Final Report
STRATEGIES
clobtloa
The U. S. Environmental Protection Agency is charged by Congress
with protecting the Nation's land, air, and water resources. Under
a mandate of national environmental laws, the Agency strives to
formulate and implement actions leading to a compatible balance
between human activities and the ability of natural systems to
support and nurture life. To meet this mandate, EPA's research
program is providing data and technical support for solving
environmental problems today and building a science knowledge base
necessary to manage our ecological resources wisely, understand how
pollutants affect our health, and prevent or reduce environmental
risks in the future.
The National Risk Management Research Laboratory is the Agency's
center for investigation of technological and management approaches
for reducing risks from threats to human health and the
environment. The focus of the Laboratory's research program is on
methods for the prevention and control of pollution to air, land,
water, and subsurface resources, protection of water quality in
public water systems; remediation of contaminated sites and
groundwater; and prevention and control of indoor air pollution.
The goal of this research effort is to catalyze development and
implementation of innovative, cost-effective environmental
technologies; develop scientific and engineering information needed
by EPA to support regulatory and policy decisions; and provide
technical support and information transfer to ensure effective
implementation of environmental regulations and strategies.
This publication has been produced as part of the Laboratory's
strategic long-term research plan. It is published and made
available by EPA's Office of Research and Development to assist the
user community and to link researchers with their clients.
E. Timothy Oppelt, Director
National Risk Management Research Laboratory
EPA REVIEW NOTICE
This report has been peer and administratively reviewed by the
U.S. Environmental Protection Agency, and approved for publication.
Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
This document is available to the public through the National
Technical Information Service, Springfield, Virginia 22161.
EPA-600/R-02/073 October 2002
Engineering and Economic Factors Affecting the Installation of
Control Technologies for Multipollutant Strategies
Prepared by:
ARCADIS Geraghty & Miller
4915 Prospectus Drive, Suite F
Durham, NC 27713
EPA Contract No. 68-C-99-201
Work Assignment 3-034
U.S. EPA Project Officer: Ravi K. Srivastava
National Risk Management Laboratory
Research Triangle Park, NC 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, DC 20460
^ÅâåçïäÉÇÖÉãÉåÃë
ARCADIS would like to acknowledge the many contributors to this
document, without whose efforts this report would not be complete.
In particular, we wish to acknowledge the advice of Chad Whiteman
of Clean Air Markets Division, Office of Air and Radiation, U.S.
Environmental Protection Agency. Helpful suggestions were received
from Kevin Culligan and Mary Jo Krolewski, both of Clean Air
Markets Division. Technical guidance was received from Dr. Ravi
Srivastava of the National Risk Management Research Laboratory,
Office of Research and Development, U.S. Environmental Protection
Agency.
ARCADIS would also like to acknowledge other organizations,
which provided research that was incorporated into the report.
These organizations include ESI International and Andover
Technology Partners.
bñÉÅìÃáîÉ=pìãã~êó
This report evaluates the engineering and economic factors of
installing air pollution control technologies to meet the
requirements of multipollutant control strategies. The
implementation timing and reduction stringency of such strategies
affect the quantity of resources required to complete the control
technology installations and the ability of markets to adjust and
to provide more resources where needed. Using the Integrated
Planning Model (IPM), the U.S. Environmental Protection Agency
(EPA) estimated the number and size of facilities that need to
install new emissions control equipment to meet the implementation
dates and emission reductions set forth in the Clear Skies Act.
This study provides an estimate of the resources required for
the installation of control technologies to obtain emission
reductions of sulfur dioxide (SO2), nitrogen oxides (NOx), and
mercury under the Clear Skies Act. More innovative control
technologies and compliance alternatives requiring fewer resources
than those considered for this study are likely to be developed
with the implementation of the Clear Skies Act. Market based
approaches reward firms for finding cost-effective measures that
exceed emission reduction targets. For example, improved scrubber
performance and the ability of some firms to switch to lower sulfur
fuels under the Acid Rain Program were reasons the cost of that
program were less than projected. The development of control
technology alternatives to selective catalytic reduction (SCR)
under the NOX State Implementation Plan (SIP) Call is another
example of how alternative solutions may require fewer resources
than the projected approach. In addition to innovative
technologies, the time allowed for installation of significant
numbers of control technologies is an important factor to consider,
especially for the near future. While it is expected that markets
for the materials and labor used in the construction and operation
of the control technologies will respond to increased demand, this
response will not be instantaneous. It is likely that the strength
of this market response will increase as time progresses. It is
expected that the market would have sufficient time to respond to
phase II of the program as the more stringent emission targets for
phase II are set for 2018. Even though this analysis looks at the
resource availability beyond 2010, these projections are of limited
value as they do not take into account this market response.
However, it is projected that there are sufficient resources
available to complete the projected control technology
installations for phase I by 2010. It should also be noted that
decreasing the amount of time provided to install control
technologies to meet a given strategy has the potential to affect
the cost of compliance as this will accelerate their
installation.
The control technologies considered by this report as candidates
to be used for multipollutant control strategies include: limestone
forced oxidation (LSFO) flue gas desulfurization (FGD) for the
control of SO2, SCR for the control of NOX, and activated carbon
injection (ACI) for the control of mercury.
Installation of LSFO presents a conservatively high estimate of
anticipated resources and time to provide additional control of SO2
emission, since LSFO systems commonly are more resource intensive
than many other FGD technologies. Conservatively high assumptions
were made for the time, labor, reagents, and steel needed to
install FGD systems. For LSFO installation timing, it is expected
that one system requires about 27 months of total effort for
planning, engineering, installation, and startup, with connections
occurring during normally scheduled outages. Multiple retrofits at
one plant would take longer to install (e.g., approximately 36
months for the
retrofit of three absorbers for six boilers). Limestone is the
reagent used in LSFO to remove SO2 from the flue gas stream. Steel
is the major hardware component for FGD systems and is used
primarily for the absorber, ductwork, and supports.
Other elements of FGD installations, such as construction
equipment requirements, are typically modest, particularly given
that systems are installed at the back end of the facility and
close to the ground. More recently, improvements in technology have
been implemented where space requirements were an issue for
construction and accommodating the FGD system, including fewer and
smaller absorbers and more efficient on-site use and treatment of
wastes and byproducts.
SCR is currently the predominant technology to be used for NOX
control and is also the most demanding in terms of resources and
time to install when compared to other NOx control technologies. It
is expected that one SCR system requires about 21 months of total
effort for planning, engineering, installation, and startup.
Multiple SCR systems at one facility would take longer to install
(e.g., approximately 35 months for seven SCRs). Ammonia and urea
are the reagents used along with a catalyst to remove NOX from the
flue gas stream. Experience in installing SCRs for the NOX SIP Call
has shown that the SCR equipment can be installed on the facilities
in the space provided. In some cases, some moving of equipment has
been necessary. One of the primary pieces of specialized
construction equipment that can be useful for SCR installations are
tall, heavy-lift cranes, and these appear to be in adequate
supply.
ACI was presumed to be the technology that would be used to
reduce mercury where dedicated mercury controls were needed.
Planning, engineering, installation, and start up of one ACI system
is only about 15 months. Multiple ACI systems at any one facility
are assumed to take longer to install (e.g., approximately 16
months for two ACI). ACI hardware is comprised of relatively common
mechanical components and is largely made of steel. An ACI system
requires much less in terms of steel, labor, or other resources to
install than either FGD or SCR technology. Therefore, the impact of
ACI hardware on resource demand is much less than that of FGD or
SCR technologies for SO2 or NOX control, respectively.
The resources required for the installation of control
technologies to achieve the emission reductions under the Clear
Skies Act were estimated and compared to their current market
availability. For the Clear Skies Act, control technology
installations have been looked at for the periods between now and
2005, 2005 and 2010, 2010 and 2015, and 2015 and 2020. For the
first period, it is assumed that all controls need to be installed
in a 31-month period. This will provide a conservatively high
estimate of the required resources because many of the necessary
control installations have already begun. For the other five
year-periods, it is conservatively estimated that all installations
will be completed within three years. However, the estimates
indicate that there is ample steel and general construction labor
to support the installation of these technologies over these time
periods. As noted above, projections beyond 2010 are of limited
value as market conditions could change significantly between now
and 2010 in response both to demand for resources for a
multipollutant program and because of other market factors. Skilled
labor requirements, specifically for boilermakers, were estimated
and have the potential to be the more limiting resource requirement
in phase I of the program. The demand for boilermaker labor due to
the NOX SIP Call over the next few years is likely to be limiting,
but through the implementation of the Clear Skies Act, additional
recruiting and training of new boilermakers would create a stronger
market for skilled labor, ultimately increasing the supply.
With regards to reagents and other consumables, it is projected
that there is sufficient supply of limestone for additional FGD
systems. It is estimated that there is also enough SCR catalyst
capacity to supply this market. Ammonia and urea supply is also
plentiful, although it is expected that NOX reduction will cause a
moderate increase in U.S. demand. Bolstered by the fact that there
is currently a worldwide excess capacity problem for suppliers of
these globally traded commodity chemicals, it is projected that
there will be an ample supply of ammonia and urea. U.S. demand for
activated carbon is expected to slightly increase as a result of
the Clear Skies Act. Activated carbon is traded on a global basis
and there is currently substantial excess capacity that can readily
provide for this increase in demand.
`çåÃÉåÃë
Acknowledgements
.........................................................................................................................iv
Executive
Summary..........................................................................................................................v
List of Figures
.................................................................................................................................xi
List of
Tables.................................................................................................................................
xii
List of
Acronyms..........................................................................................................................
xiii
Chapter 1 Background
............................................................................................................1
Chapter 2 SO2 Control Technology Retrofits
.........................................................................3
2.1 System
Hardware................................................................................3
2.2
Reagents..............................................................................................6
2.3 Construction Equipment
.....................................................................7
2.4 Installation
Time.................................................................................7
2.5
Labor.................................................................................................11
2.6 Space
Requirements..........................................................................12
Chapter 3 NOX Control Technology Retrofits
......................................................................15
3.1 System
Hardware..............................................................................15
3.2 Catalyst and
Reagents.......................................................................18
3.3 Construction Equipment
...................................................................19
3.4 Installation
Time...............................................................................20
3.5
Labor.................................................................................................22
3.6 Space
Requirements..........................................................................24
Chapter 4 Mercury Control Technology
Retrofits................................................................26
4.1 System
Hardware..............................................................................26
4.2 Reagent
.............................................................................................29
4.3 Construction Equipment
...................................................................30
4.4 Installation
Time...............................................................................30
4.5
Labor.................................................................................................32
4.6 Space
Requirements..........................................................................34
Chapter 5 Synergies of Combinations of Control Retrofits on a
Single Unit.......................35
5.1 SCR and FGD (Scrubber) Installations
............................................35
5.2 Mercury Control Technology and Scrubber
Installations.................35
5.3 Mercury Control Technology and SCR
Installation.........................36
Chapter 6 System Resource Availability
..............................................................................37
6.1 System
Hardware..............................................................................40
6.2
Labor.................................................................................................41
6.3 Construction Equipment
...................................................................46
6.4
Reagents............................................................................................47
6.5 Creation of Jobs under Clear Skies Act due to Control
Technology
Installations.......................................................................................53
Chapter 7 Conclusions
..........................................................................................................54
Chapter 8 References
..........................................................................................................58
Appendix A Implementation Schedules for Control Technology
Installations ..................... A-1
iáëÃ=çÑ=cáÖìêÉë
2-1. Gas path for coal-fired boiler with FGD
..................................................................................4
2-2. Major components of Wet FGD
..............................................................................................5
3-1. Gas path for coal-fired boiler with SCR, ESP, and
FGD.......................................................15
3-2. SCR installation at 675 MWe AES Somerset Station
............................................................16
3-3. Plate and honeycomb catalyst
................................................................................................17
4-1. Gas path for coal-fired boiler with SCR, ACI, and ESP
........................................................27
4-2. Gas path for coal-fired boiler with SCR, ACI, ESP, and
FF..................................................27
4-3. Simplified schematic of ACI
system......................................................................................28
6-1. U.S. construction employment and unemployment
...............................................................42
6-2. Boilermaker Demand Under Clear Skies Act (32 GWe of FGD
Installations) ......................44
6.3. Boilermaker Demand Under Clear Skies Act (10 GWe of FGD
Installations) ......................45
6.4. Cumulative SCR Catalyst Demand Compared To Cumulative
Production Capacity............50
6-5. SCR installations on coal-fired plants in
Germany................................................................51
iáëÃ=çÑ=q~ÄäÉë
2-1. Estimated Resources Needed for Single and Multiple FGD
Retrofits...................................13
3-1. Estimated Resources Needed for Single and Multiple SCR
Retrofits ...................................24
4-1. Estimated Steel Requirement for 500 MWe ACI
System.......................................................29
4-2. Estimated AC Injection Rates for a 500 MWe Boiler
............................................................30
4-3. Estimated Man-hours for Supply of an ACI System for a
500 MWe 0.6 % Bituminous
Coal Boiler Coal with ESP (Example 1 from Table
4-4)33...................................................32
4-4. Estimated Performance and Resources Needed for Single
ACI Retrofit ...............................33
4-5. Estimated Performance and Resources Needed for Single
and Multiple ACI Retrofits........34
6-1. a) FGD Retrofits, b) SCR Retrofits, c) ACI
Retrofits............................................................38
6-2. Estimated Steel Required for Multipollutant Initiative
..........................................................40
6-3. Estimated Annual Construction and Boilermaker Labor
Required for Clear Skies Act........41
6-4. Estimated Annual Boilermaker Demand Created by the Clear
Skies Act .............................46
6-5. Crushed Limestone Sold or Used By U.S.
Producers............................................................47
6-6. Estimated FGD Limestone Consumption and U.S. Production
.............................................48
6-7. SCR Catalyst Capacity for Coal-fired Boilers
.......................................................................48
6-8. Estimated Annual SCR Catalyst Demand Resulting from
Clear Skies Act
and NOX SIP Call
.................................................................................................................49
6-9. Projected AC Demand Due to Multipollutant Initiative
.......................................................52
7-1. Estimated Resources Needed for Installation and
Operation of Technologies ......................57
iáëÃ=çÑ=^Åêçåóãë
AC Activated carbon
ACI Activated carbon injection
CAAA Clean Air Act Amendments
DCS Distributed control system
ESP Electrostatic precipitator
FF Fabric filter
FGD Flue gas desulfurization
GWe Gigawatt (electric)
IPM Integrated Planning Model
LSD Lime spray dryer
LSFO Limestone forced oxidation
MEL Magnesium enhanced lime
MWe Megawatt (electric)
NAAQS National Ambient Air Quality Standards
PJFF Pulsejet fabric filter
PLC Programmable logic controller
SCR Selective catalytic reduction
SIP State Implementation Plan
TVA Tennessee Valley Authority
`Ãœ~éÃÉê=N _~ÅâÖêçìåÇ
In response to continuing concerns about emissions from electric
generating units, further reductions of emissions of multiple
pollutants from electric power sector are being considered. Because
the largest portion of emission reductions are expected to come
from the coal-fired electricity-generating segment of the electric
power sector, this report considers environmental improvement for
coal-fired electricity generating power plants. Strategies enabling
the control of multiple pollutants (multipollutant control
strategies) from these plants have recently been receiving
increased attention.
Currently, power plants are required to reduce emissions of
nitrogen oxides (NOX) and sulfur dioxide (SO2). The revisions of
the National Ambient Air Quality Standards (NAAQS) aimed at
reducing haze may require electric utility sources to adopt
additional control measures. In addition, the U.S. Environmental
Protection Agency (EPA) has determined that the regulation of
mercury emissions from coal-fired power plants is appropriate and
necessary. Concurrently, legislation has been proposed in previous
and current Congresses that would require simultaneous reductions
of multiple emissions, and the Administration's National Energy
Policy recommends the establishment of "mandatory reduction targets
for emissions of three main pollutants: sulfur dioxide, nitrogen
oxides, and mercury."
The administration's multipollutant proposal, a far reaching
effort to decrease power plant emissions, was introduced as the
Clear Skies Act in the U.S. House of Representatives on July 26,
2002 and in the U.S. Senate on July 28, 2002. This legislation is
intended to reduce air pollution from electricity generators and
improve air quality throughout the country. The Clear Skies Act is
designed to decrease air pollution by 70 percent through an
emission cap-and-trade program, using a proven, market-based
approach that could save consumers millions of dollars. The Clear
Skies Act calls for:
â– Decreasing SO2 emissions by 73 percent, from current emissions
of 11 million tons to a cap of 4.5
million tons in 2010, and 3 million tons in 2018, â– Decreasing
NOx emissions by 67 percent, from current emissions of 5 million
tons to a cap of 2.1
million tons in 2008, and to 1.7 million tons in 2018, and â–
Decreasing mercury emissions by 69 percent by implementing the
first-ever national cap on mercury
emissions. Emissions will be cut from current emissions of 48
tons to a cap of 26 tons in 2010, and
15 tons in 2018.
Therefore, it is timely to review the engineering and resource
requirements of installing control technologies for multipollutant
control strategies.
This report analyzes the resources required for installing and
operating retrofit control technologies for achieving reductions in
multiple pollutants from coal-fired power plants in the United
States. It examines the control technology's hardware, reagents,
availability of the needed construction equipment, time required to
implement at plants with single and multiple installation
requirements, and the availability of labor needed for
installation. The control technologies considered in this report
include limestone forced oxidation (LSFO) wet flue gas
desulfurization (FGD), selective catalytic reduction (SCR), and
activated carbon injection (ACI) for the control of SO2, NOX, and
mercury, respectively.
The report is organized into eight chapters and one appendix.
Chapter 1 provides general background information on emission
control technologies. Chapter 2 analyzes the SO2 control technology
resource
requirements by providing information on control technology's
hardware and reagents, the construction equipment necessary to
install a control technology, time required to implement this
control technology at plants with single and multiple installation
requirements, and the amount of labor needed to install the control
technology. Chapters 3 and 4 review, in the same fashion, the
resource requirements of installing NOX and mercury control
technology, respectively. Chapter 5 focuses on synergistic
combinations of control retrofits on a single unit. Chapter 6
examines the availability of resources necessary for the
installation of SO2, NOX, and mercury control retrofit technologies
for the timing and emission reductions proposed under the Clear
Skies Act. Conclusions are presented in Chapter 7 and references in
Chapter 8. Appendix A is located at the end of this report. It
provides implementation schedules for single and multiple control
technology installations.
`Ãœ~éÃÉê=O plO=`çåÃêçä=qÉÅÜåçäçÖó=oÉÃêçÑáÃë
In this chapter, the resource requirements to retrofit FGD
systems to remove SO2 are examined for existing coal-fired electric
utility boilers. The FGD technology most commonly installed in the
U.S. and worldwide is LSFO. Material, labor, and construction
equipment resource estimates presented in this chapter are for LSFO
systems and are a conservative estimate compared to less resource
intensive magnesium enhanced lime (MEL) or lime spray dryer (LSD)
technologies. Typically, MEL and LSD technologies rely upon
increased reactivity of reagents with flue gas and require fewer
resources for installation. Advances in FGD technology, design,
materials, and expertise available for retrofit installations made
over the last decade form a sharp contrast to earlier retrofit
systems. Technology to remove SO2 is anticipated to continue along
current trends and rely heavily on wet FGD and other advanced
technologies. This chapter examines the experience and issues for
the retrofit installation of LSFO technology.
The chapter focuses on the resources needed for typical or
normally constrained wet FGD, specifically LSFO, retrofit
installations. Wet FGD retrofit technology generally provides a
conservatively high estimate of most resources. However, it is
likely that other SO2 removal technologies, as well as upgrades or
enhancements to existing FGD systems, will compete in the market
under a multipollutant strategy. Upgrades to existing FGD systems
would include a case-by-case examination of the absorber tower,
flue gas inlet, absorber gas velocity, reagent preparation, upgrade
pumps, and potential changes to some internals, the type of
reagent, and to the chemical processes to increase performance.
Scrubber upgrades were not considered in this analysis in order to
provide a conservative estimate of the resource demand for a
multipollutant strategy. This is because upgrades to existing
retrofits will generally consume fewer resources than full
retrofits regardless of the technology.
OKN póëÃÉã=e~êÇï~êÉ
The wet FGD process operates by reacting SO2 in the flue gas
with a reagent in an absorber. FGD systems are typically positioned
after the particulate control device. FGD retrofits are positioned
downstream, typically at the back end of the facility, and are not
intrusive to the boiler. Typical configuration is shown in Figure
2-1. At the typical unit, hot corrosive flue gases leaving the
particulate control device (149- 182 ºC) are cooled, or quenched,
before entering the main absorber device. Quenching cools and
saturates the flue gas with absorber slurry. Quenching can occur in
a prescrub area or more commonly an area integral to the absorber.
After quenching, the less corrosive flue gases entering the
absorber decrease to temperatures of 49 - 66 ºC with pH values
between 5 and 6.5. Some higher efficiency boilers may have
increased flue gas velocities and can result in corrosive flue gas
"blow through" to the absorber. The chemical reactions that occur
with the limestone reagent form a corrosive environment requiring
many of the system components to be corrosion and abrasion
resistant. The quenching area is typically a highly corrosive
environment and the reagent slurries are highly abrasive. The
handling and processing of the reagent, commonly limestone, is
often done onsite, as is the treatment of the effluent as waste or
processing into a saleable product (e.g., gypsum handling
facility).
FGD
Stack
Disposal
Figure 2-1. Gas path for coal-fired boiler with FGD.
The major systems and components of a wet FGD limestone reagent
system include:
Reagent Feed
â– Limestone conveying and crushing (e.g.,
ball mills) equipment
â– Slurry preparation tank and reagent feed equipment
SO
2
Removal
â– Absorber tower or reactor (tray optional)
â– Absorber slurry recirculation/reaction tank and optional
air sparger (forced oxidation operation)
â– Mist eliminator wash system
â– Slurry bleed
â– Pipes, pumps, and valves
â– Gas reheaters
Flue Gas Handling
â– Ductwork
â– Support steel
â– Fans, blowers, and dampers
Waste/By-Product
â– Dewatering system (settling tanks/vessels,
hydrocyclones, and/or vacuum filters)
â– Stacking equipment
Major wet FGD system components are shown in Figure 2-2. The
hardware and equipment to support wet FGD technology involves five
major systems. Two systems are primarily responsible for the direct
scrubbing and handling of flue gases, and three systems are
involved with delivery of reagents, processing of wastes (air,
solids, and water), and the processing of wastes into saleable
by-products. Typically, the greatest hardware requirements involve
the systems for SO2 removal, primarily the absorber vessel, and
flue gas handling, particularly ductwork and support steel. Much of
the equipment for these systems will be unique to the site and
project requirements, although the equipment
specifications may be repeated if multiple absorbers are
involved. Typically an FGD retrofit can use an existing chimney or
stack.
The material used in the largest quantity for an FGD retrofit,
aside from reagent, is steel. In general, the amount of structural
steel for a typical FGD retrofit system is equal to or less than
the steel requirements for a SCR retrofit of the same size.1 The
reduced requirement for structural steel is due to the FGD absorber
usually being self-supporting, weighing less, and being installed
closer to the ground. In contrast, a typical SCR installation is
heavier, elevated, and adjacent to the boiler. The majority of
structural steel in modern FGD installations is dedicated to
ductwork and supports. Other steel may be needed to reinforce
existing steel at a facility. In addition to structural steel,
additional light, or gallery, steel may be used in the limestone
preparation area and for the processing of waste or byproducts
(e.g., buildings). Modern FGD systems are more attuned to the
corrosive SO2 scrubbing environment and therefore increasingly
utilize fiberglass, rubber lined steel, and alloys in construction.
In addition, because of existence of corrosive zones, preference is
often given to the use of large-sized sheets that minimize
welding.2 Large-sized sheets are used to fabricate the absorber
vessel, the ductwork, and supports. Particularly over the past
decade, there has been greater availability of plate steel for FGD
projects due to the global sourcing of carbon steel.
Total steel requirements for retrofitting a typical 500 mega
Watt, electric (MWe) FGD system are in the range of 1000 to 1125
tons of steel, or between 2.0 and 2.25 tons of steel per MWe. This
range assumes approximately 80 percent of the structural steel is
for ductwork and supports and 20 percent is required for
miscellaneous steel such as reagent conveying equipment, buildings,
and solids handling systems. An assumption of 1125 tons of steel is
a conservatively high estimate since 500 MWe FGD retrofit
installations have been completed with as little as 250 to 375 tons
of steel, or 0.5 tons of steel per MWe..1,3 Often a single absorber
will serve multiple boilers and reduce much of the steel that would
be required if absorbers had been fed by individual boilers.
Currently, the installed maximum single absorber capacity in the
U.S. is 890 MWe being fed by 2 boilers at Tampa Electric's Big Bend
Station. It is likely that two 450 MWe boiler units will use a
common single absorber with commensurate reduction in required
steel from efficiencies gained by common areas. For example, a 900
MWe system is estimated to use approximately 2000 tons of steel, or
about 2.2 tons of steel per MWe, rather than a combined maximum of
2025 tons for a situation in which the boilers required separate
absorbers.
In general, a better understanding of the chemical processes in
an FGD system allows designers to maximize mass transfer in a
minimum amount of space.4 More efficient designs can also reduce
the amount of steel needed for the absorber and ductwork. Other
advanced design factors reduce steel requirements, including the
virtual elimination of redundant absorbers, the ability to
down-size absorbers without sacrificing performance (e.g., by
increasing flue gas velocity, advanced placement of spray nozzles,
enhancing limestone characteristics), and material changes
including wallpapering with alloys and utilizing fiberglass.5 A new
generation of wet FGD systems, pioneered in the mid to late 1990's,
improved mass transfer, which resulted in the usage of more compact
absorbers that are estimated to require 50 percent less material,
compared to an older generation of wet FGD systems.5 In addition,
typical MEL absorber units need less steel due to the use of
smaller absorbers enabled by shorter residence time requirements
than for LFSO systems.==In this report, the estimate of typical
LFSO FGD system hardware requirements provides a conservatively
high estimate of installation resources compared to other SO2
control technologies.
The majority of hardware required for FGD systems is commonly
available. Storage tanks, nozzles, and piping for the reagent
storage and delivery system are also common and therefore widely
available. The major hardware for an FGD system includes the flue
gas duct system, limestone storage (including loading and conveyer
equipment), gypsum dewatering and wastewater treatment, gypsum
storage, piping, valves, pumps and tanks, electricity supply,
controls, instrumentation, pipe bridges and cable channels, and
foundations and buildings as needed. FGD systems also include
hardware such as inlet fans, dampers, absorber internals,
recirculation pumps, and oxidation blowers that are commonly used
in other large industries. Because this hardware is used
extensively throughout industry, availability should not be an
issue, except that supply of this type of equipment needs to be
integrated into the overall project schedule so it does not cause
bottlenecks. Early FGD systems were designed with separate
quenching, or prescrubber, systems to cool the flue gas coming off
the particulate control device. Modern systems take the hot flue
gas directly into the absorber, where quenching occurs. Limestone
and gypsum handling also includes milling, conveying, and
wastewater treatment systems.
OKO oÉ~ÖÉåÃë
Wet FGD systems require a continuous feed of reagent to remove
SO2. Generally, if pH of scrubbing liquor falls below a range of
5.0 to 6.0, additional reagent is required to maintain the
reactivity of the absorbent. Limestone is the most commonly used
reagent, with the quantity of its consumption depending primarily
on coal sulfur levels. For example, even in the range of 3 percent
to 4 percent coal
sulfur levels, a 4 percent sulfur coal can increase consumption
by about one-third.6 It is not uncommon for modern FGD systems to
achieve 97 percent utilization of limestone.7 Generally, higher
percent utilization equates to higher reactivity of the reagent,
and, therefore, less reagent is needed to achieve a given level of
SO2 removal. The production of gypsum requires a minimum of 92
percent limestone utilization.
A 500 MWe plant uses roughly 25-32 tons per hour of limestone. A
coal with 4 percent sulfur, conservatively, will require 32 tons*
of limestone per hour, or 0.064 tons per MWe per hour. This
estimate is based on an 85 percent load factor, 10,500 Btu/kWh,
stoichiometry of 1.1 for limestone, SO2 removal rate of 95 percent,
and a minimum purity of 95 percent for CaCO3. Any enhancement of
the reagent (i.e., magnesium or buffering agents) will reduce the
amount of limestone needed. MEL sorbent consumption for a 4 percent
sulfur coal is approximately 17-18 tons per hour.6 These estimates
are conservatively high given that recent FGD systems are operating
at near stoichiometric levels8 and additives are commonly used to
achieve higher SO2 removal, particularly to enhance the performance
of existing retrofits. Limestone stoichiometry is the number of
moles of Ca added per mole of SO2 removed. Typically the required
stoichiometry of a wet FGD limestone system is in the range of 1.01
to 1.1, with 1.01 to 1.05 typical for modern wet FGD systems.8 A
stoichiometry of 1.03 is typical when the FGD process is producing
gypsum by-product, while a stoichiometry of 1.05 is needed to
produce waste suitable for a landfill. Grinding limestone to an
ultrafine particle size improves dissolution rate of limestone in
the slurry and can decrease the size requirement of the reaction
tank.8
OKP `çåëÃêìÅÃáçå=bèìáéãÉåÃ
The construction equipment required for typical FGD
installations is standard construction equipment - welders,
excavation equipment, concrete pouring equipment, cranes, etc.
Crane requirements for FGD technology retrofits are generally site
specific, although these requirements are generally less demanding
than requirements for SCR retrofits. Generally, FGD systems tend to
be constructed closer to the ground compared to SCR technology
retrofits. Lift at a site rarely exceeds 30 meters (100 feet) and
100 tons. The use of modular and fabricated absorbers shifts much
of the construction off-site, reducing the need for specialized
cranes and equipment. The usefulness and appropriateness of using
cranes in an FGD installation is dependent on several factors,
including the ability to physically place a crane on site or
adjacent to the site (e.g., on a barge) and the use of modular
construction of major FGD technology components. Prefabrication has
been used since the early 1990's, notably on two large retrofit
projects: the 1300 MWe Zimmer Station and the 2600 MWe Gavin
station. Both facilities had limited lay-down area to perform the
retrofit installation. In these two retrofits, the absorber modules
were fabricated in two pieces, shipped by barge, and assembled on
site. Often modular units can be transported via barges or trucks
to the site for assembly. Component modularization and
prefabrication off-site can reduce the amount of time cranes are
needed on a site, as well as provide opportunities to reduce
project schedules and construction costs and to concentrate jobs
locally at the prefabrication facility.
OKQ fåëÃ~ää~Ãáçå=qáãÉ
Implementation of an SO2 control technology at a plant involves
several activities contingent upon each other. These activities may
be grouped under the following phases of an implementation project:
(1)
* EPA cost modeling for wet scrubber installations estimate 29
tons of limestone required to achieve 95 percent SO2 removal while
burning a 4 percent sulfur coal.
conducting an engineering review of the facility and awarding a
procurement contract; (2) obtaining a construction permit; (3)
installing the control technology; and (4) obtaining an operating
permit.
Modular construction can minimize field labor and construction
time on a site by prefabricating at a shop and then transporting
large sections, such as ductwork or absorber modules, by barge or
truck. For example, the 550 MWe boiler at Kansas City Power &
Light Company's Hawthorne Generating Station required rebuild and
NOX, SO2, and PM simultaneous control retrofits. To expedite
placing the facility back into use, large sections were fabricated
off site and transported by barge to the site. Shop fabrication has
also been used outside of the U.S. For example, ten absorber
modules handling a combined capacities of 2000 MWe and 3000 MWe at
two facilities were installed during December 1995 (order placed)
through March 2000. The ten absorbers were mostly installed
sequentially with startup of the units staggered over 22 months.
The 2000 MWe FGD systems at the Taean facility in South Korea were
fabricated off-site in three modules, shipped by barge, and then
assembled on-site.
FGD installation plans and experience have been extensive in the
U.S. and abroad. To date, there have been approximately 94 GWe of
scrubber capacity built on coal-fired power plants in the US. Over
200 GWe of capacity has been built worldwide.
Exhibits A-1 and A-2 in Appendix A depict the timelines typical
to complete a single absorber module and a three absorber-module
installation of FGD, respectively. The three absorber-module
installation assumes each absorber module can treat up to 900 MWe
of boiler capacity. Currently, approximately 900 MWe of single
absorber capacity has been successfully installed in the U.S.
However, greater absorber capacities are being offered outside of
the U.S.9 While the sum of the time estimated to complete
individual tasks generally exceeds the overall estimated
installation time, the overall installation schedule accounts for
overlap in these tasks. These timelines also indicate that
completion of some of the activities is contingent upon completion
of some other activities. In general, the FGD implementation
schedule appears to be driven primarily by the pre-hookup
construction activities. Multiple absorber installations will
typically add a few months to the implementation schedule,
particularly to connect additional absorbers during scheduled
outages. Prefabrication of absorber modules can reduce the overall
construction schedule. The major phases of the implementation
schedule are discussed below.
båÖáåÉÉêáåÖ=oÉîáÉï
As shown in Exhibits A-1 and A-2 in Appendix A, in the first
phase of technology implementation, an engineering review and
assessment of the combustion unit is conducted to determine the
preferred compliance alternative. During this phase, the
specifications of the control technology are determined, and bids
are requested from the vendors. After negotiating the bids, a
contract for implementing the SO2 control technology is awarded.
The time necessary to complete this phase is approximately four
months.
`çåëÃêìÅÃáçå=mÉêãáÃ
Before the actual construction to install the technology can
commence, the facility must receive a construction permit from the
applicable State or local regulatory authority. The construction
permit process requires that the facility prepare and submit the
permit application to the applicable State or local regulatory
agency. The State or local regulatory agency then reviews the
application and issues a draft approval. This review and approval
process is estimated to take about six months. The draft
construction permit is then made available for public comment.
After any necessary revisions, a final construction permit is
issued. The estimated time to obtain the construction permit is
approximately=nine months10 but can vary with State and local
permitting procedures as well as other interests in the
project.
`çåÃêçä=qÉÅÜåçäçÖó=fåëÃ~ää~Ãáçå
In the second phase, the control technology is installed. This
installation includes designing, fabricating, and installing the
control technology. In addition, compliance testing of the control
technology is completed in this phase. Since FGD technology is not
invasive to the boiler, most of the construction activities, such
as earthwork, foundations, process electrical, and control tie-ins
to existing items, can occur while the boiler is in operation. The
time needed to complete this phase of an implementation project is
about 23 months.
An important element of the overall control technology
implementation is the time needed to connect, or hook up, the
control technology equipment, particularly in relationship to the
planned outage time for the unit. On average, it takes about four
to seven weeks to connect FGD.1, 3 For example, the Homer City and
Centralia facility FGD retrofit connections were performed during
the scheduled outages in approximately five weeks. Based on
experience in Germany in response to a compliance directive, a
significant quantity of SO2 and NOX control installations were
performed within outage periods consisting of less than four
weeks.11 Electricity generating facilities often plan the
connection to occur during planned outages to avoid additional
downtime. Additional downtime leading to loss of a unit's
availability to supply electricity is atypical for FGD technology
installations.3,12 Because peak electricity demand generally occurs
during the summer months (May through September), typically control
connections occur during months of other seasons, notably the
spring or fall.10 For example, FGD connections to the two Centralia
units were performed outside of the peak electricity demand period.
Sources located where peak demand does not occur during the summer
months may be less time-constrained to connect the FGD controls.
However, FGD connections for single and multiple systems can
typically be performed during planned outage times. Multiple
systems normally are installed in sequence and overlapping to
maintain a high level of activity at the site. Installation of the
control device hookup on a sequential basis usually involves an
overlap of compliance testing of FGD system on one unit with hookup
of an FGD system with the next unit. The total implementation time
for sequential hookup for multiple systems is estimated at between
32 months for two absorber modules and 36 months for three absorber
modules. Although not as common, multiple systems installed at a
single facility can be performed simultaneously. Generally,
scheduled outages will govern which method can be used for multiple
FGD system installations.
léÉê~ÃáåÖ=mÉêãáÃ
Facilities will also need to modify their Title V operating
permit to incorporate the added control devices and the associated
reduced emission limits. In some States, an interim air-operating
permit may need to be obtained until the Title V permit is
modified. The operating permit modification process consists of
preparation and submission of the application to the applicable
State or local regulatory agency. This process can occur
simultaneously with the processing of the construction permit
application.10 The process of transitioning from the construction
permit to the operating permit varies among States, but the
application review process is estimated to take between 9 - 11
months. The Title V operating permit must also be made available
for public comment. The Title V operating permit is then not made
final until compliance testing on the control device is completed.
Therefore, the total estimated time to modify the Title V operating
permit is about 17 months, plus the additional time to complete
compliance testing.
Based on the estimated time periods needed to complete each of
the four phases described above, the estimated time period to
complete the implementation of a single FGD installation is
conservatively 27 months. For the Clear Skies Act, EPA's
projections reflect that the majority of FGD installations will
involve a single absorber unit installation per plant; however, the
maximum projected number of scrubbers retrofitted at any facility
is three absorber modules serving six boilers with a maximum of
2400
MWe of boiler capacity. Changes in FGD technology and
reliability have resulted in planning for smaller and fewer
absorbers per retrofit installation. For example, the Zimmer and
Gavin station FGD retrofits performed in the early 1990's both
involve three absorbers on each 1300 MWe unit. If these retrofits
were being planned today, these stations' two 1300 MWe units would
likely require only two absorbers per unit rather than three.6
Average FGD installation times have commonly been within 24-27
months. For example, recent FGD retrofit systems installed at Homer
City (September 2001) and Centralia (July 2001) were both completed
within approximately 24 months.13 Although these FGD system
installations are considered typical, the Homer City FGD retrofit
installation was performed during the same time frame as the
installation of three SCR units. Both of these units provide more
recent insight into the ability and scheduling to install FGD
systems during a period of high demand for SCR installations.
One factor that can increase the time to install a scrubber is
competition for resources with other emission control projects.
During the first time period analyzed (through the end of 2005),
EPA projects that a large number of SCR's will be installed to meet
the requirements of the NOX SIP Call. However, SCR installations
designed to comply with the NOX SIP Call are generally already into
the installation process or, at a minimum, into the engineering
phase of the project.13 Furthermore, construction has already begun
or been completed for 4 GWe of the scrubbers that EPA projects will
be built by 2005 under current regulatory requirements. Typically
the overall engineering, fabrication, and construction resources
would remain the same as the scenario analyzed above, with the
exception that these resources are reallocated over an extended
schedule. One estimate is that, as demand for installation
resources increase for FGD and other air pollution control
installations, planned FGD retrofit installations could be between
30 and 42 months1 while another source estimates FGD installations
at 36 months.14 It should be noted, however, that some recent
contracts have been signed to install scrubbers between now and
2005 that would be installed in less than 36 months. For instance,
a contract to install a scrubber on a 500 MWe unit at the Coleman
Station in Kentucky is scheduled to be completed in early 2004
(approximately 24 months after the contract was announced, which is
several months shorter than the installation schedule set forth in
Exhibit A-1). This suggests that labor demands to install SCR's for
the NOX SIP Call may not lead to increased installation time for
scrubbers.
Single-unit FGD installations have occurred in as little as
20-21 months9, and multiple FGD systems have been installed within
36 months. In addition, owners of new, or "greenfield," power
generation facilities often request 24 months for completion of
these projects, including installation of the boiler, FGD system,
and SCR. Primarily as a cost cutting option, more relaxed
installation schedules of up to 36 months for a single FGD retrofit
installation may be planned, but are not common. Despite changes in
overall installation schedules, efficient utilization of labor and
sequencing the installation during planned outages will continue to
be planning issues. In summary, the total time needed to complete
the design, installation, and testing at a typical 500 MWe facility
with one FGD unit is 27 months, 32 months at a facility with two
boilers being served by a single absorber module, and approximately
36 months at a facility with three absorber modules (six boiler
units). For the multiple installation of three absorber modules at
one plant (six boiler units), an additional four months may be
needed to schedule the outage for the FGD hookup outside of the
high electricity demand months. Typically, multiple absorbers will
be installed sequentially with some overlap to conserve and
schedule continuous use of labor, as well as keep associated
installation costs down.
OKR i~Äçê
The installation of an FGD system requires a significant amount
of labor. Approximately 80 percent of the labor is for construction
of the system, and 20 -25 percent of the labor is for engineering
and project management. The installation of the FGD control
technologies may require the following types of labor:
â– general construction workers for site preparation and storage
facility installation;
â– skilled metal workers for specialized hardware and/or other
material assembly and construction;
â– other skilled workers such as electricians, pipe fitters,
millwrights, painters, and truck drivers; and
â– unskilled labor to assist with hauling of materials and
cleanup.
A typical turnkey 500 MWe unit FGD system retrofit requires
380,000=man-hours, or approximately 200 person-years, of which 20
percent, or 72,000 man-hours, are dedicated to engineering and
project management,3 and roughly 40 percent of man-hours are for
boilermakers.14 The labor required to install an absorber vessel
and ductwork is a major portion of the system installation
man-hours. Generally, construction labor is proportional to the
amount of steel used in the system. The greatest labor requirement
occurs for FGD on a single unit (i.e., 500 MWe), and additional
efficiencies in incremental labor occur when scheduling multiple
units at one facility, particularly when combining multiple boilers
into a single absorber. In general, large numbers of boilermakers
have been used in this industry; however, it is not expected that
this demand will impact other industries. A more thorough
discussion of boilermaker labor demand is given in Chapter 6.
There are some efficiencies that result when multiple systems
are installed at one site. In engineering alone, there is a 10-15
percent savings in engineering and project management labor
commonly realized when installing multiple units of similar design.
In addition, other increases in project management and labor
productivity and efficiencies in using resources and equipment can
occur with multiple system installations on one site. While
multiple systems on one site are common, the number of required
systems to serve large MWe of capacity has been decreasing. For
example, in the past FGD systems for 2,600 MWe stations included
six absorbers; however, today these systems would likely be
designed for four absorber systems, or approximately 650 MWe of
boiler capacity per absorber.6 Using the methodology described for
900 MWe of capacity, today a six-absorber system could serve as
much as 5,400 MWe of capacity, or more than double the capacity
served in installations in the early 1990's.
A reasonable estimate of multiple FGD installations at one site
includes 380,000 man-hours for the initial 500 MWe of capacity (or
760 man-hours/MWe) and an additional 500 man-hours per MWe, up to a
total of 900 MWe, for any combination into a single common
absorber. Therefore, a 900 MWe system requires the initial 500 MWe
at 380,000 man-hours, and the second 400 MWe at about 273,200, for
a combined labor requirement of 653,200 man-hours, or approximately
300 person-years, or the equivalent of about 725 man-hours per MWe.
As another example, a 1400 MWe system retrofit using 2 (700 MWe)
turnkey systems requires 700,000 man-hours, or only 500 man-hours
per MWe.6 Generally, extending FGD installation schedules may
reduce the number of persons on a job at one time but will not
reduce the overall labor requirement.
While, it is likely that installation of multiple systems will
benefit from economies of scale to reduce labor requirements, the
range for man-hours per MWe for multiple systems is bounded by 500
and 725 man-hours per MWe. It is also clear that boiler capacities
of at least 900 MWe can be served by a common absorber, and a
minimum 10 percent reduction in engineering and project management
labor will result from multiple absorbers being installed at a
single site. For example, procurement contracts only need to be
negotiated once, and common site issues need only be addressed
once. Therefore,
653,200 man-hours is a conservative estimate of labor required
to install FGD at a 900 MWe facility. Because no additional
efficiencies in engineering and project management are assumed for
larger installations, multiple 900 MWe absorber systems each add
another 653,200 man-hours. A 2,700 MWe facility requires
approximately 1,960,000 man-hours for the retrofit installation.
This produces a conservatively high estimate bounding the
uncertainties of labor and how many boilers or units will be
combined into a single absorber.
The above estimates of labor are conservative particularly given
efficiencies realized in recent retrofit installations. In many
cases, portions of retrofit construction can be performed off-site,
particularly with modular designs. For example, at the Gavin
scrubber retrofit installation the absorber modules were fabricated
at the vendor's shop and then shipped by barge to the site for
hookup. To take advantage of off-site fabrication requires that a
shop and facility are available and collocated with an adequate
shipping and transport facilities (i.e. water accessible facility
and a barge). When the requirements are met the fabrication has the
potential to employ and retain a skilled work force as well as
opportunity to save time and reduce field labor requirements.
Based upon the discussion from sections 2.1 through 2.5, the
total resources needed for a single 500 MWe FGD retrofit and
multiple FGD retrofits are shown in Table 2-1.
OKS pé~ÅÉ=oÉèìáêÉãÉåÃë
Generally 1-acre on-site will allow the installation of an FGD
retrofit.1 The need for additional space for support systems ranges
from no additional space needed to 2.5 acres typical for up-front
reagent processing and 1 acre for dewatering when reagent
processing and dewatering operations are selected as part of the
FGD system design. Space issues also include the positioning of the
FGD after the particulate control device and before the stack. This
area of the power unit is generally referred to as the "back
end"
- an area where there is typically ample space for retrofit
installations.
The FGD retrofit on the Cinergy's Gibson Unit 4 is an example of
an extremely space limited retrofit.1 Gibson Unit 4 is a 668 MWe
inhibited oxidation limestone FGD retrofit designed for 92 percent
SO2 removal and completed in late 1994. In the case of this
scrubber retrofit, the congestion at the site did not allow for a
clean pick by a standard sized crane. With up-front planning, one
module was raised by less conventional means (jacking
construction), allowing for the second module to be constructed
using more conventional methods. Because of the difficulty due to
congestion of the site, this retrofit required additional time and
labor, but worked within space constraints. This method of jacking
construction has been used in other retrofits. The wet FGD retrofit
at the Bailly Generating Station, Units 7 & 8, is an example
where a full service system (single limestone absorber for combined
528 MWe capacity, 2-4.5 percent sulfur coal, >95 percent SO2
removal) was able to significantly reduce space requirements while
also decreasing cost by about one-half and creating no new waste
streams. Much of the success of this public/private project
(DOE/operator and vendors) was due to a more compact and
multi-functional (pre-quenching, absorption, and oxidation)
absorber vessel that used a co-current flow design. As a result,
the FGD system required only modest space requirements.15 In most
locations connection space is not a problem since there is usually
adequate space behind the flue gas stack to perform the scrubber
retrofit. If connection space is limited, additional ductwork may
be necessary.
q~ÄäÉ=OJNK==bëÃáã~ÃÉÇ=oÉëçìêÅÉë=kÉÉÇÉÇ=Ñçê=páåÖäÉ=~åÇ=jìäÃáéäÉ=cda=oÉÃêçÑáÃëK
= råáÃë låÉ=_çáäÉê qïç=_çáäÉêë páñ=_çáäÉêë
`~é~ÅáÃó=oÉÃêçÑáà kìãÄÉê=çÑ=^ÄëçêÄÉê=jçÇìäÉë eÉ~Ã=o~ÃÉ
`~é~ÅáÃó=c~Ã…Ãçê
mÉêÅÉåÃ=oÉÇìÅÃáçå
Ä
råÅçåÃêçääÉÇ=plO= Ã…
råÅçåÃêçääÉÇ=plO= Ä
`çåÃêçääÉÇ=plO= Ã…
`çåÃêçääÉÇ=plO=
plO=oÉÇìÅÉÇ=éÉê=óÉ~ê=Ä
plO=oÉÇìÅÉÇ=éÉê=óÉ~ê=Å
pÃÉÉä
=
i~Äçê
====mêçà ÉÅÃ=aìê~Ãáçå
====båÖáåÉÉêáåÖLmêçà ÉÅÃ=jÖÃK
====`çåëÃêìÅÃáçå
====qçÃ~ä=i~Äçê
====qçÃ~ä=i~Äçê
iáãÉëÃçåÉ=`çåëìãéÃáçåÄ
iáãÉëÃçåÉ=`çåëìãéÃáçåÅ
jbi=iáãÉëÃçåÉ=bèìáî~äÉåÃÃ…
~
Ãçåë NINOR OIMMM SIMMM
ÃçåëLóê NORIORM OORIQRM STSIPRM ÃçåëLóê ORRIMMM QRVIMMM
NIPTTIMMM ÃçåëLóê NQOIRSM ORSISMU TSVIUOQ
~
==råÇÉê=ÃÜÉ=`äÉ~ê=pâáÉë=^Ã…ÃI=bm^=ãçÇÉäáåÖ=éêçà ÉÅÃë=ÃÃœ~Ã=ÃÜêÉÉ=éä~Ã¥Ãë=ïáÃÃœ=ëáñ=ÄçáäÉêë=É~Ã…Ãœ=ïáää=ÄÉ=êÉÃêçÑáÃÃÉÇ=ïáÃÃœ=cdaK==qÜÉ
ã~ñáãìã=Ã…~é~ÅáÃó=~Ã=~åó=çåÉ=çÑ=ÃÜÉ=ÃÜêÉÉ=éä~Ã¥Ãë=ï~ë=OIQMM=jtÉ
Ä==P=éÉêÅÉåÃ=ëìäÑìê=Åç~ä
Ã…==Q=éÉêÅÉåÃ=ëìäÑìê=Åç~ä
Ç==j~ó=êÉèìáêÉ=~Ã¥=~ÇÇáÃáçå~ä=Ñçìê=ãçåÃÜë=Ãç=~îçáÇ=ëÅÜÉÇìäáåÖ=çìÃ~ÖÉë=Ñçê=ÅçåÃêçä=Üççâìé=ÇìêáåÖ=éÉ~â=ÉäÉÅÃêáÅáÃó=ÇÉã~åÇ=ãçåÃÜëK
Often absorbers can be designed to accommodate site-specific
requirements. Smaller absorbers, use of common absorbers for
multiple boilers, and technology advances that supplant the need
for redundant absorbers, have decreased the footprint needed for a
modern FGD retrofit installation. Where space for the FGD
installation is an issue, reducing the overall absorber size can be
accomplished by using multiple absorber trays (within one absorber)
and improving mass transfer with the use of a fan.16 Improved
absorption at higher velocities has contributed to smaller, more
compact absorbers. For example, designers are continually improving
absorber efficiencies by increasing absorber gas velocities in the
range of 5 m/s (15 feet/s) and greater. Velocities of 6.1 m/s (20
feet/s) have been demonstrated.8 By contrast, earlier systems'
design capacities were based on absorber flue gas velocities of 3
m/s (10
feet/s). In addition to smaller absorbers, single absorbers
commonly serve multiple boilers, reducing the overall footprint of
the FGD retrofit. To date, a single absorber has been successfully
installed to serve up to 900 MWe of capacity in the U.S. while even
larger absorber modules (i.e. 1000 MWe and greater) are now being
offered for purchase overseas.
Space for an FGD installation may also include areas for reagent
processing and treatment of the waste or byproduct. Complete
limestone processing (delivery, crushing, slurry preparation,
reagent feed equipment, etc.) requires as much as 2 to 3 acres;
however, this space is a one-time requirement and does not increase
with increasing FGD capacity being served. Conservatively, when
on-site reagent processing is selected, an additional 2.5 acres for
an entire facility will be sufficient. While limestone processing
can be performed at the facility, purchased powdered limestone is
an option that also reduces or eliminates the requirement for
on-site reagent preparation and other equipment, as well as the
space these processes would occupy. Ultra-fine limestone has been
demonstrated as an optional enhancement over typical limestone
reagent feed.17 In areas where the ability to deliver limestone on
a continuous basis during winter months may be limited, storage of
limestone may be needed. For example, a 30-day supply of limestone
to feed a 500 MWe FGD system (95 percent control efficiency, 85
percent capacity, 4 percent sulfur coal) will require approximately
40 by 40 m (120 by 120 foot) storage area to handle approximately
23,000 tons of limestone.
Traditionally, FGD systems have produced a solid waste product
that can be sent to a landfill, or an increasingly attractive
alternative is to treat the byproduct for the manufacture and sale
of gypsum. If dewatering is required, typically 1-acre will be
needed for an entire facility regardless of the amount of FGD
capacity being served. One approach to improving sorbent
utilization is recycling the spent sorbent for multiple exposure to
the SO2 in the flue gas. The result is less unreacted sorbent and
smaller quantities of end product.18 Improved performance and
alternative reagents are becoming common. By the mid-1990's, at
least one FGD vendor was supplying a system that took advantage of
a water treatment system's precipitated calcium and magnesium
carbonates that produced a high quality, fine calcium and magnesium
carbonate FGD reagent. In addition to reducing the facility's
dependence on limestone, this process also reduced equipment
required for limestone handling and milling.4
More efficient use of water in modern systems has almost
completely removed the need for dewatering and containment ponds.
Typically, purge streams are used if the wastewater contains high
levels of chlorides. However, usually water is either evaporated
from the system or remains in the by-product or waste. Techniques
for wastewater minimization or elimination are commonly available.
For example, many FGD systems repeatedly cycle the cooling tower
blow-down before being treated in the wastewater system. As a
result, the wastewater has a high solids content as well as high
alkalinity for improved performance. Since large amounts of water
are evaporated during this cycling, this method also benefits from
reduced effluent that requires treatment by a wastewater
system.16
While water treatment of FGD effluent was once a concern,
contemporary FGD systems are much more effective in limiting
production of waste water and can achieve zero, or near-zero,
wastewater discharge.9 Many of the wastewater advances being used
outside of the U.S., including conserving blowdown in the absorber
vessel primarily for chloride control, are now being used or
considered in the U.S. For example, the 446 MWe Hunter Unit 3
(operated by PacifiCorp) installed a wet FGD limestone reagent
system in 1983; and, by use of mechanical draft cooling towers, the
plant is zero-discharge for waste water. The FGD system operates at
0.12 lb SO2/MMBtu and is designed for 90 percent SO2 removal. An
additional example of zero wastewater discharge is the 446 MWe
Craig Units 1 & 2 (installed 1980) that are designed for 85
percent SO2 removal and also employ limestone reagent and
mechanical draft cooling towers.
`Ãœ~éÃÉê=P
klu=`çåÃêçä=qÉÅÜåçäçÖó=oÉÃêçÑáÃë
In this chapter, retrofit of SCR will be assessed for coal-fired
electric utility boilers that would be affected by a multipollutant
regulation. SCR is the NOX control technology that is expected to
have the greatest impact on future utility boiler NOX emissions and
is the most difficult NOX control technology to install. It is,
therefore, the most important NOX control technology to understand
from both a NOX reduction and resource requirement perspective.
PKN póëÃÉã=e~êÇï~êÉ
The SCR process operates by reacting ammonia with NOX in the
exhaust gas in the presence of a catalyst at temperatures of around
315 to 370 ºC. For most applications, this temperature range makes
it necessary to locate the SCR reactor adjacent to the boiler -
immediately after the boiler and before the air preheater as shown
in Figure 3-1. An infrequently used alternative approach is to
locate the SCR after the FGD. This approach, however, increases
operating costs, as it requires additional heating of the gas. By
locating the SCR reactor as in Figure 3-1, it is often necessary to
install the catalyst reactor in an elevated location, which may
result in a structure hundreds of feet tall. Figure 3-2 shows the
configuration of the SCR that was retrofit onto AES Somerset
Station, a 675 MWe boiler already equipped with an electrostatic
precipitator (ESP) and wet FGD system. In this common installation,
the SCR reactor is installed on structural steel that elevates it
above existing ductwork and the ESP (designated "precipitator" in
the Figure 3-2). In the lower right corner of Figure 3-2, an image
of a person provides a perspective of the size of the SCR
installation.
Figure 3-1. Gas path for coal-fired boiler with SCR, ESP, and
FGD.
The SCR system reduces NOX through a reaction of ammonia and NOX
in the presence of oxygen and a
catalyst at temperatures around 315 to 370 °C (600 to 700
ºF). The products of this reaction are water
vapor and nitrogen. The catalyst is mounted inside an
expanded section of ductwork and is configured
for the gas to pass through it as in Figure 3-2.
The major components of an SCR system include:
â– Ammonia or urea storage
â– Ammonia vaporization system (if aqueous ammonia is used)
â– Urea to ammonia converter (if urea is used)
â– Ammonia or urea metering and controls
â– Dilution air blowers
â– Ammonia injection grid
â– Catalyst
â– Catalyst reactor, ductwork and support steel
â– Catalyst cleaning devices (soot blowers, sonic horns, etc)
â– Instrumentation
Except for the catalyst, most of the material/equipment used to
assemble an SCR system is either standard mechanical and electrical
components (pumps, blowers, valves, piping, heaters, pressure
vessels, temperature and pressure sensors, etc.) or is largely
manufactured for other power plant applications and has been
adopted for use in SCR systems (cleaning devices such as soot
blowers or sonic horns, gas analyzers, etc.). The catalyst,
however, is a specialized product designed specifically for this
purpose.
The catalyst is typically a ceramic material that, in most
cases, is either extruded into a ceramic honeycomb structure or is
coated onto plates, as shown in Figure 3-3. The catalyst is
assembled into modules at the factory. The modules are shipped to
the site and installed into the SCR reactor in layers. Each layer
of catalyst is comprised of several individual modules that are
installed side-by-side.
The material used in the largest quantity, aside from a catalyst
or reagent, is steel. The amount of steel required for an SCR in
the range of 300-500 MWe is about 800 to 1200 tons,20 or about 2.4
to 2.6 tons per MWe. About 4,000 tons of steel is necessary for
retrofit of two 900 MWe units (1,800 MWe total),20 or about 2.2
tons per MWe. The steel used for an SCR includes large structural
members, plates, and sheets. These steel pieces are used to
fabricate the catalyst reactor, the ductwork, and the support
steel. There is typically less of a requirement for corrosive
resistant alloys for an SCR installation when compared to a
scrubber installation. Steel is also needed for boiler
modifications. In this case, large pieces of steam piping or other
large steel boiler components may need to be replaced. The catalyst
reactor is often fabricated on-site. Sections of the catalyst
reactor and ductwork may be fabricated off-site and shipped in
pieces to the site for final assembly, or they may be fabricated
on-site into subassemblies and lifted into place during
erection.
If more than one boiler at a facility is to be retrofit with
SCR, then some, but not all, equipment can be made common. For
example, it may be possible, and is probably preferable, to have a
common ammonia
or urea storage facility. Reagent storage is probably the only
major equipment item that lends itself to sharing between adjacent
boilers. Therefore, there is some gained efficiency in the use of
equipment at a site with multiple units. However, this gain in
efficiency is generally small compared to the total project. The
major synergy will be in construction equipment and in labor, as
will be discussed in Sections 3.3 and 3.5, respectively.
PKO `~Ã~äóëÃ=~åÇ=oÉ~ÖÉåÃë
An SCR system requires an initial and ongoing supply of
catalyst. It also requires reagent. The reagent can be ammonia or
urea. Most facilities to date have used ammonia; however, urea is
becoming an increasingly popular reagent due to its inherent safety
and the recent availability of systems to convert urea to ammonia
on-site.
`~Ã~äóëÃ
The amount of catalyst required for an SCR system is directly
proportional to the capacity (or gas flowrate) of the facility, if
all other variables are equal. The actual amount of catalyst for
any specific plant depends upon several parameters; in particular,
the amount per MWe (measured in m3 per MWe) for a given level of
reduction and lifetime will fall within a general range. Therefore,
it is possible to make an estimate of how much catalyst would be
necessary to retrofit a particular facility or a large number of
facilities if the total capacity is known. It is assumed that most
SCR systems to be retrofit onto electric utility boilers will be
designed for about 90 percent reduction. For most boilers, this
level of reduction may initially require about 0.90 to 1.3 m3 of
catalyst for each MWe of coal-fired boiler capacity.18,21,22 For
example, a 500 MWe plant would be expected to have about 450 to 650
m3 of catalyst. The amount of catalyst for a particular situation
will vary somewhat depending on the catalyst supplier and the
difficulty of the application. At the 675 MWe AES Somerset Boiler,
90 percent NOX reduction was achieved with SCR using 897 m3 of
plate catalyst,19 or about 1.33 m3 per MWe. This unit fires 2.5
percent sulfur coal. At each of the 745 and 755 MWe Montour Units
1&2, 671 m3 of ceramic catalyst were used,22 or about 0.89 m3
per MWe. This unit fires 1.5 percent sulfur coal that can have
arsenic levels as high as 100 ppm (limestone injection is used to
reduce gaseous arsenic concentration in the furnace). The amount of
catalyst will tend to be lower in situations that are less
challenging, such as with lower sulfur coals or situations expected
to have lower gaseous arsenic concentration (gaseous arsenic is a
catalyst poison that originates in the coal; it will reduce the
lifetime of the SCR catalyst). Hence, less than 0.90 m3 per MWe may
be sufficient in some cases.
The catalyst is typically loaded in three or more layers. This
permits replacement of sections of the catalyst as activity is
reduced. The advantage of this approach is that it permits lower
overall catalyst usage over the economic lifetime of the plant.
Normally, room for an extra layer is provided, so a fourth layer
can be added, if necessary. At the first catalyst addition
(typically, after about 24,000 operating hours), the fourth layer
will be filled or half filled. Once the SCR reactor is full, layers
of catalyst are replaced after catalyst activity drops to a minimum
level. At the first catalyst replacement, new catalyst will replace
the original first layer; at the next catalyst replacement, new
catalyst will replace the original second layer, and so on. EPA
modeling projections conservatively assumed that one layer of
catalyst is replaced for every 15,000 - 20,000 hours of operation
for coal-fired units. Therefore, after the initial installation,
there is a need to replace roughly one fourth of the total catalyst
reactor volume every 24- 32 months or so - or conservatively about
1/8 of the installed volume should be replaced each year for the
coal-fired installations.
The catalyst may also be regenerated rather than replaced.23
This will reduce the amount of new catalyst that must be purchased.
However, due to the limited experience with this method, it will be
assumed that the catalyst is replaced according to the catalyst
management plan.
oÉ~ÖÉåÃë The amount of reagent consumed in the SCR process is
directly proportional to the amount of NOX reduced. Although
ammonia is the chemical that actually participates in the chemical
reaction, some suppliers have developed equipment to convert urea
to ammonia on-site. According to one supplier of urea-to-ammonia
converters, each mole of urea within the conversion system is
converted to two moles of ammonia.24 For example, reducing one
pound of NOX will require roughly 0.176 kg of ammonia or about
0.312 kg of urea. This includes a ½ percent increase in reagent
demand due to ammonia slip and a five percent increase to account
for a small amount of nitrogen dioxide (NO2) in the flue gas.
Therefore, for any given plant size, the amount of catalyst and
reagent consumption can be estimated. For a 500 MWe plant reducing
NOX from 0.50 lb/MMBtu to 0.05 lb/MMBtu and 85 percent capacity
factor (this is conservatively high for most coal boilers),
approximately 3,400 tons/yr of ammonia (anhydrous equivalent) or
about 6,100 tons/yr of urea (as 100 percent urea) would be needed.
The same 500 MWe plant would have around 450-650 m3 of catalyst
with roughly 120-160 m3 replaced about every three years. This is,
if a third of the initial catalyst loading must be replaced, on
average, every 15,000 to 20,000 operating hours, then 0.015 to
0.0289 cubic meters per MWe per 1000 hours must be replaced.
PKP `çåëÃêìÅÃáçå=bèìáéãÉåÃ
Construction equipment needed for installation of an SCR
includes standard construction equipment - welders, excavation
equipment, concrete pouring equipment, cranes, etc. In some cases,
installers may use tall-span heavy-lift cranes. These cranes are
capable of lifting heavy loads, as much as 100 tons or more,
several hundred feet. The advantage of this crane type is realized
when lifting assembled sections of catalyst reactor or other large
pieces high off the ground. If lower capacity cranes are used,
smaller pieces must be lifted, which means that less
pre-fabrication is possible and more assembly must be done in
place. Less pre-fabrication could lengthen the necessary boiler
outage somewhat. Although the availability of the largest cranes is
reported to about 60 or more, about 12 new cranes can be supplied
every six months.25 It has been reported that, in some cases, it
has been necessary to go further away from the plant to source
cranes with adequate lift and reach capacity. In other cases,
engineers found that by changing the design/fabrication method to
meet the available crane, the project could be managed with lower
capacity cranes (lifting smaller pieces).26,27 If more than one
boiler is retrofit at one facility, then the crane can be used for
both boilers, saving cost and time when compared to boilers
retrofit separately. It is important to note that in many cases the
erection method is not limited by the available crane, but is
limited by the access to the plant (For example, can large sections
be delivered by barge, rail, or roadway?) and by the available
lay-down area for material and construction equipment on site. At
many facilities, there is inadequate area to prefabricate large
sections. In some instances, transportation routes to the facility
do not permit transporting large, pre-assembled equipment to the
site. In such cases, it will not be possible to do much
pre-assembly, and a smaller, less expensive crane may be adequate.
As a result, the type of crane that is best for a particular SCR
installation frequently is not the largest crane available. The
crane selected for a project will be determined as part of an
overall construction plan developed to optimize all of the
available resources - labor, material, and equipment - for a
particular project.
The need to lift material to high elevations is a result of the
location of the SCR - often above existing ductwork and adjacent to
existing equipment. Figure 3-2 provided one good example of this.
It may be necessary to move existing equipment, such as the air
preheater, in order to accommodate the addition of the SCR reactor.
As a result, every retrofit is a custom fit. However, engineers
have been very innovative when installing these systems, even on
facilities that apparently had little room available for the SCR.
Hence, the physical size of the technology has not been
limiting.
PKQ fåëÃ~ää~Ãáçå=qáãÉ
Implementation of a NOX control technology at a plant involves
several activities contingent upon each other. These activities may
be grouped under the following phases of an implementation project:
(1) conducting an engineering review of the facility and awarding a
procurement contract; (2) obtaining a construction permit; (3)
installing the control technology; and (4) obtaining an operating
permit.
Exhibit A-3 in Appendix A depicts the timeline expected for
completing a single unit installation of SCR. Completion of some of
the activities is contingent upon completion of some other
activities. For example, construction activities cannot commence
until a construction permit is obtained. In general, the SCR
implementation timeline appears to be driven primarily by the
engineering activities (i.e., design, fabrication, and
construction).
båÖáåÉÉêáåÖ=oÉîáÉï
As shown in Exhibit A-3 in Appendix A, an engineering review and
assessment of the combustion unit is conducted in the first phase
of technology implementation to determine the preferred compliance
alternative. During this phase, the specifications of the control
technology are determined, and bids are requested from the vendors.
After negotiating the bids, a contract for implementing the NOX
control technology is awarded. The time necessary to complete this
phase is approximately four months for SCR.
`çåëÃêìÅÃáçå=mÉêãáÃ
Before the actual construction to install the technology can
commence, the facility must receive a construction permit from the
applicable state or local regulatory authority. The construction
permit process requires that the facility prepare and submit the
permit application to the applicable state or local regulatory
agency. The state or local regulatory agency then reviews the
application and issues a draft approval. This review and approval
process is estimated to take about six months. The draft
construction permit is then made available for public comment.
After any necessary revisions, a final construction permit is
issued. The actual time needed will depend on the size and
complexity of the project and the local procedures for issuing a
permit. Exhibit A-3 in Appendix A shows that nine months are
allowed for the construction permit. This is expected to be ample
time. In one case, only about 4-5 months were needed for obtaining
the construction permit,26 and only six months were needed to
obtain the construction permit for retrofit of two 900 MWe boilers
in another case.21 Shorter periods for construction permit
authorization would allow earlier commencement of construction
activities and could potentially shorten the overall schedule.
`çåÃêçä=qÉÅÜåçäçÖó=fåëÃ~ää~Ãáçå
In the second phase, the control technology is installed. This
installation includes designing, fabricating, and installing the
control technology. In addition, compliance testing of the control
technology is also completed in this phase. Most of the
construction activities, such as earthwork, foundations,
process
electrical and control tie-ins to existing items, can occur
while the boiler is in operation. The time needed to complete this
phase of an implementation project is about 17 months for SCR.
An important element of the overall control technology
implementation is the time needed to connect, or hook up, the
control technology equipment to the combustion unit because the
boiler typically must be shut down for this period. SCR connection
can occur in a three to five week outage period.28 In some cases
longer outages are needed. When Babcock & Wilcox retrofitted
the 675 MWe AES Somerset boiler, the outage began on May 14, and
the boiler was returned to service on June 26 - about a six-week
outage.19 One major SCR system supplier in the U.S. stated that
they would want in the range of one to two months of boiler down
time and have never required more than two months.27 Difficulty is
increased as the extent of boiler modifications necessary to fit
the SCR into the facility is increased. A German SCR system
supplier installed SCR on a significant portion of the German
capacity within outage periods consisting of less than four
weeks.11 Based upon outages in this time range for SCR connection,
electricity-generating facilities would normally be able to plan
the SCR connection to occur during planned outages to avoid
additional downtime. Some facility owners have been innovative in
their construction plans to minimize down time. At the Tennessee
Valley Authority's (TVA's) 700 MWe Paradise Unit 2, it was
necessary to demolish the existing ESP with the unit on line. TVA
installed a construction bypass to send gas from the air preheater
outlet directly to the FGD, while the ESP was being demolished and
the SCR reactor erected in its place.11 However, in more difficult
retrofits, down time might be impacted in a significant way. In
some cases it may be desirable to plan a brief outage in advance of
the hook-up to install structural steel through sleeves placed in
existing equipment, such as the ESP, or to relocate existing
equipment that would otherwise interfere with erection of the SCR.
This permits erection of the catalyst reactor above existing
equipment while the unit is on line.26 However, because an SCR
project is expected to extend close to two years (see Exhibits A-3
and A-4 in Appendix A), it should be possible to incorporate this
work into planned outages, which would have occurred regardless of
whether an SCR was to be installed.
léÉê~ÃáåÖ=mÉêãáÃ
Facilities will also need to modify their Title V operating
permit to incorporate the added control devices and the associated
reduced emission limits. In some states, an interim air-operating
permit may need to be obtained until the Title V permit is
modified. The operating permit modification process consists of
preparation and submission of the application to the applicable
state or local regulatory agency. As shown in Exhibit A-3 in
Appendix A, this process can occur simultaneously with the
processing of the construction permit application. The process of
transitioning from the construction permit to the operating permit
varies among states and appears to be somewhat unclear due to the
infancy of the Title V operating permit process. Nonetheless, based
on discussions with several states, the application review process
is estimated to take approximately 9-11 months. The Title V
operating permit must also be made available for public comment.
Following public comment, the Title V operating permit is not made
final until compliance testing on the control device is completed.
Therefore, the total estimated time to modify the Title V operating
permit is about 17 months, plus the additional time to complete
compliance testing.10
Based on the estimated time periods needed to complete each of
the four phases described above, the estimated time period to
complete the implementation of SCR on one combustion unit is about
21 months. This time period is shown in Exhibit A-3 in Appendix A.
However, depending upon the specifics of the project, the time
needed could vary by a couple of months. For example, at AES
Somerset station, the time to complete the retrofit from the point
of contract award was nine months.19 Assuming four months of work
prior to contract award, a total elapsed time of 13 months would
have been necessary to retrofit this 675 MWe boiler. Another
facility, Reliant Energy's Keystone plant, has
two 900 MWe, 8-corner, T-fired combustion engineering units that
burn approx 1.5 percent sulfur bituminous coal. Reliant intends to
reduce the NOX from a baseline of 0.40 lb/MMBtu to 0.04 lb/MMBtu.
The permit to construct was received in approximately six months.
The time from placing the order to completion of commissioning
activities is 46 weeks for both units. However, preliminary
engineering was accomplished earlier. Even if preliminary
engineering and contract negotiation took as long as six to eight
months, the total time for completing two 900 MWe units would be
about 17 to 19 months.21 For the New Madrid plant, units 1 & 2
(600 MWe each), the specifications were released to turnkey
contractors in February 1998, the project specification was
released in March 1998, the contract was awarded on June 26, 1998,
and the first unit was in operation by February 2000. In this
project, an option for a second unit was available (and was
exercised), and air preheaters were replaced.29 Therefore, 21
months should be a reasonable, and in some cases a conservative
estimate of the total time necessary to retrofit a single utility
boiler.
Under the Clear Skies Act, EPA does not expect that SCR will be
implemented at every facility. For those plants where EPA projects
SCR retrofits will occur, EPA's projections reflect that these
facilities will typically have 1 to 4 boilers retrofit per site.
However, for one facility, seven SCR retrofits are projected to be
installed by 2020. Exhibit A-4 in Appendix A examines a schedule
for retrofitting a facility with multiple (seven) SCR retrofits.
This examines the installation of the control device hook-up on a
sequential basis. Installation is staggered by two to three months
between sequential units to enable more efficient utilization of
manpower and project management than if multiple units were
connected at one time. This approach also assures that at least
about 83 percent of the plant capacity is available at any given
time (only one boiler is shut down), and during most of the time
there is no impact to the plant availability at all. This approach
requires a total time of 35 months for seven SCR retrofits. An
alternative approach might be to schedule outages to avoid any
outage during high electricity demand periods. This might extend
the total elapsed time by about four months. However, because there
is a substantial amount of work that can be accomplished with the
boiler on line, the additional time would be much less than the
number of high electricity demand months that are accommodated by
this approach. Another alternative approach would involve retrofit
of more than one unit at a time during low-demand periods and
avoiding any outage during high demand periods. This alternative
could result in a faster project completion, but would have less
even labor utilization, which is an important cost-benefit
tradeoff.30
In summary, the total time needed to complete the design,
installation, and testing at a facility with one SCR unit is about
21 months; at a facility with multiple SCR (seven) units, total
time is approximately 35 months. Based on these timelines, it is
estimated, in principal, that the NOX controls needed to comply
with a multipollutant strategy can be met provided that: (1) an
adequate supply of materials and labor is available, and (2) the
control technology implementation process begins at least about 35
months prior to the date controls must be in place. However,
ideally, longer than 35 months would allow for all of the retrofits
to occur over a period of several years so that facility owners can
properly plan outages and suppliers can properly plan for resource
availability.
PKR i~Äçê
The installation of an SCR system requires a significant amount
of labor. Most of the labor is necessary for the construction of
the facility. However, engineering and project management labor are
also needed for the project. The total construction labor for an
SCR system of 500 MWe is in the range of 333,000 to 350,000
man-hours.22,27 Typically, approximately 40-50 percent of the labor
is for boilermakers.31 However, the percent of labor for
boilermakers will vary from one project to another, with 40-50
percent
being an average for several projects.32 Some projects require a
higher degree of boiler integration and less erected steel and,
therefore, have a higher percentage of boilermaker labor. Other
projects require extensive steel erection with less boiler
integration and will, therefore, have a lower percentage of
boilermakers versus other trades. For a 500 MWe plant, the
construction labor would be about 340,000 man-hours, of this
roughly 136,000-170,000 man-hours would be boilermaker activity.
Engineering and project management are about 5 percent of the total
cost, while construction is about 50 percent of the total cost.19
If labor rates for engineering and project management is 50 - 100
percent greater than construction labor, then about 17,000 to about
28,000 man-hours of engineering and project management are needed
for the project. Total labor man-hours of construction and
engineering labor are then about 365,000 man-hours for a single 500
MWe unit.
Construction man-hours are approximately proportional to the
tons of steel fabricated. As noted earlier, the material needed for
multiple boiler installations is generally not reduced
significantly over the projects if they were installed separately.
However, if more than one system is installed at a site, some
significant efficiencies result.
When there are multiple units retrofit at one site only, one
mobilization is needed for all of the boilers, only one
construction supervisor is required, and equipment is more
efficiently used. As a result, 15-20 percent efficiencies can be
realized from these activities and can be planned into the
project.22 Long-term projects, such as retrofits of more than one
unit at one site, also lend themselves to additional efficiencies
from learning curves. Learning curves result from productivity
improvements over the duration of the project. Productivity
measures the actual man-hours used versus those planned. A
productivity value over 100 indicates that fewer man-hours are
needed to accomplish the goal than expected. A labor productivity
value of 110 means that 10 percent more work was accomplished for
the level of labor expended than if a productivity value of 100 was
achieved. There are examples of productivity improvements of 9 to
19 percent during the project due to additional efficiencies gained
from learning curves.30 If only 10 percent or less improved
efficiency results from planned reduced labor and from productivity
improvements that occur after the project commences, the labor for
each additional 500 MWe plant might be reduced from 340,000
man-hours to about 310,000 man-hours, or about 2,170,000
construction man-hours for seven 500 MWe units at one plant.
For a site with multiple units, the total engineering and
project management man-hours are likely to be significantly less
than the total if each unit were addressed separately. This is
because there will be many common site issues that need to be
addressed and engineered only once. Procurement contracts need to
be negotiated only once, and only one project management team is
needed over the duration of the contract. However, it is difficult
to say how much engineering will be reduced, because adjacent units
may be very similar or very different. One approximation is to make
total engineering, project management, and testing proportional to
the project duration. Thus, a seven-unit facility would require
about 42,000 man-hours of engineering and project management.
Based upon the discussion from sections 3.1 through 3.5, the
total resources needed for a single 500 MWe plant and a site with
seven 500 MWe plants is shown on Table 3-1.
q~ÄäÉ=PJNK==bëÃáã~ÃÉÇ=oÉëçìêÅÉë=kÉÉÇÉÇ=Ñçê=páåÖäÉ=~åÇ=jìäÃáéäÉ=p`o=oÉÃêçÑáÃë
= råáÃë låÉ=_çáäÉê pÉîÉå=_çáäÉêë
`~é~ÅáÃó=oÉÃêçÑáà eÉ~Ã=o~ÃÉ `~é~ÅáÃó=c~Ã…Ãçê mÉêÅÉåÃ=oÉÇìÅÃáçå
råÅçåÃêçääÉÇ=klu `çåÃêçääÉÇ=klu kl =oÉÇìÅÉÇLóê
u
=
pÃÉÉä
`~Ã~äóëÃ=J=áåáÃá~ä=Ñáää
qáãÉ=ÄÉÃïÉÉå=Ñáääë
`~Ã~äóëÃ=J=êÉéä~ÅÉãÉåÃ=~ääçï~åÅÉ
^ããçåá~=pâáÇë
^ããçåá~=pÃçê~ÖÉ
=
i~Äçê
====mêçà ÉÅÃ=aìê~Ãáçå
====båÖáåÉÉêáåÖLmêçà K=jÖÃK
====`çåëÃêìÅÃáçå
====qçÃ~ä=i~Äçê
=
^ããçåá~=`çåëìãéÃáçå EÉèìáî~äÉåÃ=^åÜóÇêçìëF
rêÉ~=bèìáî~äÉåÃ
ãçåÃÜë ON PRÄ ã~Ã¥JÜçìêë ORIMMM QOIMMM ã~Ã¥JÜçìêë PQMIMMM
OINTMIMMM ã~åJÜçìêë PSRIMMM OIONOIMMM
ÃçåëLóê PIQMM OPIMMM
ÃçåëLóê SINMM QOITMM
~
=råÇÉê=ÃÜÉ=`äÉ~ê=pâáÉë=^Ã…ÃI=bm^=ãçÇÉäáåÖ=éêçà ÉÅÃë=ÃÃœ~Ã=ÃÜÉ=éä~Ã¥Ã=ïáÃÃœ=ÃÜÉ=ã~ñáãìã=åìãÄÉê
çÑ=p`o=êÉÃêçÑáÃë=áë=çåÉ=éä~Ã¥Ã=ïáÃÃœ=ëÉîÉå=ÄçáäÉêë=~åÇ=~=Ã…~é~ÅáÃó=çÑ=OIQMM=jtÉK
Ä
=j~ó=êÉèìáêÉ=~Ã¥=~ÇÇáÃáçå~ä=Ñçìê=ãçåÃÜë=Ãç=~îçáÇ=ëÅÜÉÇìäáåÖ=çìÃ~ÖÉë=Ñçê=ÅçåÃêçä=Üççâìé
ÇìêáåÖ=éÉ~â=ÉäÉÅÃêáÅáÃó=ÇÉã~åÇ=ãçåÃÜëK
PKS pé~ÅÉ=oÉèìáêÉãÉåÃë
An SCR system for a coal-fired boiler may have a negligible
impact on the footprint of the boiler. This is because the SCR is
frequently installed in an elevated position near the boiler and
well off of the ground. The choice of installing the SCR reactor
near the ground level or elevated well above ground level depends
upon which configuration is viewed as most cost effective while
considering installation cost and operating cost. Locating the SCR
in an elevated location near the boiler economizer and air
preheater is frequently done to minimize the length of ductwork
(with the associated pressure loss) and because no additional real
estate is necessary for the SCR reactor. When this type of
installation is performed, the SCR reactor is installed atop a
steel structure that must be erected above existing equipment, such
as the electrostatic precipitator. This is an approach that is
frequently used because engineers have developed cost effective
methods to install the SCR reactor while addressing potential
interferences from existing equipment. Section 3.4 of this document
discussed how brief outages, in advance of the outage to connect
the SCR, were taken to address interferences and permit SCR
reactor
construction with the unit on line. In some cases, however, the
preferred approach has been to locate the SCR reactor on the ground
near the boiler and to route the ductwork to and from the SCR
reactor. This is the approach that was taken on the retrofit of
PSNH Merrimack Unit 2, the first retrofit of a coal-fired boiler in
the United States. In this case there was a large amount of space
near the boiler to permit this approach. Regardless of where the
SCR reactor is located, ductwork from the economizer outlet to the
SCR reactor and back to the air preheater inlet must be
accommodated. In cases where space for this ductwork was extremely
limited, the air preheater was relocated. However, relocation of
the air preheater(s) usually is not necessary. Only a few
installations have required the relocation of the air
preheater.
The other item that must be located is the reagent storage
system. This usually does not take up as much room as the SCR
reactor itself. However, the storage and unloading system must be
located near rail or truck access to permit delivery of reagent. In
some cases, long piping is run from the storage and unloading area
to the SCR reactor. In these cases, the piping may be insulated and
heat traced to prevent condensation of the ammonia vapor.
`Ãœ~éÃÉê=Q jÉêÅìêó=`çåÃêçä=qÉÅÜåçäçÖó=oÉÃêçÑáÃë
Under a multipollutant control scenario, mercury emissions would
be controlled from coal-fired power plants by equipment that
reduces emissions of other pollutants (e.g., scrubbers and SCR) and
the use of sorbent injection. Other methods are being investigated
(such as oxidation and scrubbing technologies), which utilize
ozone, barrier discharge, and catalyst and/or chemical additives in
combination with existing technologies. To the extent that other
technologies are developed, these would provide more options for
compliance, so their introduction would serve to reduce issues
related to resource requirements of installing controls. Similarly,
with regard to sorbent injection, sorbents other than activated
carbon (AC) may ultimately prove to be superior for this
application in terms of cost or collection efficiency performance
and may reduce the likely demand for ACI from what is projected
here. Nevertheless, all of the sorbent-based approaches use similar
hardware to inject sorbent as ACI. Therefore, the assumption of ACI
as a control method will provide a fairly representative indication
of the demand for hardware and construction resources regardless of
which sorbents are used in the market. The assumption of ACI as a
mercury control method will be more conservative with regard to
sorbent consumption since it will assume that all of the facilities
installing sorbent injection for mercury control require AC.
QKN póëÃÉã=e~êÇï~êÉ
The AC is typically injected at the lowest temperature available
that is upstream of a particle-collecting
device because experience has found that mercury collection
is most efficient at lower temperatures. On
a boiler equipped with an ESP or a fabric filter (FF) for
particle collection, the configuration would look
as in Figure 4-1. Collection of mercury is somewhat more
efficient when a FF is used for particle
collection because of the higher gas-sorbent contact in the
filter cake. Another approach is to have
injection downstream of an ESP, which would collect most of
the coal fly ash, and upstream of a fabric
filter (FF), which would mostly capture sorbent. This
approach is shown in Figure 4-2. The advantages
of this approach are that greater mercury capture occurs
because of the additional mercury capture that
can occur on the FF filter cake; and, because the ash is
largely separated from the sorbent, more efficient
sorbent utilization is possible through sorbent recycling.
This approach could be implemented through
addition of a Pulse Jet Fabric Filter (PJFF) when ACI is
installed.
The ACI System consists of the following components, as shown in
the simplified schematic of Figure
4-3:
â– A silo for storing the sorbent
â– A metering system for metering the amount of sorbent
injected into the ductwork - typically a rotary
metering valve â– A pneumatic or mechanical conveying system for
moving the sorbent to the injection location
â– An injection system for dispersing and distributing the
sorbent in the boiler ductwork. For many facilities, injection of
sorbent will occur after the air preheater and upstream of the ESP
or FF. This injection system is principally made from piping that
may split off to manifolds for injecting in multiple locations.
Special nozzles or other hardware are generally not required.
â– A blower to provide a carrying medium
â– Associated piping for the blower and the distribution
system
â– A humidification system may be used in some cases to reduce
temperature and improve mercury capture. The humidification system
will typically consist of water spray injectors (possibly air
atomized) located upstream of the ACI injectors, a grid for the
spray injectors, and a water supply system that will include
pumping and metering systems.
â– A control system that may utilize a programmable logic
controller (PLC) or may be accommodated by the plant distributed
control system (DCS)
Stack
Activated Carbon Sorbent Storage Silo
From air preheater Exhaust Gas Duct
To ESP Or FF
Regardless of boiler size, an ACI system will require the same
equipment. The principal differences will be the size of the
sorbent storage silo, the size of the metering and conveying
system, and the size and number of injectors for the sorbent
injection system.
There are also several other combinations that may be used,
including combinations of ACI with spray dryer and FF and
combinations of ACI with FGD.32 The various combinations will be
discussed further in Section 4.2. In each of these combinations,
the actual equipment associated with the ACI system is similar. The
particular combination of equipment chosen for mercury reduction at
a particular facility is largely determined by the existing
equipment and conditions at the facility. However, most facilities
are currently equipped with ESPs, and some are equipped with FFs.
Thus, the most likely scenario for application of ACI is in a
configuration with ESP or FF.
Most existing facilities have ESPs for particle emission control
and do not have any SO2 removal technology. Therefore, injection of
sorbent, and possibly water for humidification, will most often be
performed downstream of the air preheater and upstream of the
electrostatic precipitator, where the gas temperature is typically
in the range of 280-300 ºF. In this part of the boiler ductwork,
there are no water wall tubes. Therefore, the mechanical interface
between the ACI system and the boiler is through the duct walls,
and high-pressure boiler tubing will not be affected by the
retrofit of ACI.
Some companies offer other sorbent-based methods for reduction
of mercury emissions; however, the equipment used is very similar
in scope to the equipment used for ACI.32
The majority of the equipment used for an ACI system is produced
from standard mechanical or electrical hardware that is sold for a
wide range of purposes. The total amount of steel is relatively
small in comparison to an SCR or an FGD for a 500 MWe plant.27 The
estimated steel requirement for a 500 MWe ACI system is indicated
in Table 4-1.33 Since the largest item contributing to the steel
requirement is the storage silo, it will be assumed that the total
steel requirement is proportional to the capacity of the unit, as
the storage requirement would be proportional to the capacity. For
multiple units at one site, all but the silo would certainly need
to be duplicated. It is possible that there might be one large silo
serving several units with more than one feed off of it, or, that
individual silos may be needed. In any event, the synergies in
reducing total steel requirement over what would be needed for
individual units is expected to be small.
q~ÄäÉ=QJNK==bëÃáã~ÃÉÇ=pÃÉÉä=oÉèìáêÉãÉåÃ=Ñçê=RMM=jtÉ=^`f=póëÃÉãPP
fÃÉã bëÃáã~ÃÉÇ=tÉáÖÜÃ=EäÄëF
NÃ’=pÃ…ÃœK=QM=máéÉI=NRMM=Ñà QIMMM OÃ’=ëÃÉÉä=ÃìÄáåÖ=NRMM=Ñà QIRMM
jáëÅ=pÃêìÅÃìê~ä=pìééçêÃ=pÃÉÉä RIMMM
mêçÅÉëë=bèìáéãÉåÃ=EãçëÃäó=ëÃÉÉäF NRIMMM pÃçê~ÖÉ=páäç POMIMMM qçÃ~ä
PRMIMMM=äÄë=ENTR=ÃçåëF
QKO oÉ~ÖÉåÃ
AC is assumed to be the principal reagent used to absorb the
mercury in the exhaust gases. Most of the information on the AC
injection requirements for a coal-fired power plant is from pilot
studies and demonstrations of ACI technology. Table 4-2 shows AC
injection rates estimated from the data provided a comprehensive
assessment of ACI under a range of scenarios.34 For example, to
achieve 80 percent mercury reduction from a low sulfur bituminous
coal using an ACI system with humidification will require a
treatment rate of about 8 lb/million acf (MMacf).34 If a pulsejet
FF (PJFF) is used downstream, the sorbent injection rate can be
reduced to about 4.6 lb/MMacf. If the facility fires high sulfur
coal and is equipped with FGD, then the estimated sorbent rate is
between 6.1 lb/MMacf to 2.0 lb/MMacf, without and with a PJFF,
respectively. For a high sulfur coal application, humidification
would not be performed due to risk of acid condensation. Table 4-2
summarizes estimated injection rates for a 500 MWe boiler under
various scenarios.34 As shown, the injection rates vary
substantially based upon the circumstances.
Because combination of SCR and FGD are expected to have high
mercury removal due to the SCR and FGD systems, those facilities
that are so equipped are not expected to add ACI systems.
There are really no synergies in consumption if multiple ACI
units are installed at one site. Therefore, the total AC
consumption at a plant will be roughly proportional to the total
plant capacity equipped with ACI.
q~ÄäÉ=QJOK==bëÃáã~ÃÉÇ=^`=fåà ÉÅÃáçå=o~ÃÉë=Ñçê=~=RMM=jtÉ=_çáäÉêPQ
bñ~ãéäÉ eÖ `ç~ä pìäÑìêI bñáëÃáåÖ ^ÇÇáÃáçå~ä fåà ÉÅÃáçå
bëÃK=^`=ê~ÃÉ=Ñçê=RMM=jtÉ=éä~åà êÉÇìÅÃáçåI éÉêÅÉåà `çåÃêçäë~
bèìáéãÉåÃÄ o~ÃáçI éÉêÅÉåà äÄLjj~ÅÑPQ äÄLÜê ÃçåëLóê=Ã…
~
=bpm=Z=ÉäÉÅÃêçëÃ~ÃáÅ=éêÉÅáéáÃ~ÃçêX=cda=Z=ÑäìÉ=Ö~ë=ÇÉëìäÑìêáò~ÃáçåX=cc=Z=Ñ~ÄêáÅ=ÑáäÃÉêX=p`o=Z=ëÉäÉÅÃáîÉ=Ã…~Ã~äóÃáÅ=êÉÇìÅÃáçå
Ä
=pf=Z=ëçêÄÉåÃ=áåà ÉÅÃáçåX=tf=Z=ï~ÃÉê=áåà ÉÅÃáçåX=mgcc=Z=éìäëÉà ÉÃ=Ñ~ÄêáÅ=ÑáäÃÉê
Ã…
=qçåëLóê=ÉëÃáã~ÃÉÇ=~Ã=UR=éÉêÅÉåÃ=Ã…~é~ÅáÃó=Ñ~Ã…Ãçê=EäÄLÜê=G=UTSM=G=MKURLOMMMF
QKP `çåëÃêìÅÃáçå=bèìáéãÉåÃ
Construction equipment needed for installation of an ACI system
includes standard construction equipment - welders, excavation
equipment, concrete pouring equipment, cranes, etc. Since an ACI
system is much smaller and uses substantially less steel than an
SCR or FGD system, cranes and other lifting equipment can be of low
to moderate lifting capacity. Blowers, the sorbent storage silo,
and other equipment will be mounted on concrete pads or
foundations. In most cases, the sorbent storage silo will be field
erected; however, for some facilities that require less sorbent, a
smaller, prefabricated silo may be installed. Steel erection and
minor excavation and concrete work is necessary for an ACI system,
and this work should not require any more than very common
construction equipment. Piping for sorbent transport will typically
be welded steel and can be erected in the field in many cases. It
should not be necessary to relocate any existing boiler equipment
to install an ACI system. Therefore, the construction effort and
need for equipment is relatively modest compared to the more
involved SCR and FGD projects.
QKQ fåëÃ~ää~Ãáçå=qáãÉ
Implementation of a control technology at a plant involves
several activities contingent upon each other. These activities may
be grouped under the following phases of an implementation project:
(1) conducting an engineering review of the facility and awarding a
procurement contract; (2) obtaining a construction permit; (3)
installing the control technology; and (4) obtaining an operating
permit.
Exhibit A-5 in Appendix A depicts the timeline expected for
completing a single unit installation of ACI. Completion of some of
the activities is contingent upon completion of other activities.
For example, construction activities cannot commence until a
construction permit is obtained. In general, the ACI implementation
timeline appears to be driven primarily by the engineering
activities (i.e., design, fabrication, and construction).
båÖáåÉÉêáåÖ=oÉîáÉï
As shown in Exhibit A-5 in Appendix A, in the first phase of
technology implementation, an engineering review and assessment of
the combustion unit, is conducted to determine the preferred
compliance alternative. During this phase, the specifications of
the control technology are determined and bids are requested from
the vendors. After negotiating the bids, a contract for
implementing the control technology is awarded. The time necessary
to complete this phase is approximately four months.
`çåÃêçä=qÉÅÜåçäçÖó=fåëÃ~ää~Ãáçå
In the second phase, the control technology is installed. This
installation includes designing, fabricating, and installing the
control technology. In addition, compliance testing of the control
technology is also completed in this phase. Most of the
construction activities, such as earthwork, foundations, process
electrical and control tie-ins to existing items, can occur while
the boiler is in operation. The time needed to complete this phase
of an implementation project is expected to be less than three
months.33
An important element of the overall control technology
implementation is the time needed to connect, or hook up, the
control technology equipment to the combustion unit. As a result of
the minimal mechanical interface between the sorbent injection
system and the boiler, retrofit of an ACI system will typically
require a fairly short outage - one week or less.33,34 This brief
outage is necessary to install injection hardware and to make any
control system connections that may be necessary between the ACI
control and the boiler control. Other equipment associated with the
ACI system can be installed with the boiler on line, as it does not
require any interfacing with the boiler and should not require
moving any essential boiler equipment.
It should be possible to complete a project in less than 4
months from receipt of order.34 If construction and operating
permits are included in the analysis, the project is likely to take
longer than would be necessary only for engineering, supply,
installation, and startup of the ACI system. This is because the
permitting activities might become the time-limiting steps. In some
localities, it is possible that the permitting activities will not
be the limiting steps. In this case, a faster execution is possible
than shown on Exhibit A-5 in Appendix A.
léÉê~ÃáåÖ=mÉêãáÃ
Facilities will also need to modify their Title V operating
permit to incorporate the added control devices and the associated
reduced emission limits. In some states, an interim air-operating
permit may need to be obtained until the Title V permit is
modified. The operating permit modification process consists of
preparation and submission of the application to the appropriate
state or local regulatory agency. As shown in Exhibit A-5 in
Appendix A, this process can occur simultaneously with the
processing of the construction permit application. The process of
transitioning from the construction permit to the operating permit
varies among states and appears to be somewhat unclear due to the
infancy of the Title V operating permit process. Nonetheless, based
on discussions with several states, the application review process
is estimated to take approximately 38 weeks (9-10 months). The
Title V operating permit must also be made available for public
comment and is not made final until compliance testing on the
control device is completed. Therefore, the total estimated time to
modify the Title V operating permit is about 12 months, plus the
additional time to complete compliance testing.10
Based on the estimated time periods needed to complete each of
the four phases described above, the estimated time period to
complete the implementation of ACI on one combustion unit is about
15 months, as shown in Exhibit A-5 in Appendix A. Since the
permitting process limits the timeline, a faster permitting process
will shorten the time necessary to install ACI on a single
unit.
Under the Clear Skies Act, EPA does not expect that ACI will be
implemented at many facilities due to the co-benefit of mercury
removal from other control technologies. For those plants where EPA
projects ACI retrofits will occur, EPA's projections reflect that
these facilities will either have 1 to 2 boilers retrofit per site.
Exhibit A-6 in Appendix A examines a schedule for retrofitting a
facility with multiple (two) ACI retrofits. This examines the
installation of the control device hook-up on a sequential basis.
Installation is staggered by one month between sequential units to
enable more efficient utilization of manpower and project
management than if multiple units were connected at one time. This
approach requires a total time of 16 months.
In summary, the total time needed to complete the design,
installation, and testing at a facility with one ACI unit is about
15 months, at a facility with two ACI units is approximately 16
months. Based on these timelines, it is estimated that, in
principle, the mercury controls needed to comply with a
multipollutant strategy can be met provided that (1) an adequate
supply of materials and labor is available and (2) the control
technology implementation process begins at least 16 months prior
to the date controls must be in place. However, ideally, longer
than 16 months would allow retrofits to occur over a period of
several years so that facility owners can properly plan outages and
suppliers can properly plan for resource availability. Erection of
a PJFF would typically take 16 to 20 months from award of contract
to start up.35 If 4 months is added for pre-contract effort and 1-2
months is provided for start up and commissioning, the total
project duration would be anywhere from about 21 months to 26
months. However, EPA's modeling under the Clear Skies Act projects
that the units installing ACI will not be installing PJFFs.36
QKR i~Äçê
The man-hours of labor estimated to be required for supply of an
ACI system are listed in Table 4-3, which includes a breakdown of
man-hours by task.33 Craft labor for installation is also
indicated.
q~ÄäÉ=QJPK==bëÃáã~ÃÉÇ=j~Ã¥JÜçìêë=Ñçê=pìééäó=çÑ=~Ã¥=^`f=póëÃÉã=Ñçê=~=RMM=jtÉ=MKSB=p=_áÃìãáåçìë=`ç~ä=_çáäÉê=ïáÃÃœ
bpm=Ebñ~ãéäÉ=N=Ñêçã=q~ÄäÉ=QJQFPP
q~ëâ j~åJÜçìêë
lÑÑJpáÃÉ=båÖáåÉÉêáåÖ=~åÇ=lÃ¥JpáÃÉ=qÉëÃáåÖ NISMM
fåëÃ~ää~ÃáçåI=ÉñÅÉéÃ=ëáäç=EáêçåïçêâÉêëI=éáéÉ=ÑáÃÃÉêëI=ÉäÉÅÃêáÅá~åëF
NIOMM bêÉÅÃáçå=çÑ=páäç=EáêçåïçêâÉêëI=éáéÉ=ÑáÃÃÉêëF OIMMM
qçÃ~ä=j~Ã¥JÜçìêëG QIUMM
GbëÃáã~ÃÉÇ=ÃáãÉ=Ñçê=ÉåÖáåÉÉêáåÖI=ÇÉëáÖåI=ÉèìáéãÉåÃ=éêçÅìêÉãÉåÃI=~åÇ=~ëëÉãÄäó=áë=S=ãçåÃÜëK
In summary, a 500 MWe boiler firing eastern bituminous coal with
0.6 percent sulfur, an ESP, and no SCR or FGD, is estimated to
provide the performance and require the resources listed in the
first column of Table 4-4, and estimates of performance and
resources needed for other types of fuels and boiler configurations
are shown in the other columns. A boiler firing subbituminous coal
and with only an ESP for particle collection and pollution control
will require the most activated carbon consumption and the most
steel for the ACI system. Table 4-5 shows the estimated performance
and resources needed for a single and multiple (two) ACI retrofit
on a 500 MWe boiler firing subbituminous coal and equipped with an
ESP. As shown, as long as at least 16 months are provided for
installation of ACI control technology,
then there should be sufficient time for the technology to be
installed. If a facility owner chose to install a Pulse-Jet Fabric
Filter (PJFF) in addition to the ACI system for the purpose of
improving sorbent utilization, the project time would necessarily
be lengthened beyond this 16-month period to allow for the
installation of the PJFF. As stated in section 4.4, the total
duration for a PJFF retrofit is estimated to be anywhere from about
21 months to 26 months, including pre-contract effort and start up
and commissioning. Since the Clear Skies Act provides much more
than 26 months of notice for any mercury control regulation, there
should be adequate time for compliance even if some facilities
install PJFFs.
q~ÄäÉ=QJQK==bëÃáã~ÃÉÇ=mÉêÑçêã~åÅÉ=~åÇ=oÉëçìêÅÉë=kÉÉÇÉÇ=Ñçê=páåÖäÉ=^`f=oÉÃêçÑáÃ
råáÃë bñ~ãéäÉ=N bñ~ãéäÉ=O bñ~ãéäÉ=P bñ~ãéäÉ=Q bñ~ãéäÉ=R
bñ~ãéäÉ=S bñ~ãéäÉ=T
kçÃÉW==^ååì~ä=ÅçåëìãéÃáçå=î~äìÉë=~êÉ=ÉëÃáã~ÃÉÇ=ìëáåÖ=UR=éÉêÅÉåÃ=Ã…~é~ÅáÃó=Ñ~Ã…ÃçêK==eçìêäó=ÅçåëìãéÃáçå=î~äìÉë=~êÉ=~Ã=Ñìää=äç~ÇK===pÃÉÉä=~åÇ=i~Äçê=Çç
åçÃ=áåÅäìÇÉ=ëÃÉÉä=~åÇ=ä~Äçê=Ñçê=~=mgccI=áÑ=~ÇÇÉÇK
q~ÄäÉ=QJRK==bëÃáã~ÃÉÇ=mÉêÑçêã~åÅÉ=~åÇ=oÉëçìêÅÉë=kÉÉÇÉÇ=Ñçê=páåÖäÉ=~åÇ=jìäÃáéäÉ=^`f=oÉÃêçÑáÃë=EìëáåÖ=bñ~ãéäÉ
R=Ñêçã=q~ÄäÉ=QKQF
kìãÄÉê=çÑ=_çáäÉêë råáÃë N O bñáëÃáåÖ=`çåÃêçäë _çáäÉê=eÉ~Ã=o~ÃÉ
cìÉä=eÉ~Ã=s~äìÉ qçÃ~ä=jtÉ qçÃ~ä=eÉ~Ã=fåéìà mä~Ã¥Ã=`~é~ÅáÃó=c~Ã…Ãçê
`ç~ä=cäçïê~ÃÉ `ç~ä=jÉêÅìêó=`çåÃÉåà råÅçåÃêçääÉÇ=jÉêÅìêó
råÅçåÃêçääÉÇ=jÉêÅìêó mÉêÑçêã~åÅÉ
cìää=iç~Ç=^`=áåà ÉÅÃáçå=ê~ÃÉ ^`=fåà ÉÅÃáçå=o~Ãáç
cìää=iç~Ç=^`=áåà ÉÅÃáçå=ê~ÃÉ ^ååì~ä=^`=êÉèìáêÉÇ eÖ=ÅçåÃêçä=iÉîÉä
`çåÃêçääÉÇ=eÖ
~
=bm^=ãçÇÉäáåÖ=éêçà ÉÅë=ÃÃœ~Ã=ÃÜÉ=éä~Ã¥Ã=ïáÃÃœ=ÃÜÉ=ã~áãìã=åìãÄÉê=çÑ=^`f=êÉÃêçÑáë=áë=çåÉ=éä~Ã¥Ã=ïáÃÃœ=Ãïç
ÄçáäÉêë=~åÇ=~=ÃçÃ~ä=Ã…~é~ÅáÃó=çÑ=VMM=jtÉK
kçÃÉW==^ååì~ä=ÅçåëìãéÃáçå=î~äìÉë=~êÉ=ÉëÃáã~ÃÉÇ=ìëáåÖ=UR=éÉêÅÉåÃ=Ã…~é~ÅáÃó=Ñ~Ã…ÃçêK==eçìêäó=ÅçåëìãéÃáçå
î~äìÉë=~êÉ=~Ã=Ñìää=äç~ÇK
QKS pé~ÅÉ=oÉèìáêÉãÉåÃë
Most of the equipment and piping associated with a sorbent
injection system is fairly small and can be easily accommodated on
any facility. The only piece of equipment that could potentially be
a challenge to locate on site is the sorbent storage silo, the
other equipment largely being piping and a blower. The storage silo
is, by far, the largest part of the ACI system. It is estimated
that a storage silo that is sized for 15 days of AC storage at full
load for a 500 MWe plant firing bituminous coal and with only an
ESP would be about 10.7 m in diameter and about 26.7 m high.33 This
sized piece of equipment, while large, should be readily
accommodated on most sites large enough for a 500 MWe boiler. For
boilers with fabric filters, the size of the silo would be less
because of the lower sorbent injection rate. Some facility
operators may choose to install a PJFF in order to reduce sorbent
consumption and to segregate carbon from the ash. In this case,
more space would be needed for the PJFF. The dimensions for a PJFF
on a 500 MWe plant would be roughly 62 feet wide x 92 feet long x
90 feet high.35
`Ãœ~éÃÉê=R
póåÉêÖáÉë=çÑ=`çãÄáå~Ãáçåë=çÑ=`çåÃêçä=oÉÃêçÑáÃë=çå=~=páåÖäÉ=råáÃ
This chapter will explore the combination of these technologies
and how deployment of more than one technology at a unit could
potentially result in improved resource utilization. It is assumed
that the ACI, FGD, and SCR would not be necessary at a single unit
because of the high mercury removal efficiencies expected through
combination of FGD and SCR. Hence, the synergies of combining all
three technologies were not explored.
RKN p`o=~åÇ=cda=EpÅêìÄÄÉêF=fåëÃ~ää~Ãáçåë
In some cases, facility owners may choose to retrofit their
plants with both SCR and FGD technology to achieve both NOX and SO2
reduction. Combination of SCR and FGD will also result in
significant reduction of mercury emissions, thereby mitigating the
need for the addition of ACI. However, both SCR and FGD are very
capital-intensive projects, which require a substantial level of
material and construction. Therefore, it is worthwhile to consider
if both SCR and FGD installations can be combined efficiently.
An SCR project involves retrofitting in the boiler and its
immediate area. Therefore, an SCR retrofit project may require
relocation of equipment in the boiler area. An FGD system is
installed farther downstream in the plant, after the ESP.
Occasionally, it is necessary to install a new smoke stack, and it
may be necessary to add more fan capacity. However, an additional
smoke stack is normally unnecessary. The FGD connection with the
facility is generally less difficult than with SCR because it does
not require modification of the boiler, just connection to ductwork
in the vicinity of the stack. As a result, the construction
activities would normally be in different locations at the plant,
reducing the interference between the two projects. The SCR might
be the limiting item on the boiler outage because of its more
complex connection. In any event, the tie-in of the SCR and the FGD
systems could be done in the same outage, and it has been confirmed
that the installation of SCR and scrubber could be performed
simultaneously without interference.37 Therefore, installing these
at the same time on a boiler is preferable to doing them separately
as they may be able to use the same outage, and project
efficiencies result from a single mobilization, a single
construction manager, and sharing of large construction equipment
for the two projects. At Kansas City Power and Light's Hawthorn
Power Station, Unit 5 was replaced (excluding turbine) in under 22
months. This included the boiler, an SCR, and an LSD/FF.38
Although, in this case the equipment did not have to be erected
adjacent to an operating boiler, the erection included demolishing
and erecting a complete boiler island and demolishing the existing
electrostatic precipitator. Hence, this was a very complex project
that was completed approximately within the time frame estimated
and shown in Exhibit A-7 in Appendix A.
RKO jÉêÅìêó=`çåÃêçä=qÉÅÜåçäçÖó=~åÇ=pÅêìÄÄÉê=fåëÃ~ää~Ãáçåë
As noted in Chapter 4, ACI entails a much smaller construction
project than either an FGD or an SCR. Moreover, the ACI is located
in a different part of the plant than FGD or SCR and activated
carbon injection occurrs in the ductwork between the air preheater
and the ESP or FF. One benefit of combining
these two projects is that the ACI hookup can be completed
during the outage for the scrubber hookup, since the installation
effort necessary for the FGD will far outweigh that of the ACI
system. A second benefit is better planning of material storage and
handling equipment. Both FGD and ACI require a substantial amount
of material (limestone and AC, respectively) and associated storage
and handling facilities. Installing both technologies at the same
time will permit better planning of material storage and equipment
locations, thereby avoiding interference. Other benefits, such as a
single mobilization, a single construction manager, and sharing of
large construction equipment for the two projects exist, but they
are not expected to make a significant difference due to the
difference in size between the portions of the combined FGD and ACI
project. Therefore, as shown in Exhibit A-8 in Appendix A, the
schedule for a combined FGD and ACI project is expected to be the
same as the schedule of an FGD project.
RKP jÉêÅìêó=`çåÃêçä=qÉÅÜåçäçÖó=~åÇ=p`o=fåëÃ~ää~Ãáçå
As noted in Chapter 4, ACI entails a much smaller construction
project than either an FGD or an SCR. The primary benefit of
combining these two projects is that the ACI hookup can be
completed during the outage for the SCR hookup, since the
installation effort necessary for the SCR will far outweigh the ACI
system. Other benefits, such as a single mobilization, a single
construction manager, and sharing of large construction equipment
for the two projects exist, but they are not expected to make a
significant difference due to the difference in size between the
SCR and ACI portions of the combined project. Therefore, as shown
in Exhibit A-9 in Appendix A, the schedule for a combined SCR and
ACI project is expected to be the same as the schedule of an SCR
project.
`Ãœ~éÃÉê=S póëÃÉã=oÉëçìêÅÉ=^î~áä~ÄáäáÃó
Having assessed the resource requirements for individual or
multiple retrofits of control technologies, this chapter will
assess the resource availability in the United States for retrofit
of control technologies for the Clear Skies Act. This analysis
considers the current availability of resources for the
construction of control technologies and does not consider any
potential increase in production of resources due to the demand
created by the Clear Skies Act. Because this effect will be more
pronounced in the period following 2010 and because other market
factors may also change over time, the longer term projections are
of less value than those out to 2010. EPA has made preliminary
estimates of the retrofits of each technology that would result
from the Clear Skies Act. Tables 6-1a, b, and c list the expected
total MWe of facilities that would be equipped with SCR, FGD, or
ACI after response to a multipollutant rule. It is important to
note that the "Current Air Quality Rule Retrofit MWe" row of the
table includes only the projected retrofits under the current air
quality rules. The control technology retrofits estimated to result
from the Clear Skies Act, including the retrofits from current air
quality rules, is listed in the "Multipollutant & Current
Retrofits MWe" row of the tables. The "Cumulative Total" MWe shown
in Table 6-1 includes facilities that currently are equipped with
the technology or are expected to be equipped with the technology
as a result of current air quality rules, such as SCRs resulting
from the NOX SIP Call as well as the projected retrofits under the
Clear Skies Act. EPA estimated that up to 72 GWe of SCR would
result from the NOX SIP Call and an additional 13 GWe from
individual state multipollutant rules with approximately 14 GWe
currently installed. However, facilities are responding to the NOX
SIP Call at this time and it is uncertain exactly how many
facilities will ultimately be equipped with SCR in 2004 when the
NOX SIP Call deadline arrives.
EPA projections estimate that it would be cost effective for
32,000 MWe of FGD retrofits to be installed under the Clear Skies
Act by 2005 even though the first phase of the SO2 cap is not in
effect until 2010. These retrofits would be early installations
that sources initiate due to the economic benefits of banking SO2
allowances. It is estimated that there are about 4,000 MWe of FGD
capacity being constructed or just recently completed. Based on
availability of resources, particularly labor, it is projected that
an additional 6,000 MWe of FGD capacity could be built for a total
of 10,000 MWe by 2005. Because the FGD estimate based on
availability of resources is much less than the amount of FGD
capacity that would be cost effective to build, EPA ran model
sensitivities constraining the amount of scrubber capacity that
could be installed by 2005 at 10,000 MWe. This estimate for the
potential number of FGD retrofits considers the resource and labor
requirements of the simultaneous installation of SCRs, which is
further discussed under the labor section (6.2) of this chapter.
The 22,000 MWe difference, between the number of FGDs which would
be cost effective to build and the estimated number based on
resources, would be pushed back a few years to be completed by
2010. Therefore, the 53,000 MWe of FGD retrofits projected to be
built by 2010 for Clear Skies and current requirements remains the
same under both scenarios. It is likely that additional FGD
retrofits could be completed by 2005, but there would be the
potential for an increase in the cost of construction due to
decreased implementation time.
A typical unit size of 500 MWe was selected for each technology.
In previous sections, capacity factors of 85 percent were assumed.
In reality, coal-fired facilities, on average, have lower capacity
factors. For example, in 1999, 39.8 percent of 786 GWe of
generating capacity in the U.S., or 313 MWe, was coal fueled. In
that same year, coal-fired U.S. plants produced about 51 percent of
3,691 billion kWh, or
1,882 billion kWh.39 This corresponds to a capacity factor of
68.7 percent (Data is from Energy Information Administration Web
Site; Capacity Factor is total MWe-h produced divided by the total
MWe-h that would be produced if the plant were run at full capacity
for 8,760 h in the year). As a result, assuming a capacity factor
of 85 percent will result in a much more conservative (high)
estimate of resources needed than is likely to be the case.
In estimating the resources necessary to put new control
technology capacity in place (labor, steel, etc.), the
"Multipollutant & Current Rule Retrofits MWe" values of Tables
6-1a, b and c are of greatest interest. For estimating the
consumables necessary for the technologies, such as limestone,
ammonia, catalyst, or activated carbon, the "Cumulative Total MWe"
value is most important. The multipollutant and total MWe of
control technology retrofits are given for 2005, 2010, 2015, and
2020. To provide a conservative estimate of required resources, the
following analysis looks at implementing the retrofits for 2005 in
31 months prior to 2005 and retrofits for 2010, 2015, and 2020
three years prior to each five-year period. For example, it
estimates the resource requirements from the period between 2005
and 2010 over three years prior to 2010 instead of five years.
Thirty-one months was used for 2005, because the analysis for 2005
was based on the projected number of retrofits needed by 2005, less
the amount of capacity installed by May 2002. It should also be
noted that most of the retrofits needed by 2005 are being installed
to meet existing requirements under the NOX SIP Call or other
regulatory requirements as opposed to the requirements of a
multipollutant program such as the Clear Skies Act.
q~ÄäÉ=SJN~K==cda=oÉÃêçÑáÃë
pÅêìÄÄÉêë OMMR OMNM OMNR OMOM
~=J=VQIMMM=jtÉ=çÑ=ëÅêìÄÄÉêë=ÅìêêÉåÃó=áåëÃ~äÉÇ
Ä=Ó=áåÅäìÇÉë=ÃÜÉ=êÉÃêçÑáÃë=Ñêçã=ãìäÃáéçääìÃ~Ã¥Ã=ëÃê~ÃÉÖó=~ë=ïÉää=~ë==ÃÜÉ=éêçà ÉÅÃÉÇ=êÉÃêçÑáÃë=ÇìÉ=Ãç=ÅìêêÉåÃ=êÉÖìä~Ãáçåë
Ã…=Ó=fmj=êÉëìäÃë=êÉÑäÉÅÃ=ÃÃœ~Ã=áÃ=ïçìäÇ=ÄÉ=ÅçëÃJÉÑÑÉÅÃáîÉ=Ãç=áåëÃ~ää=POIMMM=jtÉ=I=ÄìÃ=Ä~ëÉÇ=çå=êÉëçìêÅÉ=~î~áä~ÄáäáÃó=áÃ=áë=ÉëÃáã~ÃÉÇ=ÃÃœ~Ã=NMIMMM=jtÉÅ~Ã¥=ÄÉ=ÄìáÃ
Ç=J=OVO=áë=ÃÜÉ=ÃçÃ~ä=åìãÄÉê=çÑ=éêçà ÉÅÃÉÇ=ìåáÃë=Ãç=ÄÉ=êÉÃêçÑáÃÃÉÇ=ïáÃÃœ=ëÅêìÄÄÉêë==Ñêçã=éêÉëÉåÃ=Ãç=OMOM
É=J=QMQ=áë=ÃÜÉ=~îÉê~ÖÉ=ëáòÉ=êÉÃêçÑáÃ=çÑ=ÃÜÉ=OVO=ìåáë=Ãç=ÄÉ=êÉÃêçÑáÃÉÇ=ïáÃÃœ=ëÅêìÄÄÉêë=EÇçÉë=åçÃ=áåÅäìÇÉ=ÃÜÉ=ÅìêêÉåÃ=VQ=dtÉ==F
Ñ=J=åìãÄÉê=áå=é~êÉåÃÜÉëÉë=êÉéêÉëÉåÃë=Åìãìä~ÃáîÉ=åìãÄÉê=çÑ=éä~Ã¥Ãë=ïáÃÃœ=?ñ?=ÄçáäÉêë=êÉÃêçÑáÃÃÉÇ=Ñêçã=éêÉëÉåÃ=Ãç=ÖáîÉå=óÉ~ê
fåÅêÉãÉåÃ~ä=kìãÄÉê=p`o ^îÉê~ÖÉ=páòÉ=råáÃ=EjtÉF
j~ñáãìã=kìãÄÉê=çÑ=råáÃë=~Ã=låÉ=mä~Ã¥Ã
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=T=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=S=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=R=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=Q=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=P=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=O=p`o=oÉÃêçÑáÃëÑ
kìãÄÉê=çÑ=mä~Ã¥Ãë=ïáÃÃœ=N=p`o=oÉÃêçÑáÃëÑ
Ö=Ó=ìé=Ãç=TO=dtÉ=çÑ=p`o=Ãç=ÄÉ=áåëÃ~ääÉÇ=Ñçê=kl=pfm=`~ää=ïáÃÃœ=NQ=dtÉ=çÑ=ÃÜÉ=ÃçÃ~ä=UR=dtÉ=~äêÉ~Çó=áåëÃ~ääÉÇI=NP=dtÉ=çÑ=áåëÃ~ää~Ãáçåë=éêçà ÉÅÃÉÇ
u
Ñçê=áåÇáîáÇì~ä=ëÃ~ÃÉ=ãìäÃáéçääìÃ~Ã¥Ã=êìäÉë
Ãœ=J=QST=áë=ÃÜÉ=åìãÄÉê=çÑ=ÃçÃ~ä=p`oë=áåëÃ~ääÉÇ=Ñêçã=éêÉëÉåÃ=Ãç=OMOM=EáåÅäìÇÉë=klu=pfm=`~ää=êÉÃêçÑáÃëFX=QRR=áë=ÃÜÉ=åìãÄÉê=çÑ=ÃçÃ~ä=p`oë=áåëÃ~ääÉÇ
Ñêçã=OMMR=Ãç=OMOM
áJ=QNP=áë=ÃÜÉ=~îÉê~ÖÉ=ëáòÉ=êÉÃêçÑáÃ=çÑ=ÃÜÉ=QST=éêçà ÉÅÃÉÇ=p`oë=EáåÅäìÇÉë=klu=pfm=`~ää=êÉÃêçÑáÃëF
q~ÄäÉ=SJNÃ…K==^`f=oÉÃêçÑáÃë
^`f OMMR OMNM OMNR OMOM
à =J=P=áë=ÃÜÉ=åìãÄÉê=çÑ=ÃçÃ~ä=^`f=áåëÃ~ääÉÇ=Ñêçã=éêÉëÉåÃ=Ãç=OMOM
â=J=QPP=áë=ÃÜÉ=~îÉê~ÖÉ=ëáòÉ=êÉÃêçÑáÃ=çÑ=ÃÜÉ=P=éêçà ÉÅÃÉÇ=^`fë=êÉÃêçÑáÃë
SKN póëÃÉã=e~êÇï~êÉ
The hardware items such as steel, piping, nozzles, pumps, soot
blowers, fans, tower packing, and related equipment required for a
typical SCR, FGD, or ACI systems installation are used in large
industries such as construction, chemical production, auto
production, and power production. Consequently, installation of
these technologies on many coal-fired utility boilers is not
expected to result in severe changes in demand for the hardware
items listed.
From Chapter 2, roughly 1,125 tons of steel is needed for a 500
MWe FGD system, which is about 2.25 tons per MWe. This is
conservatively high since there are some significant synergies
possible when there are multiple units on site. In particular, two
boilers with 900 MWe of capacity require approximately 2.1 tons per
MWe. From Chapter 3, an SCR for a coal-fired utility boiler
requires roughly 2.5 tons of steel per MWe for the typical size.
From Chapter 4, a 500 MWe facility will need about 175 tons of
steel to install an ACI system, or about 0.35 tons per MWe.
Estimated steel requirements for the projected retrofit MWe are
shown in Table 6-2 assuming that the retrofits occur over 31 months
prior to 2005 and over three years prior to 2010, 2015, and 2020.
For retrofits starting in 2005 facility owners are likely to have
more than three years to complete this work as many of these
retrofits have already begun. These time periods were chosen to
show that even under short periods of time, no significant impact
to U.S. steel supply is expected.
Census Bureau data on=U.S. steel shipments in 2000 was
approximately 108,703,000 tons, and imported steel was 30,993,000
tons for a total demand of about 140 million tons. An assumed
growth rate of US steel demand was chosen at 3 percent, a typical
number for growth in GDP. For each increment of time, the impact to
US steel demand was less than one tenth of one percent. Even if
there were no growth in the US steel production and imports from
2000 out to 2020, the amount of steel needed to complete the
retrofits for the Clear Skies Act would still be less than one
tenth of one percent of US production including imports.
Similarly, available supplies of piping, nozzles, pumps, soot
blowers, fans, and other related standard component necessary for
SCR, FGD or ACI installations are not expected to present
constraints on the ability of facilities to install the technology.
SCR catalyst is the only specialized piece of equipment that is
needed. Catalyst is discussed in Section 6.4.
q~ÄäÉ=SJOK==bëÃáã~ÃÉÇ=pÃÉÉä=oÉèìáêÉÇ=Ñçê=jìäÃáéçääìÃ~Ã¥Ã=fåáÃá~ÃáîÉ
vÉ~ê OMMR OMNM OMNR OMOM
kçÃÉW=^ëëìãÉë=ÃÃœ~Ã=~ää=êÉÃêçÑáÃë=çÅÅìê=çîÉê=PN=ãçåÃÃœ=éÉêáçÇ=éêáçê=Ãç=OMMR=~åÇ=~=PJóÉ~ê=éÉêáçÇ=éêáçê=Ãç=OMNMI=OMNRI=~åÇ=OMOMK==pÃÉÉä
ÇÉã~åÇ=ï~ë=ÇÉÃÉêãáåÉÇ=Ñêçã=ÃÜÉ="jìäáéçäìÃ~Ã¥Ã=~åÇ=`ìêÉåÃ=oìäÉ=oÉÃêçÑáë=jtÉÒ=êçï=çÑ=q~ÄäÉ=SJN~IÄIÃ…K==qÜÉëÉ=ÉëÃáã~ÃÉë
áåÅäìÇÉ=ÃÜÉ=êÉÃêçÑáë=ÇìÉ=Ãç=ÃÜÉ=kl=pfm=`~ääI=ÅìêêÉåÃ=êìäÉëI=~åÇ=ëÃ~ÃÉ=êìäÉëK==qÜÉ=ëÃÉÉä=ÇÉã~åÇ=Ñçê=p`o=ï~ë=~Çà ìëÃÉÇ=Ñêçã=UR=dtÉçÑ
uÇÉã~åÇ=áå=OMMR=Ãç=TN=dtÉ=Ãç=~ÅÅçìåÃ=Ñçê=NQ=dtÉ=çÑ=p`oI=ïÜáÅÜ=Ãœ~îÉ=ÄÉÉå=ÅçãéäÉÃÉÇ=Äó=j~ó=OMMOK
SKO i~Äçê
The installation of the SCR, FGD, and ACI control technologies
will require the following types of labor:
• general construction workers for site preparation and storage
facility installation;
• skilled metal workers for specialized metal and/or other
material assembly and construction; • other skilled workers such as
boilermakers, electricians, pipe fitters, painters, and truck
drivers; and
• unskilled labor to assist with hauling of materials, placing
of catalyst elements, and clean up.
From Chapter 2, it takes roughly 760 man-hours of labor per MWe
of FGD built. Chapter 3 showed that about 700 man-hours of labor
per MWe are required for an SCR system on a coal-fired boiler, and
Chapter 4 showed that roughly 10 man-hours of labor are needed per
MWe for an ACI system. Using these factors and the expected
retrofits, the labor requirement for SCR, FGD, and ACI retrofits
can be determined and are shown in Table 6-3. These estimates do
not take into account any synergies or efficiencies realized from
retrofitting multiple units on a site, as are described in Section
2.5 and 3.5, or from a combination of technologies, as described in
Chapter 5. Roughly 50 percent of an SCR project man-hours and 40
percent of an FGD project man-hours are for boilermakers.40 There
is little data on ACI breakdown of labor; however, a conservative
level of 50 percent is assumed. Using these rates and assuming the
above mentioned construction periods, it is possible to estimate
the number of fully employed laborers and boilermakers. The results
are shown in Table 6-3. The actual annual requirement for labor
would be less if the estimated number of retrofit installations
were evenly distributed over the full five-year increment of time
instead of the conservative three-year increment.
q~ÄäÉ=SJPK
bëÃáã~ÃÉÇ=^ååì~ä=`çåëÃêìÅÃáçå=~åÇ=_çáäÉêã~âÉê=i~Äçê=oÉèìáêÉÇ=Ñçê=`äÉ~ê=pâáÉë=^Ã…Ã=Eã~Ã¥JóÉ~êëF
vÉ~ê OMMR OMNM OMNR OMOM
M ^`f=_çáäÉêã~âÉêë MNN
qçÃ~ä=i~Äçê NMISTR NOIROR PIURO UIURO qçÃ~ä=_çáäÉêã~âÉêë RIOMM
RITRM NISMN QIMMN
kçÃÉW=^ëëìãÉë=ÃÃœ~Ã=~ää=êÉÃêçÑáÃë=Ñçê=OMMR=çÅÅìê=çîÉê=ÃÜáêÃóJçåÉ=ãçåÃÜë=éêáçê=Ãç=OMMR=~åÇ=çîÉê=~=ÃÜêÉÉJóÉ~ê=éÉêáçÇ=Ñçê=É~Ã…Ãœ=ÑáîÉJóÉ~ê
áåÅêÉãÉåÃ=~ÑÃÉê=OMMR=ÃÜêçìÖÜ=OMOMK==p`o=ä~Äçê=~åÇ=ÄçáäÉêã~âÉê=ÇÉã~åÇ=ïÉêÉ=~Çà ìëÃÉÇ=Ñêçã=UR=dtÉ=çÑ=ÇÉã~åÇ=áå=OMMR=Ãç=TN
dtÉ=Ãç=~ÅÅçìåÃ=Ñçê=NQ=dtÉ=çÑ=p`o=ïÜáÅÜ=Ãœ~îÉ=ÄÉÉå=ÅçãéäÉÃÉÇ=Äó=j~ó=OMMOK
Figure 6-1 shows a summary of construction worker labor
available in the United States. The data shows steady growth in
construction industry employment at the national level during the
1992 to 2000 period. Employment in the construction sector grew by
49.1 percent (4.1 percent annualized) over the period compared to
21.7 percent (2.0 percent annualized) for the economy as a whole.
The unemployment rate of 6.4 percent in 2000 compares to 4.0
percent for the whole economy.41
Figure 6-1. U.S. construction employment and unemployment
(Source: Bureau of Labor Statistics).
The available construction labor in the United States, about 6.7
million, will provide a large labor pool for the trades that are
not unique to the power industry, such as iron and steel workers,
pipe fitters, and electricians. In other words, that the estimated
demand of under 20,000 full-time workers represents only about 0.3
percent of the current total labor pool.
Boilermakers are a skilled labor source that is fairly unique to
utility work. Sixty percent of the demand for boilermakers in the
construction division is from the utility industry.31 Other
industries requiring boilermaker labor include refinery (13
percent), chemical (6 percent), paper (7 percent), and metals (6
percent). These are the industries where boilers and high-energy
vessels are most likely to be found. Retrofit of equipment on
utility boilers often requires a significant number of boilermakers
due to the integration that is needed with the boiler that often
requires modification of steam piping or other boiler equipment.
Also, in response to the increase in demand for boilermakers over
the last few years, their ranks have increased from 15,444 active
members in 1998 to 17,587* members in 2000 - an annualized growth
rate of 6.7 percent.31 Employment level also increased during this
time from 69.8 percent to 81.8 percent (employment level is equal
to the total man-hours worked in the year divided by total active
members time 2080 h/yr). During much of the 1990's the number of
active boilermakers had been declining due to very low employment
levels resulting from very low activity in the utility power plant
construction business.30 Therefore, the increased activity of the
last few years has been a welcome change to boilermakers.
Several sources have mentioned that the availability of
boilermakers has been tight for the SCR projects underway for the
NOX SIP Call.42, 43 However, where shortages have been experienced
in manning SCR construction projects with adequate numbers of
boilermakers, manpower planning had been done with short notice.44
Many boilermakers travel to work sites that are out of their local
area. A large project may require mobilization of several hundred
boilermakers to a site, which will frequently require pulling
* It is assumed that this number is for journeyman boilermakers
and does not include persons in apprenticeship programs.
members from other parts of the country. In the current,
competitive environment for utilities, power plant owners are
reluctant to provide much advance notice of when outages will
occur. Therefore, in some cases contractors must find manpower on
very short notice. The boilermaker's union (The International
Brotherhood of Boilermakers, Iron Ship Builders, Blacksmiths,
Forgers and Helpers) attempts to provide the necessary manpower to
the contractors. However, with very short notice, it is sometimes
difficult to move the manpower to the site in the short time
desired. Nevertheless, the union has been successful in providing
sufficient manpower to the project sites where they have had
adequate advance notice.44 Therefore, although there is little
slack in the availability of boilermakers, better coordination may
have avoided the labor shortage problems.
It is worthwhile to consider the expected future state of the
supply of boilermakers. The total number of members in the
boilermaker's construction division is currently about 24,000
journeymen and apprentices.44 The union has about 4000 members in
Canada. These numbers are up from the 1990's when a severe drought
of work for boilermakers caused many boilermakers to seek other
lines of work. Due to the current workload, the boilermaker ranks
are growing. However, the average age of the work force is about
48. Because of the aging workforce and because of the anticipated
demand for work at power plants, the union has made it an objective
to have at least 28,000 members in the construction division by
2005, or at least a 5.3 percent annual growth rate. The
boilermaker's union is working to recruit new members into their
apprenticeship programs, which takes four years to complete. Also,
skilled workers from other trades may choose to work as a
boilermaker, so a shorter apprenticeship may be possible, depending
upon the experience and skill level of the individual. For example,
iron and steelworkers who had been boilermakers in the past could
move back into boilermaker work very quickly. Since, boilermakers
earn somewhat more than ironworkers,42 it is reasonable to expect
that with increased job stability in the boilermaker trade, some
ironworkers might choose to move to the boilermaker trade for the
higher pay, especially if they had worked as boilermakers in the
past. The iron and steelworkers union has 150,000 members.45 Even
without prior boilermaker experience, some of these iron and
steelworkers could choose to move to boilermakers with much less
than a full four-year training requirement because of their
knowledge and skill level. In addition, the boilermaker's
shipbuilding division has about 30,000 members45 who, depending
upon industry conditions, could move over to the construction
division quickly.
As noted earlier, the number of boilermakers dropped quickly
during the 1990s when little work was available. Conversely,
increasing demand for boilermakers that would result from a
multipollutant rule should stimulate more workers to enter the
trade. The overall employment outlook for boilermakers should be
quite good, considering the work created by a multipollutant
initiative and the work on new power plants that is projected over
the next 20 years. As stated in the National Energy Policy (May
2001):
Over the next 20 years, the United States will need 1300 to 1900
new power plants. Electricity demand is expected to increase at a
rate of 1.8 percent per year over the next 20 years, creating the
need for 393,000 MWe of generating capacity. At a 1.5 percent
growth rate that number is reduced by between 60,000 to 66,000 MWe
to about 330,000 MWe of new generating capacity.
A large quantity of new generating capacity, consisting mostly
of gas combined cycle units, has been built within the last several
years. Since 1998, close to 200 GWe of new capacity have been built
or is currently under construction with an even larger quantity
being proposed. This excess in capacity is projected to create an
overall reserve margin greater than 25 percent in the US over the
next few years. By comparison, this is a significant increase in
the reserve margin since it dipped below 10 percent in the late
1990's. As the demand in electricity grows, the need for new
generating capacity will not be felt
until the excess capacity is worked off. Assuming new capacity
will be needed when the reserve margin approaches 15 percent, it is
expected to push back the need for additional capacity beyond 2005
and in some regions as late as 2010.
Due to the installation of SCR units for the NOX SIP Call, a
significant percentage of the boilermakers who are currently
working in the utility industry would be needed to complete those
retrofits by 2004. Integrated Planning Model (IPM) projections
indicate that it would be cost effective to install 32 GWe of
scrubbers by 2005 in addition to the projected SCR installations;
however, boilermaker labor is not expected to be sufficient to meet
this demand even if their numbers grow at the projected 5.3 percent
annual growth rate. Figure 6.2 shows the boilermaker labor
requirements out to 2010 assuming 32 GWe of scrubbers and 85 GWe of
SCR installations are installed by 2005 and compares the demand to
the supply of labor.
Assuming that the boilermaker membership grows at a 5.3 percent
growth rate out to 2005, it is estimated that there will be
sufficient new boilermaker labor to complete approximately 10 GWe
of scrubber retrofits by 2005. Considering that 4 GWe of scrubber
capacity is either being built or recently constructed, it is
conservatively assumed than an additional 6 GWe could be completed
by 2005. This estimate of 10 GWe of scrubber retrofits by 2005 was
determined from the difference between the boilermaker labor hours
available from the boilermaker membership working in the electric
utility industry and the labor hours needed to complete control
technology retrofits and other electric utility projects. This
estimate is supported by the number of orders of FGDs for 2001 and
projected orders through 2002 by the electric utility industry,
which totals over 11 GWe46, and over 13 GWe of announced scrubbers
which are scheduled to start up by 2005. Orders for scrubbers, such
as the recent order for the Coleman Station in Kentucky, are
continuing to be received in spite of the concerns raised about the
availability of boilermaker labor during the simultaneous
installation of SCRs for the NOX SIP Call. The other electric
utility projects that boilermakers work on include such projects as
routine maintenance at operating plants and new plant construction,
which account for approximately 13,500,000 man-hours of
boilermaker labor per year.30 Figure 6-3 compares the available
boilermaker labor to the demand from the electric utility industry
which includes the retrofits from the Clear Skies Act.
Since boilermakers earn more money than most other craft
trades42 and the demand for boilermakers should be steady and
increasing, it is reasonable to expect that the growth in
boilermaker numbers experienced these last few years should
continue for many more years. To assess the impact of this, it was
assumed that the boilermakers in the U.S. continued to grow at the
5.3 percent pace that the International Brotherhood of
Boilermakers, Iron Ship Builders, Blacksmiths, Forgers, and Helpers
has set as a minimum growth target. Based upon the estimates of
Table 6-3 and the assumed growth rates, the annual boilermaker
demand created by the Clean Skies Act can be estimated and is shown
in Table 6-4. Table 6-4 was derived considering that 14 GWe of SCRs
have already been installed for the NOX SIP Call, so the remaining
71 GWe of SCR and 10 GWe of scrubber installations were considered
for 2005. According to Table 6-4, if the retrofit of the FGD, SCR,
and ACI systems for 2005 occur over thirty-one months prior to 2005
and over a three-year period for each five-year increment after
2005 to 2020, the maximum demand would be about 23 percent of the
journeyman boilermakers or about 19 percent for journeymen and
apprentices combined.
Considering only the boilermakers who are currently in demand by
the utility industry, the demand would be about 38 percent of the
journeyman boilermakers or about 31 percent for journeymen and
apprentices combined. These percentages of demand are expected to
be experienced prior to 2010, but with growth in the boilermaker
numbers out to 2010, the percent of boilermakers affected drops
off. The number of boilermakers in demand for retrofit
installations under the Clear Skies Act is spread fairly evenly out
to 2010 when the demand begins to decrease. However, there may
still be significant demand for boilermakers after 2010 from other
power plant construction programs.
q~ÄäÉ=SJQK==bëÃáã~ÃÉÇ=^ååì~ä=_çáäÉêã~âÉê=aÉã~åÇ=`êÉ~ÃÉÇ=Äó=ÃÜÉ=`äÉ~ê=pâáÉë=^Ã…Ã
vÉ~ê OMMM OMMR OMNM OMNR OMOM
kçÃÉ=NW=^=RKP=~åÇ=SKT=éÉêÅÉåÃ=ÖêçïÃÃœ=ê~ÃÉ=áå=ÄçäÉêã~âÉê=à çìêåÉóãÉå=~åÇ=~ééêÉ
Ã¥ÃáÅÉë=áë=~ëëìãÉÇ=çîÉê=ÃÜÉ=éÉêáçÇK==fÃ¥=êÉ~
óI=ÃÜáë=ÖêçïÃÃœ
ê~ÃÉ=ïçìäÇ=éêçÄ~Ääó=Çêçé=çÑÑ=ëçãÉ=ÃáãÉ=~ÑÃÉê=OMNM=ìåäÉëë=ÃÜÉêÉ=ïÉêÉ=çÃÜÉê=ÇÉã~åÇJÖÉåÉê~ÃáåÖ=ÉîÉåÃëK==qÜÉ=ã~ñáãìã=ÖêçïÃÃœ
ê~ÃÉ=~ëëìãÉë=ÃÜÉ=ÄçáäÉêã~âÉê=ãÉãÄÉêëÜáé=ãÉÉÃë=ÃÜÉ=ìåáçå=Öç~ä=çÑ=OUIMMM=áå=OMMR=Ñêçã=OMMO=äÉîÉäë=çÑ=OQIMMM=~åÇ=ÃÜÉå=Öêçïë=~Ã
RKP=éÉêÅÉåÃ=ÃÜÉêÉ~ÑÃÉêK
kçÃÉ=OW=fÃ=áë=ÅçåëÉêî~ÃáîÉäó=~ëëìãÉÇ=ÃÃœ~Ã=~ää=ÅçåëÃêìÅÃáçå=Ñçê=êÉÃêçÑáÃë=Äó=OMMR=çÅÅìê=çîÉê=ÃÜáêÃóJçåÉ=ãçåÃÜë=éêáçê=Ãç=OMMR=~åÇ=çîÉê=~=ÃÜêÉÉJ
óÉ~ê=éÉêáçÇ=Ñçê=É~Ã…Ãœ=ÑáîÉ=óÉ~ê=áåÅêÉãÉåÃ=~ÑÃÉê=OMMR=ÃÜêçìÖÜ=OMOMK==p`o=ä~Äçê=~åÇ=ÄçáäÉêã~âÉê=ÇÉã~åÇ=ïÉêÉ=~Çà ìëÃÉÇ=Ñêçã=UR
dtÉ=çÑ=ÇÉã~åÇ=áå=OMMR=Ãç=TN=dtÉ=Ãç=~ÅÅçìåÃ=Ñçê=NQ=dtÉ=çÑ=p`o=ïÜáÅÜ=Ãœ~îÉ=ÄÉÉå=ÅçãéäÉÃÉÇ=Äó=j~ó=OMMOK
The actual impact on the demand for boilermakers could be lower
for several reasons. Due to the longer increments of time that the
Clear Skies Act provides facility owners to comply than was assumed
in this analysis, installation of these technologies will extend
over more than three years, spreading out the demand. As stated
earlier, this analysis does not consider any of the synergies or
efficiencies that have been demonstrated to occur on multiple unit
retrofits or multiple-technology retrofits. The boilermaker
population has been growing at a faster rate- 6.7 percent annually
- in recent years than the union's minimum target of 5.3 percent
that was assumed. Therefore, the number of boilermakers may
actually grow more quickly than what was assumed. This analysis
also neglects overtime, which would reduce the demand for workers
somewhat.
SKP `çåëÃêìÅÃáçå=bèìáéãÉåÃ
Most of the construction equipment necessary for the
installation of SCR, FGD, and ACI technology is standard
construction equipment that is used for most construction
activities. The piece of equipment that is not standard that may be
needed for SCRs and possibly for FGD systems is a tall-span
heavy-lift crane. These cranes are necessary to lift heavy pieces
(sometimes over 100 tons) several hundred feet and are not needed
for all projects. When the largest piece to be lifted is determined
from the construction plan, the necessary crane can be determined.
In some cases, the available crane or the crane pricing may limit
the largest piece to be lifted, and the construction plan may be
modified to accommodate a smaller crane by lifting smaller pieces.
In many instances, the best crane for the job is
not the largest because the large cranes are very expensive to
rent (one size up could double or triple the monthly charges for
renting the crane30). As a result, it may be more cost effective
overall to use a smaller crane and lift smaller pieces. This may
lengthen the installation time slightly, but it will reduce crane
rental fees. Therefore, an economic trade off must be assessed for
each project.
As discussed in Section 3.3, utility engineers reported that
while installing SCRs for the NOX SIP Call, crane availability has
been an issue that can be accommodated with proper planning. The
construction plan could be modified to employ the available or most
cost-effective crane. Therefore, sufficient supply of construction
equipment is expected to be available for installing air pollution
control equipment.
SKQ oÉ~ÖÉåÃë
The major groups of reagents considered in this Section include
limestone for FGD systems, SCR catalyst, Ammonia/Urea, and AC for
ACI systems.
iáãÉëÃçåÉ=Ñçê=cda=póëÃÉãë
Limestone is used for a wide range of purposes in the United
States. Overall limestone usage increased 22 percent over the four
years from 1995 to 1999 (annualized growth of 5.1 percent). Table
6-5 shows the production of crushed limestone sold or used by U.S.
producers.
q~ÄäÉ=SJRK==`êìëÜÉÇ=iáãÉëÃçåÉ=pçäÇ=çê=rëÉÇ=_ó=rKpK=mêçÇìÅÉêë
qçÃ~ä=rëÉvÉ~ê EÃÜçìë~åÇ=ÃçåëF
NVVV NIMUMIMMM NVVU NIMRMIMMM NVVT NIMNMIMMM NVVS VRSIMMM NVVR
UUQIMMM
pçìêÅÉW=rKpK=dÉçäçÖáÅ~ä=pìêîÉóI=jáåÉê~äë=vÉ~êÄççâI=sçäìãÉ=fK=jÉÃ~äë=~åÇ=jáåÉê~äëI
`êìëÜÉÇ=pÃçåÉI=NVVR=J=NVVVK==ÃœÃÃéWLLãáåÉê~äëKìëÖëKÖçîLãáåÉê~äëLéìÄëLÅçããçÇáÃóL
ëÃçåÉ|ÅêìëÜÉÇLáåÇÉñKÃœÃã
ä
As noted in Chapter Two, 500 MWe plant firing 4.0 percent sulfur
coal and equipped with LSFO FGD technology will use about 32 tons
per hour of limestone, or about 240,000 tons/yr (about 0.064
tons/MWh), and limestone consumption for MEL technology would be
less.6 Using an LSFO consumption rate is conservatively high, and
Table 6-6 shows expected consumption rates if all projected FGD
retrofits were LSFO technology and operated at 85 percent capacity
factor. The row "Multipollutant & Current Rule FGD Limestone
Consumption (tons)" provides an estimate of the limestone
consumption for the projected retrofits due to the multipollutant
strategy and current air quality rules. The row "Cumulative FGD
Limestone Consumption (tons)" provides an estimate of the limestone
consumption for the cumulative total number of FGD installations,
which includes 94 GWe of current installations. As shown, the
impact to total U.S. production for the multipollutant strategy
remains less than 2 percent out to 2020 while the overall demand
from all installed FGD remains less than 4 percent out to 2020.
p`o=`~Ã~äóëÃ
SCR catalyst is a critical part of the SCR system that is
manufactured on a worldwide basis by some of the largest companies
in the world. Manufacturing is largely in the United States,
Europe, and Japan, and the worldwide capacity is used to support
worldwide sales. The current and planned capacity of SCR catalyst
supply available to the U.S. market for coal-fired boilers is
nearly 90,000 m3/yr. Table 6-7 shows the results of a survey of
major suppliers of SCR catalyst to coal-fired boilers. The
suppliers provided EPA their current capacity and the capacity that
will be on line in the year 2002. The estimated capacity of other
suppliers of catalyst to coal-fired boilers that could not be
reached in time for this study is listed also. Suppliers that have
offered catalyst for coal applications in the past but currently
focus strictly on gas and oil -fired applications were not
included. However, it is recognized that these companies could
shift their product mix if the market conditions justified it, so
the capacity value shown could be quickly increased if
manufacturers simply changed product focus.
The current capacity was originally built overseas to meet
overseas demand or was subsequently built to meet U.S. demand for
catalyst spurred by the NOX SIP Call and the build up of gas
turbine power plants in the U.S. Except for a moderate demand for
replacement catalyst, much of this capacity will be available after
2004 because these large demand peaks will have mostly passed.
Because most of the companies that supply catalyst are divisions of
very large companies with the resources to rapidly expand their
manufacturing capacity to meet increases in market demand, it is
reasonable to assume that this manufacturing capacity could be
expanded if the market demand justified it. In fact, recent
capacity expansions provide strong evidence of this.
q~ÄäÉ=SJTK==p`o=`~Ã~äóëÃ=`~é~ÅáÃó=cçê=`ç~äJÑáêÉÇ=_çáäÉêë
`~é~ÅáÃó=ÃóéÉ `~é~ÅáÃó=s~äìÉ
`ìêêÉåÃ=`çåÑáêãÉÇ=`~é~ÅáÃó RRIPMM=ãPLóê
kÉï=`~é~ÅáÃó=`çãáåÖ=lÃ¥=iáåÉ OOIMMM=ãPLóê qçÃ~ä=`çåÑáêãÉÇ=`~é~ÅáÃó
TTIPMM=ãPLóê ^ÇÇáÃáçå~ä=bëÃáã~ÃÉÇ=`~é~ÅáÃó NMIMMM=ãPLóê
qçÃ~ä=bëÃáã~ÃÉÇ=`~é~ÅáÃóG UTIPMM=ãPLóêG
GqÜáë=çåó=áåÅäìÇÉë=Ã…~Ã~óëÃ=ã~åìÑ~Ã…ÃìêÉë=ïÜç=ÅìêêÉåÃó=ëìééäó=Ã…~Ã~óëÃ=Ñçê=Åç~äJÑáêÉÇ=~ééäáÅ~áçåëK
páÖåáÑáÅ~Ã¥Ã=~ÇÇááçå~ä=Ã…~é~Åó=ë=~î~ä~ÄäÉ=Ñêçã=ëìééäáÉë=ÃÃœ~Ã=ã~ó=Ãœ~îÉ=çÑÑÉêÉÇ=Ã…~Ã~óëÃ=Ñçê=Åç~ä
~ééäáÅ~áçåë=áå=ÃÜÉ=é~ëÃI=ÄìÃ=ÅìêêÉåÃó=ÑçÅìë=çå=Ö~ë=~åÇ=çä=JÑáêÉÇ=~ééäáÅ~áçåë=~åÇ=ÅçìäÇ=éçÃÉåÃá~ó
ëÜáÑÃ=Ã…~é~ÅáÃó=Ãç=~=Åç~ä=éêçÇìÅÃK
Currently, the equivalent of approximately 100 GWe of coal, oil,
and gas-fired capacity worldwide utilizes SCR technology. At these
worldwide installations, the volume of SCR catalyst in use is
estimated to be approximately 55,000 to 95,000 m3.10 Assuming that
one-twelfth of the current catalyst is
replaced each year on average, the annual demand for replacement
SCR catalyst is approximately 5,000 to 8,000 m3/yr. Note that the
estimate for the current annual demand is quite conservative since
the catalyst replacement rate on oil- and gas-fired combustion
units is likely to be less frequent than one-twelfth of the
catalyst per year. By 2005, an additional 85 GWe of coal-fired SCR
capacity is expected to be on line in response to the NOX SIP Call
and recently promulgated State rules (this includes anticipated SCR
retrofits under the state rules for Missouri, Connecticut, and
Texas). Assuming conservatively that one-eighth of the catalyst is
replaced each year on average for coal-fired units, the annual
demand for replacement SCR catalyst would increase by 12,600 m3/yr
by 2005. Adding the current annual replacement demand from
worldwide installations to the projected annual replacement demand
under the Clear Skies Act would yield a total of 17,600 - 20,600
m3/yr demand for replacement catalyst by 2005.10
The estimated annual demand for catalyst from the Clear Skies
Act, which consists of the demand due to new installations and
annual replacement is shown in Table 6-8. The highest catalyst
demand will occur by 2010. From Table 6-7, the estimated capacity
of catalyst supply is 87,300 m3/yr. Considering the initial fill
demand of 26,000 m3/yr from 65 GWe of SCR installations and
replacement demand of 22,300 m3/yr from 150 GWe of cumulative SCR
installations plus the worldwide catalyst replacement demand of
between 5,000 and 8,000 m3/yr, the annual excess capacity is
estimated to be 31,000 to 34,000 m3/yr. A more conservative
approach to determining if there is sufficient catalyst supply to
meet the demand from the Clear Skies Act is demonstrated in Figure
6-4. It compares the current cumulative production capacity for SCR
catalyst to the cumulative annual demand for SCR catalyst from the
total SCR installations in 2005 and 2010. This approach assumes
that the annual production of catalyst continues at the current
level of 87,300 m3/yr and starts accumulating in May 2002. If all
SCR systems were loaded with catalyst in just a one year period
prior to 2005 and 2010 instead of spreading out the loading over
several years, there would be sufficient accumulated supply to meet
the increased demand.
q~ÄäÉ=SJUK==bëÃáã~ÃÉÇ=^ååì~ä=p`o=`~Ã~äóëÃ=aÉã~åÇ=êÉëìäÃáåÖ=Ñêçã=`äÉ~ê=pâáÉë=^Ã…Ã=~åÇ=klu=pfm=`~ää
vÉ~ê OMMR OMNM OMNR OMOM
`~Ã~äóëÃ=Ñçê=åÉï=áåëÃ~ää~Ãáçåë=~I=ãPLóê PPIMMM OSIMMM NIOMM
NSIMMM oÉéä~ÅÉãÉåÃ=Ã…~Ã~äóëÃ=ÄI=ãPLóê NOISMM OOIPMM OOIUMM =OUITMM
qçÃ~ä=Ã…~Ã~äóëÃI=ãPLóê QRISMM QUIPMM OQIMMM =QQITMM
~
=fÃ=áë=~ëëìãÉÇ=ÃÃœ~Ã=áåëÃ~ää~Ãáçåë=Äó=OMMR=çÅÅìê=çîÉê=PN=ãçåÃÜë=éêáçê=Ãç=OMMR=~åÇ=çîÉê=~=ÃÜêÉÉJóÉ~ê=éÉêáçÇ=éêáçê=Ãç=OMNMI=OMNRI=~åÇ=OMOMK==p`oÃ…~Ã~äóëÃ=ÇÉã~åÇ=ï~ë=~Çà ìëÃÉÇ=Ñêçã=UR=dtÉ=çÑ=ÇÉã~åÇ=áå=OMMR=Ãç=TN=dtÉ=Ãç=~ÅÅçìåÃ=Ñçê=NQ=dtÉ=çÑ=p`o=ïÜáÅÜ=Ãœ~îÉ=ÄÉÉå=ÅçãéäÉÃÉÇ=Äó=j~ó
OMMOK Ä
=qÜÉ=êÉéä~ÅÉãÉåÃ=Ã…~Ã~
Ã=ï~ë=ÉëÃáã~ÃÉÇ=Ä~ëÉÇ=çå=ÃÜÉ=éêçà ÉÅÃÉÇ=åìãÄÉê=çÑ=p`o=áåëÃ~ä~áçåë=ÖáîÉå=áå=q~ÄäÉ=SJNÄ=~åÇ=ÇçÉë=åçÃ=áåÅäìÇÉ=ÃÜÉ
Ã…~Ã~äóëÃ=êÉéä~ÅÉãÉåÃ=ÇÉã~åÇ=Ñêçã=ÅìêêÉåÃ=NMM=dtÉ=çÑ=ïçêäÇïáÇÉ=p`o=áåëÃ~ää~ÃáçåëK
Utility power plants are already installing SCR catalyst for the
purpose of NOX SIP Call compliance in 2004. As shown in Figure 6-4,
the cumulative demand from the Clear Skies Act plus the worldwide
demand can be met with the total cumulative confirmed capacity.
Consequently, adequate capacity of SCR catalyst supply is available
to satisfy the demand that may result from the projected
installations. Of course, as demonstrated by the catalyst
suppliers, if more capacity was desirable to satisfy the market, it
could be added given sufficient lead time for the construction of
the catalyst production facility.
The ability to retrofit a large number of SCR systems over a
short period of time was exemplified in Germany during the late
1980s. Figure 6-5 shows the number of systems installed over an
eight-year period, with most of these systems (97 of 137) installed
during two consecutive years (1989-1990). This pattern of
installations exhibits that the catalyst market demonstrated the
ability to respond to the surge in demand resulting from a dramatic
increase in SCR installations.
^ããçåá~=~åÇ=rêÉ~
The installation and operation of SCR systems is not expected to
be constrained by the future availability of ammonia or urea. The
production of anhydrous ammonia in the U.S. in 2000 was
approximately 17,400,000 tons (equivalent anhydrous) with apparent
consumption of 22,000,000 tons and about 4,600,000 met through net
imports, as shown in a 2001 edition of U.S. Geological Survey
Minerals Commodity Summaries. Ammonia demand is directly
proportional to the tons of NOX reduced. The increased ammonia
demand from a multipollutant rule is estimated to increase to about
1,040,000 tons per year by 2020. This 4 percent increase in demand
over a nearly 20-year period can easily be met. Moreover, the U.S.
and worldwide ammonia business is struggling because of slumping
domestic demand and increased global capacity for the product and
other nitrogen fertilizers derived from it, such as urea.
Nevertheless, more capacity is scheduled to come on in the U.S.
during the near future. In addition, 1.2 million tons of capacity
is being built in Trinidad and Venezuela. Algeria and the former
Soviet Union have also added significant capacity. Another problem
is the withdrawal of China as an importer of ammonia. China
traditionally bought 3 to 6 million tons of urea annually (which is
produced
from ammonia), but in 1996, the country launched a drive to
become self-sufficient in urea, a move that has displaced 1.9 to
3.7 million tons of ammonia.47 Based on these estimates, the
ability to supply of ammonia will continue to exceed its demand,
even with the additional demand from newly installed SCR
systems.
SCR systems can also use urea as a reagent, and it is becoming
preferred to ammonia in many cases because of its safety. Urea is a
commonly available chemical with approximately 11,760,000 tons of
domestic annual production capacity.48 For SCR purposes, this adds
effectively another 6.7 million tons of ammonia annually available
as SCR reagent.* Additionally, U.S. urea manufacturers and
distributors routinely trade within a 130,000,000 tons worldwide
annual production capacity.10 Based on total world urea trade,
increased demand due to a multipollutant regulation would be well
under 2 percent of world trade if all SCRs used urea rather than
ammonia. And, like ammonia, the urea market is currently
experiencing an oversupply situation. Urea prices have fallen
precipitously since China, formerly a major buyer, decided to
strive for self-sufficiency. From 1994 to 1997, China opened nine
new urea plants and raised its domestic production by 50 percent.
U.S. producers knew China would bring on the new, more-efficient
plants, but they did not expect that country to continue running
its smaller, less-efficient ones.48 Thus, it is expected that this
worldwide supply will provide additional flexibility in meeting any
significant increases in demand. Since urea production is performed
on a worldwide basis, plants producing urea would be able to expand
their capacity if needed. Based on these considerations, adequate
urea supply is expected to be available for the SCR systems.
60
50
40
30
20
10
# of Installations Per Year
* It takes about 1.76 lbs of urea to make one lb of ammonia
reagent in a urea to ammonia conversion.
^`=Ñçê=^`f=póëÃÉãë
AC is produced in the United States and abroad for filtration
and other manufacturing purposes. Total AC usage in the United
States was 182,887 tons/yr in 2000, as given in the U.S. Census
Bureau Summary Current Industrial Reports for the Inorganic
Chemical Industry. Capacity in the U.S. is equal to 465 million
pounds/yr, or 233,000 tons.49 Both of these numbers include both
granular and powdered carbon, powdered being preferable to granular
for ACI applications. U.S. demand is projected to grow to 454
million pounds, or about 227,000 tons, in 2004.49 However, large
underutilized capacity overseas will provide a ready supply of
potential imports, which will tend to limit price increases for
most grades.49 The competition from Chinese and South-East Asian
producers remains strong.50 Chinese exports quadrupled from 53,230
tons in 1995 to 224,331 tons in 1997, with the average product cost
dropping by 16 percent to 660/ton. Therefore, growth in demand
experienced in the 1990's has not been reflected in the value of
the market due to over-capacity and the continued rise in Asian
exports. AC producers are concentrating increasingly on the
Asia-Pacific region to exploit growing markets and take advantage
of lower production costs; reported capacity expansions of over
15,000 tons/yr are all planned for Asia-Pacific and Russia.50
According to Norit, the largest supplier of AC for air pollution
control purposes, there is currently adequate excess capacity to
accommodate significant growth in the demand (tens of millions of
pounds/yr, or roughly tens of thousands of tons/yr).51 However,
depending upon how much growth occurs as a result of regulation,
additional capacity may be necessary. It would take 2-3 years to
add a plant; and this would only be done after a regulation was put
in place, the technical advantages of ACI for mercury removal were
proven relative to other approaches, and a clear time-line for
compliance was mandated.51 Therefore, even if a multipollutant
strategy implementation causes a large increase in demand for AC,
provided that the timing of compliance was clear and far enough in
the future, adequate supply of AC should be assured.
EPA estimated that, of the total 1,300 MWe to be retrofit by
2020 with ACI, all of that capacity would have existing fabric
filters. 36 As mentioned before, EPA's modeling indicates that none
of the total MWe of ACI retrofits will include a PJFF.36 AC usage
nationally for mercury control from power plants should be roughly
proportional to the total MWe of coal-fired facilities that are
equipped with the technology (this assumes an average capacity
factor of 85 percent and other assumptions of Tables 4-4 and 4-5).
Table 6-9 shows the results of this analysis. Based upon this
analysis, it is possible that existing excess capacity in AC
production could adequately address the increased demand for AC.
And, even if ACI is more broadly used than anticipated by EPA (more
than 1,300 MWe), it is clear that with at least 2-3 years of
preparation time to build more production capacity the AC industry
can accommodate any additional demand.
q~ÄäÉ=SJVK
mêçà ÉÅÃÉÇ=^`=aÉã~åÇ=aìÉ=Ãç=jìäÃáéçääìÃ~Ã¥Ã=fåáÃá~ÃáîÉ=EjtÉ=î~äìÉë=Ñçê=êÉÃêçÑáÃ=~êÉ=Ä~ëÉÇ=çå=bm^Ûë
ÉëÃáã~ÃÉë=Ñêçã=fmjF
vÉ~ê OMMR OMNM OMNR OMOM
bpm=H=^`fI=Ãçåë=éÉê=óÉ~êG M M M M cc=H=^`fI=Ãçåë=éÉê=óÉ~ê M M
RRM TOM qçÃ~äI=Ãçåë=éÉê=óÉ~ê M M RRM TOM
GléÉê~áçå=oÉëçìêÅÉëI=éêçà ÉÅÃÉÇ=~ååì~ä=~Åî~ÃÉÇ=Ã…~êÄçå=ÇÉã~åÇ
SKR`êÉ~Ãáçå=çÑ=gçÄë=ìåÇÉê=`äÉ~ê=pâáÉë=^Ã…Ã=ÇìÉ=Ãç=`çåÃêçä=qÉÅÜåçäçÖó
fåëÃ~ää~Ãáçåë
The Clear Skies Act is expected to create jobs for those
directly involved in the retrofit of facilities. These have been
estimated in Section 6.2 of this document. In addition to the jobs
that are directly created by this activity, jobs will be created
indirectly as a result of the economic activity that is stimulated
by additional discretionary income workers will have. Workers that
are directly employed on these clean air projects will purchase
consumer goods and services, which will stimulate additional
economic activity. To account for these indirect effects of
economic activity, economists use economic multipliers that are
related to worker's marginal propensity to consume. Economic
multipliers of 2 to 3 are often used.52 Using the lower multiplier
of 2 and the total labor estimates of Table 6-3, 25,000 additional
jobs may be created through indirect economic activity (2 times the
peak direct labor level of 12,500 workers indicated in Table 6-3).
This effect does not consider the additional job-gain potential
from U.S.-based equipment suppliers that export to other countries
the clean-air technology know-how they will gain from these
clean-air programs.
`Ãœ~éÃÉê=T `çåÅäìëáçåë
This report evaluated the resources necessary to comply with the
Clear Skies Act for which EPA estimated, by using the IPM, the
number, and size of facilities that will have to install new
hardware. The control technologies considered by this report as
candidates to be used for this multipollutant control strategy
include:
â– LSFO for the control of SO2; â– SCR for the control of NOX; and
â– ACI for the control of mercury.
Based upon the IPM-generated information from EPA and the
characteristics of the technologies listed above, the total
resources needed to comply with the multipollutant control strategy
were estimated and compared to the available resources. The
availability of resources was based on their current market demand
and does not reflect the increased production capacity that a
multipollutant strategy may create. It is likely that the market
for materials, labor, construction equipment, and other resources
used in the construction and operation of air pollution control
technologies would respond by increasing production to meet demand
where needed.
Installation of wet FGD, specifically LSFO, presents a
conservatively high estimate of anticipated resources and time to
provide additional control of SO2 emissions. LSFO systems commonly
are more resource intensive than many other FGD technologies.
Conservatively high assumptions were made for the time, labor,
reagents, and steel needed to install FGD systems. Although FGD
installations are time and labor intensive, they are typically
planned and installed within normally scheduled outages. It is
expected that one FGD system requires about 27 months of total
effort for planning, engineering, installation and startup. Modern
FGD systems typically use fewer and smaller absorbers and
increasingly control greater amounts of generating capacity using
common absorbers fed by multiple boilers. Under the Clear Skies
Act, three absorber systems for six boilers are anticipated to
handle 2,400 MWe of capacity. The estimate of labor includes
planning and engineering, general labor, and skilled boilermakers.
Construction of absorbers off-site is one way that projects can
control project resources, schedules, and labor.
Steel is the major hardware component for FGD systems.
Structural steel is used primarily for the absorber, ductwork, and
supports, and secondarily in miscellaneous components including
reinforcement of existing structures at a facility. FGD systems are
installed on the back end of a facility, are usually built close to
the ground, and do not require the amounts of structural steel
generally associated with elevated installations such as SCR. By
comparison, the conservative estimate of the amount of steel
required for a full FGD system is less than or equal to that
required for an SCR retrofit. Corrosion and abrasion resistant
materials are increasingly being applied with success in modern FGD
systems to improve reliability and long-term performance. The total
demand for additional FGD installations will be modest and is
expected to be well within the anticipated steel capacity, even
with demands from other applications.
Construction equipment requirements for FGD installations are
typically modest, particularly given that systems are installed at
the back end of the facility and close to the ground. However,
experience has indicated that project planning can surmount even
difficult situations (e.g., prefabrication and jacking up
components). Experience has also shown that specific site issues,
while often a planning challenge, have not prevented installations
of FGD systems. More recently, space requirements for construction
and accommodating the FGD system have been addressed with the
implementation of improvements in technology, including fewer and
smaller absorbers and more efficient on-site use and treatment of
wastes and byproducts.
Limestone was used as an estimate of reagent for FGD systems.
Experience indicates that the quantity of limestone is
conservatively high compared to other enhanced reagents such as
fine-ground limestone and MEL. Even with the assumption that all
new FGD capacity will require limestone, the amount of limestone
needed as a reagent is projected to be within availability of
supply.
SCR is the technology that will primarily be used for NOX
control. Since it is also the most demanding in terms of resources
needed for installation, it was assumed to be the only technology
used for NOX control. SCR systems are primarily made from steel,
standard mechanical hardware, and catalyst. Conservatively high
assumptions were made for steel, catalyst, reagents, and the labor
and equipment necessary to install the systems projected by the IPM
that result from a multipollutant control strategy. The amount of
ammonia or urea reagent needed can be estimated with good
confidence as constituting a small portion of available supply.
Experience in installing SCRs for the NOX SIP Call has shown
that the SCR equipment can be installed on the facilities in the
space provided. In some cases, moving of equipment has been
necessary, but this has not proved to be limiting. The only
specialized construction equipment that can be useful for SCR
installations are tall, heavy-lift cranes. These appear to be in
adequate supply and are not essential, since the erection plan can
be modified to accommodate the use of smaller cranes, which are
frequently more economical. The only specialized labor necessary
for SCR installations are members of the boilermakers trade, and
estimates of boilermaker demand were made. It is expected that one
SCR system requires about 21 months of total effort for planning,
engineering, installation, and start-up. Experience has shown that
many installations have been completed in much shorter times.
Therefore, 21 months appears to be somewhat conservative.
ACI was presumed to be the technology that would be used to
reduce mercury where dedicated mercury controls were needed because
the hardware is representative of most sorbent injection
technologies. Also, other sorbent-based approaches in development
may prove in time to be preferable to ACI, making the use of ACI
only a conservative assumption. ACI hardware is comprised of
relatively common mechanical components and is largely made of
steel. An ACI system requires much less in terms of steel, labor,
or other resources to install than either FGD or SCR technology.
Therefore, the impact of ACI hardware on resource demand is much
less than that of FGD or SCR technologies for SO2 or NOX control,
respectively. The only piece of equipment of any consequence in
terms of size is the storage silo, and this piece of equipment is
not so large as to pose a problem with regard to location for most
facilities. Planning, engineering, installation, and start-up of an
ACI system is only about 15 months and could be done in much less
time if administrative matters, such as permitting, occur more
quickly than assumed. Figures for consumption of AC were based on
prior, peer-reviewed EPA work, and conservative operating
conditions were assumed.
In summary, this study found that the expected demand for
resources resulting from a multipollutant control strategy could be
met. However, the market is expected to adjust to changes in both
the demand for resources under a multipollutant program and other
market factors. For this reason, the longer term
projections are of less value than those for the 2005 and 2010
time period. Table 7-1 shows a summary of resource demand and its
effect on current supply. In Table 7-1 the Supply Basis may be
current U.S. demand, capacity available to U.S. users, or other
basis as appropriate and described in the table notes. For all
resources needed for installation, it is assumed that these
resources are required over a 31-month period prior to 2005 and a
three-year period prior to 2010, 2015, and 2020. This is a very
conservative assumption because the most complex FGD installations
will require three years while the actual available time to
complete the projected retrofits for each period, except prior to
2005, is five years.
As shown in Table 7-1, there is ample steel and general
construction labor to support the installation of these
technologies, assuming a 31-month period of installation prior to
2005 and a three-year installation prior to 2010, 2015, and 2020.
Moreover, the demand assumptions do not consider any efficiencies
that can be achieved at multiple unit installations or
installations of multiple technologies at a site. As discussed in
this document, these efficiencies can reduce steel requirement
somewhat and labor needs substantially. Demand for boilermaker
labor is significant when compared to the boilermaker labor supply
basis. However, most boilermakers (60 percent) work in the electric
power industry, so it should not be surprising that the percentage
is high. It should also be considered that the value in this table
assumes conservatively high proportion of boilermaker labor and
that the boilermaker trade grows at its minimum target rate of 5.3
percent. In fact, the boilermaker trade has been growing at about
6.7 percent annually in recent years due to the improving
employment prospects for boilermakers.
There is also ample SCR catalyst capacity to supply this market.
SCR catalyst manufacturing is almost entirely dedicated to power
generating applications. Thus, it should not be surprising that
demand for initial fill and periodic replacement catalyst should
account for a significant portion of the supply basis. Moreover,
the U.S. market for catalyst is currently larger than all of the
other national markets combined. With regard to reagents and other
consumables, clearly there is ample supply of limestone for
additional FGD systems, especially in light of conservatively high
assumptions that were used to make these estimates. Ammonia and
urea supply is also plentiful, although it is expected that NOX
reduction will cause a modest increase in U.S. demand. In fact,
there is currently a worldwide excess capacity problem for
suppliers of these commodity chemicals that are traded globally.
Although U.S. demand for activated carbon is expected to increase
by a small amount as a result of a multipollutant strategy,
activated carbon is traded on a global basis, and there is
currently substantial excess capacity that can readily provide for
this increase in demand. Suppliers have also indicated that new
plants could be brought on line within 3 years, if needed, to
satisfy increased demand. Additionally, there are other
technologies under development that potentially could reduce
activated carbon demand from what is estimated here.
q~ÄäÉ=TJNK==bëÃáã~ÃÉÇ=oÉëçìêÅÉë=kÉÉÇÉÇ=Ñçê=fåëÃ~ää~Ãáçå=~åÇ=léÉê~Ãáçå=çÑ=qÉÅÜåçäçÖáÉë
fåëÃ~ää~Ãáçå=oÉëçìêÅÉë == =
G=oÉëçìêÅÉ=ÇÉã~åÇ=Ñêçã=p`o=ï~ë=~Çà ìëÃÉÇ=Ñêçã=UR=dtÉ=çÑ=ÇÉã~åÇ=áå=OMMR=Ãç=TN=dtÉ=Ãç=~ÅÅçìåÃ=Ñçê=NQ=dtÉ=çÑ=p`o=ïÜáÅÜ=Ãœ~îÉ=ÄÉÉå=ÅçãéäÉÃÉÇ=Äó
j~ó=OMMOK
`Ãœ~éÃÉê=U oÉÑÉêÉåÅÉë
1. Personal Communication with R. Telez, Babcock & Wilcox,
August 2001.
2. DOE/EPRI: Materials & Components Newsletter, No. 141,
August 1, 1999.
3. Personal Communication with P. Croteau, Babcock Borsig
Power, August 2001.
4. Feeney, S., Gohara, W.F., Telez, R.W., "Beyond 2000: Wet FGD
in the Next Century," ASME International Joint Power Generation
Conference, October 1995.
5. Fukasawa, K., "Low Cost, Retrofit FGD Systems," IEA Coal
Research, London England, September 1997.
6. Personal Communication with D. Foerter, Institute of Clean
Air Companies, August 28-29, 2001.
7. Wu, Z., "Materials for FGD Systems," IEA Coal Research,
London England, January 2000.
8. Srivastava, R. K., Jozewicz, W., "Controlling SO2 Emissions:
A Review of Technologies," EPA-600/R-00-093 (NTIS PB2002-101224),
Research Triangle Park, NC, December 2000.
9. Personal Communication with S.Kumar, FLS miljo, September
2001.
10. U.S. Environmental Protection Agency, "Feasibility of
Installing NOX Control Technologies by May 2003," September
1998.
11. Bultmann, A., Watzold, F., "The Implementation of National
and European Legislation Concerning Air Emissions from Large
Combustion Plants in Germany," UFZ-Centre for Environmental
Research Leipzig-Halle, August 2000.
12. North American Electric Reliability Council (NERC), "Impact
of FGD Systems: Availability Losses Experienced by Flue Gas
Desulfurization Systems," NERC Generating Availability Trend
Evaluations Working Group, July 1991.
13. Personal Communication with J. Bushman, Alstom Power,
February 19, 2002.
14. Klingspor, J.S., Brown, G.N., "Techniques for Improving FGD
System Performance to Achieve Ultra-High SO2 Removal Efficiencies,"
In the Proceedings of the U.S. EPA-DOE-EPRI Combined Power Plant
Air Pollutant Control Symposium: The Mega Symposium and The
A&WMA Specialty Conference on Mercury Emissions: Fate, Effects,
and Control, Chicago, IL, August 20-23, 2001.
15. Advanced Flue Gas Desulfurization Demonstration Project,
DOE/NETL project fact sheet,
http://www.lanl.gov/projects/cctc/factsheets/puair/adflugasdemo.html
16. Feeney, S., Gohara, W.F., Telez, R.W., "Beyond 2000: Wet
FGD in the Next Century," Presented at the ASME International Joint
Power Generation Conference, October 1995.
17. CT-121 Chiyoda Thoroughbred 121 Flue Gas Desulfurization
Process, http://www.chiyodacorp.com/select business, environmental
preservation, Chiyoda Thoroughbred 121 (CT-121) FGD process
18. Claussen, R. L., Martin, C. E., Smith, P. V., Oberjohn, W.
J., Weber, G. F., "Engineering and Design Guidelines for Duct
Injection Retrofits," In the Proceedings of the 1993 SO2 Control
Symposium, Boston, MA, 1993.
19. Nischt, W., Woolridge, B., Hines, J., Robison, K.,
"Selective Catalytic Reduction Retrofit of a 675 MWe Boiler at AES
Somerset," Presented at the ICAC Forum 2000, Washington, DC, March
23-24, 2000.
20. Siemens, Kat Treff '97, Washington, D.C. June 3-5,
1997.
21. Personal Communication with J. Urbas, Reliant Energy,
August 13, 2001.
22. Glaser, R., Licata, A., Robinson, T., "The SCR Retrofit at
the Montour Steam Electric Station," Electric Power Generation
Association Meeting, Hershey, PA, October 24-25, 2000.
23. Hartenstein, H., Servatius, P., Schluttig, A., "Lifetime
Extension of SCR De-NOx Catalysts Using SCR-Tech's High Efficiency
Ultrasonic Regeneration Process," Presented at the Coal-Gen
Conference, Chicago, IL, July 25-27, 2001.
24. Personal Communication with M. Gialanella, Hamon Research
Cottrell, August 7, 2001.
25. Air Daily, "For Now, Labor Capable of Meeting SCR Demand,"
Clear Air Regulations and Markets, Vol. 8, No. 138, July 18,
2001.
26. Personal Communication with a utility engineer that asked
to remain anonymous, August 6, 2001.
27. Personal Communication with J. Bushman of Alstom, August 8,
2001.
28. Babb, B., Angelini, E., Pritchard, S., "Implementation of
SCR System at TVA Paradise Unit 2," Presented at the ICAC Forum
2000, Washington DC, March 23-24, 2000.
29. Cochran, J., Hellard, D., Rummenhohl, V., "Design and
Initial Startup Results from the New Madrid SCR Retrofit Project,"
Presented at the ICAC Forum 2000, Washington DC, March 23-24,
2000.
30. Hines, J., Kokkinos, A., Fedock, D., "Design for
Constructability - A Method for Reducing SCR Project Costs," In the
Proceedings of the U.S. EPA-DOE-EPRI Combined Power Plant Air
Pollutant Control Symposium: The Mega Symposium and The A&WMA
Specialty Conference on Mercury Emissions: Fate, Effects, and
Control, Chicago, IL, August 20-23, 2001.
31. Personal Communication with T. Licata, Babcock Borsig
Power, February 20, 2002.
32. Personal Communication with T. Licata, Babcock Borsig
Power, August 3, 2001.
33. Personal Communication with C. Martin of ADA Environmental
Solutions, August 14, 2001.
34. Personal Communication with M. Durham, ADA Environmental
Solutions, August 3, 2001.
35. E-mail from Rich Miller, Hamon Research-Cottrell, March 19,
2002.
36. Personal Communication with Chad Whiteman, U.S.
Environmental Protection Agency, February 22, 2002.
37. Personal Communication with John Bushman, Alstom, July 10,
2001.
38. Burnett, G., Tonn, D., Redinger, K., Snyder, R., Varner,
M., "Integrated Environmental Control On the 21st Century's First
New Coal-Fired Boiler," In the Proceedings of the U.S. EPA-DOE-EPRI
Combined Power Plant Air Pollutant Control Symposium: The Mega
Symposium and The A&WMA Specialty Conference on Mercury
Emissions: Fate, Effects, and Control, Chicago, IL, August 20-23,
2001.
39. Electricity production, Energy Information Administration
web site
http://www.eia.doe.gov/cneaf/electricity/epav1/elecprod.html .
40. Personal Communication with Tony Licata, Babcock Borsig
Power, February 20, 2002.
41. Bureau of Labor Statistics Website, Industry at a Glance,
Construction, http://www.bls.gov/iag/ iag.construction.htm.
42. "Market Gyrations Make Hitting Targets for Skilled Crafts
an Art", Power Magazine
, vol. 146, no. 1, January/February 2002, pp.
28-32.
43. Hines, J., Jones, W., Wasilewski, K., "Reliant Energy SCR
Construction Implementation Plan", presented at ICAC Forum 2002,
Houston, February 12-13, 2002.
44. Personal Communication with Ande Abbot, International
Brotherhood of Boilermakers, Iron Ship Builders, Blacksmiths,
Forgers and Helpers, February 5, 2002.
45. Personal Communication (2) with Ande Abbot, International
Brotherhood of Boilermakers, Iron Ship Builders, Blacksmiths,
Forgers and Helpers, February 22, 2002.
46. Institute of Clear Air Companies, "Equipment Market
Forecasts," Issue Number 20, September 2001.
47. Chemexpo.com web site, November 29, 1999,
http://www.chemexpo.com/news/newsframe.cfm?
framebody=/news/profile.cfm.
48. Chemexpo.com web site, December 13, 1999,
http://www.chemexpo.com/news/newsframe.cfm?
framebody=/news/profile.cfm.
49. Chemexpo.com web site, April 23, 2001,
http://www.chemexpo.com/news/PROFILE010423.cfm.
50. Roskill Information Services Website,
http://www.roskill.co.uk/acarbon.html
51. Personal Communication with Bob Thomas, Norit Americas,
September 6, 2001.
52. Institute of Clean Air Companies, White Paper, November
2001.
^ééÉåÇáñ=^
fãéäÉãÉåÃ~Ãáçå=pÅÜÉÇìäÉë=Ñçê=`çåÃêçä=qÉÅÜåçäçÖó=fåëÃ~ää~Ãáçåë
Exhibit A-1: Single FGD
Exhibit A-2: Three FGD Modules on Six Units
Exhibit A-3: Single SCR
Exhibit A-4: Seven SCRs
Exhibit A-5: Single ACI
Exhibit A-6: Two ACIs
Exhibit A-7: Single FGD and SCR
Exhibit A-8: Single FGD and ACI
Exhibit A-9: Single SCR and ACI
iÉÖÉåÇW
Major Task - Dark Gray
Subtask - Light Gray
Key Completion Point - Black
A-2
A-3
A-4
A-5
A-6
A-7
A-8
A-9
A-10
A-11
