport
Lower Freeport
Brookville
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Middle
Kittanning
Lower
Kittanning
A. Covallaro et al.. Sulfur
Raw
10.6
13.1
10.9
17.4
8.7
9.8
9.3
9.5
10.4
7.6
9.3
10.4
9.0
6.6
17.7
17.9
16.1
3.5
7.2
Reduction
%Ash
Cleaned
8.0
7.9
8.5
9.6
5.3
5.9
3.9
4.1
4.1
3.2
3.9
4.0
5.2
3.9
5.1
5.6
6.3
2.7
3.5
Potential of the
(Btu/lb)
Raw
13023
12622
13002
11963
12829
13346
13539
13380
13234
13644
13497
13242
13407
13208
11583
11598
11827
13652
13515
Coals of the United
Cleaned
13410
13377
13336
13091
13303
13916
14345
14179
14164
14294
14300
14188
13967
13590
13367
13345
13195
13870
14063
States, Rl 8 118 (Pi
% Btu Recovery
90
79
90
72
92
93
92
90
91
90
91
93
94
95
77
88
69
97
92
ttsburgh. Po^ 1 1 c
Deportment of Interior, Bureau of Mines, 1 976)
Notes; Selected washability indices for all data shown here: specific gravity of 1.4, crushing to 3/8 inch.
This list includes those of the 57 Ohio samples in Rl 81 18 that produced no more than 3.1 Ib SO, per million Btu «**
cleaning (assuming no sulfur retention in the boiler). £ aft*r
Values of sulfur content and heating value are given on a moisture-free basis. Values of lbSO,/|Q^ Btu
comparable on both a moist and moisture-free basis. 2
are
54
-------
the Btu recovery of each sample. Table 11 also shows the raw coal's moisture
content, the raw coal's sulfur content (total and pyritic), and the SC^ emissions
from the raw and cleaned coal (assuming all coal sulfur is emitted as SCX from
the boiler stack). Table 12 lists the ash content and heating values of the
samples both before and after PCC.
The Btu recovery from 14 of the 19 samples equalled or exceeded 90 percent.
The Btu recovery from the other five samples ranged from 69 to 88 percent.
Again, these data are important, since fuel loss can account for a relatively
significant cost factor in PCC.
The major coal beds listed in Tables 11 and 12 are (in order of current levels of
production) Pittsburgh, Middle Kittanning, Sewickley (or Meigs Creek), and
Waynesburg. We note that within each bed - even within the same county or
mine _ coal and washability characteristics can vary significantly. On the
averoge, the Pittsburgh seam contains the highest-sulfur coal (estimated at
£iQ
5.8 percent on a moisture-free basis). For this reason, the cleaning of supplies
from the Pittsburgh seam cannot be expected to produce significant quantities of
SO2 compliance coal for Sammis. The Sewickley (Meigs Creek) seam also has a
high sulfur content (estimated on the average as 5.5 percent).50 Furthermore,
the ash content is relatively high (about 12 to 20 percent). A number of the
Meigs Creek coals are washed (for example, in the large Georgetown Preparation
Plant); but, as is true for almost all PCC plants today, the facilities are designed
for ash removal, not sulfur removal. While the very high sulfur content may
make it impossible for Meigs Creek coals to be cleaned to SO2 compliance levels
for use by themselves, PCC will be able to reduce the sulfur content of these
coals sufficiently to increase their use in compliance-coal blends.
The Bureau of Mines washability data are "point data" from producing coal beds.
BOM has not attached values of associated coal reserves or coal production to
these data. In order to estimate the quantity of coal reserves represented by the
washability samples, EPA's Office of Research and Development has developed a
model - the Reserve Processing Assessment Model (RPAM) - to produce over-
lays of BOM reserves data and analytical data and to match the overlays with
the BOM washability data. The objective of the model - which is still in the
55
-------
process of being tested is to estimate the quantity of coal reserves (in terms
of both weight and energy content) capable of meeting alternative S0? emission
standards. The estimates are made for raw coal and also for coal that has been
cleaned with alternative levels of PCC. We have used this model to compute
cumulative percentages (by weight) of Ohio reserves that would meet a range of
conceivable SOj standards. The results are shown in Figure 8.
According to Figure 8, about 44 percent of the total Ohio coal reserves (raw)
would produce average emissions of less than 4.46 Ib 502 per million Btu, and
about 20 percent of the raw Ohio reserves would fall under a 3.1 Ib limit (which
would allow a comfortable margin below Sammis's 4.46 Ib limit for sulfur
variability). The separate curves in Figure 8 show the reserves' increased
availability vis-a-vis these $©2 standards when alternative levels of PCC are
applied. The curve coded by a "2," for example, indicates that an intensive level
of cleaning (3/8 inch and specific gravity of 1.3) produces a fairly dramatic
increase in the reserves capable of meeting the 3.1 Ib $©2 standard- up to
about 58 percent from the corresponding raw-coal availability of about
20 percent.
While the RPAM results depicted in Figure 8 are useful in relating SOj standards
to coal availability for various levels of PCC, a caveat is in order regarding their
use: their accuracy is not known. Although the data used are the best that are
publicly available, they contain inherent errors that are not easily quantifiable.
Furthermore, there are errors in the matching, or overlay, processes used in
(I) partitioning reserves among the analytical point data and (2) partitioning the
overlaid reserves and analytical information among the washability point data.
The results of the RPAM matching process, where reserves are distributed
uniformly among analytical samples within the union of a county and coal bed
are, however, identical to those of a BOM matching process using the logis*ic
function. When either process is applied to the same data, 61 percent of the
Ohio reserves are found to contain more than 3 percent sulfur. Both matching
procedures make the tenuous (but unavoidable) assumption that the distribution
of the data is reasonably representative of the coal reserves.
56
-------
Figured
Available Ohio Coal Reserves for Alternative
SO2 Standards and Levels of PCC
inn 7 * H H M « 6 6 till I
w T H * 6 I I 1 2 2 N N M
7146 |N22»4N3NS
74 1N244.431* 0
90
30
70
60
«- 50
O
I.
5
30
10
Total tonnage: 21 .1 x I09 tons , J * * , ., " T 1 " ' * i ' ! ' ' a
76 I \ < 333)00
6 II** 3 00
H 2 1 4 3390
7 Z I * 330
6 2 * 3 5 0
21 )00
7 2 I * 35
624 0
* ISO
7 21 30
624 3
2 H 330
X 10
4 3
r 21 so
624 3
1 3 0
T » 3
621 SO
24 3
41 3*0
H 2
41 ISO
7 2 0
6 41 15
72 5
41 30
6 2 S
7 41 30
624 90
72 1 3
41 30
N 4 35
H I 330
41 33
2 «, 3)00
7 H 1 30
24 1 330
761 30
24 33
41 3«0
2 7 H M H N
2ND HMO
1.8 3.6 5.4 7.2 9.0
Emission Standard (!b SO2/106 Btu)
Source; EPA's Reserves Processing Assessment Model (RPAM). See text.
Note; The codes represent raw coal and alternative levels of PCC;
0; raw coal
1: 1 fe inch at 1.6 specific gravity (s.g.)
2: 3/8 inch at 1.3 s.q.
3: 1.6 s.g. on sink of 3/8 inch, 1.3 s.g.
4r "Homer City clean" (stringent level of PCC to meet New Source
Performance Standard limitations see discussion at end of
Section 3.3)
57
-------
It is important also to note the discrepancy between the BOM data and recent
deliveries of Ohio coal. These deliveries indicate that a higher percentage of
high-sulfur Ohio coals are being mined than would be expected from the BOM
data if those data are, in fact, representative of the state's reserves. In contrast
to the value of 61 percent derived from the BOM data (see preceding paragraph),
about 80 percent of the Ohio coal delivered in 1977 contained more than three
percent sulfur (see Figure 5). Whether or not these recent deliveries are
themselves representative of Ohio reserves is not known.
The PCC performance data discussed in this section came mainly from labora-
tory tests. What information has been reported on sulfur removed by operational
PCC plants? Unfortunately, very little. Some, however, was presented in the
Versar study. Here we present the Versar data on a plant using coal from the
Middle Kittanning seam in Ohio. The level of cleaning represented is one in
which the coal is crushed to 3/8 inch; material greater than 3/8 inch is processed
in a jig or dense-medium vessel, and material smaller than 3/8 inch is not
processed. Among the six tabulated samples, the reduction of $©2 emissions
ranges from about 25 to 40 percent:
Test Number Ib S0?/106 Btu
(Ohio PCC Plant, = % Reduction by PCC of
Middle Kittanning Coal) Raw Coal Product Coal lbSO2/IO Btu
1 7.4
2 6.4
3 7.1
4 8.3
5 7.3
6 6.3
4.8
4.6
4.9
4.9
4.9
4.8
35.1
28.1
31.0
41.0
32.9
23.8
A new PCC process being developed for commercial use, the Otisca process, is a
closed-cycle, heavy-media system that uses an organic medium (as does the BOM
in its washability tests). American Electric Power plans to build a 125-ton-per-
hour demonstration plant for the Otisca process in Beverly, Ohio, using Ohio coal
with a high sulfur content (9-10 Ib S02 per million Btu) to produce a product coal
with emissions of just under 7 Ib SO0 per million Btu. It is reported that as much
CO
as 90 percent of the ash may be removed.
58
-------
Although we have been discussing PCC in terms of sulfur removal, we do not
underestimate its value as a technology for ash removal, traditionally its most
important objective and one that is especially critical for Sammis in light of the
plant^ past and present noncompliance with Ohio's emission limitation of O.I Ib
particulates per million Btu. Depending on the level of cleaning, PCC can
remove from about 15 to 75 percent of the ash content (or even 90 percent, if
the claims for the Otisca prove correct). Lower particulate emissions are not
the only benefit of ash removal. The process also results in a product of higher
heating value and less variability. Moreover, ash removal reduces boiler
unavailability caused by the fouling or slagging problems associated with
constituents of the ash.
To date, the largest plant designed and constructed largely for the removal of
sulfur from steam coals is the multistream system at the Homer City Generating
Station power complex in Homer City, Pennsylvania. This plant is scheduled to
process 5.2 million tons of coal per year. Although it is still in the "shakedown"
stage and uses Pennsylvania rather than Ohio coal, we present the design
performance parameters as an example of the capabilities of an advanced sulfur-
removal PCC system (see Table 13). The Homer City plant will produce two
streams - one for a higher percentage of sulfur removal to serve Unit 3, which
is regulated by the New Source Performance Standards (see "Homer City clean"
on Figure 8), and one for units I and 2, which are regulated by State Implemen-
tation Plan limitations. The product coal is to be transmitted by conveyor to the
plants. To eliminate the "blackwater" problem of earlier PCC plants, a closed-
circuit system will be used.
3.4 The Potential Consumption of Cleaned Ohio Coal at Sammis
The preceding section indicated that cleaned Ohio coals in many cases can serve
as SO2 compliance coals for units 5-7, for which SO2 standards are less stringent
than they are for units 1-4 (4.46 rather than 1.61 Ib SO2 per mi lion Btu). That
section also suggested that PCC can, to some extent, enhance the prospects of
59
-------
Table 13
Homer City PCC Plant: Performance Design Values
Weight Recovery (%)
Btu Recovery (%)
Heating Value (Btu/lb)
Ash (Wt. %)
Sulfur (Wt. %)
Sulfur (lbS/!06Btu)
Sulfur Removal (Wt. %)
Units 1 & 2
56.2
61.6
12,550
17.75
2.24
1.78
52.6
Unit3
24.7
32.9
15,200
2.84
0.88
0.58
91.8
Refuse
19.1
5.5
3,400
69.7
6.15
18.3
Source: "The Environmental Award," Power (November 1978), p. 214.
60
-------
burning Ohio coal in units 1-4. Given the assumptions regarding sulfur varia-
bility, the effective emission limits for units 1-4 (which represent about 30 per-
cent of Sammis's nameplate capacity) range from about 1.2 to 1.4 Ib S00 per
53
million Btu (see Section 3.1 and the Appendix). For the Lower Kittan-
ningCoshocton County coals listed in Table 10, this range of emissions is not
attainable, even with complete removal of pyritic sulfur. However, the BOM
washability data for Ohio coals (see Table 11) and the data on available
representative low-sulfur coals (Table 6) are somewhat more encouraging, indi-
cating that some of the Ohio coals may attain $©2 compliance levels with deep
cleaning even for the relatively stringent $©2 standards of units 1-4. Despite
these indications, in the remainder of this analysis we shall focus our attention
upon the more feasible prospect the use of cleaned coals to satisfy a large
fraction of the requirements of units 5-7, which consume more coal than
units 1-4 and are much less demanding with respect to S02 emissions. We shall
surmise that units 1-4 will use low-sulfur, non-Ohio coal from Southern
Appalachian states. According to the list in Table 6 such coals will be available
over the expected lifetimes of units 1-4. (Of course, cleaning the non-Ohio, low-
sulfur coals would somewhat enhance the prospects for using cleaned Ohio coals
in units 1-4).
As shown in Table I, units 5-7 account for 68 percent of Sammis's total
nameplate capacity. In 1977 the coal used by these units almost 2.5 million
tons came to 64 percent of Sammis's coal consumption. Also shown in Table I
are the relatively low capacity factors of these (newer and larger) units, which
have consistently experienced serious operational problems. If these problems
were alleviated and the yearly capacity factor were to increase to 60 percent,^
units 5-7 would increase their annual coal consumption from the 1977 rate of
2.5 million tons to almost 4 million tons.
Reflecting different assumptions about coal-sulfur variability, the effective
emission limit of units 5-7 will be from about 3.2 to 3.7 Ib S02 per million Btu
(the higher end of the range corresponds to the assumption that cleaned coals,
with lower RSD, are used). If we assume a 30 percent decrease of S02 emissions
61
-------
by PCC, these limits imply potential emissions from uncleaned coal of 4.6 to
5.3 Ib S02 per million Btu (2.4 to 2.8 Ib coal sulfur per million Btu)."^ If we
assume a 45 percent decrease of SC^ emissions by PCC (achieved by half the
BOM Ohio washability samples in Table 11), then the effective limits imply
potential SCX emissions of 5.8 to 6.7 Ib SOj Per niillion Btu from the uncieaned
coal (3.1 to 3.5 Ib coal sulfur per million Btu). Looking at Table 4 for the levels
of the Ohio coals delivered to Sammis in May 1978 (keeping in mind that some of
the listed coals may have been washed to some extent), we see that all these
levels fall within the allowable range when 45 percent sulfur removal is assumed.
Most fall within the allowable range when 30 percent sulfur removal is assumed.
Looking also at the potential SO9 emissions from the list of Ohio coals in Table 7
Cf £-
(and multiplying by 0.95), we see that most of these coals fall within or
slightly exceed the range specified above for 45 percent sulfur removal by PCC.
It appears, then, that a significant fraction of the SC^ compliance coals required
by units 5-7 can be met by cleaned Ohio coals. What that fraction will be will
depend on the raw-coal characteristics and level of PCC. To the extent that
cleaned Ohio coals cannot meet all the $©£ compliance needs of units 5-7, non-
Ohio, lower-sulfur coals will have to be mixed with the cleaned Ohio coal. (We
continue to assume that Sammis's adopted strategy for $©2 compliance will be to
burn low-sulfur coals.)
The technology of coal blending can be fairly sophisticated and highly auto-
mated. For example, the Navajo Mine in the Four Corners area of New Mexico,
which supplies about 2.5 million tons per year of highly variable coal, ensures
uniformity of product by use of a blending system that includes ten separate
storage piles of crushed coal, each built with a specially designed stacker. A
running inventory is automatically indicated while a pile is being built, to allow
for the adjustment of loading schedules. Reclaiming the coal from the piles also
involves special equipment. Sometimes special storage equipment, such as silos,
are used.
Blending may be performed at the mine, at a preparation plant, at a coal
transhipping terminal, or as part of the user's coal handling system. The
characteristics of the blended product must, of course, be compatible with the
user's facilities.
62
-------
The cost of a large, automated blending system was estimated in 1977 to be
about $1.50 per ton for a 4-milliorv-ton-per-year blending operation. This
translates to about 6 cents per million Btu for Ohio coals.
Becaus- of the space limitations at Sammis, we assume that only two streams
will be combined for use in units 5-7: the low-sulfur coal stored primarily for
units I -A, and the cleaned high-sulfur coal. The combined product must result in
SO? compliance. We assume further that these two streams will be combined by
means of only "ordinary mixing," and not by a sophisticated "blending" system.
Sarnmis has reported plans to store the low-sulfur coal for units 1-4 in a new
CO
coal-pile area served by a new conveyor belt system. The need for this second
coal pile is based not only on the different SOj emission requirements of the two
sets of units but also on the need for additional cod storage.
Assuming fixed-ratio mixing of two coal streams at Sammis, we ask now what
fraction and what quantity of cleaned Ohio coals can be used at Sammis
units 5-7. To answer we apply the formula:
Emax/(2 0.95) = SL(U2.l7RSDL)(l-fH)
I + 2.l7RSDH)(fH),
where th? factor 2.17 corresponds to the normal variate (the number of standard
deviations between the mean and allowable maximum E ) for two exceed-
ances per month and 99.87 percent confidence level (see Appendix), and:
fH is the fraction of cleaned high-sulfur coal;
E is the emission limit for units 5-7 (4.46 Ib
max SO, per million Btu), and E /(2^0.95)
is ^the corresponding coal-suffSr content,
assuming 5 percent retention of coal sulfur
during combustion;
S. and S|_| are mean values of Ib sulfur per million Btu
in the low-sulfur coal and high-sulfur coal
(after cleaning), respectively; and
RSD, and RSDH are the RSDs for the low-sulfur and cleaned
L n high-sulfur cods (0.15 and 0.08), respec-
tively, assuming the low-sulfur coal is not
cleaned.
63
-------
Solving for f,, (the fraction of cleaned high-sulfur coal in the mix), we have:
fH = EmQX/2 0.95 - SL( I + 2. 1 7 RSD,_)
SH(I 4 2.17 RSDH) - SL(I + 2.17 RSDL)
We observe that mixing is necessary (f^-j^ 0 only if the maximum probable
emissions from the "high-sulfur" coal exceed EmQx, that is:
Sw( I + 2. 1 7 RSDW)> E 12 - 0.95.
n rt max
Or substituting the values given for RSDH and EmQX,
SH(U74)>4.46/2-0.95,or
Ib sulfur per million Btu.
We now apply the formula for fH to determine the fraction of cleaned high-
sulfur coal. Values of f|_j are shown in Table 14 for two values of S. ,
corresponding to mean emissions from the low-sulfur coal of 1.2 and 1.4 Ib SO9
per million Btu (representing the previously described range of allowable mean
emissions from units 1-4).
For S, equal to 1.2, the fraction of cleaned Ohio coal ranges from 0.52 (when S,,
L- H
corresponds to the high mean emission level of 6.1 Ib $©2 per million Btu) to 1.0
(when S,. = 2.0, corresponding to the lower - but in many cases attainable
mean emission level of 3.8 Ib S02 per million Btu). For S^ = 1.4, the fraction of
clean high-sulfur coal is not very different: 0.49 when S,, corresponds to mean
emissions of 6.1 Ib $©2 per million Btu, and 1.0 when S., corresponds to values
not exceeding 3.8 Ib 5O2 per million Btu (the upper limit is, of course,
independent of S, ).
To translate these fractions into actual annual quantities of cleaned Ohio Coal
for units 5-7, we recall that these units burned almost 2.5 million tons in 1977,
and that, if the capacity factor of these units were increased to about
60 percent, the annual consumption would be almost 4 million tons per year.
64
-------
Table 14
The Allowable Fraction of Cleaned High-Sulfur Coal
at Sammis Units 5-7°
Low-Sulfur Coal
Mean Emissions
(lbS02/IObBtu)
Cleaned High-Sulfur Coal
Mean Emissions
(lbS02/l(TBtu)
Fraction, f.
1.2
1.2
1.2
1.2
1.4
1.4
1.4
1.4
° Based upon
b c r*»* Q
0.63
0.63
0.63
0.63
0.74
0.74
0.74
0.74
the formula and
nro x/nlilAc nf c
3.8
4.0
5.0
6.1
3.8
4.0
5.0
6.1
assumptions
nlfnr rr\n+»iV
2.0
2.1
2.6
3,2
2.0
2.1
2.6
3.2
described in the text.
t flK CltlflW r\Ap ml II inn
1. 00
0.93
0.68
0.52
1.00
0.92
0.66
0.49
D*..\ : *u~.
IflW-sulfur and high-sulfur coals. The mean emissions listed are based on
the assumption that five percent of the sulfur in these cools is retained in
the boiler ash.
65
-------
3.5 The Costs of Coal Cleaning
A utility will perceive the net monetary cost of PCC, for specifed PCC levels
and coals, in terms of two main sets of factors. The first set translates into an
incremental price of delivered coal, which reflects mainly (I) the capital and
operating costs associated with the PCC plants, and (2) the loss of Btu during
PCC, and (3), to a lesser extent, other items such as reduced transportation costs
and reduced payments for miners' benefits. The second set of factors relates to
the combustion of cleaned rather than raw coal at the power plant (for given
environmental regulations and operating conditions): burning cleaned rather than
raw coals generally results in monetary benefits (often difficult to quantify)
having to do with the pulverizers, the boiler, and the particulate control
equipment, and storage and disposal requirements. At Sammis, the use of PCC
may reduce the need for constructing new barge-unloading facilities; and at
Conesville (see Section 3.2), it may reduce the need for building rail facilities.
Further, any reduction in sulfur variability resulting from PCC will increase the
coal purchaser's options and therefore his bidding position. Finally, there is the
argument that PCC may reduce the need for unemployment or welfare payments
by enhancing the competitive position of locally produced coal.
In the following section we discuss estimates of the unit cost of PCC for a high-
sulfur eastern coal, considering several different levels of cleaning. In Sec-
tion 3.5.2 we discuss the (generally advantageous) effects that the removal of
mineral matter by PCC may have upon various power-plant operations. Finally,
in Section 3.5.3 we discuss the factors that must be balanced in order to
determine the point at which the use of cleaned Ohio coals at Sammis units 5-7
will be economically competitive with the use of non-Ohio, naturally low-sulfur
coals. To the extent possible, we quantify these factors but, unfortunately,
many of the data needed for a definitive comparison are not available.
3.5.1 Estimated Costs of Cleaning High-Sulfur Eastern Coal
The unit cost of producing cleaned coal will be the sum of processing costs
(including the disposition of refuse) and the value of the combustible material
66
-------
lost during processing. Some credit will accrue when PCC takes place at the
mine, because of lower payments for miners' benefits (the tonnage of the coal
received per energy content is reduced by PCC, and miners' benefits are based
upon tonnage sold).
To suggest approximate PCC costs for Ohio coal, we present here engineering
estimates developed by Versar, Inc., for "high-sulfur Eastern coal" (see Table 15);
these estimates were prepared for EPA as background material relating to
studies on a New Source Performance Standard for industrial boilers. We caution
that the processing costs will depend on the PCC site, the PCC process, and the
coals used. No generalizable PCC cost model has yet been developed, and, as
mentioned earlier, experience with intensive sulfur removal is still limited.
Table 15 shows the estimated costs associated with five levels of PCC applied to
a coal with about 12,000 Btu per Ib, 23 percent ash, and 3.4 percent sulfur
(2.8 percent pyritic sulfur). The main performance parameters the reduction
in weight and energy, the reduction in ash and sulfur contents, the increase in
heating value and required ancillary energy are shown. The annualized cost of
preparation is $2.00 per ton of product for Level 2 cleaning (which, in this
example, reduces pounds of SC^ per million Btu by 17 percent, and ash content
by 15 percent). The cost is $6.00 per ton of product for the two-stream,
intensive Level 5 process (similar to the Homer City design), which reduces SO2
emissions by 75 percent and 85 percent, respectively, in the two output streams.
We note that the levels of sulfur removal depicted for the higher PCC levels are
unreal 1sticalIy high for most Ohio coals: the percentage of pyritic sulfur in the
total sulfur of Ohio coals is rarely as high as 82 percent, the value that applies to
the example in Table 15.
The annual ized cost presented in Table 15 is the sum of first-year operating and
maintenance (O&M) costs and a fixed annual capital charge (based here on a
10 percent discount rate, a 20-year plant life, and four percent for taxes,
insurance, and G&A). By not levelizing O&M costs (significant for PCC), the
costs are underestimated, since O&M cost escalation is not accounted for.
67
-------
Table 15
00
Annual Physical Coal Cleaning Casts (1978 $) far a High-Sulfur Eastern Coala
(8,000-ton-per-day plant)1*
Levels of Cleaning
Yield: wt. %
Recovery: % energy
Btu content of clean coal (Btu/lb)
Weight % ash reduction
% Ib SO2/I06 Btu reduction
Hourly output, clean coal, tons/hr
Total turnkey costs, $
Land cost, $
Working capital, $
Grand total capital investment, $
Total annual costs (excluding coal cost), $
Cost of preparation (excluding coal cost),
$/ton of clean coal
Average energy requirement, Kw (10 Btu/hr)
1
98
100
1 1 ,974
4
3
603
3,962,000
120,000
170,800
4,252,800
1,572,400
0.80
250 (0.8)
2
85
92
12,678
15
17
523
9,506,400
180,000
365,200
10,051,600
3,377,500
1.99
650 (2.2)
3
75
85
13,265
51
53
462
16,634,400
264,000
555,600
17,454,000
5,409,200
3.60
1,000 (3.4)
4
70
87.5
14,674
68
69
431
19,010,400
720,000
714,300
29,444,700
6,635,300
4.74
1,300 (4.5)
5C
78
92
13.852
75 and 52
75 and 58
480
28,989,600
480,000
933,800
30,403,400
9,393,100
6.02
2,300 (7.9)
Source; Versar, Inc., Individual Technology Assessment Report for Physical and Chemical Coal Cleaning and Low
of NSPS for Industrial Boilers. Draft Report, vol. ^Springfield. Vo.. 1979).
Sulfur Cool in Support
a Raw coal characteristics include: Heating value = 11,740 Btu/lb; weight % ash = 23.4; weight % total sulfur = 3.4; weight % pyritic
sulfur = 2.8s Ib SO2/I06 Btu = 5.79.
Based on 13 hr/day, 250 days/year operation.
c The plant will generate two produci streams: a very high Btu stream, and a middlings stream. The heating value applies to the
combined product.
Based on first-year operating costs and annualized investment costs (10 percent discount rate, 20-year PCC plant life, and 4 percent
of depreciable investment for taxes, insurance, and G&A).
-------
In Table 16 we calculate three sets of costs for Levels 2, 3, and 4: (I) the
process costs (see Table 15); (2) the value of the Btu loss that occurs during
processing, based on an assumed current raw-coal cost of $1 per million Btu; and
(3) a credit for reduced miners' payments based on the reduced tonnage per Btu
after PCC. Interestingly, the vaiue of Btus lost during PCC (even for the
relatively efficient PCC processes indicated in Table 15) is comparable in each
case to the annualized PCC cost. The credit for miners' benefits is relatively
small, from 0.04 to 0.09 dollars per million Btu in the first year.
The values for each set of costs in Table 16 (in dollars per million Btu) are given
in terms of (I) current costs, and (2) levelized costs based on a 20-year period,
reflecting the escalation of costs and the cost of capital. (To levelize the
processing cost, the levelized O&M cost is added to the annualized capital cost.)
Based upon the assumptions indicated in the table, the total first-year costs for
Levels I, 2, and 3, respectively, are 0.16, 0.31, and 0.29 dollars per million Btu;
and the total levelized costs, again for the three levels respectively, are 0.26,
0.49, and 0.45 dollars per million Btu.
3.5.2 Cost Advantages of Burning Cleaned Coal
To what extent does PCC reduce operation and maintenance (O&M) costs at a
particular power plant? Work is proceeding on this complex question, but
estimates reported so far are tentative and not readily applied to specific plants.
Nevertheless, ./e mention several general observations and some recent esti-
mates to give some idea of how the removal of mineral matter by PCC may
affect plant costs.
The removal of mineral matter is expected to have its greatest effect on:
(I) furnace-wall slagging and fouling; (2) pulverizer wear; (3) convection pass
fouling; (4) coal handling and storage; (5) ash handling, storage, and disposal; and
(6) particulate control devices.
69
-------
Table 16
Summary of the Cost of Producing Cleaned Coa|a
Level of PCCb
2
Current
Cost
Level ized
CosT
3
Current
Cost
Level ized
Cost^
4
Current
Cost
Level ized
Cost*
Processing Cost"
$/ton 1.99 3.60
$/IO*Btu 0.078 O.i07
Btu loss
fractional lossb 8/92 15/85
$/!06Btud 0.87 0.157
Miners' Benefits
Increase in Btu/lb6 938 1 ,526
Change in $/ 1 Q6
Btu paid*
Total ($/ 10° Btu) 0.161 0.257
(0.004) (0.007)
0.136
4.74
0.182
0.162 0.219
12/88
0.176 0.318
0.136 0.246
2,963
(0.007) (0.013)
0.305 0.487
(0.009) (0.016)
0.289 0.449
Values pertain to a unit of PCC product coal.
See Versar's values in Table 15.
c Costs are levelized on the following basis: AIJ cases represent a 20-year period and a discount rats of 11.5 percent per
annum. Base costs for coal and miners' benefits are multiplied by 1.81, representing escalations of 7.5 percent per
annum. Base costs for the O&M costs for PCC are multiplied by 1.66, representing escalations of 6.5 percent per annum
The O&M costs are the difference between the annual costs and the product of (I) the total capital investment anH
(2) the factor of 0.15 (see Table 15).
^ Assuming a coal price of $1.00/10 Btu for the value of rejected Btu.
e Miners' benefits are $1.39 per ton of coal sold (1978 National Bituminous Wage Agreement of the United Mine Workers)
70
-------
Tentative "typical" cost benefits have recently been put forward, but with the
emphatic reminder of "the necessity to analyze each potential situation inde-
eg
pendently. Most significant among the estimated benefits are the following,
expressed in units of cleaned coal:
Typical Benefit
Benefit Area ($/ton) ($/IQ6Btu@ 13,000 Btu/lb)
Ash Disposal 0.20 0.008
Boiler Availability 0.40 0.015
Boiler Efficiency 0.70 0.029
Boiler O&M 0.50 0.027
Slagging and fouling tendencies will change as a result of PCC, usually (but not
always) favorably, largely because of the reduced quantity of ash but also
because of the selected reduction of some chemical constituents. In particular,
the removal of iron (in the pyrite removed by PCC) will generally result in less
slagging and fouling. However, this effect may be offset somewhat if the
product coal is contaminated by iron-containing materials, such as magnetite,
used in dense-media PCC.
In the case of Sammis, the low capacity factors of units 5-7 reflect serious O&M
problems attributed partly to poor coal quality. Referring to these problems, a
consultant to Ohio Edison reported that "the poorer quality coal on the market
today as compared to coal commonly available when the plant was designed, is
aggravating plant problems." In light of this situation, coal ash removal may
play an important role in improving plant availability at Sammis. To the extent
that PCC can effectively increase boiler capacity or make it unnecessary to add
new boiler capacity, its contribution to plant economics will be especially
valuable (new large coal-fired plants may require a capital investment of as
much as one million dollars per megawatt).
In a recent report on TVA's experience and analysis, the increase of rated plant
capacity by PCC was determined to have a high monetary worth: $3.02 per ton
71
-------
of product coal ($0.13 per million Btu for a heating value of 13,000 Btu/lb).^'
This value was determined for a high-ash, high-sulfur, western Kentucky coal
(20.5 percent ash and 7.0 percent sulfur before PCC; 10.5 percent ash and
4.5 percent sulfur after PCC; and a raw-coal heating value of 10,400 Btu per
pound).
Because PCC increases a coal's heating value (as illustrated in Table 15),
cleaning can lower the costs on a unit-energy basis of transporting coal. In
the case of Sammis, this effect may not be significant if cleaned Ohio coals are
compared with uncleaned out-of-state coals. Although cleaning would permit
the use of considerably more Ohio coal, and the Ohio coals originate much closer
to Sammis (up to about 150 miles, but usually within 50 miles) than do the out-
of-state, low-sulfur coals (see Table 6), the Ohio coals are hauled to Sammis by
truck, whereas the eastern Kentucky and southern West Virginia coals are hauled
mainly by barge and, in some cases, also by rail. Barge rates are considerably
lower than truck rates (in many cases by a factor of about one-tenth); rail rates
are also lower. Furthermore, the Southern Appalachian coals are often lower in
ash content than are the Ohio coals (see Tables 4 and 6). The increased use of
Ohio coals, therefore, will not obviously result in direct coal transportation
savings. If, however, the comparison is made between transporting cleaned coal
and transporting raw coal within Ohio, the transportation savings may be
significant: about $0.35 to $0.40 per ton for an average 50-mile haul, assuming a
truck rate of $0.06 per ton-mile and a post-PCC weight loss per energy unit of
approximately 15 percent. Recalling our discussion of Sammis's restricted coal
choices, we note that the comparison between hauling cleaned and uncleaned
Ohio coal applies realistically to only a small fraction of the coal that Sammis
will burn; the amount of Ohio coal burned at Sammis must be severely reduced
unless it js cleaned.
One last point perhaps the most important must be made in connection with
transportation costs: the cleaning of Ohio coals can significantly reduce the
need for constructing new barge unloading facilities at Sammis.
72
-------
PCC may have an important effect upon the plant's participate collection
devices, especially since Sammis must install new devices baghouses or
electrostatic precipitators (ESPs) to bring about the dramatic increase in
particulate emissions needed to comply with the Ohio standard of O.I Ib
particulates per million Btu (see Section 2.2). With cleaned coal, the reduced
particulate loading into the collection device can significantly reduce the
required capacity and consequently the cost of the new control devices.
The reduction in the cost of an ESP is roughly half the reduction in required
capacity; if, for example, PCC reduces ash content by 50 percent, the cost of
the control device can be expected to be reduced by roughly 25 percent
roughly $0.01 per million Btu of cleaned coal.
Although the efficiency of an ESP is sometimes reduced by lower coal-sulfur
levels, it should not be affected by the sulfur removal that would result from
PCC for Sammis units 5-7, which would burn coal of about 3.5 Ib sulfur per
million Btu. At this level there would be adequate concentrations of SO^ in the
flue gas to ensure proper conductivity in the ESP. Furthermore, it is not unlikely
that competitive bidding for contracts to build Sammis's new particulate
collection facilities will result in similar cost estimates for ESPs and fabric-
filter baghouses and the efficiency of baghouses does not depend on coal sulfur
content.
All the cost estimates mentioned in this section are for current costs. To derive
levelized costs, the current costs must be multiplied by a levelization factor
(such as 1.81, the factor applied to the coal costs shown in Table 16).
We have mentioned here some of the cost advantages of burning cleaned coal at
the power plant. In some cases there are also advantages in producing the coal.
When it is known that the coal will be cleaned, it is sometimes possible to use
cruder - and cheaper - mining methods, as, for example, in mines where
partings are difficult to remove with conventional processes. This effect will
become more important as lower-quality seams are mined, especially by under-
ground mining methods.
73
-------
3.5.3 Costs of Cleaned Ohio Coal versus Out-of-State
Low-Sulfur Coal far Unite 5-7
Whether Sammis units 5-7 burn cleaned Ohio coal or low-sulfur, non-Ohio coal
will depend largely on the comparative effective costs of the two options. To
make the comparison, one must examine each option's major cost factors, which
will involve a number of considerations, including alternative sources of the coal,
types of PCC facility and operation, and institutional agreements. Applicable
cost factors include:
Cleaned Ohio Coals Low-Sulfur, Non-Ohio Coals
Coal prices, f.o.b. mine Coal prices, f.o.b. mine
PCC production costs
Transportation (by truck) Transportation (by barge and
possibly by rail)
Contractual arrangements Contractual arrangements
Benefits at the power plant due
to removal of mineral matter
(see Section 3.5.2)
We consider first the difference between the raw-coal prices of the cleaned Ohio
coals and the low-sulfur, out-of-state coals (keeping in mind the allowable sulfur
levels for units 5-7), and then we compare this difference with the costs incurred
by the PCC process. Again, we caution that generalizations are risky. Because
of the anticipated changes in coal purchases by Ohio's utilities, the market price
of lower-sulfur coals, which has been depressed lately, is expected to escalate
faster than the market price of Ohio (high-sulfur) coals. It is largely in
anticipation of increased demand and higher prices for their supplies that
producers of lower-sulfur, Southern Appalachian coal are currently reluctant to
enter into long-term contracts.
With this caveat in mine, we look at Figure 9, which shows a "best fit" curve
drawn through points representing a set of eastern coal prices (early 1979) as a
function of potential SO2 emissions. The function represented by the curve
for which the coefficient of determination is only 0.565 is one in which one
-------
en
140
130
09)
1201
c "ol
CL
901
T
Eastern Coal Prices as a Function of Sulfur Content
Producing Dlstricti
I. Central Pennsylvania
2. Western Pennsylvania
3. Ohio ( '
7. Southeatlem West Virginia and parts of Virginia
8. Southern Wesl Virginia, eastern Kentucky,
northern Tennessee, and ports of Virginia
9. Western Kentucky
10. Illinois
II. Indiana
0 1.0 7.0 3.0 4.0 5.0
LbS02/l06Btu
Source; Coal prices and characteristics from Cool Week, 9 April 1979.
6.0
y = 127 x
r2 = 0.565
-0.14
7.0
-------
price is more sensitive to changes in sulfur content at the lower levels of sulfur
content. From the curve we read current coal prices corresponding to the range
of average SC^ emissions allowed for units 5-7, which is 3.2 to 3.8 Ib SO-, per
million Btu (as determined in Section 3.1 on the basis of the limit of 4.46 Ib SO,
per million Btu and stated assumptions about sulfur variability). This range
implies that, for a 30 percent reduction of $©2 emissions by PCC, allowable
average emissions from the raw coals must be 4.6 to 5.3 Ib SO^ P^r million Btu,
and that, for a 45 percent reduction of emissions by PCC, allowable average
emissions from the raw coals must be 5.8 to 6.7 Ib
We read the raw-coal prices for the uncleaned coal and then find the difference
between these prices and those for naturally low-sulfur coals that correspond to
(I) 30 percent SO2 reduction by PCC, and (2) 45 percent SO, reduction by PCC.
We then multiply these differences by a levelization factor and compare the
resulting differences (levelized savings attributable to PCC) with the levelized
costs of PCC. The PCC costs for 30 percent and 45 percent SO2 reduction are
found from Table 16 by linearly interpolating between Level 2 (1 7 percent S07
reduction) and Level 3 (53 percent 862 reduction). This comparison between the
savings in raw-coal costs by using PCC and the expenses incurred in producing
cleaned coals is summarized in Table 17.
According to Table 17, the costs of PCC outweigh the savings that are due only
to the difference in costs of the raw coal used with and without PCC the PCC
costs are about $0.25 per million Btu higher. We note that, although we have
assumed here that the maximum reduction in SOj emissions is 45 percent, a
greater reduction is possible in some cases. The new R & F PCC plant near
Cadiz, for example, is designed to reduce SO2 emissions from Ohio coals by
about 80 percent (see Section 2.3.4).
If we add to the raw-coal savings the savings in the power-plant operations
mentioned in Section 3.5.2, then the savings from using PCC in fact outweigh me
costs of PCC. The savings, in dollars per million Btu, attributed to increased
plant capacity by TVA (0.13) and to lower boiler O&M by PEDCo (0.027), when
multiplied by the levelization factor of 1.81, equal 0.28 dollars per million Btu.
76
-------
Table 17
Summary of Costs versus Savings with PCC
NoPCC
30% Sulfur Removal by PCC 45% Sulfur Removal by PCC
Allowable raw-coal S(X
emissions from units3-7
Savings in levelized
raw-coal costs with PCC
(from Figure 9)
Level ized costs for PCC
production (interpolated
from values in Table 16)
Difference between
PCC costs and raw-coal
costs savings
3.2-3.7 IbSCy 106Btu
4.6-5.4 lbSO2/106Btu
5.8-6.9 IbSCy 106Btu
0.06 x 1.81 =0.11 $/!06Btu 0.09 x 1.81 = 0.16 $/!06 Btu
0.34$/l06Btu
0.24$/l06Btu
0.43$/106 Btu
0.27$/l06Btu
-------
This more than cancels the debit of 0.25 dollars per million Btu indicated above
where PCC costs are compared only with raw-coal savings. Further, to the
extent that low-sulfur coals increase more rapidly in price than high-sulfur coals,
the raw coal savings from PCC will increase. Other savings postulated for the
burning of cleaned coals related to such factors as ash disposal, boiler
efficiency, particulate controls, pulverizer wear and capacity, and, in the case of
Sammis, elimination of the need to build new barge unloading facilities are not
added here to the cost-savings side of the equation.
We emphasize that the cost analysis presented here suggests the kind of
procedure that must be followed in order to determine the economic competi-
tiveness of using cleaned Ohio coals. Many of the illustrative values we have
used, however, must be replaced by hard data, to which we do not have access.
Ohio Edison can be expected to have many of these data; other data, however
can be determined only empirically, and still others can be determined from
negotiations in the marketplace (or on the basis of existing contracts).
Judging from what we have been able to show, it appears that PCC for units 5-7
may be economically justifiable. Given this conclusion and given the
indications that PCC can minimize the negative effects that a switch away from
Ohio coals will have on the state's coal industry it appears prudent to
recommend that the needed information be acquired for a decisive economic
analysis of the cost of PCC for Ohio plants in general and Sammis in particular.
The importance of the consequences that will result from a decision to adopt or
reject PCC dictates that such an analysis be carried out.
3.6 Institutional Barriers to Implementing PCC in Ohio
In the preceding section we showed that PCC may represent an economically
viable 502 comP''ance strategy for Sammis units 5-7. And in earlier sections we
indicated that PCC can prevent the shutdown of a large fraction of Ohio's mines
(80 percent of which may close as a result of S02 limitations). There are
78
-------
however, several barriers obstructing the adoption of PCC in Ohio. We mention
the major barriers here.
First, many Ohio mines are too small to produce the input required by the
smallest economically feasible PCC plant (about 100 tons per hour). Next, the
coal preparation engineers and contractors needed to build advanced-technology
PCC plants are often in short supply. Further, PCC plants will be subject to
environmental constraints (the effects of which include the need for closed-
circuit water systems and, in some cases, the prohibition of thermal dryers).
Finally, there will be a time lag of at least 2.5 to 3 years between inception and
completion of a PCC plant.
The time lag in the production of cleaned coals raises a barrier against the use of
PCC (one mentioned in discussing the Conesville plant, in Section 3.2): What will
utilities do to meet their $©2 standard while waiting for PCC to become
available? If they burn out-of-state, low-sulfur coal during the interim, many of
the potential producers of cleaned Ohio coal may suffer irreversible financial
problems and loss in production capability. If, on the other hand, the utilities
burn noncomplionce coal from Ohio, they will need a temporary waiver of
environmental standards.
While these barriers are not insurmountable, they do exist and some effort will
be needed before they can, in fact, be overcome.
79
-------
APPENDIX
SULFUR VARIABILITY AND A COMPARISON OF THE EFFECTIVE
AND MANDATED S02 EMISSION LIMITATIONS
This appendix develops a number of topics alluded to in Section 3.1 ("Average
Coal-Sulfur Values in Relation to SC^ Emission Limits and Coal-Sulfur Varia-
bility") and should be read in conjunction with that section. The topics include:
Effect of lot size on the relative standard deviation (RSD)
of coal sulfur content
36
Effect of coal cleaning on RSD
Effect on RSD of the number of daily exceedances
allowed per month
Effect of Lot Size on RSD
In a recent study for EPA, PEDCo examined the variation of RSD with lot size
on the basis of sulfur-content measurements from a data set representing coals
with less than one percent sulfur. Computed RSDs were compounded statisti-
cally for the case where unit trains (8,400 tons) were sampled at the rate of four
per week to obtain RSDs for lot sizes representing a unit train for periods of one
week, one month, three months, six months, and one year that is, for lot sizes
from 8,400 to more than one million tons.
PEDCo's results are shown as data points on the solid curve of Figure A-1
where RSDs for sulfur (percentage by weight) are plotted against the log of
lot size (or averaging period). The curve indicates a decreasing function that
approaches zero as the lot size exceeds several million tons. For smaller lot
sizes, a straight-line extrapolation was used:
RSD = 0.289 - 0.0341 log^T,
where T represents the lot size in tons. Extended to the very small lot size of
50 Ibs (a typical core size), the RSD according to this formula reaches the very
high value of 0.34.
80
-------
The curve indicates that the RSD for a lot size of 10,000 tons approximately
the daily average coal feed for Sammis is about 0.15. By contrast, the RSD
for 100 tons of coal (the daily consumption of a small industrial boiler of about
100 million Btu per hour at 100 percent capacity) would be twice as high 0.30.
The curve of RSD versus the logi/jT in Figure A- 1 is based on data sets for coals
with less than one percent sulfur and for which RSD decreases with increasing
lot size. In some instances in the PEDCo study the RSD remained unchanged or
even increased slightly with increasing lot size. These instances were attributed
to either "aberrations in data or the fact that samples were not truly representa-
tive of the entire lot."
It should be stressed that a coal user must empirically determine the RSD for
any particular coal he will be using. In this regard the PEDCo report explicitly
states:
It is emphasized that the values [see Figure A-Q are based on a
collection of coal data made available by selected companies.
Each company using the approach presented herein is urged to
use its own data in estimating the variability of sulfur content
for specified times or tonnages.
It is impossible to generalize about the RSDs of either raw or cleaned coal, even
for coals from a single county and seam. While the overall trend of the data
analyzed in the abovementioned report does support the intuitive notion that
RSD decreases as lot size increases, in a local sense it may not be possible to
verify any functional relationship or even trend between RSD and lot size. One
important reason for aberrations in the trend is that the RSD per unit weight
may in fact be different among different types of coal, as suggested by the data
in Table 7 in Section 3.1. In particular, the RSD of washed coal is almost always
smaller than that of the corresponding raw coal (we shall say more about this
later).
Another report for EPA, this one by Versar,67 computed RSDs from measured
values of percent sulfur and heating value. The values representing various
81
-------
RSD of Sulfur Content versus Averaging Period (Lot Size)
00
Q
§
1
Is
8
u
"E u
II
O CO
4)
0.24
0
0.20
0.16
0.12
0.08
0.04
0
48
33 tons
(3 tours for
25=MW plont)
Tons of Coal
480 4,800
48,000
480,000
4.8 x 10
.x 600 tons (3 hours for 500 MW plant)
"-». t
Region of extrapolation
4,800 tons (I day for 500 MW plant)
8,400 (I unit train for 500 MW plant)
3 Hours 1 Day 7
Averaging Period (500 MW Plant)
30
90 180 360
Source: PEDCo, Inc., Preliminary Evaluations of Sulfur Variability in Low-Sulfur Coals from Selected
Mines, EPA-45013-77-044, prepared for U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, North Carolina (Cincinnati, Ohio,
November 1977).
Note; Values represent Appalachian coals containing up to one percent sulfur. They are estimated
to apply to coals wifh up to 1.5 percent sulfur.
-------
raw and cleaned coal types and both mechanical and manual sampling were
provided mainly by coal companies but also by two EPA studies. The emphasis
was on comparing RSDs of coals before and after coal cleaning.
To examine the functional relationship of RSD and lot size, the Versar study took
25 data sets (each comprising from 23 to 26 samples and representing a raw or
cleaned coal from a given county and bed), partitioned them into three or four
lot-size ranges, and constructed 25 plots of RSD versus lot size, each with three
or four points. Two further plots one of 20, the other of 13 points were
constructed, each an aggregate of computed RSDs for all data sets from a
particular coal seam and state. These representations did not indicate a trend of
decreasing RSD with increasing lot size. We offer two explanations for the
absence of this trend. First, the sample sets in the Versar study were relatively
small: the aggregated plots represented ranges of 3.5 to 26, and 1.3 to 8.1,
thousand tons. (By contrast, the range of lot sizes in the PEDCo study
represented several orders of magnitude.) A marked trend of decreasing RSD
with increasing lot size may be apparent only over a wide range of lot sizes.
Second, there is the heterogeneity of coals within a seam: different coals may
have different RSDs (per lot size). This fact emerged as a number of multiple-
valued RSDs when RSDs of different data sets within a coal seam and state were
plotted against lot size. Thus, it is obvious that RSD is not a function of lot size
alone. In particular, assuming that RSD does not change with the mean sulfur
content implies that there is less sulfur variability for low-sulfur than for high-
sulfur coals (since RSD is the ratio of the standard deviation to the mean sulfur
content). No available data do, in fact, substantiate this assumption.
The Effect of Coal Cleaning on RSD
An important result of the Versar analysis concerns the reduction of RSD of Ib
SO2 per million Btu as a result of coal preparation. A straight-line fit to data
points from nine coal-preparation plants, each operating on a different seam,
relates the RSD of the uncleaned coal (RSD^) to the RSD of the cleaned coal
(RSDC):
RSDc=: .836 RSD^ - .051.
83
-------
Although it must be stressed that the actual reduction of RSD with physical coal
cleaning will have to be tested against the specific raw-coal type and cleaning
68
process used, it is interesting to examine the implications of this equation.
For an uncleaned coal with an RSD equal to 0.15 the RSD that can be deduced
from the PEDCo plot for Sammis's approximate 24-hour coal use, and also the
value used in a number of EPA reports the RSD of the cleaned coal would be
only 0.08.
Since organic sulfur is bound to the coal and not generally removed by PCC,
while the pyritic sulfur is associated with the incombustible (ash-producing)
material and js removed by PCC, a lower value of the RSD of Ib SC^ per million
Btu suggests that, in a raw coal, the values of RSDs (Ib SC^ per million Btu) are
higher for pyritic sulfur than for organic sulfur. While reported coal sulfur
values are not generally separated into pyritic and nonpyritic sulfur, the washa-
bility data prepared by the Bureau of Mines do include mean values and standard
deviations of measurements of both pyritic and total sulfur for raw and cleaned
coals within a county and coal bed. In examining these data to compare the
RSDs of pyritic and organic sulfur (Ib SC^ per million/Btu), we found that,
indeed, the raw-coal RSDs were usually higher for pyritic sulfur than for total
sulfur. Further, as can be seen in Table A-1, when these data are aggregated on
a regional basis, the raw-coal RSDs for pyritic sulfur are consistently higher than
those for total sulfur.
Additional data and results of analyses pertaining to RSDs will become available
from a number of ongoing EPA studies of sulfur variability.
Effect on RSDs of the Number of Daily Exceedances Allowed per Month
Implied in the computations of relative standard deviation for the cases listed in
Table 8 (Section 3.1) is the assumption that the 24-hour S02 standard will never
be exceeded. If, in fact, one or more exceedances per month will be permitted,
84
-------
Table A-1
Total Sulfur and Pyritic Sulfur Content:
Comparison of Variability
Sulfur
Number of
Coal Region Samples
Content of Raw Coal
Pyritic Sulfur (%)
Mean % Sigma
N. Appalachian
S. Appalachian
Alabama
Eastern Midwest
Western Midwest
Western
Source: Joseph
of the
227
35
10
95
44
44
2.01
0
0
2
3
0
.37
.69
.29
.58
.23
A. Cavallero et al.f
United States, Rl 81
1.
0.
0.
1.
3
4
8
0
1.9
0.3
Sulfur
RSD
0.65
1.08
1.16
0.44
0.53
1.3
Reduction
Total Sulfur (%)
Mean*
3.01
1.04
1.33
3.92
5.25
0.68
> Sigma
1.6
0,
0
1
2
0
.6
.9
.2
.3
.3
RSD
0.
0.
0.
0.
0,
53
58
68
31
,44
0.44
Potential of the Coals
18 (Pittsburgh, Pa.: U.S.
Department
of the Interior, Bureau of Mines, 1976).
Note: RSD is the ratio of the standard deviation (sigma) to the mean.
85
-------
a higher mean SC^ value will be acceptable. Since Sammis may be permitted to
exceed its 502 s^an<^arc^ *wo t'mes Per rnonth, we have also computed the
average sulfur level taking this variance into account for one of the cases
described in Table 8: RSD of 0.15, S02 standard of 4.46 Ib S02 per million Btu,
confidence level of 99.87 percent, and a normal probability distribution.
The mean, m, for this case when no exceedances are allowed was shown to be
3.07, using:
4.46 Ib - m = 3m 0.15,
where 0.15 is the RSD and the factor 3 is the normal variate (the number of
standard deviations above the mean) corresponding to a 99.87 percent confidence
level (a probability of .0013 that the 24-hour standard of 4.46 Ib S02 per million
Btu will never be exceeded).
To find the average SCU level when two exceedances per month are permitted,
we first compute the single-day probability, P, of meeting the requirement that
the probability of three or more violations occurring during a 30-day period is
.0013 (corresponding to a 99.87 percent confidence level). To do so, we sum the
fn
probabilities of (I) no violations, (2) one violation, and (3) two violations:
.0013 = I - (3Q°) P° (I - P)30 + (30) P1 (I - P)29 * (32°) P2 (I - P)28,
.9987 = (I - P)30 + 30P (I - P)29 + 435 P2 (I - P)28.
or
Through iteration, the value of P is found to be about 0.013. We now apply the
normal variate corresponding to P = 0.013, which is 2.23 (for the same case, but
with no exceptions, the normal variate is 3.0, corresponding simply to 0.0013 or a
99.87 percent confidence level). Therefore, for the Sammis limit of 4.46 Ib S00
86
-------
per million Btu (units 5-7) and an RSD of 0.15, allowing two exceptions per
month implies:
4.46 Ib-m = 0.15m -2.23,
or
m = 3.34 Ib S02 per million Btu,
which can be compared with the lower allowable mean of 3.07 Ib SO2 per million
Btu when no exceedances are permitted for the 24-hour standard of 4.46 Ib SO?
per million Btu, given the same confidence level (99.87 percent).
The result yielded by a similar computation for three exceptions per month was
not very different: 3.36 instead of 3.34 Ib S02 per million Btu.70
87
-------
NOTES
I. William R. Forsyth, Manager of the Production Fuel Department of Ohio
Edison Company, in affidavit in Case No. C-2-78-786, 12 March 1978.
2. Ibid.
3. ICF, Potential Impacts on the Ohio Coal Market; Ohio Utility Compliance
with Applicable Air Emission Limitations, Section 125 Study, reporf
submitted to U.S. Environmental Protection Agency, Office of Planning
and Evaluation, December 1978.
4. Ohio Edison, 1978 Annual Report (Akron, Ohio).
5. Federal Register 41, no. 168 (27 August 1976).
6. Daniel Bodor, superintendent of the W.H. Sammis Station, in affidavit for
Case No. C-2-8-786, 24 September 1978.
7. Ibid.
8. Jack Crittenden, Commonwealth Associated Inc. of Jackson, Michigan, in
affidavit for Case No. C-2-78-786, 21 September 1978.
9. See n. 6.
10. See n. 6.
II. Henry Modetz, power generation specialist, EPA Region V, Air Enforce-
ment Branch, Enforcement Division, speaking as affiant in Civil Action
No. 02-78-786, 22 September 1978.
12. Personal communication with F. Richard Kurzynske, engineer, Environ-
mental Protection Agency, Region V, Chicago, Illinois, 30 May 1979.
13. "Amended Motion for Preliminary Injunction," submitted by James E.
Rattan, assistant United States attorney, to U.S. District Court, Columbus,
Ohio, 15 June 1979.
14. Federal Register 41, no. 168 (27 August 1976): 36332-33.
15. Ibid.
16. Federal Register 43, no. 32 (15 February 1978): 6646.
17. Personal communication with Bertram Fry, legal counsel, Environmental
Protection Agency, Region V, Chicago, Illinois, 19 March 1979.
88
-------
18. See n. 6.
19. See n. I.
20. Temple, Barker, & Sloane, Inc., Ohio Section 125 Studyt Regional
Economic Impact Analysis, report prepared for U.S. Environmental Protec-
tion Agency, EPA Contract No. 68-01-4905 (Wellesley Hills, Mass.,
I 4 December 1978).
21. Federal Register 43, no. 25 (28 December 1978): 60652-61.
22. Energy Users Report. 15 March 1979, p. 14.
23. Environmental Reporter, 3 November 1978, p. 1247.
24. Richard A. Stuble et al., Deep-Core Investigation of Low-Sulfur Coal
Possibilities in Southeastern" ONo, Rl No. 81 (Columbus. Ohio; O~hTo
Division of Geological Survey, 1971).
25. Federal Energy Regulatory Commission, Annual Summary of Cost and
Quality of Electric Utility Plant Fuels, 1977 (Washington, D.C., 1978).
26. U.S. Bureau of Mines, Reserve Base of U.S. Coals by Sulfur Content,
Eastern States, 1C 8680, PB-243 031 (Pittsburgh, Pa., May 1975).
27. U.S. Department of Energy, Coal - Bituminous and Lignite in 1976,
DOE/EPA 0118/1(76) (Washington, D.C., 18 December 1978).
28. 1977 Keystone Coal Industry Manual (New York, New York: McGraw-Hill,
1977), p. 334.
29. Personal communications with (l)Weldon Fulgum, engineering manager,
R & F Coal Company, Cadiz, Ohio, 30 June 1979, and (2) Natie Allen, Jr.,
chief of Fossil Fuels Planning Branch, Tennessee Valley Authority,
Chattanooga, Tennessee, 12 June 1979.
30. Coal Week. 25 April 1977, p. 9.
31. See n. I.
32. See n. 12.
33. Ibid.
34. The data are based on information supplied by coal brokers, mine repre-
sentatives, and a consultant: Norman Kilpatrick, director of the Surface
Mining Research Library, Charleston, West Virginia.
35. Coal Week. 8 January 1979. Pittsburgh is about 55 miles from Sommis.
89
-------
36. Ratio of the standard deviation to the mean.
37. PEDCo, Inc., Preliminary Evaluations of Sulfur Variability in Low-Sulfur
Coals from Selected Mines, EPA-450/3-77-044, prepared for U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina (Cincinnati, Ohio, November 1977).
38. Versar, Inc., SO*, Emission Reduction Data from Commercial Physical Coaj
Cleaning Plants and Analysis of Product Sulfur Variability, Draft Final
Task 600 report under EPA Contract No. 68-0202199, submitted to Fuel
Process Branch, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina (Springfield, Virginia, 18 October 1978).
39. Draft memorandum from Kenneth Schweers, ICF, to Robert Fuhrman, U.S.
Environmental Protection Agency, 25 October 1978.
40. For units 1-4: 1.61 Ib SO2/I06 Btu. For units 5-7: 4.46 Ib SO2/I06 Btu.
41. Coal Outlook, 19 February 1979.
42. Personal communication with Jock Apel, vice president of Environmental
Affairs, Columbus and Southern Ohio Electric Company, Columbus, Ohio,
19 March 1979.
43. Cavallero et al., Sulfur Reduction Potential of the Coals of the United
States. Bureau of Mines Rl 8118 (Pittsburgh, Pa.: U.S. Department of the
Interior, Bureau of Mines, 1976).
44. For example, one can show that, for the fairly high moisture content of
30 percent,
(Ib S02/I06 Btu)M=*(lb S02/I06 Btu)MF 1.034,
where M = with-moisture basis, and MF = moisture-free basis.
45. In fact, the percentage increase in heating value that can result from PCC
spans a considerable range, depending upon the coal and the cleaning
process used. A recent report for the Electric Power Research Institute
showed a range of U to 26 percent; see M.K. Buder and K.L. Clifford et
al., The Effects of Coal Cleaning on Power Generation (San Francisco,
Calif.: Bechtel National, Inc., 1978).
46. See n. 42.
47. See n. 43.
48. Crushing to 3/8 inch; specific gravity of 1.4.
49. See n. 24.
90
-------
50. ibid.
51. See n. 38.
52. Sprn Ruggeri, American Electric Power Company, in presentation to the
"Front End Coal Cleaning Conference" sponsored by Pennsylvania Electric
Company and New York State Electric and Gas, Pittsburgh, Pennsylvania,
8 November 1978.
53. Assuming coal-sulfur RSDs from 0.08 to 0.15, normal or lognormal
probability distributions, two exceedances per month, and a confidence
level of 99.87 percent.
54. In 1976 the median capacity factor of all U.S. plants with capacity ratings
exceeding 500 MW was 59 percent.
55. Assuming 5 percent retention of sulfur in the boiler ash.
56. Again, assuming 5 percent retention in the boiler ash.
57. Peter J. Phillips, "How Blending Improves Coals' Quality," Coal Mining and
Processing, October 1977.
58. Acurex Corporation, JACA Corporation, and Professional Management,
Inc., Engineering Study for Ohio Coal Burning Power Plants. Final
Report 78-311, prepared foF U.S. Environmental Protection Agency,
Division of Stationary Source Enforcement (Mountain View, Calif.; Fort
Washington, Pa.; and Cincinnati, Ohio; March 1979).
59. PEDCo Environmental, Inc., Cost Benefits Associated with the Use of
Physically Cleaned Coal. Draft report under EPA Contract No 68-02-2603,
submitted to James Kilgroe, U.S. Environmental Protection Agency,
Industrial Engineering Research Laboratory, Research Triangle Park, North
Carolina (Dallas, Texas, 25 May 1979).
60. See n. 4.
61. Peter J. Phillips and Randy M. Cole, "Economic Penalties Attributable to
Ash Content of Steam Coals," paper presented at the AIME Coal Utiliza-
tion Symposium, New Orleans, Louisiana, 18 February 1979.
62. Personal communication with Richard A. Chapman, Teknekron Research,
Inc., regarding his ongoing research for "Evaluation and Assessment
Methodology for Collecting Fly Ash from the Combustion of Low-Sulfur
Coal," EPA Contract No. 68-02-2652.
63. The prices shown in Figure 9 are multiplied by a levelization factor of 1.81
(see Table 16); the SO^ values in Figure 9 are multiplied by 0.95 to allow
for 5 percent sulfur reduction during combustion.
91
-------
64. Weldon Fulgum. See n. 29.
65. For a more detailed discussion, see Teknekron, Inc., An Evaluation of
Institutional, Economic and Social, Regulatory and Legislative Barriers to
Investment in Physical Cod Cleaning as a Sulfur Dioxide Emissions Control
Strategy, prepared for U.S. Environmental Protection Agency, Industrial
Environmental Research Laboratory, Research Triangle Park, North
Carolina (Berkeley, California, 15 December 1977).
66. PEDCo, Inc., Preliminary Evaluations of Sulfur Variability in Low-Sulfur
Coals from Selected Mines. EPA-450/3-77-044, prepared for U.S. Environ-
mental Protection Agency, Office of Air Quality Planning and Standards,
Research Triangle Park, North Carolina (Cincinnati, Ohio, November 1977).
67. See n. 38.
68. The Versar report cautions that values of both the product-coal RSDs and
the reduction in RSD resulting from cleaning "are valid only for analysis of
individual preparation plants and should not be aggregated on a seam or
regional basis."
69. The factors (vj are "binomial coefficients," the number of ways that 30
objects and k objects can be arranged. These factors are equal to
30!/k! (30 - k)!; values can be found in standard statistical tables.
70. The method used here was adapted from an internal Environmental
Protection Agency working memorandum from John W. Melone, Statistical
Evaluation Staff, to Paul Stolpman, Office of Policy Analysis, 17 October
1978.
92
-------
TECHNICAL REPORT DATA
(Please read /inductions on the reverse before completing)
REPORT NO.
3PA-600/7-80-009
3. RECIPIENT'S ACCESSION NO.
. TITLE AND SUBTITLE
Sammis Generating Station: Meeting SO2 and
Participate Standards with Cleaned Ohio Coals
. REPORT DATE
January 1980
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
31adys Sessler
8. PERFORMING ORGANIZATION REPORT NO.
"PERFORMING ORGANIZATION NAME AND ADDRESS
Teknekron Research, Inc.
"nergy and Environmental Systems Division
2118 Milvia Street
Berkeley, California 94704
10. PROGRAM ELEMENT NO.
E HE 62 3 A
11. CONTRACT/GRANT NO.
68-02-3092, Task 3B
2. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TVPE OF REPORT AND PERIOD COVERED
Task Final; 3-7/79
14. SPONSORING AGENCY CODE
EPA/600/13
5. SUPPLEMENTARY NOTES jERL-RTP project officer is James D. Kilgroe, Mail Drop 61,
S1S/541-2851.
16. ABSTRACT
rep0rt discusses the background and issues related to the control of
air pollutants emitted by a large coal-burning plant in eastern Ohio. The plant not
only has had a history of severely exceeding Ohio's State Implementation Plan (SIP)
particulate emission limit, but also its SO2 emissions have exceeded the limit of
Ohio's forthcoming SIP. An important issue is the extent to which compliance with
the SIP will promote the plant's switching from Ohio coals to Southern Appalachian
coals (which produce fewer particulate and SO2 emissions) and the consequent disrup-
tion to Ohio's coal mining industry. Addressing this issue, the report examines the
plant's historical coal usage, the production and characteristics of Ohio and Southern
Appalachian coals, the relevance of coal-sulfur variability, and the feasibility and
implications of producing and burning cleaned Ohio coals as a strategy for complying
with Ohio's SIP. The report discusses factors that will affect the relative economics
of burning cleaned Ohio coals at the plant. The report indicates that, by burning
cleaned Ohio coals , the plant's largest and newest units (constituting 60% of the
plant's total capacity) can increase their consumption of Ohio coal by 50-100%,
depending on the characteristics of the coals and the cleaning processes used.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Coal
Combustion
Emission
Coal Preparation
Sulfur Oxides
Dust
Aerosols
Coal Mining
Industrial Processes
Pollution Control
Stationary Sources
Particulate
Coal Cleaning
13B
2 ID
21B
14B
081
07B
11G
07D
13H
'DISTRIBUTION STATEMENI
Release to Public
19. SECURITY CLASS (ThisReport}
Unclassified
21. NO. OF PAGES
100
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
93
------- |