Escolar Documentos
Profissional Documentos
Cultura Documentos
1026742
Final Report, December 2012
NOTE
For further information about EPRI, call the EPRI Customer Assistance Center at 800.313.3774 or
e-mail askepri@epri.com.
Electric Power Research Institute, EPRI, and TOGETHERSHAPING THE FUTURE OF ELECTRICITY are
registered service marks of the Electric Power Research Institute, Inc.
Copyright 2012 Electric Power Research Institute, Inc. All rights reserved.
Acknowledgments
Steven L. Reid, P. E.
Industrial Environmental Systems, Inc.
John C. Sowizal, P. E.
Industrial Chimney Engineering Co., Inc.
iii
Product
Description
For the past 14 years, the design of wet stacks around the world has
been guided by the original EPRI Wet Stacks Design Guide (1996).
Since that time, the number of wet stack installations has grown
considerably, and a wealth of practical real-world operating and
maintenance experience has been obtained. The laws of physics have
not changed, and most of the information presented in 1996 is just as
valid today as it was when originally published. What has changed is
the power-generation industrys experience in using this information
and the day-to-day operation of wet stacks. Much had been learned
over the intervening years about the design and operation of wet
stack systems, and it had become clear that some updating of the
recommendations made in the original guide were needed. This
document, the Revised Wet Stack Design Guide, has been prepared to
present this updated information and to provide the powergeneration industry with the latest state-of-the-art information for
favorable wet stack design and operation. Much of this document
will be familiar to those who have read the original design guide.
Some sections of the original guide have been reused with only minor
changes; others have been significantly revised; and new sections
discussing the industrys experience with wet stack operation and
maintenance have been added. The outline of the guide has been
rearranged to be easier to use and follow, bringing the reader through
the entire wet stack design process, from the fundamentals of droplet
collection and liquid-film flow to the stacks final design,
commissioning, and operation. This new document strives to
thoughtfully update the original guide and to provide the industry
with the definitive reference needed by the engineers and designers
responsible for the specification, design, and implementation of
effective wet stacks.
Background
A wet stack is a chimney, stack, or flue that exhausts saturated flue
gas downstream from a wet-scrubbing process, such as a wet flue gas
desulfurization (WFGD) system. All recently designed and
constructed WFGD systems have installed wet stacks. Although the
technology is relatively mature, there are a number of technical issues
that utilities must address to achieve a successful installation. This
guide provides answers to these questions, whether the installation is
new or retrofit.
Objectives
To provide background information and updates of previously
published information
Approach
Investigators collected the information from a literature survey, the
in-house expertise of contractors, phone contacts with vendors, a
utility advisory committee, and a limited number of site visits. They
collated and summarized the information to produce the report,
which the advisory committee also reviewed.
Results
The information in the guide covers the design process and
operational issues for both new and retrofit wet stack installations.
Important issues addressed include system design for favorable wet
operation, stack liquid discharge, plume downwash, stack-liner
geometry, gas velocity in the liner, and liquid-collection devices and
drainage. In addition, the report also provides a guide to developing a
wet stack specification.
EPRI Perspective
Because most new FGD systems include wet stacks, it is imperative
that accurate, reliable information is available. This guide contains
the most up-to-date information, and it should be useful for
personnel responsible for wet stack design, specifications, or
operation. Care must be taken to use these recommendations with
good engineering judgment and consideration for site-specific
installations.
Keywords
Air-emissions control
Flue gas desulfurization (FGD)
Wet stacks
Wet scrubbers
SO2 control
Stack liquid discharge (SLD)
vi
Abstract
In 1996, the Electric Power Research Institute (EPRI), in a tailored
collaboration with New York State Electric & Gas (NYSEG),
retained Burns & McDonnell and DynaFlow Systems to prepare a
design guide for wet stacks. The purpose of this guide was to provide
the utility industry with information and recommendations
concerning the design and specification of wet stacks. Since that
time, the number of wet stack installations has grown considerably,
and a wealth of practical real-world operating and maintenance
experience has been obtained. This document, the Revised Wet Stack
Design Guide, has been prepared to present this updated information
and to provide the power-generation industry with the latest stateof-the-art information for favorable wet stack design and operation.
This new document strives to thoughtfully update the original guide
and to provide the industry with the definitive reference needed by
the engineers and designers responsible for the specification, design,
and implementation of effective wet stacks.
vii
Table of Contents
Section 1: Background and Objectives .....................1-1
1.1 Preface to the Revised Wet Stack Design Guide ............ 1-1
1.2 Introduction ............................................................... 1-1
1.3 Scenarios for Wet Stack Utilization .............................. 1-3
1.4 Important Considerations for Wet Stack Design and
Operation ....................................................................... 1-3
1.4.1 Mist-Eliminator Carryover ................................... 1-4
1.4.2 Deposition of Entrained Liquid Droplets ................ 1-4
1.4.3 Condensation .................................................... 1-5
1.4.4 Liquid Re-entrainment ......................................... 1-6
1.4.5 Washing of Wet Fans ........................................ 1-7
1.4.6 Stack Liquid Discharge (SLD) ............................... 1-7
1.4.7 Corrosion/Chemical Attack ................................ 1-8
1.4.8 Plume Downwash .............................................. 1-8
1.4.9 Icing Potential ................................................... 1-9
1.5 Contents of the Design Guide .................................... 1-11
1.5.1 Information About Key Issues ............................ 1-11
1.5.2 Knowledge of the State of the Art ...................... 1-11
1.5.3 Overview of Project Implementation ................... 1-12
Section 2: Wet Stack Design Fundamentals ..............2-1
2.1 Introduction ............................................................... 2-1
2.2 Droplet and Liquid-Flow Fundamentals .......................... 2-3
2.2.1 Flow of Droplets Suspended Within a Gas
Flow ........................................................................ 2-3
2.2.2 Liquid-Film Flow on Surfaces and Reentrainment Fundamentals ........................................... 2-4
2.2.3Gas-Flow in Wet Duct Systems ............................. 2-5
2.2.4Gas Flows in Wet Stack Systems .......................... 2-7
2.3 Sources of Liquid in a Duct/Wet Stack System ............ 2-12
2.3.1 Mist Eliminators and Absorbers ......................... 2-12
2.3.2 Liner Condensation .......................................... 2-20
2.4 Liquid-Film Behavior on the Liner Wall ........................ 2-22
2.5 Liner-Wall Discontinuities .......................................... 2-28
2.6 Re-entrainment ......................................................... 2-31
2.7 Recommended Liner-Gas Velocities ............................ 2-32
2.8 Stack Liquid Discharge ............................................. 2-33
ix
xii
List of Figures
Figure 2-1 Liner Inlet Gas-flow Patterns Close-Coupled
Absorber......................................................................... 2-7
Figure 2-2 Typical Stack-Inlet Arrangements .............................. 2-8
Figure 2-3 Gas-flow Patterns Through a Standard 90 Elbow ...... 2-9
Figure 2-4 Liner-Inlet Gas-flow Patterns Side-entry Breech ....... 2-10
Figure 2-5 Liner-Inlet Gas-flow Patterns 3-Miter Cut BottomEntry Elbow ................................................................... 2-11
Figure 2-6 Common Mist-Eliminator Arrangements................... 2-14
Figure 2-7 Comparison of Gas-Path Lengths Through a Mist
Eliminator as a Function of Angle ..................................... 2-15
Figure 2-8 Typical Chevron-Type Mist-eliminator Pressure
Drop vs. Gas-Velocity Curve ............................................ 2-17
Figure 2-9 Typical Chevron-Type Mist-Eliminator Carryover
vs. Gas-Velocity Curve .................................................... 2-17
Figure 2-10 Typical Mist-Eliminator Collection Efficiency vs.
Droplet-Size Curve ......................................................... 2-18
Figure 2-11 Lower Liner Droplet- and Liquid-Film Flow
Patterns Side-Entry Breech............................................. 2-23
Figure 2-12 Liner-Floor Liquid-Film Flow Patterns Side-Entry
Breech .......................................................................... 2-24
Figure 2-13 Lower-Liner Droplet- and Liquid-Film Flow
Patterns Bottom-Entry Elbow .......................................... 2-25
Figure 2-14 Wetting vs. Nonwetting Surfaces ......................... 2-26
Figure 2-15 Liquid-Film Flow Patterns on a Smooth, Vertical
Wetting Surface vs. Vertical Gas Velocity ......................... 2-28
Figure 2-16 Liquid Flow over 1/8-in (3.1-mm) Weld Beads ...... 2-29
Figure 2-17 FRP Liner-Can Joints Internal-Taper
Requirement .................................................................. 2-30
Figure 2-18 Plume Downwash............................................... 2-36
xiii
xv
List of Tables
Table 2-1 Recommended Stack-Liner Velocities for Wet
Operation ..................................................................... 2-33
Table 3-1 Estimated Ranges of Liquid Flows in the Wet
Duct/Stack of a Typical 550-MW Unit ............................... 3-5
Table 3-2 Liner Material of Construction ................................. 3-21
Table 3-3 Physical Properties and Chemical Requirements
Of Acid-Resistant Brick1 .................................................. 3-22
Table 3-4 Recommended Stack Velocity Range for StacklinerDiameter Sizing ..................................................... 3-34
Table 4-1 Phase II Wet Stack Design Process ............................ 4-4
Table 4-2 Estimated Ranges of Flows in the Wet Duct/Stack
of a Typical 550-MW Plant ............................................... 4-8
xvii
1-1
The 1990 Clean Air Act Amendments (CAAA) required utilities to reduce
emissions of SO2. Many utilities have added WFGD systems to comply with the
Phase I requirements, whereas a lesser number have implemented dry-scrubber
technologies. Phase II of the program went into effect in 2000, further reducing
SO2 emissions and increasing the number of plants affected. As part of their
Phase II Compliance Plan, many plants with partial flue gas bypass systems have
decided to eliminate the bypass and scrub 100 percent of their flue gas. Many
other utilities have installed new or retrofit FGD systems, which typically use wet
stack operation because of reduced operating and maintenance costs.
The design of ducts and stacks for wet operation must address several issues that
were not present in unscrubbed or reheated gas stack designs. Some of the
important issues to consider in the design of a wet duct/stack system include
Corrosion/chemical attack
Stack height
Liner-breach geometry
The purpose of this revised guide is to provide the utility industry with
information and recommendations concerning the specification and design of
wet stacks. However, these recommendations should not be used without
applying good engineering judgment and consideration for site specifics.
Operating conditions, design conditions, and economics all play important roles.
The goals of the Revised Wet Stack Design Guide are to
Identify the parameters and options that will lead to favorable wet operation.
Give specific recommendations regarding wet stack design for new and
retrofit installations.
It is assumed that those who will use this guide have a general familiarity with
stack designs and FGD systems. This guide is intended for personnel who are
considering wet stacks and/or are responsible for designs, specifications, or
operation of wet stacks.
1-2
A new FGD system that will convert an existing dry stack to wet operation
The first scenario involves adding a new FGD system that incorporates a wet
scrubber and a new wet stack. In such a scenario, the designer is given the most
freedom regarding plant layout, liner material, and stack-geometry selections. A
state-of-the-art wet stack design should be used that is based on the latest
analytical, experimental, and field data available. The goal is to produce an
effective wet stack design that minimizes both installation and operating costs.
The second scenario involves converting an existing dry stack to wet operation.
In such a scenario, the most important design consideration is whether or not the
existing liner and ductwork material can withstand the reduced temperatures and
wet conditions typical of wet stack operation and whether the liner-gas velocity
will result in droplet re-entrainment from the stack walls. If not, the existing
chimney liner must be modified or a new stack installed. Because of the outage
time required to modify an existing stack, it is often more economical to build a
new stack for scrubbed gas and use the existing stack for emergency gas bypass
around the scrubber as needed.
1.4 Important Considerations for Wet Stack Design and
Operation
The liquid inside a stack with a WFGD system exists as droplets from misteliminator carryover and as the moisture content of the gas flow entering the
stack liner. The vapor content is usually the maximum that can occur when the
flue gas is saturated with water vapor at the operating temperature. If there is an
induced-draft (ID) fan between the absorber and the stack, the fan-temperature
rise could result in a lower-than-saturation vapor content. What happens to the
vapor and liquid content before the gas reaches the top of the liner defines how
much liquid there is and in what form it is discharged from the stack. The major
gas/liquid flow processes that may lead to the SLD are described in the next four
subsectionsincluding the major features of the SLD. A description of the
liquid-flow balance in wet stacks is provided in [1], pp. 4-3 to 4-5. Liquid-flow
rates typical for wet ducts and stacks are given in Section 3.4, Wet Duct
Operation, Table 3-1.
1-3
1-5
Surface discontinuities and protrusions, such as weld seams, fiberglassreinforced plastic (FRP) joints, and mortar joints in brick liners that disrupt
gas and liquid flow locally.
Vanes and baffles that cause gas-flow separation and recirculating liquid.
Gas-flow patterns that drag liquid along the surface to re-entrainment sites.
Strong vortex patterns that can pick water up in the core of the vortex from
horizontal or vertical surfaces.
Thermal-expansion joints
Stack floor
collector system. Droplets over 100 m do not evaporate outside of the stack, and
they will probably reach the ground level.
1.4.5 Washing of Wet Fans
Occasionally, ID fans will be located between the absorbers and the stack of a
utility power plant when retrofitted with an FGD system. These fans usually
require periodic or continuous washing to prevent solids buildup on the fan
impellers, which can cause fan-rotor imbalance. All of the liquid sprayed into the
fan inlet leaves the fan impeller as droplets. Most of the liquid will be propelled
by the high centrifugal-force field to the fan scroll, where it will deposit. Most of
this liquid will re-entrain back into the gas flow unless a properly designed highvelocity scroll-liquid collector is installed. Preferably, fans should be located
upstream of the absorber to avoid these problems. Alternatively, they should be
located such that there are a sufficient number of duct turns and/or duct lengths
between the fan outlet and the stack inlet to provide opportunities for liquid
collection.
1.4.6 Stack Liquid Discharge (SLD)
The stack liquid discharge (SLD) is also known as rainout or acid-mist fallout. It
is important to note that all wet stacks have some amount of SLD. However,
SLD is only a problem if the droplets are large enough to be detectable at ground
level near the stack.
The amount of liquid discharged at the top of the stack is a result of gas- and
liquid-flow processes that take place in the ductwork and stack system between
the discharge of the absorbers and the top of the stack liner. The source of the
liquid discharge can be categorized by the liquid-flow process and by the droplet
sizes as follows:
Liquid droplets carried from the absorber to the top of the stack liner by the
gas stream without depositing along the gas-flow path. Typically, these are
droplets less than 50 m in diameter, the majority of which can travel
through the ducts without significant deposition. Droplets of this size, when
ejected from the stack, will typically evaporate before reaching the ground.
Liquid deposited and condensed on the liner, which flows either upward or
downward, depending on the local gas velocity. Liquid flows upward in those
areas of the stack where the gas velocity is higher than the flow-reversal
velocity for the specific liner material used. This phenomenon can be a
significant problem in tapered liners, in which the upper portion of the liner
1-7
can experience high gas velocities. The droplet sizes re-entrained at the top
of the liner, where velocities are highest, can be quite large (3006000 m),
depending on the liner-top geometry. The upper section of a stack choke is
usually in this mode of upward liquid flow.
The quantity and location of the fallout at a given plant is a function of the
droplet-size distribution of the discharged liquid and the atmospheric
conditionssuch as ambient temperature, wind, relative humidity, and
turbulence level. The very small droplets of the bulk condensation make the gas
plume white and visible. The plume is cooled as it mixes with the surrounding
air. This cooling results in condensation on the small droplets in the plume. The
increase in diameter is only 1 to 10 m; therefore, the droplets of the bulk
condensation can evaporate before reaching the ground and cannot be detected.
The mixing with the drier air away from the stack causes droplet evaporation.
The balance between the cooling and mixing processes defines the length of the
visible plume. Therefore, it is a function of the ambient air temperature and the
air relative humidity.
One of the major objectives in developing an effective wet stack design is to limit
SLD to a minimal acceptable level. The goal is to limit the droplet size that will
exit the stack. If the droplets are small enough, they will evaporate before hitting
the ground. However, if the droplets are large, they will land on nearby surfaces
such as plant structures, equipment, and cars. The wet stack survey indicates that
the fallout of liquid droplets, if noticeable, usually occurs on the downwind side
of the stack within a half-mile radius.
1.4.7 Corrosion/Chemical Attack
With wet stack operation, corrosion is less severe than with reheat, but it still
must be considered. Corrosion can be resisted by means of proper material
selection. The floors of the ductwork should be sloped to provide proper drainage
of the corrosive liquid after shutdown of the unit. The exposed materials of the
ductwork and the stack liner must be resistant to chemical attack. Several types of
material are available for both the scrubber-outlet duct and the stack liner. The
materials should be selected according to their corrosion resistance to the
chemical properties of the flue gas downstream of the absorber. Refer to Section
3.9.1, Materials of Construction, for more information.
1.4.8 Plume Downwash
Designing for a low gas velocity at the stack exit can decrease the amount of
SLD. However, this type of design increases the potential for plume downwash
at elevated wind velocities. During downwash episodes, the reduced staticpressure zone generated along the downwind side of the stack shell and liner
extensions can draw the flue gas into a downwash pattern along the stack shell.
In this process, saturated flue gas comes in contact with the liner extension, stack
hood, and stack shell. This contact can lead to deterioration of the stack
construction materials because of exposure to acid in the flue gas. It also increases
the potential for ice formation on the top of the stack. A cross wind at the top of
1-8
the stack will deflect the plume from its vertical path. As the ratio of vertical
plume momentum to horizontal wind momentum [(V2 ) FLUEGAS / (V2)WIND]
falls below a value of about 2.0 for a single-flue free-standing stack, the plume
may become partially entrained in the low-pressure zone that is formed on the
downwind side of the stack. This phenomenon is known as plume downwash.
To reduce the frequency of downwash, the height of the liner(s) can be increased
above the shell, or the momentum ratio can be increasedeither by reducing the
liner diameter or by installing a choke at the top of the liner.
For a multiple-flue stack, the equivalent momentum ratios for initiation of plume
downwash are higher, and they vary significantly with the wind direction relative
to the axis of the liners. The extent of downwash can also be more pronounced,
because of the larger size of the shell needed to enclose the multiple liners.
The interaction of a plume with a cross wind is a function of the following
parameters:
Liner-extension diameter
Shell diameter
Liner-extension height
Shape of hood/cap
Roof geometry
Number of liners
Wind direction, velocity, and air density at the top of the stack
of the stack. Whether ice forms on the top of the stack depends on the
temperature of the stack surface, the temperature of the mixture of saturated flue
gas and cold ambient wind, and on whether water vapor will condense out of the
mixture. Ice formation is most likely at plants where below-freezing temperatures
are common and where they last for extended periods of time.
The locations for potential ice formation are discussed below.
The hood for a
single-flue stack
The roof of a
multiple-liner stack
The stack roof usually does not directly touch the liners, which
permits expansion and movement of the liners. The roof often
receives heat from the annulus of the stack, which prevents ice
buildup and melts any snow that deposits there. Therefore,
accommodations must be made for the drainage of liquid from the
roof area. Preferably, drains should be located on the inside of the
stack shell to keep them from freezing. Alternatively, they will need
to be heat-traced. Depending on the specific geometry and
weather conditions, some icing can occur on the roof, and
therefore may require heat-tracing to eliminate excessive ice
buildup.
Railings
The metal railings at the top of the shell will quickly cool to the
temperature of the mixture of flue gas and ambient air, and they
will be the first surfaces to become iced; in other words, these
railings are the most likely places for ice to form.
A metal platform on the outside of the stack near the top can also
be exposed to low concentrations of flue gas as a result of
downwash. Because it is metal and is exposed to the ambient and
flue gas mixture temperature, it may also experience occasional ice
buildup. Such platforms are the second most likely places for ice to
form.
Shell
1-10
Icing can often lead to serious buildups that can fall to the ground. However,
when the icing is occurring, the platform near the top of the stack, the railings,
and possibly the roof, may be slippery. If icing is observed at the top of the stack,
care should be taken in the area around the base of the stack because of the
possibility that the ice could become dislodged and fall to the ground.
The potential for icing can be reduced by employing the following steps:
1. Install a liner extension above the shell to minimize the potential for plume
downwash.
2. Select a stack-liner discharge velocity that minimizes plume downwash over
the expected operating range of the unit at the existing local wind conditions
and that is consistent with other design objectives.
3. During cold ambient temperature conditions with high winds, run the unit at
near full load. (Employing this operating procedure may be natural under
such conditions, because more power is consumed in below-freezing
weather.)
4. Use heated annulus air or heat-tracing to heat the stack hood, roof, or other
areas where ice forms on the top of the stack above the freezing point.
1.5 Contents of the Design Guide
The main contents of the design guide are intended to provide the industry with
information concerning the design and specification of wet stacks.
1.5.1 Information About Key Issues
A main objective of the Revised Wet Stack Design Guide is to provide the latest
information about key issues. For example, the design guide contains discussions
on gas velocity, materials for liner construction, liquid-collection devices and
drainage, continuous-emissions-monitoring systems, and flow modeling. This
guide is intended to provide background information and identify the parameters
that need to be considered in the development of an effective wet stack design.
The key to the successful design of a wet stack is an understanding of the choices
that lead to favorable wet stack operation at an early stage in the design process.
Often, simple low-cost design changes can result in significantly improved wet
operation. But these changes need to be identified early in the design process.
This design guide focuses on the options that should be considered, identifies
existing constraints, and describes the effects that variations in the parameters
have on the design of a wet stack.
1.5.2 Knowledge of the State of the Art
Another main objective of this Revised Wet Stack Design Guide is to summarize
the latest technology available regarding wet stacks. This design guide is a
follow-up to EPRI report CS-2520, Entrainment in Wet Stacks, published in
1982 [1]; EPRI report TR107099, Wet Stacks Design Guide, published in 1996
[2]; and EPRI report TR-109380, Guidelines for the Fluid Dynamic Design of
1-11
Power Plant Ducts, published in 1998 [4]. The Revised Wet Stack Design Guide
contains a presentation of the latest information available, updating what has
been learned in the laboratory and from field installations over the past 14 years.
1.5.3 Overview of Project Implementation
The Revised Wet Stack Design Guide is intended to provide the wet stack designer
with information and recommendations concerning the design and specification
of wet stacks. A systematic process is provided in Section 4 to assist the designer
in identifying the important issues that need to be addressed in the design
process.
The guide is presented in two main sections: 1) wet stack fundamentals and 2)
the design process. These sections provide the designer with the information and
background that lead to a bid specification and identify what to specify for work
scope, what information to provide in the specifications, and what to expect in
terms of state-of-the-art operation.
1-12
Liquid droplets carried from the absorber to the top of the stack liner by the
gas stream without depositing along the gas-flow path. Only small droplets
(10100 m) travel through the ducts without complete deposition. This
means that most conventional scrubber designs, which have horizontal ducts
connecting the absorber to the stack, will result in the deposition of most of
their larger droplets on the internal duct and stack-liner surfaces.
Droplets formed by adiabatic condensation. These represent a large liquidflow rate but are made up of very fine droplets (<10 m), which have a
negligible effect on the SLD detectable at ground level.
Liquid deposited and condensed on the liner, forming a liquid film that will
flow upward in those areas of the stack where the gas velocity is higher than
the flow-reversal velocity for the liner material used (see Section 2.2.2). If
this liquid film reaches the stack outlet, it will be re-entrained back into the
gas flow. The droplet sizes re-entrained at the top of the liner, where
velocities are highest, can be quite large (3006000 m), depending on the
liner-top geometry. Droplets of this size will impact ground-level surfaces in
the vicinity of the liner because they will not be able to evaporate before
reaching the ground. The upper section of the stack choke, if installed, is
usually operating in this mode of upward liquid flow.
All plants operating with a WFGD system will have a white plume. This plume
is a result of the very small droplets formed both by adiabatic condensation and
by additional condensation as the water vapor in the plume is cooled by mixing
with the surrounding air. This cooling process results in additional water-vapor
condensation on particulates and on small droplets within the plume. The
increase in diameter is small (only 110 m). At this size, these newly formed
droplets will follow the local gas-flow patterns; they should evaporate before
reaching the ground as a result of mixing with the drier air away from the stack.
The balance between the plume-cooling and turbulent-mixing processes defines
the length of the resulting visible plume. Therefore, the visible plume length is
primarily a function of the ambient air temperature and the air relative humidity.
The quantity and location of the fallout at a given plant is a function of the
droplet-size distribution of the discharged liquid and the local atmospheric
conditions, such as ambient temperature, wind, relative humidity, and turbulence
level. Large droplets (>300 m) discharged from the top of the stack will typically
reach the ground under most circumstances. Mid-sized droplets (100300 m)
can make it to the ground before evaporating, depending on the local weather
conditions. For example, droplet deposition is typically greater on hot humid
days than on dry days because of the suppressed droplet-evaporation rates. Small
droplets (10100 m) typically evaporate before hitting the ground.
As stated previously, all stacks will have occurrences of SLD. One of the major
objectives in developing an effective wet stack design is to limit SLD to a
minimal acceptable level through proper system design and the installation of an
effective liquid-collection/drainage system. The goal of this system is to collect
and remove deposited liquid from the system before it has an opportunity to reentrain back into the gas flow and to limit the size of droplets that will ultimately
reach the stack outlet. If the droplets are small enough, they will evaporate
before hitting the ground. However, if the droplets are large, they will land on
plant structures, equipment, and cars. A survey of plants with wet stacks found
that the fallout of liquid droplets, if noticeable, usually occurs within a half-mile
radius of the stack.
2-2
2-3
Surface discontinuities and protrusions that disrupt gas and liquid flow
locally, allowing the gas flow to get under the liquid.
Vanes and baffles that cause gas-flow separation, recirculating liquid flow,
and re-entrainment.
Gas-flow patterns that drag liquid along the surface to re-entrainment sites.
Strong vortex patterns that can pick water up in the core of the vortex from
horizontal or vertical surfaces.
Duct-junction corners
Stack floor
2-5
The gasses exiting the absorber will be fully saturated and will contain suspended
droplets. These gasses will flow through the ducting, coming in contact with the
duct walls and with any internal structures within the duct. Liquid will be
continuously condensing from the gas onto these surfaces, forming a liquid film
that will flow downward under gravity and in the direction of the local gas flow.
Care needs to be taken to ensure that the gas-flow patterns moving this liquid
film are favorable for wet operation, minimizing conditions such as flow
separations or recirculation zonespotentially resulting in the re-entrainment of
liquid back into the gas-flow. Also, to the extent possible, the gas-flow patterns
should be managed in such a way as to move the resulting liquid films to drainage
points for removal from the system.
Turning vanes are typically used to optimize gas-flow patterns, reduce pressure
losses, and/or promote liquid collection. An economic assessment must be made
with respect to the value of turning vanes. Although desirable from a pressureloss standpoint, the corrosion-resistant materials required for installation in a wet
environment are very expensive, and the installation cost may not justify the
savings gained by the reduced operating cost. An economic study should be
performed to quantify the life-cycle cost or savings of such an installation.
Some turning vanes may be specifically required to reduce or eliminate SLD.
Such a situation might occur if multiple wet ducts combine immediately
upstream of the stack entrance. In such a situation, vane installation will be
required to ensure that the gas flow entering the stack is favorable for wet
operation and, possibly, to minimize swirling gas flow within the liner. Another
situation would be one in which the absorber is close-coupled to a stack, as in
Figure 2-1. Without vane installation, the gas flow would enter the stack with a
high vertical component, minimizing the ability of the stack to generate the
lower stack-liner flow profiles required for favorable wet operation and liquid
collection. Although this phenomenon is most prominent in close-coupled
arrangements utilizing a side-entry breech, it can also occur in close-coupled
units with bottom-entry elbows. In such a situation, a turning vane would be
required to ensure that the gas flow will enter the stack horizontally. Vanes that
are specifically required for the minimization of SLD should be installed as
recommended.
A more detailed discussion of wet duct design is presented in Section 3.6 of this
guide.
2-6
Droplet
Deposition
High on Liner
Wall
Droplet
Deposition
Low on Liner
Wall Liquid
Collection is
Enhanced
Absorber Outlet
Absorber Outlet
Figure 2-1
Liner Inlet Gas-flow Patterns Close-Coupled Absorber
2-7
Liner
Absorber Outlet
Liner
Absorber Outlet
Figure 2-2
Typical Stack-Inlet Arrangements
As a uniform gas profile flows through a 90 elbow, the gas flow will want to
continue in a straight line. This tendency will result in the formation of a region
of high gas velocity along the back wall of the liner for a side-entry breech
arrangement, and along the outside radius of the bend for a bottom-entry elbow
arrangement. The resulting gas velocity along the outer radius of the elbow can
be 50100% higher than the average gas velocity entering the elbow. When the
gas hits the back wall of the elbow, it splits in two, and each half flows
circumferentially around toward the inside radius of the bend. This flow
generates two counter-rotating secondary vortices, which form on the
downstream side of the turn. If the curvature radius of the elbow is small enough,
a flow-separation zone will also form along the downstream side of the inner
radius of the turn. These phenomena are shown in Figure 2-3. The strength of
the secondary vortices and the size of the separation zone increase with the
sharpness of the turn and with the velocity of the gas entering the elbow.
2-8
Higher than
Average Axial
Velocity
Along Outside
Radius of Bend
Secondary
Flows
Secondary
Flows
SECTION AA
Figure 2-3
Gas-flow Patterns Through a Standard 90 Elbow
For favorable wet operation, strong secondary vortices and a large region of flow
separation are desirable. For this reason, the side-entry breech, defined by a sharp
90 turn, is considered the most favorable design for wet operation. The flow
entering the liner through the breech generates strong robust secondary flows, in
the form of two large counter-rotating vortices, in the lower liner. A large gasrecirculation zone is formed just above the breech in the liner as the gas turns
from the vertical into the liner. These secondary flows, and the recirculation zone
detailed in Figure 2-4, are relied upon to move the collected liquid film on the
walls of the liner circumferentially out of the main gas flow and into the quiet
zones for drainage from the system. Experience has shown that the secondary
flows generated within a liner using a side-entry breech remain stable and welldefined even at reduced flows, ensuring effective wet operation over a wide range
of unit operating loads.
2-9
Uniform Vertical
Gas Flow
Flow
Reattaches
SECTION CC
High Gas
Velocity Along
Rear Wall
Flow
Separation
Upward Bias to
Flow Through
Breach
Gas Flow
B
Flow
Separation
High Gas
Velocity Along
Rear Wall
SECTION BB
Well
Centered
Stagnation
Point
A
SECTION AA
Chaotic Flow
Along Liner Floor
Secondary Flow
Vortices
Figure 2-4
Liner-Inlet Gas-flow Patterns Side-entry Breech
The use of bottom-entry elbows has become very popular with the increased use
of FRP liners. Bottom elbows are very similar to standard duct elbows and,
because of their size, are usually fabricated from liner-can sections spun and
assembled on site. Typically, bottom-entry elbows consist of three to four miter
cuts, forming a 90 elbow with the center-line turn radius to internal-diameter
ratio of between 0.8 to 1.5. This design is very clean in terms of fluid dynamics,
and it results in low system-pressure loss. When flue gas flows through an
unvaned bottom-entry elbow, two counter-rotating secondary-flow vortices are
generated, as detailed above. Because the change in direction from horizontal to
vertical is generally gradual, the intensity of the secondary-flow vortices is not as
great as that generated in the more abrupt side-entry breech arrangement.
Additionally, because the turn is more gradual, a region of flow separation may or
may not be generated along the downstream side of the inside radius of the
elbow. If a region of flow separation is generated, it will be smaller and less
intense than that generated in the side-entry breech arrangement. It has been
observed in flow-model studies of bottom-entry elbow arrangements that two
smallbut very intenseregions of flow recirculation are generated on both
sides of the elbow along the last miter cut (Figure 2-5). These recirculation zones
2-10
are formed because the change in direction from horizontal to vertical is abrupt
in these locations, and the incoming horizontal gas flow over shoots the inside
radius of the turn. Although small, these recirculation zones are problematic,
because they can be sites of significant liquid re-entrainment.
Uniform
Vertical Flow
Flow
Separation
Flow
Reattaches
Flow
Separation
SECTION CC
High Gas
Velocity
Along
Rear Wall
B
High Gas
Velocity Along
Rear Wall
SECTION BB
Gas Flow
SECTION AA
High Intensity
Gas Recirculation
Along Duct Wall
at Seam
Secondary Flow
Vortices
Figure 2-5
Liner-Inlet Gas-flow Patterns 3-Miter Cut Bottom-Entry Elbow
Condensation on the duct surfaces upstream of the stack inlet, which adds
liquid onto the duct surfaces.
Installation of liquid collectors on duct walls and ceilings, vanes, baffles, and
structure, which collect liquid from these surfaces and guide the liquid to the
duct floor of stack-bottom drains for removal from the duct/stack system.
The geometry of the ducts between the absorbers and the stack inlet; the
number of vanes, baffles, and liquid collectors in the ducts; and the
effectiveness of the mist eliminators and duct liquid collectors will determine
the amount of liquid, droplet sizes, and distribution of liquid in the stack
breeching duct at the stack inlet.
present in the gas stream that were re-entrained from the mist-eliminator blades,
the mist-eliminator support structure, and areas of solids buildup on the mist
eliminator. During the wash cycles, washing the mist eliminator can also increase
the amount of liquid carried with the gas flow downstream of the mist eliminator
because of increased liquid load from the washed section.
A field evaluation of the mist-eliminator performance can be made by
performing visual inspections. If carryover problems are noticeable, SLD will
probably occur if effective liquid collectors are not installed in the duct and the
stack. The mist-eliminator selection, operation performance, and possible
problems are described in [3].
2.3.1.1 Mist-Eliminator Design and Arrangement
The mist-eliminator system is the first and most important element of the liquid
collectors in a wet duct/stack system. The selection of the best mist eliminators to
match the absorber and duct design will result in reduced liquid load in the
duct/stack system and less elaborate duct and stack liquid-collection systems. In
FGD service, the function of the mist eliminators is to remove entrained
scrubbing slurry from the flue gas before it travels out of the absorber into the
downstream ductwork and stack. The mist eliminators are located downstream of
the absorber spray zones, where the flue gas contacts the scrubbing slurry that
removes sulfur dioxide. The slurry content of the liquid carryover can cause solids
buildup in the outlet ductwork and stack, high particulate emissions, and
corrosion. The amount of liquid that the flue gas carries into the mist eliminator
will depend on many factors, including (but not limited to) the droplet size, the
gas velocity, the liquid-to-gas ratio, and the washing method.
There are many mist-eliminator designs, available from numerous manufacturers.
But the chevron type of collector is almost exclusively used by WFGD systems
on coal-fired units. Mist eliminators are manufactured from many different
corrosion-resistant materials, ranging from alloys to FRP and various extruded
plastics.
Mist eliminators can be oriented toward any number of gas-flow directions, from
horizontal to vertical (Figure 2-6). Most modern WFGD systems utilize a
vertical gas-flow orientation or a singularly supported roof top or diamond
design.
Older weir-type scrubbers utilize a horizontal gas-flow mist-eliminator design.
Depending on the separation efficiency required, a number of stages of mist
eliminators in series are used. In utility power-generation applications, two stages
are typically used. In situations in which high liquid loading are experienced, a
bulk entrainment separator may be placed in front of the first mist-eliminator
stage. This device is typically a series of baffles used to limit the liquid loading
entering the mist eliminator.
2-13
Gas Flow
Drainage through
Incoming Gas Flow
2nd ME Stage
Gas Flow
1st ME Stage
1st ME Stage
2nd ME Stage
Drain Box
2nd ME Stage
Drain Box
Drainage through
Incoming Gas Flow
Gas Flow
1st ME Stage
Comparison of Flow
Paths Through ME
135
90
Perpendicular
137
Direction of
Gas Flow
Through Mist
Eliminator
95
ME Side Elevation
Figure 2-7
Comparison of Gas-Path Lengths Through a Mist Eliminator as a Function of Angle
2-15
The amount of liquid-droplet carryover into the wet duct/stack system from the
mist eliminators can be minimized by
1. Selecting the most suitable mist-eliminator elements for the operating
conditions of the installation
2. Using two-stage mist elimination
3. Using a vertical or angled mist-eliminator orientation
4. Using gas velocities with a safe margin under the critical re-entrainment
velocity for the mist-eliminator blades selected
5. Achieving a sufficiently uniform velocity profile into the second-stage misteliminator to keep all regions below the critical re-entrainment velocity
6. Selecting effective washing techniques and procedures to prevent solids
buildup on the mist eliminators, without entraining wash spray in the gas
flow
7. Establishing inspection and maintenance procedures for the mist eliminators
8. Maintaining good pH control to reduce solids precipitation in the mist
eliminators
EPRI Report GS-6984, FGD Mist Eliminator System Design and Specification
Guide [3] describes all essential aspects of mist-eliminator design, application,
operation, and maintenance necessary for an effective mist-elimination system.
The measured laboratory performance of several commercially available mist
eliminators is also furnished in this report.
Information regarding the performance of mist eliminators as a function of gas
velocity and inlet-liquid load is available from the mist-eliminator manufacturers,
and it should be obtained early in the system-design process. Performance is
given as a function of the gas velocity immediately upstream of the mist
eliminator. Figures 2-8 and 2-9 present typical vertical gas-flow chevron-type
mist-eliminator performance curves for pressure drop and carryover versus gas
velocity, respectively. Mist-eliminator collection efficiency versus droplet size can
also be supplied from the manufacturer if desired. This latter performance curve
is used to identify the maximum droplet size expected to pass through the mist
eliminator without being collected. (See Figure 2-10 for an example.)
For design and estimating purposes, a typical mist-eliminator manufacturers
carryover guarantee is 0.02 gr/acf (46 mg/m3) can be used.
2-16
0.300
0.250
0.200
0.150
0.100
0.050
0.000
0.00
2.00
4.00
6.00
8.00
10.00
12.00
Velocity (ft/s)
14.00
16.00
18.00
Figure 2-8
Typical Chevron-Type Mist-eliminator Pressure Drop vs. Gas-Velocity Curve
Figure 2-9
Typical Chevron-Type Mist-Eliminator Carryover vs. Gas-Velocity Curve
2-17
20.00
Figure 2-10
Typical Mist-Eliminator Collection Efficiency vs. Droplet-Size Curve
The location of the absorber outlet too close to the outlet of the mist
eliminators
A nonuniform liquid loading from the absorber into the mist eliminators
Mist eliminators have no moving parts. After they are properly installed, they
will operate at design conditions without any further attention. Because there
are no moving parts, the major potential operating problems are the plugging
of the mist eliminator with insoluble solids and/or the corrosion and
degradation of the materials of construction. The usual operating-pressure
decrease across the mist eliminator when clean is generally 0.20.5 in (5.1
12.5 mm) of water column per chevron stage.
Mist eliminators are typically broken down into multiple smaller segments
for on-line washing. Each segment is washed sequentially in series. A
minimum wash frequency for each segment of 1.5 min/h is recommended.
More frequent washing is always better. The pressure of the wash-water
nozzles is generally 2545 psig (0.170.31 MPa). Mist-eliminator spray
nozzles are typically 90120 full-cone nozzles. Most manufacturers
recommend a 150200% overlap of the mist-eliminator spray-wash pattern.
A nozzle to mist-eliminator spacing of 23ft (0.60.9 m) should be used.
2-19
During the wash cycle, the liquid within the mist-eliminator section being
washed will be significantly higher, and increased carryover into the downstream
ducting from this section will be expected. A properly designed wet duct/stack
system will be able to handle this additional carryover without increased potential
for SLD.
For a detailed discussion on the effectiveness of mist-eliminator systems, refer to
[3]. To maintain effectiveness and to prevent sludge buildup, it is very important
to keep mist eliminators clean by washing them according to the manufacturers
specifications.
2.3.2 Liner Condensation
The most significant source of liquid in a wet duct/stack system is condensation
of water vapor from the saturated gas flow exiting the WFGD absorber. This
condensation comes from two main sources: 1) thermal condensation on the wall
as a result of heat transfer from the flue gas to the outside air through the liner,
insulation, annulus air, and concrete shell and 2) adiabatic condensation in the
bulk of the gas flow because of the expansion of the saturated gases as a function
of the pressure decrease along the height of the stack.
Detailed analytical calculations must be completed to predict the quantity of
liquid produced by condensation in the absorber outlet duct and in the stack.
These computer models are based on the equations presented and discussed in
EPRI Report CS-2520, Entrainment in Wet Stacks [1]. The accuracy of the
condensation calculation is estimated to be 10%. The rate of liquid
condensation is a function of ambient air temperature, wind velocity, and linerdesign variables. Calculations are performed for a selected worst-case
atmospheric condition to estimate the maximum quantity of liquid condensation
expected in the liner of the subject unit. Determining the expected quantities of
liquid condensation in the duct and liner is necessary for the development of an
effective liquid-collection system, because it will affect the number, sizing, and
location of the various liquid-collection gutters and drains.
To perform the condensation calculations, the following information is required:
Detailed design drawings of the system ductwork from the absorber outlet to
the stack inlet, including material specifications and thicknesses
Detailed design drawings of the stack shell and liner, including the number
and orientation of liners relative to true North, materials of construction,
material thickness versus elevation, annulus air-gap thicknesses versus
elevation, location and size of stack-shell vents to atmosphere
Expected flue gas conditions: temperature, flow rate, constituents, watervapor content
Plant elevation
the saturated gas causes water vapor to condense on fine droplets or on small
solid particles within the bulk of the gas. Droplets formed by this type of
heterogeneous condensation will be very small (<10 m). Only a very small
fraction of this adiabatic condensation will reach and be deposited on the liner
surface by the turbulent-deposition process. Estimates of this fraction range from
2.5 to 5%. The expected percentage of droplets coming in contact with the liner
wall decreases with increasing liner diameter, because of the greater distance a
droplet would need to travel from the center of the liner to the liner wall.
Historically (as stated above), estimates of 2.55% have been used, and these
percentages have been anecdotally supported by observations from operating
units, which can monitor the liquid-drainage rate from their liquid-collection
systems. Condensed droplets that do not encounter the liner wall are discharged
from the top of the liner, which creates the familiar white plume associated with
wet stack operation. These droplets do not represent a problem, because their
size is very small, and they will evaporate before reaching the ground.
The stack effect is the hydrostatic-pressure difference between the inside and
outside of the stack liner. This difference is due to the lower density of the heated
gas in the stack compared to the density of the ambient air. Therefore, the stack
effect is calculated by multiplying the density difference of the gas in the stack
and the ambient air times the height of the stack. Net stack draft is the operating
pressure at the inlet of the stack liner, and it is calculated by subtracting the fluid
friction, the inlet losses, and the exit losses from the stack effect.
For wet stack design, the worst-case conditions are the maximum stack gas-flow
rate resulting in the highest friction loss, in combination with the hottest and
most humid ambient conditions. These conditions can normally be expected to
produce a positive gas pressure at the stack inlet. The stack effect at worst-case
conditions is generally only 0.20.3 in Wg (0.050.07 kPa) for wet stack
operationa small amount compared to the friction and exit losses at full load.
2.4 Liquid-Film Behavior on the Liner Wall
Because the flow characteristics are so different, the discussion of the liquid-film
behavior on the liner wall must be broken down into two sections: the lower liner
and the upper liner. The lower liner is defined as that portion of the liner located
below a point four liner diameters above the roof of the breech duct or bottomentry elbow. The upper liner is everything above that location.
As discussed previously, the gas-flow patterns in the lower liner are threedimensional and complex, and they are used for the majority of the liquid
collection in the stack. The flow entering the liner through the breech or bottomentry elbow generates secondary flows in the form of two large counter-rotating
vortices in the lower liner. These secondary flows are relied upon to drag the
liquid film on the walls of the liner circumferentially out of the high-velocity gas
flow into low-velocity or separated regions where it can be drained from the
system.
2-22
Uniform
Downward Liquid
Film Flow
From Upper Liner
Circumferential
Liquid Film Flow
Downward
Liquid Film
Flow in Flow
Separation
Upward Liquid
Flow
Droplet Impact
Zone
Reentrainment
Gas Flow
Downward Film
Flow Below
Impact Zone
Downward Flow
Along Breach
Opening
REAR WALL OF
LINER
Figure 2-11
Lower Liner Droplet- and Liquid-Film Flow Patterns Side-Entry Breech
2-23
Some liquid will make it to, and collect on, the liner floor. If the liner floor is flat,
the collected liquid will pool and slosh around until it finds a drain point.
Typically, two vortices can be observed on the flat floor of the liner, their centers
located approximately 1/3 of the way across the liner and even with the side walls
of the inlet breech duct (Figure 2-12). Under some conditions, these vortices can
be sufficiently strong to re-entrain liquid off of the floor and back into the gas
flow.
X
Liquid Pooling on
Liner Floor
X/3
Gas Flow
Vortex Flows on
Liner Floor
Reentrainment
Liquid Film Flow
Gas Flow
Liquid
Reentrainment
from Vortex Flow
Vortex Flows on
Liner Floor
Figure 2-12
Liner-Floor Liquid-Film Flow Patterns Side-Entry Breech
For a bottom-entry elbow, the droplets entrained in the incoming gas-flow will
typically impact in a zone starting on the outside radius of the elbow at an
elevation corresponding approximately to the roof of the inlet duct and extending
vertically approximately 11.5 liner diameters (Figure 2-13). Similar to the sideentry breech, finer droplets will impact along the top of the impact zone, and
larger droplets will impact along the bottom. Secondary gas flows generated by
the 90 elbow into the liner will drag a large portion of the liquid collected on the
2-24
rear wall of the liner upward. Then it will be dragged circumferentially around to
the inside radius of the inlet elbow, where it will encounter a lower vertical gas
velocity. It will then flow downward until it either re-entrains back into the gas
from the top of the duct, or it will flow circumferentially downward on either
side of the inlet duct toward the bottom of the elbow. In this latter case, there is a
high probability that the liquid will be shredded off of the wall by recirculation
zones located along the walls of the elbow near the inside radius. Liquid that
makes it to the floor of the inlet elbow will tend to pool there, because the
horizontal gas velocity along the floor is not high enough to drag the liquid back
up the outside radius of the elbow.
Reentrainment
Uniform
Downward
Liquid Film
Flow From
Upper Liner
Circumferential
Liquid Film Flow
Upward Film
Flow
Downward
Flow in Flow
Separation
Circumferential
Liquid Film Flow
Droplet
Deposition
Zone
Downward
Film Flow
Gas Flow
Droplet Reentrainment
from High Intensity Gas
Recirculation
REAR WALL OF
LINER
Figure 2-13
Lower-Liner Droplet- and Liquid-Film Flow Patterns Bottom-Entry Elbow
Stacks are typically designed so that condensed liquid on the liner wall will flow
downward for collection. The stack-design velocity is the average gas velocity in
the liner defined as the volume flow rate divided by the area of liner. As discussed
above, for both inlet arrangements, the flue gas that enters the liner impacts the
wall opposite the inlet and forms a region of locally high gas velocity along the
rear wall of the liner or outer radius of a bottom-entry elbow. The gas velocity is
usually high enough to drag the deposited liquid vertically up the liner wall until
the secondary-flow vortices are strong enough to pull the liquid circumferentially
around and out of the high-velocity region. If the vertical gas velocity is not
sufficiently high, the deposited liquid will either move very slowly upward or not
2-25
move at all. Both of these situations are undesirable, because as more liquid is
deposited in this region, the liquid-film thickness will increase until the surface
becomes unstable, wavelets are formed, and droplets will begin re-entraining
back into the gas flow. There is little chance that many of these droplets can be
recollected before exiting the liner as SLD. If this situation occurs, liquidcollection devices in the form of V- shaped gutters and other types of liquid
collectors will need to be installed in the droplet-impact region to mechanically
help the liquid move circumferentially into the secondary-flow vortices. These
devices will be discussed in more detail in Section 3.11.8.
Liquid-film flows in the upper liner should be better behaved. In this region, the
gas-flow profile moving up the liner is more uniform, and the liquid film formed
as a result of thermal and adiabatic condensation should flow downward without
obstruction.
The motion of a vertical liquid film exposed to a gas flow is a function of many
variables. The most important of these are the surface tension of the liquid being
used, the gas velocity near the film, the roughness of the surface, and the liquidfilm thickness (which is itself a function of the condensation rate, the filmdrainage rate, and the height over which the condensation takes place).
To understand liquid-film flow on a liner surface, it is best to begin by looking at
the ideal case of a smooth wetting surface. A wetting surface is defined as a
surface upon which a liquid will form a film, as opposed to beading up after being
spread out (Figure 2-14). A wetting surface is referred to as being hydrophilic,
whereas a nonwetting surface is referred to as being hydrophobic. To be effective,
a stack-liner material must be hydrophilic. Some materials, such as FRP, can
transform from one surface characteristic to the other after exposure to the
corrosive conditions within the liner.
Wetting Surface
Figure 2-14
Wetting vs. Nonwetting Surfaces
2-26
Non-wetting Surface
2-27
0-50 ft/s
(0-15.2 m/s)
55-65 ft/s
(16.7-19.8 m/s)
Gas Flow
70-85 ft/s
(21.3-25.9 m/s)
>90 ft/s
(>27.4 m/s)
Figure 2-15
Liquid-Film Flow Patterns on a Smooth, Vertical Wetting Surface vs. Vertical Gas
Velocity
When the liquid film flows over a horizontal discontinuity, there is a potential for
the upward- flowing flue gas to get under the liquid, resulting in the formation of
droplets. A portion of these droplets will be re-entrained back into the gas flow
and will exit the liner as SLD.
The major areas of concern in an alloy liner are the horizontal weld beads joining
the plates together. Vertical weld beads will have little to no impact on droplet
re-entrainment. But horizontal weld beads on alloy liners can cause liquid reentrainment in the stack at gas-flow velocities as low as 40 ft/s (12 m/s) if not
properly made. The height of the horizontal weld beads is critical, and their
height should be the minimum possible that does not exceed the standard for
good weld quality: 1/8 in (3.1 mm). Grinding all horizontal beads flush to the
base material is ideal. However, laboratory studies have demonstrated that, for
gas velocities up to 50 ft/s (15.2 m/s), a 1/8-in (3.1-mm) bead height on a butt
joint will result in minimal levels of droplet formation (assuming the bead is
smoothly rounded). Slight increases above this velocity5560 ft/s (16.718.3
m/s)resulted in liquid holdup along the joint and the start of droplet reentrainment back into the gas flow. At velocities above 65 ft/s (19.8 m/s), droplet
re-entrainment becomes more pronounced. Figure 2-16 details liquid-flow
patterns over a 1/8-in (3.1-mm) weld bead as a function of gas velocity.
Droplets
Reentrained From
Surface
Liquid Flows
Over Weld Bead
Wave/Rivulets
Forming on Surface
Gas Flow
Weld Bead
Liquid Starts to
Holdup Over
Weld Bead
With some
Reentrainment
Gas Velocity
<50 ft/s (15.2 m/s)
Film Thickening
Due to Liquid
Holdup
Droplets
Reentrained
From Weld Bead
Gas gets Under
Liquid at Weld
Bead and Waves
Down Flow of
Liquid Reduced or
Stopped
Gas Velocity
Gas Velocity >65
55 - 60 ft/s (16.8-18.3 m/s)
ft/s (19.8 m/s)
Figure 2-16
Liquid Flow over 1/8-in (3.1-mm) Weld Beads
The rings of an FRP liner should be joined together with material mostly on the
outside of the liner. Only a minimal inward protrusion with smooth transitions is
recommended. FRP liner-can joints are typically lap-type joints with interior and
exterior overlays (Figure 2-17). A 6:1 to 10:1 taper should be used from the
stiffener ringtop and bottomback to the base-liner material, with smooth
transitions.
2-29
Liner Wall
Detail 1
Inside Flue
Interior
Reinforcing Band
(Minimum Thickness
Possible)
Minimum Smooth
1:6 Taper Back To
Liner Wall
(1:10 Preferred)
Outside
Flue
Exterior
Reinforcing
Band
Detail 1
Figure 2-17
FRP Liner-Can Joints Internal-Taper Requirement
Special attention must be paid to brick liners. Every effort must be made to
ensure the proper radial alignment of the bricks and to strike the mortar flush
with the inside surface of the liner. If the maximum projection or offset between
bricks on the interior surface of the liner exceeds 1/8 in (3.1 mm), re-entrainment
may occur from the liner surface. Extruded or depressed mortar joints can be
significant sources of liquid re-entrainment back into the gas flow. The
construction should be clean, without mortar smeared on the inside surface of the
liner, because this could also create a location for droplet re-entrainment.
For borosilicate-block lining systems, such as the PennGuard, every effort
must be made to ensure the proper radial alignment of the blocks and to strike
the adhesive/mastic flush with the inside surface of the liner. The surface of the
block should be clean and free from smeared adhesive/mastic material. If the
maximum projection or offset between bricks on the interior surface of the liner
exceeds 1/8 in (3.1 mm), re-entrainment may occur from the liner surface.
Extruded or depressed adhesive/mastic material can be a significant source of
liquid re-entrainment back into the gas flow.
2-30
2.6 Re-entrainment
The liquid on the duct and liner surfaces produced by deposition and
condensation flows in the form of film or rivulets is governed by gravitational,
surface-tension, and gas-shear forces. The gas-shear force can shear the liquid off
of the surface, causing re-entrainment. This phenomenon is the most frequent
source of SLD and fallout of liquid droplets in the vicinity of the stack.
Re-entrainment most frequently occurs because of one or more of the following:
Surface discontinuities and protrusions that disrupt gas and liquid flow
locally.
Vanes and baffles that cause gas-flow separation, recirculating liquid flow,
and re-entrainment.
Gas-flow patterns that drag liquid along the surface to re-entrainment sites.
Strong vortex patterns that can pick water up in the core of the vortex from
horizontal or vertical surfaces.
Duct-junction corners
Stack floor
2-32
Table 2-1
Recommended Stack-Liner Velocities for Wet Operation
*Laboratory testing of coating material is recommended to finalize liner velocity for favorable wet
operation.
A number of different liner-coating materials have been suggested for wet stack
operation, ranging from gunite to spray-on polymers. Most of these coatings
have not been evaluated for determination of their recommended liner velocity
for favorable wet operation. The favorable velocity for coatings is highly
dependent upon the smoothness and wetting nature of the material after it has
been applied within the liner. Laboratory testing of the coating system should be
performed before a liner velocity is finalized for a specific material and
application. A velocity of 55 ft/s (16.8 m/s) is recommended for materials that
result in a smooth, discontinuity-free surface within the liner.
2.8 Stack Liquid Discharge
Stack liquid discharge (SLD) can occur at any time, even from a well-operating
stack. Disturbances in the operating conditions of the plant and/or absorber as
well as ambient weather conditions can initiate the start of an SLD event.
However, SLD is only a problem if the droplets are large enough (>100 m) to
be detectable at ground level near the stack.
The amount of liquid discharged at the top of the stack is a result of gas- and
liquid-flow processes that take place in the ductwork and stack system between
the discharge of the absorbers and the top of the stack liner. The source of the
liquid discharge can be categorized by the liquid-flow process and by the droplet
sizes as follows:
2-33
Liquid droplets carried from the absorber to the top of the stack liner by the
gas stream without depositing along the gas-flow path. Only small droplets
(10100 m) travel through the ducts without complete deposition. This
means that most conventional scrubber designs, with horizontal ducts
connecting the absorber to the stack, will result in deposition of most large
droplets on the duct surfaces.
Liquid deposited and condensed on the liner, which flows upward in those
areas of the stack where the gas velocity is higher than the flow-reversal
velocity for the liner material. The droplet sizes re-entrained at the top of the
liner, where velocities are highest, can be quite large (3006000 m),
depending on the liner-top geometry. The upper section of a stack choke is
usually in this mode of upward liquid flow.
The quantity and location of the fallout at a given plant is a function of the
droplet-size distribution of the discharged liquid and the local atmospheric
conditionssuch as ambient temperature, wind, relative humidity, and
turbulence level. The very small droplets of the bulk condensation make the gas
plume white and visible. The plume is cooled as it mixes with the surrounding
air. This cooling results in condensation on the small droplets in the plume.
These droplets are typically 110 m in diameter and therefore do not fall to the
ground and cannot be detected. The mixing with the drier air away from the
stack causes droplet evaporation. The balance between the cooling and mixing
processes defines the length of the visible plume. Therefore, it is a function of the
ambient air temperature and the air relative humidity.
One of the major objectives in developing an effective wet stack design is to limit
SLD to a minimum acceptable level. The goal is to limit the droplet size that will
exit the stack. If the droplets are small enough, they will evaporate before hitting
the ground. However, if the droplets are large, they will land on nearby surfaces
such as plant structures, equipment, and cars. The wet stack survey indicated that
the fallout of liquid droplets, if noticeable, usually occurs within a half-mile
radius of the stack.
If SLD is experienced, note the time of day, location of the fallout relative to the
stack liner, and local weather conditions. If the location of the fallout is known,
rolling aluminum foil out on the ground in 26-ft (0.61.8-m) squares for a
period of time sufficient to collect numerous drops will be extremely helpful.
Both the droplet size and the quantities of droplets can be obtained from the foil.
This information, along with information on plant operations at the time of the
event, will be critical in evaluating the source and possible cause of the rainout
event. If SLD is a continuous problem, some stack sampling may be required.
Knowing the sizes of the droplets entrained within the gas flow inside of the liner
can be very helpful in identifying the source (or sources) of the discharge.
Additionally, an inspection of the liquid-collection system should be performed
as early as possible. The inspection should be performed and photo-documented
before anything inside of the wet ducts and stack are disturbed in any way. Often,
clues to the source of the discharge problem can be as simple as a deposit on a
wall or marks on the liner surface.
2.9 Plume Downwash and Icing
2.9.1 Plume Downwash on Stack
A cross wind at the top of the stack will deflect the plume from its vertical path.
As the cross wind increases, the plume may become partially entrained into the
low-pressure region behind the stack windscreen. This phenomenon is known as
plume downwash.
Designing a liner for favorable wet operation will decrease the potential for SLD.
However, such a design increases the potential for plume downwash at elevated
wind velocities (Figure 2-18). During downwash episodes, saturated flue gas
comes in contact with the liner extension, stack hood, and stack shell. This
contact can lead to deterioration of the stack construction materials because of
exposure to acid in the flue gas. It also increases the potential for ice formation
on the top of the stack. The potential deterioration problem is a long-term
situation caused by continual exposure to SO2, which has been evidenced in
recent years by the deteriorating conditions of tombstones, stone buildings, and
the top of stack shells. The icing problem can occur at below-freezing conditions
all winter long, every winter, and creates a potential danger to people and
property.
The most important fluid-dynamic parameter is the momentum ratio between
the stack exit flow and the cross wind. This ratio defines the relative influence of
each flow on the other and is determined by the densities and velocities of the
flue gas and the wind. This momentum ratio (MR) is defined as:
Figure 2-18
Plume Downwash
The second most important ratio is the shell outer diameter to the liner inside
diameter at the top of the shell. A larger shell- to liner-diameter ratio results in
increased downwash potential at the same momentum ratio.
Plume downwash is most likely to occur during reduced unit load operation, in
which stack-discharge velocities are reduced, and during high wind conditions.
Plume downwash is least likely to occur under the opposite conditions of high
unit load operation and low wind velocity.
For a multiple-flue stack, the equivalent momentum ratios for initiation of plume
downwash are higher, and they vary with the relative wind direction. This
difference in momentum ratios occurs because the shell- to liner-diameter ratio is
larger for multiple-flue stacks. The reduced static pressure in the wind vortices
generated off the stack shell and liner extensions can draw the flue gas into a
downwash pattern along the stack shell. In the process, the saturated flue gases
that are drawn into the vortices come into contact with the roof and sides of the
stack liner and shell.
The interaction of a plume with the cross wind is a function of the following
parameters:
Liner-extension diameter
Shell diameter
Liner-extension length
Roof shape
Number of liners
2-36
Wind direction, velocity, and air density at the top of the stack
The momentum ratio (MR) is the most important fluid-dynamic force ratio that
controls plume downwash. The individual parameters within this ratio have the
following importance for wet stack plume downwash:
1. Flue gas density (flue gas) is almost constant for saturated flue gas at about
0.066 lb/ft3 (1.06 kg/m3) at sea-level ambient pressure.
2. Wind density (wind) varies with ambient temperature at sea-level ambient
pressure from 0.090 lb/ft3 (1.44 kg/m3) at -20F (-29C) to 0.071 lb/ft3
(1.14 kg/m3) at 100F (38C) (an approximately 25% spread).
3. Reduced ambient pressure for plants at high elevations above sea level
reduces density by 1020%, but it changes both density values by the same
amount.
4. The flue gas discharge velocity (Vflue gas) is at the highest stack elevation and is
set by the liner or choke discharge diameter and the low-load to high-load
flue gas volume flow rate. This value can range from 20 to 120 ft/s (637
m/s) for a wet stack plume discharge, including partial-load operation.
The wind-velocity magnitude, direction, and frequency of occurrence varies
significantly at different geographical locations, but this information is usually
available at weather stations and airports located near existing or new powerplant sites. The information needed is usually referred to as the "wind-rose" data
in graphical or tabular form, and it includes wind velocity and frequency as a
function of direction. The quantity of cumulative data is usually supplied on a
yearly basis, but the data is also available on a monthly (or more frequent) basis if
time of year is important.
When reviewing the wind data, a judgment must be made when selecting a
worst-case wind magnitude and cumulative frequency of occurrence (to eliminate
unusually high windstorms of short duration that seldom occur).
Items 1, 2, 3, and 5 above can usually be set for a specific plant site. Item 4, the
flue gas discharge velocity, must then be selected to satisfy the requirements for
Plume downwash
Wind-rose data
On liner extension
On roof or hood*
On stack shell
24 days (0.51%)*
48 days (12%)
*The extent of downwash and the cumulative time of downwash marked are suggested
values at which to start the evaluation of downwash and its effect on stack-discharge
velocity.
For all stack-top geometries discussed below, the stack-shell diameter should be
as small as practicable to contain the number of liners.
2-38
Single-Flue Stacks
1. For an angled hood and no liner extension:
-
MR1 = 35
MR2 = 23
2. For a flat or angled roof and a short liner extension of 0.5 liner diameter:
-
Two-Flue Stacks
1. For a flat roof plus a liner extension of about one liner diameter:
Best wind direction (prevailing-wind direction in line with the axis of the
two flues)
-
MR1 = 1.52.0
MR2 = 1.11.5
MR1 = 3.55
MR2 = 34.5
2. For a minimum liner extension of 0.75 liner diameters, the above MR values
will remain the same for the best wind direction, but the above MR values
for the worst wind direction will rise by 1020%.
3. For a taller liner extension of 1.5 extension diameters, the above MR values
for the best wind direction will remain the same, but the above MR values
for the worst wind direction will drop by 1020%.
4. A chamfered stack top (2 sides) will slightly increase (by1020%) the best
wind direction MR values, but will reduce (by 3050%) the worst wind
direction MR values.
Three- and Four-Flue Stacks
The best wind direction MR values are expected to be similar to those for
single-flue stacks.
The worst wind direction MR values are expected to be similar to those for
two-flue stacks.
the lowest cumulative time for plume downwash for a specific combination of
flue gasdischarge velocity and stack-top geometry.
These plume-downwash guidelines can be used to obtain preliminary estimates
of the stack-top design to limit the extent and duration of plume downwash:
1. Select plume downwash criteria at a specific unit load.
2. Select a basic stack-top design except for flue gasdischarge velocity.
3. Obtain the wind-rose data and select reasonable design wind velocities.
4. Select the appropriate MR-guideline value that is compatible with steps 1
and 2.
5. Calculate the corresponding flue gasdischarge velocity, flue gasdischarge
area, and the preliminary stack-top geometry.
6. Compare the results in Step 5 to values compatible with good plume
dispersion, stack liquid collection and drainage, and construction cost.
7. Repeat steps 1 through 6 until satisfactory results are obtained for all
important criteria by making appropriate compromises.
The final stack-top geometry should be optimized by CFD flow-model studies,
particularly for multiple-flue stacks for which downwash-minimization is
important.
2.9.2 Stack Top Icing
Whether ice forms on the top of the stack depends on the temperature of the
stack surface, the temperature of the mixture of saturated flue gas and cold
ambient wind, and on whether water vapor will condense out of the mixture. Ice
formation is most likely at plants where below-freezing temperatures are
common and where they last for extended periods of time.
The locations for potential ice formation are discussed below:
The hood for a
single-flue stack
2-40
The roof of a
multiple-liner stack
The stack roof usually does not directly touch the liners, which
permits expansion and movement of the liners. The roof can
receive heat from the stack annulus. Liquid must be drained from
the roof area. Preferably, drains should be located on the inside of
the annulus to keep them from freezing.
Railings
The metal railings at the top of the shell will quickly cool to the
temperature of the mixture of the flue gas and the ambient air, and
they will be the first surfaces to become iced. In other words, these
railings are the most likely places for ice to form.
A metal platform on the outside of the stack near the top can also
be exposed to low concentrations of flue gas. Because it is metal
and is exposed to the ambient- and flue gasmixture temperature, it
may also experience occasional ice buildup. Such platforms are
the second most likely places for ice to form.
Shell
Icing can result in significant ice buildups along the top of the stack. These ice
buildups can dislodge and fall, potentially causing serious damage to people and
equipment on the ground.
The potential for icing can be reduced by employing the following steps:
1. Select a stack-liner discharge velocity that minimizes plume downwash over
the expected operating range of the unit at the existing local wind conditions
and that is consistent with other design objectives.
2. During cold ambient temperature conditions with high winds, run the unit at
near full load. Employing this operating procedure may be natural under
such conditions, because more power is consumed in below-freezing weather.
3. Use heated annulus air or electrical heating elements to heat the stack hood,
roof, or other areas where ice forms on the top of the stack.
4. Do not use insulation on sloped rain hoods constructed of a corrosionresistant material.
5. Use a parapet wall instead of metal railings around the top of the liner shell.
A parapet wall is usually an extension of the concrete shell above the roof.
There is normally less icing on a concrete parapet than on a metal railing,
and the parapet wall will contain ice that could fall from the liner or liner
extension.
2-41
3-1
the gas flow. Critical re-entrainment velocities in the ductwork are difficult to
determine and are not available.
If the liner velocity is lower than the range of recommended maximum values,
the liquid re-entrainment rate is lower and the need for additional liquid
collectors is reduced.
A single-stack liner that serves multiple FGD modules may produce large
volumes of liquid for drainage. Drainage-pipe diameters need to be sized
accordingly. If the liquid collected from multiple FGD modules is returned to
the FGD process, the volume of liquid may be too large to be returned to a single
FGD module without significant changes to its operational characteristics.
A flow-model study should be performed to evaluate liquid re-entrainment and
liquid collection in the ductwork and liner, based on the anticipated gas
velocities. If adequate liquid collection cannot be accomplished at the anticipated
velocities, a new liner and/or ductwork sized for a lower gas velocity should be
considered.
3.4 Wet Duct/Stack Operation
Initially, water enters the duct and the stack by droplet and water carryover, with
the saturated flue gas passing through the mist eliminators of the scrubbers. The
liquid accumulates on surfaces by deposition of liquid droplets and by vapor
condensation from the saturated flue gas.
Mist-eliminator carryover is usually characterized by fine droplets that can pass
through the mist eliminator without being collected and by larger droplets reentrained from deposits and during washing. The finer droplets will follow the
gas flow and will generally not be collected within the system before exiting the
stack. However, because these droplets are very small, they will evaporate on the
outside of the liner before reaching the ground. The larger droplets (both in size
and quantity) re-entrained from partially plugged mist eliminators and during
mist-eliminator washing, will generally be deposited on the surface within the
ductwork and the lower liner.
Most of the liquid suspended in the gas will be deposited on the walls and
internal structures (turning vanes, dampers, trusses) by impingement, caused by
the departure of the trajectory of a liquid droplet from the streamline flow in a
curved gas-flow field. Some deposition also occurs because of turbulent
deposition, caused by the turbulence-induced motion of small droplets in a
direction perpendicular to the main gas-flow direction. The liquid on the walls
and internal structures will be in the form of films, rivulets, attached droplets,
and (in some cases) large pools on horizontal surfaces. The gas-shear force can
drag the liquid through the system (even upwards against gravity and wall
friction) if the gas velocities are high enough; it can also shear off and re-entrain
droplets from the liquid flowing on the walls and internals. Some small droplets
may negotiate the gas-flow path from the scrubber outlet to the top of the stack
without contacting a solid surface.
3-3
During normal operation, the largest contributor to the liquid flow on a liner
surface is condensation, rather than deposition of the droplet carryover from the
mist eliminators. The condensation on the liner cannot be measured in a
laboratory-scale model of the stack. Therefore, analytical calculations are required
to define the amount of condensed liquid-flow rate in the stack.
For brick liners, the pressurizing-air leakage through the liner increases the
bulk condensation by additional cooling. Thermal condensation also
increases because of the colder annulus air used for pressurization.
Having more than two liners in a shell increases the total wall-condensation
rate by about 5%. This increase is mainly attributable to a higher
condensation rate on the liner extension above the shell.
Estimated Flows
Mist-eliminator carryover:
Duct-wall condensation
Stack-wall condensation
3
3
All the droplet-producing sources are controlled primarily by the local gas
velocity. The two basic mechanisms for producing liquid droplets are the
shear force caused by the local gas velocity and the dynamic pressure of the
gas flow at local flow disturbances on the solid surfaces. The particle-size
range for re-entrained droplets is primarily 3006,000 m.
3-5
1. Mist-eliminators: The primary source of liquid in stacks is the misteliminator carryover from normal operation, partially plugged operation, and
washing cycles. If a WFGD retrofit to an existing unit is made with nonideal
duct/stack conditions and a wet stack is used, one needs to specify the best
possible mist eliminator. The quoted efficiency and carryover from mist
eliminators is typically much lower (0.020.04 gr/acf [4692 mg/m3]) than
the limited field data, which may go as high as 0.172.00 gr/acf (3904600
g/m3). The wash cycle of the mist eliminator can produce short periods of
increased liquid-droplet discharge. EPRI has conducted field tests on mist 3-6
eliminator carryover drop sizes and loading levels. EPRI has also conducted
tests on many mist-eliminator modules.
2. Surface discontinuities in duct walls: The liquid films of rivulets flowing in a
horizontal, inclined, or vertical passage will be partially entrained at surface
discontinuities, such as a step-up, a step-down, and sharp bends. The
amount of liquid entrained is a strong function of the local geometry and the
gas-flow disturbance caused by the discontinuity. Liquid-droplet production
at wall discontinuities has been observed in both laboratory models and field
wet stack systems, but it has not been measured quantitatively.
3. Thermal-expansion joints: These duct components represent a special
discontinuity because they usually have a capacity to store liquid. Thermalexpansion joints represent a potential liquid re-entrainment area in the ducts
and in the stack, although the expansion joint is often covered. Expansion
joints can act as liquid collectors on the walls, and they should have drains
installed to remove collected liquid to reduce the possibility of reentrainment. Expansion-joint drains are only practical at duct pressures near
ambient pressure downstream of ID fans. At locations where negative
pressure exists in the duct, traps are needed to prevent air leakage into the
duct. Other re-entrainment control methods may also be needed.
4. Internal duct components: Necessary internal components such as mounting
plates, louver-damper blades, and turning vanes collect liquid by inertial
deposition, and re-entrainment takes place on or downstream of these
components.
5. Special flow patterns: Gas-flow passages of power plants are designed to be
functional and low-cost, but they are usually aerodynamically crude. Flow
separations, reattachments, and secondary flows are encountered at several
locations in most power-plant duct systems. The liquid droplets and films
collect in separated zones, forming a large pool or a heavy stream that
entrains more easily than a thin film at lower gas velocity. An increase of gas
velocity from a load change, or the addition of an FGD module on-line, can
produce re-entrainment from unprotected pools.
6. Roughness of surfaces: Liquid entrainment by gas streams from rough
surfaces is proportional to the gas velocity and the roughness of the surface,
starting at some critical threshold velocity that varies with the surface
characteristics. Because large surfaces with thousands of square feet are
involved, the integrated effect of surface roughness is very important
particularly in the vertical stack flow. Discontinuities and other sources of
surface roughness can be controlled or eliminated by design and careful
installation, but practical acid-resistant liner materials have to be accepted
and installed as produced.
7. Fan washing: If ID fans are included downstream of wet scrubbers, they may
have an intermittent on-line wash system for cleaning deposits off of fan
impellers. When in use, this system will produce entrained droplets.
3-7
Maximum deposition and separation of the liquid on the duct walls, baffles,
turning vanes, and stack liners
Good collection and protected drainage of the liquid from the duct and stack
system
In addition, special care is needed if an ID fan located in the wet portion of the
system is used. See Section 3.8.2 for guidelines on these components.
The effect of velocity and velocity head on pressure loss and the costs
associated with the fan-energy requirement
The upper limit of the velocity that starts to entrain liquid droplets from the
walls and wall-related discontinuities
The movement of liquid films, rivulets, and droplets along the wall because
of gravity and gas-drag forces
The addition of liquid collectors will add a small amount of additional pressure
loss as a result of the contraction and expansion of the gas flow to get past the
liquid collectors.
For wet ducts with saturated gas flow downstream of WFGD modules, the ductdesign velocity levels are limited by high gas-velocity droplet re-entrainment
values. Suggested maximum design-area average-velocity values for different wet
duct design situations are presented below. The design-area average velocities
used can be any value less than these. It is necessary that liquid collectors and
drains be designed, optimized, and evaluated by an experimental two-phase
laboratory-model test for each installation.
3-8
1. 65 fps (20 m/s): This maximum velocity level should be used if the following
duct features are present:
-
Any duct with an upward slant of less than 30, and in which most of the
duct is horizontal or sloped downward. Some sections with vertical,
upward, smooth duct slants (6090) could also be included if the
velocity for the stack design is compatible.
2. 50 fps (15 m/s): This maximum velocity level should be used if the
following duct features are present:
-
If higher velocities must be used, then the re-entrained liquid could still be
collected by the installation of more extensive baffles and liquid collectors in the
stack bottom.
3-9
The following duct components have desirable features for inclusion in wet duct
sections:
Vertical to horizontal elbows with large vanes that can act as impingement
and collection surfaces
Ducts that are horizontal or sloped downward, in which water will easily flow
on surfaces in the direction of the gas flow
The use of a few large vanes in bends that can have vertical or inclined liquid
collectors on the outlet edges of the vanes. The vanes may also have extended
trailing edges with inclined trailing-edge collectors for improved collection
and drainage.
Expansion joints are natural collection regions and must have drains installed
to prevent overflow and re-entrainment. Drains must be designed to prevent
flow of air into the flue gas duct. The flexible portion of the expansion joint
must not protrude into the duct during plant operation.
The following duct components have undesirable features for inclusion in wet
duct sections:
Ducts that slope upwards on the floor or roof, because liquid can flow back
against the gas flow and be re-entrained.
Internal pipe trusses, gussets perpendicular to the gas flow, louver dampers,
and structural beams located near the stack entrance where liquid can
impinge and be re-entrained as larger droplets.
Unprotected natural liquid-pool areas on the floors where gas flow can reentrain liquid from the pool surfaces. These areas can be protected to prevent
re-entrainment.
A large number of small vanes in bends, because they would require liquid
collectors on each vane.
3-10
In a field installation, liquid collectors and drains will include some of the
following types of devices, fabricated out of corrosion-resistant materials:
To optimize the selection and design of these liquid collectors and drains
for a specific installation, experimental two-phase flow-model tests are
recommended. These tests will provide the designer with a high level of
confidence for successful operation without significant SLD.
3.6.1.3 Internal Structures
Ideally, the inside of the ducting used for wet operation should be as smooth as
possible, with a minimum of discontinuities to act as sites for droplet reentrainment back into the gas flow. However, some structures are often required,
and their negative impact on wet operation must be understood andto the
extent possibleminimized.
To the extent possible, all structural supports should be located on the exterior of
ducts used for wet operation. Internal bracing is often installed in rectangular
ducts on retrofit units that originally operated dry. Horizontal bracing is also
commonly used to support the tall side-walls of breech openings on FRP liners
with side-entry breeches. These internal support structures are not favorable for
wet operation because they provide additional impingement surfaces and reentrainment sites for liquid droplets. Liquid collectors should be designed for
these braces, which reduce the re-entrainment of large droplets.
Internal trusses, if necessary, should be designed to direct liquid collected on
them to the side-walls and duct floor (Figure 3-1). Trusses that direct collected
liquid to the center of the duct should be avoided. Internal truss work is typically
fabricated from 36-in (75150-mm) piping attached to the duct surfaces using
gusset plates. Liquid deposited or condensed on the trusses will flow around to
the back or downwind side of the truss and along the length of the trusseither
downward because of gravity, or across the duct if the support is horizontal.
3-11
LIQUID FILM
FLOW TO WALL
DUCT WALL
LIQUID FILM
FLOW TO
FLOOR
Good Truss
Arrangement
LIQUID FILM
FLOW TO
CENTER OF
DUCT
DROPLET
REENTRAINMENT
FROM CENTER OF
DUCT
Poor Truss
Arrangement
LIQUID FILM
FLOW TO
FLOOR
Figure 3-1
Good and Poor Internal Duct-Truss Arrangements for Wet Operation
3-12
Internal duct-support trusses are often located at the end of duct sections, and as
such, can be located immediately upstream and downstream of an expansion
joint. If the truss members are attached to the floor using a gusset plate, care
must be taken to ensure that the expansion joint is protected from a gas-flow
separation zone formed downwind of the gusset plate(s) (Figure 3.2). This
recirculation zone can be very intense, and it can easily rip water out of the
expansion-joint cavity if the joint is not drained. If such an arrangement exists, it
is highly recommended that a 34-ft (0.91.2-m)-wide metal plate be placed
over the joint. The plate should be centered immediately behind the gusset plate
for protection from the gas-recirculation zone. The plate will need to be
fabricated from a corrosion-resistant material and will need to be secured in such
a manner that it will not hinder the relative motion between the adjacent duct
sections.
3-4 DIA PIPE
GUSSET PLATE
DUCT FLOOR
RECIRCULATION ZONE
FORMED BEHIND GUSSET
PLATE
GAS FLOW
GUSSET PLATE
EXPANSION
JOINT
Figure 3-2
Internal Duct-Truss Gusset Plate
3-13
3-14
3.8 Fans
Occasionally, ID fans will be located between the absorbers and the stack of a
utility power plant. In this type of application, the ID fans will operate wet. Only
a few existing FGD systems were designed for wet fans, primarily because of the
high cost associated with the requirements for corrosion-resistant materials and
the inherent problem of rotor imbalance caused by scale buildup. For most
installations, wet fans should be considered only if they can be constructed of
corrosion-resistant alloy and if a spare fan can be installed to permit a regular
maintenance program of cleaning and rebalancing.
Wet ID fans will require periodic or continuous washing to prevent solids
buildup on the fan impellers, which can cause fan-rotor imbalance. All of the
wash liquid sprayed into the fan inlet leaves the fan impeller as droplets. Most of
the liquid will be propelled by the high centrifugal-force field to the fan scroll,
where it will deposit but immediately re-entrain as a result of the high gas-stream
velocity. Most of the washing liquid will escape the fan discharge as droplets
entrained in the gas stream, causing additional liquid load in the ducts and in the
stack. The fan-wash method can be optimized for effective washing with the
minimum wash-flow rate. Two-dimensional computer calculations are used to
optimize the fan washing. The duct and stack liquid-collection system must be
designed to account for the extra liquid carryover from the wet fan-wash system.
Fans are used with WFGDs for system pressurelevel control and for the
absorber-pressure loss. The fans are either ID fans or booster fans added to the
ID fan to produce the required pressure rise. Large centrifugal fans are used most
frequently, but axial fans are also used. These fans are mostly located upstream of
the absorber, but sometimes the design places them downstream of the absorbers.
3.8.1 High-Gas-Temperature Fans
The fans located upstream of the absorber are operating with high gas
temperatures and a lower gas density, which means that a higher-head-rise fan
must be used to produce the pressure loss of the FGD system. They are
downstream of the electrostatic precipitators or bag-houses and encounter only
limited fly-ash erosion. Axial fans are more sensitive to erosion than centrifugal
fans. No fan-liquid collectors are needed.
3.8.2 Wet Fans and Fan-Liquid Collectors
Fans located downstream of WFGDs operate with saturated gas at adiabatic
temperature level (wet fan). The head rise of the wet fan produces a higher
pressure rise with the lower temperature and a higher-density gas. The gastemperature rise produced by the fan reduces the relative humidity below
saturation in the ducts and the stack. The rate of liquid condensation in the stack
is reduced or eliminated completelyan important beneficial effect.
The carryover from the mist eliminators usually contains water droplets, fly ash,
and scrubber sludge, which cause solids buildup on the blades of the fan
3-15
impellers. Wet fans have to be washed to keep the impellers in balance. The fanwash spray has to be designed to work effectively with the specific fan-blade
shape and speeds, spraying periodically or continuously. The fan wash is less
sensitive to the fan-blade design of the axial fans.
To minimize the SLD, liquid collectors are needed. Fans are good liquid
separators, but the liquid has to be captured in the high-velocity gas flow in the
scroll, and in the ducts and the stack downstream of the fan. The centrifugal fans
can be equipped with scroll-liquid collectors especially designed for the high gas
velocity in the scroll. The ducts and the stack may require the usual liquidcollection system with wet fans, depending on the temperature riseabove
saturation remaining in the gas entering the stack.
If an ID fan is located in the wet portion of the duct system, the following
steps should be taken:
As much liquid should be collected and drained ahead of the fan as possible.
An on-line periodic fan-wash system should be installed and used to keep the
fan impeller clean and in balance. This system needs to be custom-designed
for the particular fan design.
Liquid collectors should be incorporated into the scroll and the downstream
ducts to collect and drain the fan-wash water.
The power input to the fan may provide a 520F (2.811.2C) reheat of the flue
gas if the liquid-droplet load and its total latent heat of evaporation is minimized.
This reheating will help reduce condensation in the downstream ducts and stack
liner.
3.9 Stack-Liner Design
A variety of factors should be considered in determining the overall geometry
configuration of a new wet stack. The stack-liner geometry should provide the
most economical configuration to accommodate the physical and design
constraints required. The stack-liner diameter should be based on the
recommended gas velocities to accommodate the liquid downflow on the liners.
Platforms, access system, and means of inspection must be considered in the
stack-diameter selection. Stack diameters can be plumb, single-tapered, multipletapered, or incorporate stepped diameter changes. Stack-liner geometry should
also be evaluated to minimize the design implications associated with wind and
earthquake loads. The major items for the preliminary stack design are discussed
in the following sectionsfrom the stack floor to the top of the stack. For
examples of stack geometry for a brick, FRP, and metal-lined stack, refer to
Figures 3-3, 3-4, and 3-5, respectively.
Steel, alloy, and FRP liners are typically, but not necessarily, limited to plumb or
constant diameters. Several different support methods are available for these
types of liners. For steel and alloy liners, a common arrangement is to support the
liner at or near the base or breeching entry. The liner is supported in
compression, and bumpers or stay rods are used to provide lateral support.
3-16
RAINHOOD
LINER
EXTENSION
PLATFORM
OBSTRUCTION
LIGHTS
REINFORCED
CONCRETE SHELL
BRICK LINER
OBSTRUCTION
LIGHTS
BREECHING DUCT
EXPANSION JOINT
LIQUID COLLECTION
GUTTERS
ABSORBER
OUTLET DUCT
LINER FLOOR
DRAINS
CONCRETE
PEDESTAL
PRESSURIZATION
FAN
FOUNDATION
Figure 3-3
Chimney with Constant-Diameter Brick Liner
3-17
LINER
EXTENSION
RAINHOOD
OBSTRUCTION
LIGHTS
PLATFORM
STAY RODS
REINFORCED
CONCRETE SHELL
LINER SUPPORT
STAY RODS
FRP LINER
OBSTRUCTION
LIGHTS
LIQUID COLLECTION
RING
LINER SUPPORT
LINER EXPANSION
JOINT
EXPANSION JOINT
LINER SUPPORT
LIQUID COLLECTION
GUTTERS
ABSORBER
OUTLET DUCT
LINER BOTTOM
ENTRY ELBOW
DRAINS
LINER EXPANSION
JOINT
FOUNDATION
Figure 3-4
Chimney with Fiberglass-Reinforced Plastic Liner
3-18
RAINHOOD
LINER
EXTENSION
PLATFORM
OBSTRUCTION
LIGHTS
STAY RODS
REINFORCED
CONCRETE SHELL
LINER SUPPORT
ALLOY LINER
STAY RODS
OBSTRUCTION
LIGHTS
PLATFORM
STIFFENERS (TYP.)
BREECHING DUCT
EXPANSION JOINT
EXPANSION JOINT
LIQUID COLLECTION RING
LINER SUPPORT
LIQUID COLLECTION
GUTTERS
ABSORBER
OUTLET DUCT
SLOPED LINER
FLOOR
DRAINS
FOUNDATION
Figure 3-5
Chimney with Metal Liner Typical for Alloy, Glass-Block, or Protective-Coating
Liner System
Tall steel and alloy liners supported in compression would need to be evaluated
for the cost of increased wall-plate thicknesses, versus providing a second support
level near the top of the chimney to support most of the liner in tension. An
expansion joint would be required above the lower support to accommodate
thermal movements.
3-19
FRP liners are generally supported with the major part of the liner in tension.
Supports are provided at two locationsone near the top and the second near
the breeching levelwith an expansion joint located above the breeching level.
Expansion joints in wet stack liners are a problem because they tend to collect
liquid, resulting in complete liquid re-entrainment at the expansion-joint cavities.
Therefore, expansion joints should be avoided or provided with liquid collectors.
If required, expansion joints should be located 1 to 1.5 liner diameters above the
roof of the breeching duct or bottom-entry elbow to ensure that the lower liner is
unobstructed for liquid collection and that liquid film flows to the liner-drainage
points.
Because of structural considerations to accommodate breeching openings and
design loads, brick liners can be plumb or tapered. Depending upon the
breeching-opening elevation and other factors, brick liners can be supported from
the stack foundation, from a concrete pedestal, or from a corbel support in the
concrete column. Chimneys with breeching openings that are less than 50 ft (15
m) above the foundation would probably use a brick liner that is supported
directly off the foundation. Breeching openings that are more than 50 ft (15 m)
above the foundation should consider utilizing a concrete pedestal to support the
lining. For tall brick liners with thick walls at the base, it is generally more
economical to use a concrete pedestal for high breeching entries.
Brick liners in high-risk seismic locations sometimes use a corbel-supported liner
to reduce the seismic loads. The liner is constructed in short segments, which are
supported from concrete ledges in the chimney, called corbels. This method of
construction is not recommended for wet stack applications. Liquid and flue gas
leaking through the brick liner can attack the concrete support system.
Additionally, it is difficult to pressurize the annular-space areas at each of the
corbel sections.
3.9.1 Liner Materials
Several types of liner materials and coating systems are available and have been
used successfully in WFGD environments. The advantages and disadvantages of
each system must be considered prior to the selection of the liner material or
coating. The following sections describe the most common materials that are
currently considered appropriate for wet stack applications. These materials have
previously been utilized with some degree of success. Other materials and
suppliers may be available and acceptable, and they should be evaluated on a
case-by-case basis. Operating conditions, design conditions, and economics all
play important roles in this decision. The advantages and disadvantages of several
different liner and coating-system options, along with estimated installation
costs, are presented in Table 3-2.
3-20
Table 3-2
Liner Material of Construction
Liner Material or
Coating
Advantages
Disadvantages
o
FiberglassReinforced Plastic
Borosilicate-Glass
Block
Cost-effective
Liquid adheres to the
porous surface.
Protective Coating
on Carbon Steel
Alloy C276
Excellent corrosion
resistance
Welding seams
Iron Contamination
Acid cleaning
Condensation
3.9.1.1 Brick
Acid-resistant brick is a solid kiln-fired brick made of clay, shale, or mixtures of
clay and shale, which conforms to the requirements of ASTM C980, Standard
Specification for Industrial Chimney Lining Brick. ASTM C980 defines three types
of brick, which are categorized by compressive strength, water absorption, and
acid solubility. Of the three types, Type III has the most stringent requirements,
allowing the least amount of water absorption and acid solubility, whereas Type I
3-21
allows the highest levels. Because of the stringent requirements, Type III brick is
the most difficult and expensive to manufacture. Because of availability
limitations for Type III brick, most brick liners for FGD applications are
constructed using Type II bricks.
The physical properties and chemical requirements for Type I, Type II, and Type
III brick, as defined by ASTM C980, Standard Specification for Industrial
Chimney Lining Brick, are presented in Table 3-3.
Table 3-3
Physical Properties and Chemical Requirements Of Acid-Resistant Brick1
Minimum Compressive Strength,
Gross Area (psi)
Designation
Type I
Average
of 10
Individual
Average
of 10
Individual
Maximum
Average
Weight Loss
by H2SO4
Boil Test
(%)
8,500
8,000
6.0
7.0
20
9,000
4.0
5.0
12
1.0
1.5
Type II
10,000
Type III
12,000
10,000
Per ASTM C980, Standard Specification for Industrial Chimney Lining Brick 1 psi = 6.895 kPa
other types of liners. However, maintenance for a brick liner can generally be
higher than that for an alloy or FRP liner. Typical maintenance for a brick liner
would include repair of brick cracks, repair or replacement of bands, maintenance
of the pressurization fans, and repair of expansion-joint seals at the rain hood and
breeching.
Brick liners should be designed and constructed in accordance with the
recommendations of ASTM C1298, Standard Guide for the Design and
Construction of Brick Liners for Industrial Chimneys.
3.9.1.1.1 Mortar Joints
Careful attention must be paid to the quality of mortar joints in a brick-lined wet
stack. All joints must be struck smooth with the brick surface in order to
eliminate (to the extent possible) horizontal discontinuities on the liner surface.
Mortar protuberances and depressions will act as sites for liquid collection and
re-entrainment back into the gas flow.
3.9.1.1.2 Annular Space and Pressurization
Sufficient space should be provided between the inside of the concrete column
and the outside of the liners to allow for inspection and maintenance over the full
height of the stack. A minimum annular space of 2 ft 6 in (0.76 m) is
recommended. If platforms, ladders, and an elevator are located on the interior of
the stack, sufficient annular space needs to be provided for these items. The size
of the annular space should account for differential movements that may occur
between the concrete column and the liner as a result of wind, seismic, or thermal
expansion. Damage to platforms or the liner may occur if sufficient space is not
provided. Access and adequate clearance for clean-out doors and continuousemissions-monitoring equipment also needs to be considered. Ports should be
accessible and provide sufficient clearance between the column and liner to insert
EPA test probes. To minimize annular-space requirements, doors can be
provided in the column. Monorails can be used to assist in installation of the
probes. In order to ensure personnel safety, the annular space should be wellventilated, especially when an interior access system is used.
Annulus pressurization of the brick-liner stacks increases the rate of condensation
on the liner surfaces. Leakage of pressurizing air through the liner in older brick
stacks can significantly increase the condensation.
Because of the porous characteristics of brick and mortar, brick liners operating
under WFGD conditions should use an annulus-pressurization system. Annulus
pressurization is not needed for other liner materials. An annulus-pressurization
system consists of fans that provide ambient air under positive pressure into the
annular space of the stack, ensuring that permeation of the flue gas and liquids
through the liner and liner cracks is minimized. The pressure created by these
fans should exceed the maximum anticipated positive pressure inside the liner by
a minimum of 1 in Wg (249 Pa). Flue gas and liquids that permeate the liner
3-23
create a highly corrosive environment and can cause damage to the liner bands
and concrete column.
Airtight seals are required at all openings in the stack and liner to minimize air
leakage from the pressurization system. Air-locking chambers should be provided
at door locations to provide safe passage into and out of the chimney.
Pressurizing fans typically have a 30,00070,000-cfm (1433-m3/s) flow rate and
a 34-in Wg (7471245-Pa) rise for a 500650-MW unit.
An adjustable damper can be provided at the top of the chimney for use when
reduced-pressurization pressure and flow rate are acceptable. Excessive annulus
pressurization could lead to air leakage through cracks in the brick into the inside
of the liner and cause increased liquid condensation in the flue gas.
3.9.1.1.3 Target Walls and Linings
Some brick liners operating under WFGD applications have experienced
differential liner growth, resulting in a permanent deflection of the liner.
Although the cause of this permanent deflection has not been determined, wet
saturated conditions and nonuniform temperature appear to be contributing
factors. EPRI has previously funded laboratory and field studies directed at
preventing this problem. The results of these studies are presented in [21].
Precautions such as installing a target wall in the bottom of the liner should be
considered. Mixing of gases with a nonuniform temperature profile should also
take place prior to entering the stack.
A target wall or target lining provides protection to the brick liner from the wet
flue gas. Target walls or target linings can be constructed partially or fully around
the inside circumference of the brick liner, and they should extend from the
bottom of the liner to 1 or 2 liner diameters above the breeching entry.
A target wall is an acid-resistant brick wall, constructed independently of the
brick liner. An air space should be provided between the target wall and the brick
liner.
A target lining is constructed directly against the inside surface of the brick liner.
The lining can be constructed using an elastomeric coating or a borosilicate-glass
block with an adhesive membrane, which provides a good moisture- and
chemical-resistant barrier.
Another method that has been considered to reduce the direct impingement of
wet gases exiting the breeching onto the interior surface of the liner is providing a
bottom entry into the brick liner. This can be accomplished by providing an alloy
breeching-thimble section that extends below the bottom of the brick liner.
Target walls may be needed for good liquid collection and reduced erosion of the
brick liners in the stack-entrance region.
3-24
material on the inside of the liner should be minimized to reduce the potential
for creating horizontal discontinuities on the liner surface, from which droplets
may be re-entrained back into the gas flow. The surface of the joint material
must be tapered back to the base-liner material with a minimum 6:1 taper (10:1
preferred).
3.9.1.3 Alloy
Stainless steel covers a wide variety of corrosion-resistant materials. Operating
parameters and environmental factors greatly affect alloy performance, especially
pH, temperature, and chloride and oxygen levels. The corrosion resistance of
stainless steel is improved by increasing its molybdenum, nitrogen, chromium,
and nickel contents. Molybdenum and, to a lesser extent, nitrogen improve
resistance to pitting. Chromium aids in the development of protective, passive
surface films and increases corrosion resistance in oxidizing environments. Nickel
assists in the renewal of damaged passive films and stabilizes the austenitic
microstructure with its improved fabricability and weldability.
On average, chloride levels in the new FGD systems will likely be much higher
than earlier FGD systems because of the requirement for zero-discharge water
systems. Designs specifying chlorides above 15,000 ppm have been fairly
common, and some designs have anticipated sustained operation above 30,000
ppm. Closed-loop operation causes chloride levels to increase, thereby increasing
the possibility of pitting and/or crevice corrosion of stainless steel. Stainless steels
that have been used successfully in scrubbers may not perform well in FGD
outlet ducts or stack liners. These locations are wetted by essentially unbuffered
mist and/or condensate. Absorption of residual SO2 can cause the pH to drop to
below 2. Because of the potential for high chlorides and low pH in the absorber
outlet duct and stack, stainless steels are not recommended. For these areas, a
nickel-alloy material such as Alloy C276 or Alloy C22 is recommended.
High-nickel alloys are an effective corrosion-resistant material in wet FGD
applications. Nickel alloys containing high percentages of chromium and
molybdenum provide good resistance to pitting and crevice corrosion in a highchloride, low-pH environment.
Although several different nickel alloys may be adequate for FGD applications,
Alloy C276 (ASTM B575, UNS N10276) and Alloy C22 (ASTM B575, UNS
N06022) have been shown to provide the best overall corrosion-resistance
characteristics. Alloy C276 contains 16% molybdenum and 15.5% chromium.
Alloy C22 contains approximately 13% molybdenum and 22% chromium. These
alloys have performed well in numerous WFGD environments and offer
approximately the same level of corrosion resistance. The two products have
successfully withstood a wide range of acid concentrations and extremely high
chloride levels, in conditions ranging from 130F (54.4C) saturated flue gas to
full system bypass.
3-26
3-28
Because of its high insulative properties and low thermal mass, borosilicate block
will minimize the quantity of thermal condensation within the liner during unit
startup. The block also provides some level of protection to the base-liner
material in the event of a stack fire. In such a situation, the lining system will
most likely need to be replaced, but the structural portion of the liner may be
undamaged.
Borosilicate-glass block is well suited for lining ductwork. However, because the
block can be easily damaged, areas that are susceptible to physical or mechanical
abuse should be protected with an abrasion-resistant coating. Internal bracing is
difficult to line with block. Membrane coatings or alloy materials are
recommended for internal braces.
3.9.1.4.1 Mastic Joints
Borosilicate block is attached to the substrate material using an
adhesive/membrane material. It is important that the liner substrate be fully
covered with the membrane material to ensure that a continuous chemical- and
moisture-resistant barrier is formed between the inside and outside of the flue. A
nominal adhesive/membrane thickness of 1/8 in (3.2 mm) is recommended
behind and between each block. A 1/16-in (1.6-mm) thick layer should be
toweled onto the liner, as well as to the sides and back of each brick, before
installation. This double-buttering technique ensures a full bond of each block to
the liner and to each other. Excess adhesive material squeezed out during
placement of the block should be struck clean. Every effort should be made to
minimize the quantity of adhesive material smeared on the exposed surface of the
block, because these locations can lead to liquid re-entrainment back into the gas
flow during wet operation.
3.9.1.4.2 Lining Inside of Concrete Shell
Borosilicate-glass block can be attached directly to the inside surface of a
concrete shell. This technique has been used to reduce the cost of stacks and also
as a method for reusing older stacks whose liners are not favorable for wet
operation. The most common application is in existing stacks with tapered brick
liners previously operating with flue gas reheat. After conversion to wet
operation, the upper portions of the liner would be operating at gas velocities in
excess of the recommended velocity for wet operation for brick-liner material. In
such a situation, the upper portion of the brick liner can be removed down to a
point at which the gas velocities are acceptable within the brick portion of the
liner, and the inside of the existing larger-diameter reinforced-concrete shell can
be used as the upper portion of the liner. In such a situation, the inside of the
upper shell would be lined with borosilicate block to protect the reinforced
concrete. This arrangement is presented in Figure 3-6, where it can be seen that
a ring collector will be required at the interface between the lower and upper liner
sections of the flue to collect condensate generated within the upper portion of
the liner. This arrangement allows the entire liner to operate at acceptable liner
velocities.
3-29
BOROSILICATE
BLOCK
ATTACHED TO
INSIDE OF
SHELL
REINFORCED
CONCRETE SHELL
LIQUID
COLLECTION
GUTTER BETWEEN
LINER SECTIONS
BRICK LINER
REMOVED TO
THIS
ELEVATION
LIQUID COLLECTION DRAIN
LINER BANDS (TYP.)
FLEXIBLE
CONNECTION
BREECHING DUCT
EXPANSION JOINT
LIQUID COLLECTION
GUTTERS
EXISTING
BRICK LINER
DRAIN
PIPES
CONCRETE
PEDESTAL
FOUNDATION
PRESSURIZATION
FAN
Figure 3-6
Borosilicate BlockLined Upper Wind Screen
evaluate in such a system. The site-specific situation and risks involved should be
evaluated when considering this system for a utility wet stack.
3.9.1.5 Liner-Protective Coatings
Many existing units utilizing reheat and bypass liners have been constructed
using bare carbon-steel plate. This liner material is not acceptable for wet
applications. Carbon steel will corrode rapidly under wet acidic conditions.
Steel liners with an existing gunite lining are also not considered acceptable for
wet operation. Gunite is porous and tends to develop cracks through which liquid
can penetrate and attack the substrate material. Conversion to wet operation
requires removal of the gunite and relining with a corrosion-resistant material.
If liquid collection can be effectively collected without modifying the existing
liner geometry, applying corrosion protection to the carbon-steel liner should be
considered. Because the available options involve extensive field construction, an
extended unit outage is required to perform this work. The expense of
performing these modifications and the lost revenue associated with the unit
outage needs to be evaluated against the construction costs for a new stack that
would require a minimal amount of unit outage.
The following modification options can provide corrosion protection to the
carbon-steel liner:
Protective coating
resistant than the polyester resins that have been used in older liner-coating
applications. This coating system should be applied in three layers. The first layer
is a primer coat, which is used to prevent abrasive blasted steel from developing
rust bloom. The lining is then trowel-applied in two 3040 mil layers. Toweling
and subsequent rolling allows the glass-flake filler in each layer to be properly
oriented to the substrate and to achieve maximum resistance to water-vapor
permeation. In order to ensure adequate coverage, the two layers of resin can be
applied using different color pigments. Historical experience indicates that this
coating system may develop permeability, and liquid may eventually seep through
and reach the carbon-steel substrate. Coating systems other than the flakeline
resin systems have been utilized, but with only a limited degree of success.
The performance of any coating is dependent upon surface preparation and
application. Strict quality-control measures are required. The coatings need to be
applied under good weather conditions or under a controlled environment. The
coating should not be applied at temperatures less than 50F (10C) or greater
than 120F (49C). In order to get optimum performance out of a coating,
application and curing conditions are critical. The steel-substrate surface requires
significant surface preparation prior to application of the coating. All sharp edges
and imperfections must be ground smooth. Degreasing the surface and then
sand-blasting to a white metal blast (SSPC SP-5) is required. Blasting should
take place only over an area that can be primed within the same day. Surface
preparation and installation should carefully follow the manufacturer's written
instructions.
Coatings have been used successfully in some wet stack applications and have
failed in others. There is some risk involved in using a coating system, and the
potential for failure should be realized. Maintenance to repair chips and blisters
should be anticipated. Total lining replacement is typically required after a life of
about 810 years. However, coatings can represent substantial savings, and they
have been included as an alternative here for cost-comparative purposes.
3.9.1.6 Liner Insulation
A wet stack with a flue gas temperature of 130F (54C) typically will not require
insulation on the exterior surface of an alloy or FRP liner, because the liner is
already operating in a wet condition. Insulation is installed on coated steel liners
to reduce the temperature gradient across the coating, which may help in
extending the life of the coating. Borosilicate-lined stacks are not insulated to
ensure that the temperature of the mastic/adhesive material between the block
and liner wall does not exceed the manufacturers recommendations.
Insulation should be considered if the condensation calculations predict a high
liquid-film thickness within the liner, which could lead to an increased potential
for SLD. A thick liquid film could occur because of a high condensation rate or
could be the result of a tall stack. Insulation should also be considered as a means
of minimizing the potential for SLD if its minimization is a high priority; for
example, such might be the case if the stack were located in a densely populated
area.
3-32
3-33
Table 3-4
Recommended Stack Velocity Range for Stack-linerDiameter Sizing
LINER MATERIAL
Liner
Velocity
Liner
Velocity
45
13.7
55
16.8
55
16.8
60
18.3
55*
16.8*
(ft/s)
Acid Brick
Radial tolerance of construction is 1/8 (3mm)
Alloy
Weld bead height <= 1/8 (3mm)
Borosilicate Block
Radial tolerance of construction is 1/8 (3mm)
Coatings
(m/s)
* Laboratory testing of coating material is recommended to finalize liner velocity for favorable wet
operation.
Prevented from re-entering the breeching area by side and top gutters on
each entrance opening
Drained to a protected sump in the bottom of the stack liner and out of the
system
Although single entrances are easiest to work with to obtain good liquid
collection and drainage, a double-entrance stack with a special center baffle for
liquid collection and drainage can also be made to work satisfactorily.
3-34
Variations can be made in all of these parameters and still achieve a satisfactorily
low level of SLD. However, more extensive liquid-collector and drain systems
will be needed, and further steps may be needed to reduce condensation in the
stack if velocities are in excess of stack-liner design values.
3.9.3.1 Side-Entry Breech
Side-entry breeches can be used on liners of any material. Because the flow
gasses entering the liner must make a sharp 90 turn to vertical after passing
through the breech opening, strong secondary-flow vortices are generated, which
are required for effective liquid collection. These secondary flows are stronger in
a liner with a side-entry breech than in one with a bottom-entry elbow.
Additionally, the liquid-collection system in the lower liner of a unit with a sideentry breech is less complexand generally performs betterthan that of a unit
with a bottom-entry elbow. For this reason, from the perspective of effective
liquid collection, a side-entry breech is preferred to a bottom-entry elbow for wet
operation.
For tall breeching openings, the vertical edges of the opening should be designed
to possess adequate stability as a column element. Multiple and single openings
under axial compression should be reinforced or proportioned to fully restore the
structural capacity of the entire cross section of the liner. Brick liners utilize brick
pilasters, and alloy liners utilize vertical jamb stiffeners, to provide vertical-edge
stability.
For brick liners, the width of the breeching opening may dictate the diameter of
the liner. The ASTM C1298 brick-liner design guidelines recommend that the
opening width not exceed one- half the internal diameter of the liner at the
opening elevation.
3.9.3.2 Breech D/W, H/W Ratios
Side-entry breeching-duct dimensions are typically rectangular, with the long
dimension in the vertical direction to minimize the structural effects on the
column and liner. It is beneficial to increase the height-to-width ratio of the
breeching by transitioning the width narrower and the height taller as the
breeching approaches the liner and by keeping the floor level. These
precautionary steps minimize the amount of liner cross-sectional area that is
interrupted at the breeching openings. Tall, thin breech openings also result in
the enhanced generation of strong secondary-flow vortices in the lower liner.
These vortices are used to push the collected liquid films to their respective drain
points. Experience has shown that the breech height (H) to width (W) ratio
(H/W) of the breeching-duct opening should be in the 23.5 range for favorable
wet operation.
Another dimensionless ratio affecting a breechs ability to generate strong
secondary-flow patterns is the liner-diameter (D) to breech-width (W) ratio
(D/W). This value should be between 1.5 and 2.5 for favorable wet operation.
3-35
3-36
STACK LINER
6-8 GAP
TO WALL
DRAIN
BREECH
BREECH
~1/3 BREECH
HEIGHT
SLOPED LINER
FLOOR INTEGRATED
INTO LINER WALL
DRAIN
~1/4 LINER
DIA.
QUIET SUMP
AREA
Figure 3-7
Sloped LinerFloor Arrangements
A good starting point for material-estimating purposes is to assume that the top
point of the sloped floor impacts the rear wall of the liner at a point 2/3 of the
way up the height of the breech opposite the opening, and that its bottom point
terminates 1 ft (0.3 m) below the floor of the breech at a point approximately 1/4
of the way across the diameter of the liner. Allowing the bottom edge of the
sloped floor to extend below the floor of the breech opening ensures that the
bottom edge of the incoming gas stream will not be peeled off and directed into
the sump area, thus ensuring that the sump remains aerodynamically quiet for
effective drainage. The sloped liner floor can be fabricated as part of the liner
wall, or it can be fabricated separately and supported off of the flat liner floor. In
the latter case, it is recommended that a 68 in (150100 mm) gap be provided
between the sloped floor and the liner wall, both to allow for differential
expansion and to allow liquid flowing down the liner wall to pass into the quiet
area behind the sloped portion of the floor.
Incorporating the sloped liner floor into the liner wall will reduce the quantity of
liner material required. This consideration can be important if the liner is being
fabricated from an expensive material such as an alloy.
3.9.3.5 Bottom-Entry Elbow
Bottom-entry elbows are fabricated from cylindrical sections, typically of the
same material and diameter as the liner, which have been cut into mitered
sections and joined together to form a 90 elbow. The number of miter cuts
defines the elbow. For example, a three-miter-cut elbow has three miter cuts; a
four-miter-cut elbow has four miter cuts (Figure 3-8). Three-miter-cut elbows
are recommended for favorable wet operation.
3-37
Figure 3-8
Miter-Cut Bottom-Entry Elbow Arrangements
It is well understood that, for FRP liners, bottom-entry elbows are a less
expensive approach. Although side-entry breeches are preferred for wet
operation, bottom-entry elbows with a properly designed liquid-collection system
can be just as effective. Their main disadvantage is that, because of the curvature
in the elbow, the liquid-collection system will require more complex shapes and
the existence of strong gas-recirculation zones exactly in the place where a key
liquid-collection ring must be located. If the ring is not properly designed, these
recirculation zones can strip liquid out of the collector back into the gas flow.
Bottom-entry elbows are commonly used on FRP liners because the elbow
segments can be fabricated from additional liner cans fabricated on site. Care
must be taken when fabricating elbows to ensure that the sections remain round.
As with liner joints, every effort should be made to minimize the joint material
on the inside of the elbow and to taper this material back to the base-liner/elbow
material with a minimum 1:6 taper (10:1 preferred) . Discontinuities in the
diameters of adjacent sections can create sites for significant levels of liquid reentrainment back into the gas flow.
3.9.3.6 Liners with Multiple Entries
Stacks can be designed to enclose a single or multiple number of liners. Each
liner can also be designed to accommodate a single or multiple number of
operating units. Economics, plant layout, ductwork layout, and the need to
perform maintenance during unit outages should be considered when evaluating
these options. The capital cost for a stack with multiple liners is typically less
than the combined costs of several stacks with individual liners. However, this
savings is partially offset by the increased costs associated with a larger
foundation, longer ductwork, and (possibly) larger pressure-rise ID fans.
Building a single large-diameter liner to serve multiple units is not
recommended, because if maintenance needs to be done on one unit, all the units
3-38
feeding into the liner will need to be taken out of service. Multiple units
operating within a common liner will also have additional design considerations
that will need to be addressed, such as gas-temperature differentials, variableflow rates, chaotic flow patterns within the liner, and continuous emissionsmonitoring-system certification. In order to monitor readings from individual
units, the continuous-emissions-monitoring system for a liner serving multiple
units would need to locate its equipment in the upstream ductwork rather than in
the stack. Another concern would be that, with one unit off-line, low-load
operations would result in a low exit velocity, causing potential downwash
problems. For good liquid collection, a one-breeching-duct-per-liner design
configuration is recommended. However, if multiple entries into a single liner
exist (or are required) consideration should be given to 1) placing one breech
opening directly above the other or 2) using a division wall within the lower liner
as a means of separating the lower liner into two separate aerodynamic zones that
can be optimized for liquid collection.
3.9.3.7 Division Wall
A division wall is a wall within the lower liner designed to separate the lower
liner into two (or more) separate aerodynamic zones. The objective is to decouple
the operation of the individual units feeding into the liner from each other so
that the flow patterns and liquid collection in each zone can be optimized
independently from the others.
Division walls typically extend vertically to a point at least one liner diameter
above the roof of the highest breech opening. The sides of the wall are typically
attached to the liner wall, although some designs provide a gap between the
division wall and the liner wall to allow for pressure equalization between the two
zones. The division wall must be able to withstand the full force of the gas jet
entering the liner through the breech openings. As such, it needs to be designed
to withstand the full pressure of the jet, even if the opposite equalizing breech is
out of service.
A significant quantity of liquid will be deposited on the division wall opposite the
breech opening, and provisions must be made to collect the resulting liquid film
and direct it to the liner wall and floor for drainage from the system. Liquid
collectors will also be required along the top edge of the division wall to prevent
the re-entrainment of liquid back into the gas flow. The wall can be placed
perpendicular to the incoming gas flow, or at an angle to it, to promote the
motion of any collected liquid films from the division wall to the liner wall for
drainage.
The design, placement, and orientation of a division wall and its associated
liquid-collection system must be optimized in a physical-flow model of the
subject unit.
3-39
Total pressure losses across expansion joints, corbel joints, and other stackliner breeching openings, expansions, or contractions
Total pressure loss from the stack breeching duct to the stack discharge plume is
the value needed for determining the ID-fan requirements for the stack. The
maximum differential static pressure between the flue gas and the stack annulus
at the same elevation is the pressure needed to design the fans for the annuluspressurization system with brick stack liners. For a wet stackconversion project,
flue gas pressure data should be collected and, after correction to saturated
nonreheat conditions, compared to model study and calculated values.
Details of the pressure-loss coefficients for various liner-entrance geometries can
be found in [4].
3.9.4 Liner-Diameter Changes
Some stacks incorporate changes in the diameter over the height of the liner.
These can be gradual, such as in the case of a tapered brick liner. Or they can be
abrupt, such as in the case of a brick liner with multiple corbels, or in a liner in
which the top sections have been removed so that the upper section of the shell
can be used as the flue. A number of brick stacks have recently been
commissioned in which the lower breech-entry sections were fabricated from
alloy to allow for effective liquid collection, which then discharged into a largerdiameter brick liner. The transition between these two sections incorporated a
ring collector to prevent any liquid flowing down the brick portion of the liner
from entering into the lower alloy portion of the liner.
Gradual changes in the liner diameter are usually characterized by a reducing
diameter in the direction of the gas flow. This characteristic results in an increase
in the liner-gas velocity with height. Attention needs to be given to ensure that
the local liner-gas velocity does not exceed the recommended velocity for the
liner material in use.
Abrupt changes in the liner diameters are typically associated with a sudden
increase in the diameter of the flue. The transition from one section to another is
3-40
an ideal location for liquid re-entrainment back into the gas flow. In these
situations, the transition should incorporate a liquid-collection gutter to drain
away the liquid from the upper section before it flows down into the highervelocity lower section. One notable exception to this rule of thumb is the stackoutlet choke, which incorporates a rapid decrease in the liner diameter in the
direction of the gas flow. Stack chokes are unique in purpose and are discussed in
detail in Section 3.9.6.2.
3.9.5 Internal Structures
Ideally, stacks should be smooth and clear over their entire height. However,
reaching this ideal is not possible, because of the need to incorporate the many
structures necessary for the proper operation of both the liner and the plant.
Access into the liner in the form of ports for such purposes as testing, continuous
emissions-monitoring, and expansion joints must be incorporated into the liner.
Every effort must be made to minimize these structures, and their placement
should take into account their potential impact on the fluid-dynamic and liquidcollection performance of the liner.
3.9.5.1 Liner-Expansion Joints
Alloy and FRP liners sometimes incorporate expansion joints within the liner if
required by the liner-suspension method. These joints are often located at a point
less than one liner diameter above the roof of a side-entry breech or are placed at
the outlet of a bottom-entry elbow. These locations are not favorable for wet
operation because they lie in the region where most of the liner liquid-collection
occurs. Liner-expansion joints are typically longer than their ductwork
counterpartsranging from 25 ft (0.61.5 m) in length and 612 in (150300
mm) in radial depth.
The liquid condensing along the stack liner flows down on the liner surface when
gas velocities are below the liquid flow-reversal velocity. This liquid will flow
downward until it encounters a liner-expansion joint. If an expansion joint is
encountered, this surface discontinuity will cause all of the downward-flowing
liquid to re-entrain back into the gas flow by the gas recirculation in the
expansion-joint cavity. Most of these re-entrained droplets will be discharged
from the stack, and many will be large enough to reach the ground before
evaporating. For this reason, the liner should ideally be designed so that linerexpansion joints are not needed. If an expansion joint is required, it should be
located at least 1 to 2 liner diameters above the roof of the breech duct or the
inlet roof of the bottom-entry elbow. This arrangement will ensure that the
surface of the lower liner, where a significant amount of the liner liquiddeposition occurs, is free from discontinuities that would lead to liquid reentrainment. The expansion joint should be protected from the downwardflowing liquid through the installation of a liner-ring collector above the joint, or
the joint should be equipped with liquid collectors especially designed for linerexpansion joints. Laboratory flow modeling should be used to design and develop
liner expansion-joint collectors that are effective at all operating loads for a given
unit.
3-41
3-42
3-43
In stacks that have chokes, some of the fine droplets entrained in the gas flow
will be deposited on the choke surface, and the liquid collected on the choke will
lead to SLD if the local gas velocities in the choke are high enough to exceed the
flow-reversal velocity. The gas velocity on the choke increases from the liner
velocity, up to an 80120 fps (24.436.6 m/s) exit velocity, which is the usual
design range for chokes. The choke exit velocity normally exceeds the flowreversal velocity. Flow-reversal velocity is the flue gas velocity at which the flow
of the liquid on the stack walls is reversed from down- to up-flow.
The magnitude of the liquid discharge is a function of the choke geometry, the
gas-velocity variation through the choke, the droplet-size distribution, and the
spatial variation of the different-sized droplets. Flow-reversal velocity is in the
range of 7090 fps for common liner materials [1]. Therefore, liquid film will be
flowing upwards over a large percentage of the choke surface for most choke
geometries. The liquid must be collected at the top of the choke and drained out
to reduce the SLD.
3.9.6.3 Rain-Hood and Roof Design
The stack rain-hood shape is designed to cover the gap between the shell and the
single liner. A hood is typically installed on a stack in which the shell to linerdiameter ratio is small. Personnel access to this area will usually be limited. A
stack cap should be fabricated from a corrosion-resistant material such as FRP
and be angled downward away from the liner to promote drainage. Stack hoods
or caps are engulfed by wet stack gas all along their perimeter, and icing will
occur in cold climates. The amount of icing can be both minimal or large and
dangerous.
Hoods are typically left over from the dry stack design and are not favorable for
wet operation. A liner extension above a shell with a roof is favorable.
A stack roof is used on stacks in which the shell to liner-diameter ratio is large
and the liner extends above the roof. In such a case, the annulus is larger, and
personnel access is allowed to the roof. For this reason, a stack roof is structurally
stronger, and it incorporates a guard rail or parapet around its perimeter. Roofs
should be provided with a slope toward drains, leading the collected liquid to the
ground.
Typically, roofs and caps are not insulated, so that heat from the annulus can be
used to eliminate icing from this area.
3.9.7 Cyclonic Flow and Continuous-Emissions Monitors
EPA 40 CFR Part 60, Appendix A, Method 1 states that the flow condition is
unacceptable if the average yaw angle is greater than 20. If potential flowmeasurement problems are significant, it is recommended that a 3-D traverse be
performed to more accurately determine the flow characteristics of each unit.
These characteristics must be discussed with the regional EPA to determine
acceptability. If unacceptable, an evaluation of flow-straightening must be done.
3-44
3-47
The actual design and optimization of the liquid collectors and drains is sitespecific work, and it depends on the absorber type, the duct/stack geometry, the
planned operating-system operating conditions, and the gas-velocity levels.
Because the resulting three-dimensional gas- and liquid-flow patterns are unique
for each unit, the liquid-collection system must be optimized for each specific
unit. A flow-model study should be performed by a flow-modeling company
experienced in state-of-the-art wet stack design to determine the optimum
location and configuration of flow controls, liquid-collection devices, and drains.
Liquid collectors and drains will be required on any unit operating with wet ducts
and a wet stack. Because the collected liquid is acidic, the material selection for
liquid collectors and drains must be evaluated. The most common material for
liquid collectors are corrosion-resistant alloy materials and FRP.
For new installations, a general system arrangement and liquid-collection system
can be designed from the start for favorable wet operation. For units converting
from dry to wet operation, a number of unique considerations must be addressed.
The amount of vapor and liquid carryover from the mist eliminator is the same
for both new and retrofit applications. However, in many retrofit applications,
the gas velocities will be 510% lower than before conversion. Nevertheless, these
velocities are still usually higher than the gas velocities that would be selected for
the design of a new wet stack for the same gas-flow rate.
Because of the amount of liquid condensation and deposition on the duct and
stack surfaces and the relatively high gas velocities, design of liquid collectors and
drains is an important part of the retrofit process.
Mist-eliminator operation and performance need to be reviewed before a
conversion to wet operation. Mist-eliminator washing procedures will need to be
optimized to limit the amount of liquid-droplet carryover. In some instances, it
may be necessary to modify or replace the mist eliminators or mist-eliminator
wash system. Mist eliminators need to be monitored, inspected, and exchanged if
damaged or if solid scaling is excessive. The pH control needs to be good to
ensure that solid precipitation in the mist eliminators and the downstream duct
work is minimized.
Some units undergoing conversion to wet operation may have turning vanes or
other flow-control devices in the gas path between the absorber outlet and the
stack. These devices are common on units with multiple absorption modules and
interconnecting outlet ducting. If this is the case, liquid-collection devices will
probably need to be installed on them to minimize the potential for re-entraining
deposited liquid back into the gas flow.
The following sections discuss the basic liquid collectors typically found in most
effective wet stack installations. These discussions are designed to provide
engineers who are considering wet stack operation a basic understanding of the
design, location, and installation details of these collectors, as well as a basic
understanding of their purpose and operation. Although exhaustive, this list is by
3-49
To be effective, the side-wall and ring gutters must be sealed to the duct wall
along their entire length. Details of basic side-wall and round-duct ring collectors
are presented in Figure 3-9.
LOCAL GAS FLOW
DIRECTION
SEALED TO
DUCT WALL
2-6 (50-150mm)
ANGLE
DUCT WALL
LIQUID FILM
FLOW
PATTERNS
SECTION AA
A
A
LOCAL GAS
FLOW
DIRECTION
10-20
6-8" (150-200mm)
GAP TO FLOOR
20
GAP OPENING
RECTANGULAR
DUCT
CIRCULAR
DUCT
Figure 3-9
Side-Wall and Round-Duct Liquid-Collection Gutters
manner as the liquid is directed onto the side-wall, minimizing the potential for
the scouring of the collected liquid back into the gas flow.
10-45
10-20
DUCT ROOF
DIRECTION OF
LOCAL LIQUID
FILM FLOW
DIRECTION OF
LOCAL LIQUID
FILM FLOW
SEALED TO
DUCT ROOF
SECTION AA
DIRECTION OF
LOCAL LIQUID
FILM FLOW
DIRECTION OF
LOCAL LIQUID
FILM FLOW
2-6 (50-150mm)
ANGLE
DETAIL 1
SECTION BB
DIRECTION OF
LOCAL LIQUID
FILM FLOW
30
DIRECTION OF
LOCAL LIQUID
FILM FLOW
DETAIL 1
Figure 3-10
Ceiling Liquid-Collection Gutters
In some instances, the local gas velocity may be opposite to the direction of the
bulk gas flow. This phenomenon is a common occurrence along the roof of
vertical transitions, particularly in the transition connecting the absorber-outlet
ducting to the stack breech. If the vertical expansion is abrupt, a region of flow
separation will be formed along the roof of the transition, and liquid collected
and/or condensed on the roof of the transition will flow downward back toward
the absorber. In such a situation, a ceiling V collector would need to be
installed, with the point of the V aiming up the slope of the roof. This V
collector will prevent the collected liquid film from reaching the bottom of the
transition roof, where it would be completely re-entrained back into the gas flow.
By orientating the V uphill, the downward-flowing liquid film will be directed
to the duct side-walls, where it will ultimately be directed to the duct floor for
collection
The size, shape, and location of ceiling collectors can only be accurately
determined through a flow-model study of the unit.
3-52
the absorber. The L-shape faces into the wind to form a collection gutter. The
gutters are typically placed on the vane or a vane extension in a slanted
orientation, angled 530 relative to the perpendicular of the gas flow. This
arrangement will direct the collected liquid film to the side-wall or the floor of
the duct, where it can drain onto the wall or floor and, ultimately, to the duct
floor. To be effective, the gutter must be sealed to the turning vane along its
entire length. Similar to the ceiling gutters, the ends of the flow-control
collectors are trimmed back at a 30 angle where they attach to the duct sidewalls. This orientation provides a means for any gas collected in the gutter to leak
out of the gutter in a controlled manner as the liquid is directed onto the sidewall. This arrangement minimizes the potential for the scouring of any collected
liquid back into the gas flow.
The gas- and liquid-film flow patterns in and around turning vanes can be very
complicated, and the development of effective liquid collectors is difficult. It is
highly recommended that a flow-model study be performed to develop and
optimize these collectors.
3.11.5 Internal Duct Supports and Expansion Joints
To the extent possible, all structural supports should be located on the exterior of
ducts used for wet operation. Internal bracing is often installed in rectangular
ducts on retrofit units that originally operated dry. Horizontal bracing is also
commonly found supporting the tall side-walls of breech openings on FRP liners
with side-entry breeches. These internal support structures are generally not
favorable for wet operation because they provide additional impingement surfaces
and re-entrainment sites for liquid droplets. Details of internal truss-support
designs for wet operation were previously discussed in Section 3.6.1.3.
Horizontal braces in side-entry breeches are a unique problem because of their
large diameter and because they are located very close to the stack liner. When
the liquid film flowing on the duct wall encounters a horizontal brace, a portion
of the film will flow onto the support and be pulled toward the center of the duct
in the recirculation zone on the back or downwind side of the truss. Liquid
pulled into this zone will flow along the back side of the brace, re-entraining back
into the gas flow along its entire length. Because of their close proximity to the
liner, many of these droplets will pass through without being recollected. To
prevent this from happening, a circular disk approximately 1012 in (250300
mm) in diameter can be added to the brace from 612 in (150300 mm) off of
the duct wall (Figure 3-11). These disks have been proven to be effective in
keeping the liquid film on the duct wall from spreading onto the support and
being pulled into the center of the duct, where it will be re-entrained back into
the gas flow.
3-54
DUCT WALL
10-12
(250-300mm)
DIA DISK
6-12
(150-300mm)
OFF WALL
GAS FLOW
Figure 3-11
Horizontal Brace with Liquid-Re-entrainmentPrevention Disk
Expansion joints (although required for duct design) are not effective for wet
operation, because they form discontinuities on the duct surfaces (wall, ceiling,
floor), from which the flowing liquid film can be re-entrained back into the gas
flow. The liquid films approaching an expansion joint should be collected and
directed to the duct floor by appropriately designed liquid-collection gutters
upstream of the joint. It is recommended that expansion joints in horizontal
ducting be drained. Liquid will naturally collect in the bottom cavity of these
joints, and they are ideal locations for eliminating collected liquid from the
system. Depending upon the design of the joint, the drain connection can be
made through the rubber boot or through a side structural member of the joint.
A 23-in (5075-mm) drain pipe should be sufficient. The drain piping will
need to pass through a seal pot or run directly back to the absorber reaction tank
to a point below the lowest liquid level to prevent gas from back-flowing up the
drain pipe, potentially resulting in liquid re-entrainment back into the gas flow.
If the expansion joint is not drained, it will fill with liquid. And any liquid
flowing along the duct floor will flow over the joint as it is pushed by the gas flow
toward the liner. Duct misalignments, or a rubber boot protruding into the gas
path, may be sufficient to generate regions of flow separation that could scour
liquid from an undrained joint back into the gas flow.
As described previously, in Section 3.6.1.4, units with internal duct supports will
typically have a brace located on either side of the expansion joint. If the truss
support attaches to the duct floor upstream of an expansion joint filled with
liquid, the recirculation zone formed on the downwind side of the attachment
gusset will scour the collected liquid out of the joint back into the gas flow. To
prevent this, a cover plate should be placed over the joint, extending
approximately 2 ft (0.6 m) on either side of the attachment gusset.
3-55
Arrows Show
Liquid Film Flow
Patterns
Upper Aerodynamic
Zone
Lower Aerodynamic
Zone
1-2 Liner
Diameters Above
Roof of Breech
Breech
Main Droplet
Impact Zone
Figure 3-12
Upper and Lower Stack-liner Aerodynamic Zones
The stack liner opposite the breech opening on a side-entry liner and along the
outer radius of stacks with a bottom-entry elbow generally receives more liquid
deposition than any other area downstream of the absorber. Effective liquid
collection in the stack-entrance zone is therefore critical for both inlet
arrangements. As described previously, when gas flows though a 90 turn, higher
gas velocities are experienced along the outer radius of the turn, and two counterrotating secondary vortices are generated at the outlet of the turn. Similar gasflow patterns are generated in the entrance region of liners as the gas flow turns
vertically up the stack. These gas-flow patterns both drag the deposited liquid
vertically up the liner and also push them circumferentially around the inside of
the liner back toward the breech opening.
3-56
The liquid-collection systems in the stack-entrance region of liners with a sideentry breech or a bottom-entry elbow are basically similar, but they have a
number of geometry-specific differences.
Figure 3-13 presents a generic liquid-collectionsystem arrangement for the
lower liner of a stack with a side-entry breech. This arrangement consists of a
T-shaped sloped-ring collector, typically 6 x 6 x 6 in (150 x 150 x 150 mm),
used to separate the upper and lower aerodynamic zones of the liner. The
upward-facing portion of this collector is used to collect liquid flowing downward
from the upper liner, whereas the downward-facing portion of the gutter is used
to help direct the vertically moving liquid film along the rear wall of the liner
circumferentially around toward the region of lower gas velocity located along the
front wall of the liner above the breech opening. The liquid film in this region
generally flows downward toward the top of the breech opening. To collect this
liquid, a gutter is located across, and extending slightly beyond, the top of the
breech opening. This gutter is typically deep, 1624 in (0.40.6 m), and
incorporates a number of strategically located vertical or angled tabs to prevent
scouring of liquid back into the gas flow. The floor of this gutter is sloped 12
from its center point to its outer ends to promote the movement of collected
liquid to drain pipes located at each end. These 68-in (150200-mm) drain
pipes run down either side of the breech opening, terminating 812 in (200300
mm) above the liner floor.
Liquid collected in the upward-facing portion of the liner-ring collector is
drained into the breech-top gutter through a drain box running down the front
wall of the liner. This drain box is sealed to the liner wall and incorporates wing
extensions to form two vertical gutters running along the sides of the box. These
gutters provide stability to the liquid film flowing circumferentially around
toward the front wall of the liner, directing it into the breech-top gutter.
To prevent the circumferentially moving liquid film from re-entraining back into
the gas flow along the sides of the breech opening, two large wing collectors are
located on either side of the breech, running from the breech-top gutter to the
bottom of the drain pipes (Figure 3-14). These wing collectors are generally L
shaped, but have sides of 8 in (200 mm) and 1824 in (450-600 mm). The
collectors are positioned such that the 8-in (200-mm) leg is attached and sealed
to the liner wall adjacent to the breech opening, and the 1824-in (450600mm) wing is roughly parallel to the liner wallangled slightly so that the
resulting opening or gap to the liner wall at the opposite edge of the wing is 46
in (100150 mm) wide. The gap to the liner wall is maintained by support rods
or straps located approximately every 4 ft (1.2 m) up the height of the gutter.
This gutter is designed to collect the liquid film flowing circumferentially around
the inside of the liner from the main droplet-impact zone on the rear wall of the
liner back around toward the breech opening. In many wet stack installations,
the breech-top gutter-drain pipes will be located inside the channel formed by
this gutter.
3-57
3-58
UPPER RING
COLLECTOR
UPPER RING
COLLECTOR
EXPANSION
JOINT
LOWER RING
COLLECTOR
BREECH TOP
GUTTER
DRAIN BOX
SECTION AA
DRAIN BOX
DRAIN BOX
BREECH TOP
GUTTER
B
BREECH
SECTION BB
WING
COLLECTOR
WING
COLLECTOR
LINER SUMP
BREECH TOP
GUTTER DRAIN
PIPE
LINER DRAIN
SECTION CC
Figure 3-13
Typical Side-Entry Breech Liquid-Collection System
Breech Top
Gutter Drain
Pipe
Breech Top
Gutter
Breech Wall
4-6 (100-150mm)
Open Gap to Wall
Wing
Figure 3-14
Side-Entry Breech Wing-Collector Design
3-59
EXPANSION
JOINT
A
ELBOW RING
COLLECTOR ANTISCOUR TABS
DRAIN BOX
LOWER RING
COLLECTOR
ELBOW RING
COLLECTOR
SECTION CC
LINER DRAIN
ELBOW RING
COLLECTOR ANTISCOUR TABS
BREECH TOP
GUTTER DRAIN
PIPE
SECTION AA
ELBOW RING
COLLECTOR
ELBOW RING
COLLECTOR
DRAIN BOX
SECTION DD
SECTION BB
Figure 3-15
Bottom-Entry Elbow Liquid-Collection System with Liner-Expansion Joint
3-60
ELBOW RING
COLLECTOR
DRAIN BOX
Upward Facing
Ring Collector
10
Liner Expansion
Joint
1.5-2 Liner
Diameters Above
Roof of Breech
Breech
Downward Facing
Ring Collector
Figure 3-16
Liner-Expansion-Joint Placement and Incorporation Within Ring Collectors
Flue
Gas
Flow
Expansion
Joint Boot
Drain
Figure 3-17
Liner-Expansion-Joint Liquid-Collector Concept
3-61
Only laboratory flow-modeling can be used to design and develop linerexpansion-joint collectors that are effective at all operating loads of a given unit.
In some units, the secondary gas vortices in the lower liner may not be
sufficiently strong to push the liquid collected on the rear wall of the liner
circumferentially around to the front wall of the liner. This phenomenon can
occur in units with either side-entry breeches or bottom-entry elbows and is
typically associated with units in which the gas flow entering the liner has a
significant vertical-velocity component. (In other words, it is not horizontal when
entering the liner.) If this situation occurs, the deposited liquid film on the rear
wall of the liner will not drain properly. And as more liquid is deposited, the
resulting liquid-film thickness will increase until droplet re-entrainment occurs
off of the films surface. To prevent this, one or more large V collectors are
placed either in and/or slightly above the droplet-impact zone on the rear wall of
the liner (Figure 3-18). These collectors are made of L-shaped material and are
typically 6 x 6 in (150 x 150 mm), with the gutter opening facing downwards.
These V collectors (sometimes referred to as V diverters) use the vertical
motion of the liquid film to mechanically push it circumferentially away from the
rear wall of the liner to a point where the secondary gas flows are strong enough
to continue pushing the film around to the front wall of the liner. A gap opening
and cover plate are provided at the bottom of the V to prevent re-entrainment
of any liquid collected on the inside of the V back into the gas flow.
All liquid collectors in the stack liner will need to be sealed to the liner wall along
their full length to prevent leakage and possible re-entrainment of liquid back
into the gas flow.
V Diverter
Main
Droplet
Impact Zone
Liquid Film
Flow Pattern
Figure 3-18
Liner Rear-Wall V Diverter
3-62
Bottom
Opening
Cover Plate
path of least resistance. Because of these factors, a large quantity of liquid will
collect at the point where the upward-flowing gas separates from the liner to go
around the gutter. The quantity of liquid in this area will increase until it starts
re-entraining back into the gas flow. Some gas and liquid will be collected in the
downward-facing gutter as a result of the momentum of the gas, but some
bypassing of the gutter and the resulting SLD should be expected.
Reduced Gas
Velocity Within
Collector
Liquid Film
Build-up
DETAIL 1
STACK LINER
GAS FLOW
DRAIN
Detail 1
Expected
DRAIN
Detail 1
Actual
Figure 3-19
Stack-Outlet Liquid Collector
An alternative design is presented in Figure 3-20, in which it can be seen that the
outer top edge of the liner has been sharpened to a knife edge. This arrangement
allows the liquid film to flip over the top edge of the liner outlet before it has a
chance to re-entrain back into the gas flow. Although effective, this collector
design is susceptible to fouling, and such a collector needs to be cleaned on a
periodic basis to ensure that deposits are not formed from which liquid could be
re-entrained.
A flow-model study of any proposed stack-outlet collector should be evaluated
and optimized in a laboratory flow model before installation at a plant.
3-64
COLLECTION
GUTTER
SHARPENED
DETAIL
2
EDGE
DETAIL 1
GAS FLOW
INSIDE LINER
STACK LINER
DRAIN
DETAIL 1
Figure 3-20
Alternate Liner-Outlet Liquid-Collector Detail
3.11.11 Drains
Liquid collected within a wet duct/stack must ultimately be drained from the
system. In general, drains can be installed in the horizontal absorber-outlet ducts,
duct-expansion joints, stack floor, liner-expansion-joint collectors, and choke
collectors. The preliminary wet stack design should plan an adequate drainage
system by considering the drainage needs discussed in this section. The final
number of drains needed and their optimum location are specified as one of the
results of the laboratory flow modeling.
The absorber-duct drains are most effective if a trench or scupper is installed
across the duct in the floor with a drain. The drains, if installed in ductexpansion joints, are selected on the basis of local conditions. Sloping the floor of
the absorber-outlet ducts either back toward the absorber or toward the liner
(preferred) eliminates water puddles after shutdown, when all the liquid from the
walls drains to the floor. The sloping floor does not alter the liquid flow on the
duct floor during operation because the gas shear dominates the liquid motion on
the floor.
Duct-floor drains may be required at locations where liquid pools can build up
and water re-entrainment from the pool may take place. The best locations of the
floor drains are specified by the wet stack flow model.
Drains located on a flat liner floor may need to have a solid cover positioned at
812 in (200300 mm) above the drain opening to provide a quiet area
underneath for effective liquid drainage. These covers are typically 3 x 3 ft (1 x
3-65
1m) and are supported over the drain opening by legs (not unlike a table). All
drainage points should incorporate a wire cage (or equivalent) over their inlet to
prevent plugging by flaking solid scale and other debris.
Drains connect points of different static-pressure levels inside the ducts and liner.
To prevent the possibility of gas back-flowing through the drain pipespossibly
resulting in droplet re-entrainment back into the gas flowthey must all be
individually pressure-balanced through loop seals or drain pots with good on-line
cleaning capability.
Because the liquid is acidic, the material selection for the drain line and the
means of final disposal must be evaluated. FRP pipe is commonly used for wet
FGD drain-line applications. FRP pipe provides good chemical resistance to
these acids, and is more economical and more readily available than stainless steel
or nickel-alloy pipes. Drain lines are sized to prevent pluggage and for ease of
cleanout. To accommodate the potential for solids plugging, drains lines should
be sized to handle a larger liquid-flow capacity than necessary. Main liner-floor
drains should be 1012 in (250300 mm) in diameter, because these are the final
drain points in the system. As such, they are points of possible single-point
failures. Seal pots or siphons should be provided as needed. These devices will
prevent air inflow under negative pressure, which would prohibit liquid discharge
through the drain line or cause droplet re-entrainment back into the gas flow.
Different drain lines can be connected only if the static pressure at the flue gas
end is the same; in other cases, pressure-loop seals or drain pots are required.
Chimney-drain lines should use pipe crosses, tees, or Ys with bolted blind
flanges for cleanout. To reduce pluggage, the drain lines should slope without
horizontal sections. Access to cleanout connections should be provided.
Outdoor FRP drain lines should have a steep-enough slope to prevent freezing.
For long, relatively flat horizontal runs of drain line, the pipe should be heattraced and insulated.
Stacks with a roof instead of a rain hood are relatively flat so as to serve as a
walking surface. Because of the tendency for liquid-fallout accumulation in this
area, discharged liquid and rainwater sometimes freeze on the roof during cold
weather. Ice formation at drain locations may prevent liquid from entering the
drains. In order to prevent this occurrence, roof slopes should be sufficient to
prevent liquid from pooling, and multiple drains are recommended. The roof
drain should be separate from the liner or choke liquid-collector drains because of
pressure differences and different disposal methods.
Liquid collected from within the ductwork or stack can be routed to the waste
sump or returned to the FGD-system process. Liquid routed to the waste sump
does not re-enter the process and is eliminated. Liquid returned to the process
can be routed to the absorber reaction tank. Liquid returned to the process may
be of sufficient volume and pH level that it could change the operational
characteristics of the FGD process. The volume and pH of the liquid returned to
the process should be considered to compensate for any operational changes that
may be needed.
3-66
A single stack liner that serves multiple FGD modules may produce large
volumes of liquid for disposal. Drainage-pipe diameters need to be sized
accordingly. If the liquid collected from multiple FGD modules is returned to
the FGD process, the volume of liquid may be too large to be returned to a single
FGD module without significant changes to its operational characteristics.
3.11.12 Post-Installation Inspections
An often-overlooked part of the liquid-collection-system installation process is
the post-installation inspection. It is important to ensure that the designed flow
controls, liquid collectors, and drains are installed properly. The objectives of this
inspection are to ensure that the liquid-collection system has been fabricated and
installed correctly; to ensure that it is free from leaks; and to identify any
deviations to the recommended design as a result of field modifications made to
ease installation. Inspection of the installation is usually scheduled for a day when
installation of the liquid-collection system is 8090% complete. Such timing
allows on-the-spot modifications to be defined, if necessary, and corrections to
be made while the construction/installation crew is still mobilized on-site.
Inspections of the liquid-collection system and drains should be performed
during the first scheduled or unscheduled plant shutdown. If any stack-emission
incidence occurs during normal operation, the need for inspection is obvious. If
SLD is not experienced, it is still important to inspect the liquid collectors to
ensure satisfactory long-term operation. Problems with solid depositions and
liquid drainage can be detected and corrected during a scheduled or unscheduled
outage. This inspection includes the stack-inlet duct, stack liner, and stack outlet,
if accessible.
3.12 Laboratory Flow-Modeling
An experimental gas-flow-model study is a valuable engineering tool used to
design and develop liquid-collection systems for wet operation.
The primary objectives of the wet stack flow model include
1. Minimizing the discharge of droplets from the top of the stack that are large
enough to reach the ground before evaporation (SLD)
2. Investigating the potential for plume downwash and for icing on the stack
liner or shell at low ambient temperatures
3. Providing velocity profile and swirl patterns at the CEM elevation that
satisfy EPA requirements for flow uniformity
Laboratory flow-modeling provides the following benefits for wet stack design:
1. Estimation of the total liquid load on the liner surface as a result of thermal
and adiabatic condensation. This estimate is based on the liner and shell
geometry, thickness, materials of construction and the flue gasflow rate and
properties over a range of expected plant-operating conditions and worst-case
ambient weather conditions.
3-67
3-68
Figure 3-21
Typical Wet Stack Physical-Flow Models
3-69
3-70
Liner Extension
Stack Shell
Figure 3-22
Typical Plume-Downwash Study CFD Model Results
3-71
3-72
If the prevailing wind direction is in line with the axis of a stack with two liners,
the extent of downwash will be significantly less than if the wind direction is
perpendicular to the axis of the liners (Figure 3-23). In other words, when the
wind is aligned with the axis of the liners, it is more difficult to push the
combined plumes over than when the wind is perpendicular to the axis and the
wind is acting on both plumes simultaneously. Therefore , it is important to
know the prevailing wind direction when developing the initial system design
and setting the relative locations of the liners.
Liners
Favorable
Liner Orientation
Unfavorable
Liner Orientation
Figure 3-23
Recommended Alignment of a Stack with Two Flues
Stacks with three or more liners will require multiple downwash calculations with
respect to the relative wind direction and individual liner-gas velocities to
establish the units downwash potential.
3.13.4 Methods of Downwash Minimization
If the actual or predicted cumulative duration and/or extent of downwash are not
acceptable, there are a limited number of methods available to reduce them:
3.13.4.1 Liner extensions
The most common method of minimizing the potential for plume downwash is
to increase the distance between the top of the windscreen and the top of the
liner in the form of liner extensions. These extensions minimize the potential for
the plume to be pulled into the low-pressure region formed on the downwind
side of the windscreen. Typical liner-extension heights range from 0.5 to 1.5
liner diameters above the top of the stack shell. Within these heights, the
extensions should be self-supporting.
3-73
3.13.4.2 Chokes
Because downwash is controlled by the plume to wind-momentum ratio,
increasing the liner-exhaust velocity will have an immediate impact on reducing
downwash. Although this goal can be accomplished by decreasing the liner
diameter, doing so would most probably result in the liner operating at a velocity
above the recommended value for wet operation, leading to a significant increase
in SLD.
Another method of increasing the liner-outlet gas velocity is the addition of a
liner-outlet choke. This option is typically considered as a last resort; it is an
expensive option with respect to both installation and operation because of the
increased fan pressure required to overcome the increased pressure drop for the
life of the plant.
3.13.4.3 Other Stack-Top Modifications
A limited number of nontraditional stack-top design modifications have been
successfully used to minimize the potential for plume downwash.
Chamfering the top of the windscreen (on 2 sides) can reduce both the onset and
extent of downwash, particularly if the chamfering is aligned perpendicular to the
worst-case prevailing wind direction.
For smaller-diameter stacks without a windscreen, a disk placed around the stack
near the outlet has been shown to reduce the onset of plume downwash. This
disk can either be solid or in the form of a walkway made of grating. The
diameter of the disk and its position relative to the stack outlet must be
optimized by means of a plume-downwash study.
Potential approaches are highly site-specific and they should be evaluated as part
of the plume-downwash CFD evaluation.
3.13.4.4 Protective Shell Coatings
If the expected cumulative duration of plume downwash events is acceptable, or
the extent of plume downwash cannot be reduced, one option is simply to live
with it and ensure that the top of the stack is adequately protected with an
effective shell-coating system.
3.13.5 Stack-Top Icing
Whether ice forms on the top of the stack depends on the temperature of the
surface, the temperature of the mixture of saturated flue gas and cold ambient
wind, and on whether water vapor will condense out of the mixture. Ice
formation is most likely at plants where below-freezing temperatures are
common and where they last for extended periods of time.
3-74
Icing usually does not cause serious ice buildups that can fall to the ground.
However, when the icing conditions are occurring, the platform near the top of
the stack, the railings, and possibly the roof, may be slippery.
The potential for icing can be reduced by employing the following steps:
1. Select a stack-liner discharge velocity that minimizes plume downwash over
the expected operating range of the unit at the existing local wind conditions
and that is consistent with other design objectives.
2. During cold ambient temperature conditions with high winds, run the unit at
near full load. Employing this operating procedure under such conditions
may be natural, because more power is consumed in below-freezing weather.
3. Use heated annulus air or electrical heating elements to heat the stack hood,
roof, or other areas where ice forms on the top of the stack.
4. Insulation is not required on sloped rain hoods constructed of a corrosionresistant material.
3-75
Perform a preliminary economic analysis for liner material and inletgeometry options for wet stack operation.
The feasibility study should address all considerations necessary to ensure that a
wet stack is allowable and that the proposed design will work. The study should
include permit considerations and plume-dispersion modeling as well as
preliminary design choices such as the absorber outlet-duct arrangement, choice
of inlet geometry, velocity limits for liner materials, and preliminary economics.
4.1.1.1 Identification of Issues (Phase I)
A number of important issues must be addressed during the Phase I feasibility
study for the design of a new or retrofit wet stack. These include the following:
Stack height
Liner-inlet geometry
Operating conditions
Model-study testing to minimize system-pressure losses and optimize liquidcollection and drainage
Outage time
Cost
A detailed discussion of each issue that is important for new or retrofit wet stacks
is included in this section of the design guide.
4-2
Capital costs
Operating costs
Maintenance costs
Outage costs
Capital costs for a chimney include the cost of the concrete shell, liner,
foundation, and of miscellaneous items such as access platforms, doors, elevator,
rain hood, pressurization system, drain system, protective coatings, and electrical
system.
Chimney-maintenance costs will vary depending upon choice of liner material.
Coated carbon-steel liners need to be recoated periodically, whereas Alloy C276
clad liners and FRP liners require little or no maintenance. Maintenance on brick
liners can include repair of cracks, replacement of liner bands, and replacement of
expansion joints. All chimneys require periodic inspection.
Outage time does not apply to a new wet stack, because the unit has not yet been
placed into operation. For a retrofit wet stack, outage time depends upon how
long it will take to tie the new absorber-inlet duct to the existing ductwork and, if
applicable, reline or replace the liner within the existing stack shell. Outage costs
can be minimized by scheduling the ductwork tie-in to take place during a
normal maintenance outage. Outage costs per day will vary depending upon time
of year (energy demand), loss of revenue, and cost of replacement power.
The number of years considered in the economic analysis should encompass the
design life of the FGD system. Initial capital costs, periodic or annual
maintenance costs, and annual energy costs need to be converted to present value
to compare the options on an equal basis.
4.1.2 Phase II Design Process
Table 4-1 describes Phase II, which is the process for the design of a wet stack.
This table provides a brief description of each step of the design process, a
designation of each step of the design process, a designation of responsibility, and
4-3
a reference to the section of this guide that provides the relevant background
information.
The design-process phase essentially takes the feasibility study and turns it into a
bid specification. During this phase, it is necessary to establish the design criteria,
define the stack and inlet-duct geometry, perform the flow-model study to
determine liquid-collection devices and drains, and prepare the bid specifications
and drawings.
Table 4-1
Phase II Wet Stack Design Process
Step
Description
Responsibility
Reference
Section 2: "Specific
Design Issues
1. Preliminary
design
2. Preliminary
design review
Engineer
Modeling company
Section 2: "Evaluation
for Wet Operation
Prepare bid
4. Preparation of
specifications for
bid specification
chimney contract.
Engineer
Section 4: "Guide to
Develop Specifications
for Wet Stacks
5. Final design
Chimney contractor
----
Note: The utility must be an active participant in all major decisions in all steps of the design
process.
These design methods are summarized in the first four subsections below.
The subsection on "Comparison of Experimental and Computer Modeling" is an
overview of the basic gas/liquid-flow processes that need to be evaluated in a wet
system using the design tools, and it summarizes the usefulness of the
experimental and computer tools to carry out the needed evaluation.
The last two subsections are a discussion of the system tests and inspections
conducted after the unit is in operation to help identify how well the final
duct/stack geometry and the installed liquid-collection system are working. If
there are wet-operation problems, the tests and inspections can identify the
causes and practical solutions.
4.1.2.1 Preliminary Design Review Based on Experience
The first two steps in Phase II are to perform a component-by-component
design and then adjust the design for liquid-collection requirements. The results
of these first two steps should be evaluated by a preliminary design review.
The preliminary duct and stack-liner design geometry for new construction must
be reviewed to assess how well all individual components will work together as an
integrated system. The second goal of the design modifications implemented at
this phase of wet stack design is to make the integrated geometry more suitable
for liquid collection. This review should be based on actual field experience with
liquid collectors designed for different power plants. This modified wet duct and
4-5
stack design is the starting geometry for the design of the liquid-collection
devices. It also optimizes gas-flow passages by experimental flow-modeling. It is
important to get the engineer and modeling company together to review the
preliminary design early in the design process, because simple changes made at
this point in the process can have a big impact on the liquid-collection
performance of the entire system.
4.1.2.2 Condensation Calculations
A larger portion of the liquid flow on the liner surfaces is caused by condensation
than by deposition of the droplet carryover from the mist eliminators. The
condensation on the liner cannot be measured in a laboratory-scale model of the
stack. Therefore, analytical calculations are required to define the amount of
condensed-liquid flow rate in the stack.
The basic description of the adiabatic bulk condensation and the thermal-wall
condensation processes is given in Section 2.3, "Sources of Liquid in a Duct/Wet
Stack System." This section discusses the methods used to predict condensation
rates in the duct and the stack and also presents calculated condensation data for
typical sample stacks.
Liquid condensation occurs continuously on all wet duct and stack-liner surfaces
because the flow gas is saturated and the liner's inside surface temperatures are
lower than the gas dew-point temperature. But the rate of condensation per unit
surface area can vary significantly, depending on geometry, materials,
temperatures, and wind speed.
The duct- and stack-condensation rate cannot be measured in experimental flow
models. Three-dimensional computer modeling can be adapted for condensation
calculations; however, such models can be complex and are unnecessary for this
essentially two-dimensional process.
A 2-D condensation analysis is satisfactory for calculating the condensation rate
that accounts for stack geometry, mass flow, heat loss, gas-psychrometric
conditions, and the ambient atmospheric-pressure variation along the height of
the stack. Assuming axial symmetry for the stack, all the variables can be
described or calculated as a function of vertical height with computer programs
developed for this purpose as design tools.
Numerous studies performed using these programs have drawn the following
conclusions:
For brick liners, the pressurizing-air leakage through the liner increases the
bulk condensation by additional cooling.
4-6
The liner material does not significantly affect the wall-condensation rate
because the thermal resistance of the liner is a small fraction of the total
thermal resistance between the flue gas and the ambient air.
Having more than two liners in a shell increases the total wall-condensation
rate by about 5%. This increase is mainly the result of a higher condensation
rate on the liner extension above the shell.
4-7
Table 4-2
Estimated Ranges of Flows in the Wet Duct/Stack of a Typical 550-MW Plant
Source of Liquid at the Stack
Estimated Flows
Mist-eliminator carryover
-
Duct-wall condensation
Stack-wall condensation
Liquid in drains
Preventing the discharge from the top of the stack of droplets that are large
enough to reach the ground before evaporation
Investigating the potential for plume downwash and for icing on the stack
liner or shell at low ambient temperatures
and stack surfaces, prevent re-entrainment, and guide the liquid to locations
where it can be drained out of the system. This development work is
conducted on a scale model of the wet duct and stack system, from the outlet
of each absorber to approximately two stack-liner diameters above the top of
the breeching duct. Even though the absorber-outlet-duct region is not a
part of the chimney design, the duct liquid-collector designs should be
developed here and provided to the owner so that an integrated complete
system of liquid collectors can be developed for the unit.
Wind-tunnel tests that evaluate downwash of the wet plume, which could
lead to stack surface deterioration, unacceptable ground-level concentrations
of SO2, or icing problems at the top of the stack. The final geometry of
chimney-cover shape and the extension of the liner will be specified as a
result of this work.
Evaluation of existing duct and stack gas-flow passages for gas/liquid flow or
wet operation. Existing surface discontinuities, duct bends, and stack-liner
4-9
surface and geometry are evaluated, with the air and water flow simulating
the field conditions.
Design of duct and stack modifications to make the geometry more suitable
for wet operation. These modifications are limited to internal changes only.
Internal changes are inserts, target walls, turning vanes, and baffles that make
the flow more favorable for liquid collection and promote the deposition of
liquid droplets entrained in the gas flow.
Evaluation of wet operation with existing internal trusses. Decide if they are
acceptable or if they need to be equipped with liquid-collection devices or
replaced with a more suitable truss design.
Definition of the best location for drains, selection of possible drain locations
within the geometric limitations, and experimental evaluation of the final
drain system.
Wind-tunnel tests to evaluate downwash of the wet plume that could lead to
stack-surface deterioration, unacceptable ground-level concentrations of
SO2, or icing problems at the top of the stack. The final geometry of the
chimney-cover shape and the extension of the liner will be specified as a
result of this work.
The applicable programs can be separated into two categories: 3-D and 2-D
computer programs.
In the following subsections, these categories of computer modeling are
compared to laboratory experimental flow-modeling from the wet stack design
point of view only.
4.1.2.4.1 3-D Computer Programs
These are commercially available programs designed for general threedimensional fluid-mechanics calculations. These codes have the potential to
handle all aspects of the gas/liquid-flow in wet stacks. But at the present stage of
development they cannot describe some of the most important wet stack flow
processes, such as re-entrainment and liquid flow on surfaces with gas flow.
Therefore, a complete wet stack analysis, with all the necessary details from the
absorber to the top of the stack, will not be possible by using 3-D computer code
in the near future.
The 3-D computer models are useful to calculate steady gas-flow patterns in the
duct and stack. They are good for major flow patterns, but they become complex
and slow-running programs when small details are included that require larger
computers (for example, the flow in the expansion-joint cavity, the flow around
trusses, and the flow near liquid collectors). The 3-D computer models are very
good at describing the trajectories of liquid droplets of all sizes and their impact
on deposition points in the duct system when combined with the gas-flow
calculations. These codes can be used to calculate deposition on the choke-cone
surface with assumed points of droplet origin. Some of the codes come with
trajectory subroutines that yield very detailed quantitative results.
The condensation on the liner of a wet stack can be described by a 3-D computer
code. But because the process is basically two-dimensional, 2-D analytical codes
are effective and are easier to use.
Several available 3-D computer codes are written for dispersion-modeling only.
The degree of complexity and accuracy of these codes varies a great deal as they
account for the terrain and atmospheric conditions over large distances from the
stack. Computerized modeling, verified by field-dispersion measurements, is the
best tool available at this time to predict the plume-dispersion process. Plume
downwash at the top of the stack could be predicted by an elaborate 3-D
computer model with a high degree of accuracy.
4.1.2.5 Full-Scale Field Tests and Inspections
After the wet stack is placed into operation, the effectiveness of mist eliminators
and downstream liquid-collection devices should be tested to verify that design
objectives have been met. EPRI Report CS-2520, Entrainment in Wet Stacks,
evaluates various methods for making carryover measurements and presents
recommendations for obtaining accurate results [1].
4-11
Additional in-place testing at the CEM level can measure the effectiveness of the
liquid-collection system. These tests can be performed as needed to establish the
liquid-flow in the stack. The following liquid-flow measurements are not
mandatory tests performed as part of the CEM system; however, they would be
beneficial in furnishing some or all of the following information:
Wherever there is a large amount of entrained liquid, the droplets may interfere
with the CEM system velocity and angle measurements. This situation shows
that liquid collection and drainage must be improved, even though the SLD is
acceptable.
4.1.2.7 Outage Time
Outage time is the amount of time necessary for an existing unit to be out of
service because of repairs, maintenance, or modifications. The outage may be
caused by an emergency (unscheduled), or it may be routine (scheduled). Outage
time applies only to an existing, operating unit. Therefore, the designer does not
need to consider outage time for a wet stack on a new unit. For a retrofit FGD
project, outage time is the time it takes to tie the new FGD-system ductwork
into the existing ductwork.
Outage time for retrofit wet stacks should be minimized for economic reasons.
Depending on unit size, energy demand, loss of revenue, and cost of replacement
4-12
power, outage time can be very expensive. To minimize the outage time required
for a retrofit wet stack project, the ductwork connections or tie-ins should be
made during a scheduled maintenance outage.
4.2 Wet Stack Bid-Preparation Process
Once the decision has been made to utilize wet stack operation, it is necessary to
prepare a bid specification for the wet stack. For a new or retrofit wet stack, the
chimney can either be covered in a separate contract or be part of the FGDsystem contract. This section of the guide addresses key issues that should be
included in a wet stack bid specification. These guidelines should not be used
without application of good engineering judgment and consideration of site
specifics.
Preparing a bid specification for a wet stack is a multi-step process. The design
criteria for a wet stack are determined on the basis of plant considerations,
applicable codes, and system-performance requirements. General arrangement
drawings that define duct and chimney geometry should also be included. A
flow-model study should be performed in order to optimize the absorber outletduct arrangement and to design the necessary liquid-collection devices and
drains. At this point, the bid specifications can be prepared. Exhibit drawings
showing the wet stack arrangement and typical design details are included with
the specifications to define the scope of work and obtain competitive bids.
4.2.1 Establish Design Criteria
Minimum standards for design materials and construction are established in the
bid specifications. The chimney contractor is normally responsible for the design
of the wet stack to meet the specified operating conditions and any applicable
codes, such as the latest edition of the American Concrete Institute (ACI 307),
Code Requirements for Reinforced Concrete Chimneys, or ASTM D5364, Standard
Guide for Design, Fabrication, and Erection of Fiberglass Reinforced (FRP) Chimney
Liners with Coal Fired Units.
New stacks are typically designed in accordance with the latest wind and seismic
sections in the applicable steel-stack or concrete-chimney codes. Dynamic wind
loads should be evaluated in relationship to the chimney's critical wind velocity
and natural frequency. Wind effects from adjacent structures and nearby
chimneys should be considered. Closely spaced chimneys can cause an
amplification of the vortex-shedding wind loads for both the new and existing
chimneys. Spacing between adjacent chimneys can be increased in order to
decrease the amplification factor for the vortex-shedding wind loads.
In addition, a subsurface investigation is usually performed to determine the most
appropriate foundation type. For example, the subsurface investigation may
require that a deep, pile-supported foundation (versus a shallow, soil-supported
foundation) be implemented.
4-13
Experimental evaluation of the absorber outlet duct and stack liner in order
to assure proper geometry and favorable wet operation
4-14
4-15
4-16
4-17
Project meetings
Preconstruction conference
Finalizing schedules
Coordination conferences
Safety meetings
Work-progress schedule
Work-progress reports
4.3.1.4 Submittals
Submittal requirements specify what items must be reviewed by the engineer
before fabrication or construction. Submittals include both compliance submittals
and miscellaneous submittals. Compliance submittals include shop fabrication
drawings, product data, and samples that are submitted by the contractor,
subcontractor, manufacturer, or supplier. Miscellaneous submittals include
technical reports, administrative submittals, certificates, and guarantees.
4.3.1.5 Temporary Utilities and Facilities
Temporary utilities and facilities are specified to address water distribution,
drainage, dewatering equipment, enclosure of work, heat, ventilation, electricalpower distribution, lighting, hoisting facilities, stairs, ladders, and access roads.
In general, the specifications usually require the contractor to furnish, install, and
maintain temporary utilities required for construction, safety, and security.
4.3.1.6 Chimney Design and Construction Requirements
This section of the bid specifications includes minimum standards for design,
materials, and construction. The contractor is normally responsible for the
detailed design of the wet stack, whether it be new or a retrofit construction.
Specific design parameters for the piling, foundation, chimney shell, and liner
and/or lining material should be included in this section so the contractor can
perform the final design. The wet stack geometry can be included in this section
of the specifications unless indicated on the drawings. The operating conditions
for the unit, such as maximum gas-flow rate, maximum temperature, and gas exit
velocity at maximum gas flow, should be specified in this section. Specific site
conditions, such as the location of nearby structures and chimneys, are considered
for vortex-shedding wind loads on the new or existing chimneys.
4-18
Fine aggregate
Coarse aggregate
Mixing water
Compressive strength
4-19
Slump
Air Content
4.3.3.2 Foundation
Foundation requirements for a new wet stack are included in the concrete section
of the bid specifications. In some cases involving a new stack, the chimney
contractor is responsible for designing the chimney foundation, including the
number and layout of piling (if required). However, the foundation is often
designed and supplied by the general contractor. Pile type and capacity is
generally determined by the architect/engineer.
4.3.3.3 Chimney Column, Construction Tolerances
The chimney column or concrete chimney shell may require special cement or
concrete additives in the mix design to facilitate rapid construction. Coatings are
usually used to withstand the potentially corrosive environment at the top of the
column. Construction tolerances for vertical alignment of center point, diameter,
and wall thickness for a new chimney column conform to ACI 307.
4.3.3.4 Brick-Liner Support Pedestal
The liner support pedestals usually are specified to meet the same concrete
requirements as the chimney column.
4.3.3.5 Roof Slab
Specifications for a concrete roof slab should consider either a corrosion-resistant
sealant, a coating applied on the concrete surface, or a sulfate-resistant cement
such as ASTM C150 Type V cement.
4.3.4 Liner
Liner materials of construction, strength requirements, construction
requirements, and ambient considerations are specified in this section.
4.3.4.1 Materials
Liner materials of construction for a new chimney liner are specified in the bid
documents. The material selected should conform to the respective ASTM
standards, unless specifically noted otherwise. Refer to Sections 2 and 3 of this
guide for liner-material discussions.
4.3.4.2 Strength Requirements
Strength requirements for various liner materials are dictated by the structural
design of the liner. For example, the yield strength of a steel or FRP liner and the
compressive strength of a brick liner should be specified.
4-20
4-21
At the choke
The size and location of the access doors may need to be evaluated and
determined by the flow-modeling company. Liquid collectors and/or liquid-film
diverters may be required around access doors to prevent droplet re-entrainment
from these areas.
4.3.9 Protective Coatings
Protective-coating systems for each particular component of the wet stack are
included in the bid specifications, if needed. Interior and exterior coating systems
may also be considered for the stack shell and liner. An exterior protective
coating may also be considered for the chimney inlet duct, if it is fabricated using
wallpapered construction or a clad-plate material.
4-22
4.3.12 Mechanical
The following mechanical sections are typically included in the wet stack bid
specifications.
4.3.12.1 Drain Piping
The specifications provide for all pipes, fittings, hangers, supports, and
accessories required to complete the chimney corrosive drain-piping systems
associated with the liquid-collection system. Specific piping material should be
specified for each application. FRP or alloy piping is normally specified for drainpipe material. Minimum piping-installation standards are usually also addressed.
Stack-liner liquid-collector drains should incorporate some easily accessible
method for monitoring or observing the drainage flow. This capability will be
useful for determining the performance (or nonperformance) of the liquidcollection system.
4.3.12.2 Pressurization-System Fans/Dampers
This section of the bid specifications applies to a concrete chimney with an acidresistant brick liner. An annulus-pressurization system is required to keep flue
gases contained within the liner. The pressurization-system section should
include furnishing and installing the required number of fans, including drive
motors, controls, and accessories.
4.3.12.3 Louvers
The bid specifications include a section requiring the contractor to furnish and
install louvers to provide ventilation between the chimney liner and shell.
4.3.13 Electrical
The electrical section of the bid specifications includes provisions to furnish and
install all electric equipment, wiring (including plant-interface wiring),
grounding, and lighting necessary for a wet stack system.
4.3.13.1 Rounding
Grounding for all new electrical equipment is included in the bid specifications.
The grounding requirements for a new wet stack are similar to the requirements
for a dry stack. Any new electrical equipment used in a wet stackconversion
project needs to be properly grounded.
4.3.13.2 Obstruction Lighting
The bid specifications include furnishing and installing obstruction lighting,
including temporary obstruction lighting during construction. Obstruction
lighting must comply with the governing standards of the U.S. Department of
4-24
4-25
Foundation loads
Shearing-stress calculations
Column deflections
Thermal stresses
Reinforcing-steel requirements
4.4.2.3 Liner
Calculations are also typically submitted for the liner design. The design loads,
operating conditions, and material properties are specified. Load combinations
applicable to the liner material are addressed per the governing code
requirements.
Information that is typically submitted with the liner design is as follows:
Loads and stresses for wind, seismic, thermal, and dead-load conditions
Liner deflections
Joint/weld specifications
4-26
numbers and shelf life are recorded for future reference. For each batch of
material, resin gel-time tests are performed in accordance with ASTM D2471,
and resin-viscosity tests are performed in accordance with ASTM D2393.
4.4.3.4 Alloy-Material Certifications
Material-certification reports are usually furnished for all alloy materials and for
the alloy-welding materials. Certification reports should verify that the chemical
composition of the material is in conformance with the ASTM or specification
requirements.
4.4.3.5 Concrete-Material Certifications/Mix Design.
Mill certificates may be submitted for the cement being utilized to confirm its
conformance with ASTM C150 and to ensure that the correct type is being
furnished. Certifications are also required for the coarse and fine aggregate
(ASTM C33) and for any admixtures (ASTM C494) that will be utilized.
Concrete-mix design proportions and compressive-strength-test results
submittals are used to substantiate the proposed mix design.
4.4.4 Construction Procedures
The Contractor is responsible for his construction means, methods, techniques,
sequences, or procedures. The Engineer or Owner usually do not dictate and
approve the Contractor's construction practice. Submittal of construction
procedures are only required to confirm that the work can be performed within
schedule and meet the intent of the Contract Documents.
4.4.4.1 Foundation Construction
Chimney foundations are large-mass concrete placements that require substantial
planning and coordination. Before placement, the Contractor, Engineer, and
Owner review and coordinate the following issues:
Concrete-delivery schedule
Adequacy of formwork
seams is more critical than the vertical weld seams with regard to minimizing
liquid re-entrainment.
For all welding processes, the Contractor prepares and submits for approval the
welding-procedure specification (WPS) and the procedure-qualification test
results (PQR). For carbon-steel welding, the welding-procedure specification
includes the nonmandatory information included in Appendix E of AWS D1.1,
in addition to the mandatory information listed in Appendix IV, Table IV-1 of
AWS D1.1. For alloy welding, the welding procedures typically address all
essential and nonessential variables of Section IX of the ASME Boiler and
Pressure Vessel Code. Only approved welding procedures are typically used.
4.4.4.5 Coating Procedures
Proper preparation and application of a protective coating is important to ensure
its corrosion resistance and longevity. The following submittals may be required
to ensure that the coating is properly installed:
Surface-preparation procedures
Curing procedures
4-30
Piles should not exceed a variation from the vertical of more than 1/4 in
(0.006 m) per foot of pile length.
The center of the pile head should not vary from plan location at the cutoff
by more than 3 inches.
4.4.5.2 Foundation
Inspection of the chimney foundation is required before and during concrete
placement.
4.4.5.2.1 Embedded Items
Placement and positioning of all embedded items should be verified. Examples of
items that are typically embedded in a chimney foundation are:
Anchor bolts
Electrical conduits
Grounding
Drains
used, the connectors are capped off to provide protection from the wet concrete.
Mud or debris is removed from the rebar before placement of the concrete.
4.4.5.2.3 Concrete Placement
Inspection during concrete placement is performed to ensure that the concrete is
being properly placed. Inspection also confirms that the concrete is vibrated after
its placement. Concrete should be placed in lift heights that permit the concrete
to be vibrated into the lower section. The Contractor should select a placement
sequence so that a minimal amount of concrete surface area is exposed at a given
time. This reduces the opportunity for the concrete to develop its initial set and
maintains a working surface that is easy to vibrate.
4.4.5.2.4 Concrete Testing
Concrete testing is usually performed by a qualified laboratory-testing company.
Before placement, the concrete is tested for temperature, slump, and air content.
Concrete cylinders are taken to perform compression-strength tests.
4.4.5.3 Concrete Column
Inspection of the concrete column is performed during construction to confirm
that the chimney is being constructed in accordance with the design drawings.
4.4.5.3.1 Dimensional Tolerances
Dimensional tolerances for the concrete column are established by ACI 307.
Vertical alignment of the columns center point is taken by plumbing down to a
reference point on the chimney foundation. The column's diameter is taped, and
measurements are taken for out-of-roundness tolerances.
4.4.5.3.2 Location of Openings
The orientation, elevation, and size of openings in the column should be
confirmed. The breeching opening should be accurately located in order to avoid
misalignment problems with the absorber outlet-duct tie-in. Because the
breeching is typically installed before the outlet ductwork, a survey crew should
position the breeching from an established benchmark that can also be used by
the absorber contractor.
4.4.5.3.3 Reinforcing-Bar Placement
Vertical reinforcing bars should be counted and bar sizes checked for each
concrete-lift height to confirm that the bar quantities match the design
requirements. Horizontal bars are checked for size and location. Splice lengths
and concrete cover should also be checked.
4-32
Projections on the inside face of the liner should not exceed 1/8 in (0.003 m).
The liner's vertical axis should neither be off the theoretical axis by more
than 0.1 percent of its height nor by 1 in (0.03 m), whichever is greater. Also,
the center of the liner should not vary by more than 1 in (0.03 m) in 10 ft
(3 m).
Neither the diameter nor the radius of the liner should differ by more than
2% from those specified.
4-33
4-34
Resin/glass ratio for interior and exterior layers per ASTM D2584
Relative humidity in the working area where the lining system is being
installed should not be greater than 90%.
Temperature of the substrate, ambient air temperature in the work area, and
curing temperature should be between 50F (10C) and 90F (32C).
4-36
4.4.5.7.6 Visual
Visual inspection of the lining system should include the following items:
No air voids should exist between the adhesive-coated block and the
adhesive-coated substrate.
on a 6-ft (1.8-m) square grid pattern. Each measurement should consist of the
average of three random readings taken near the center.
4.4.5.8.5 Temperature and Humidity
Inspection should be performed to ensure that the coating system is being stored
and installed under the proper temperature and humidity conditions. During the
installation of the coating, the Contractor should measure and record air
temperature, substrate temperature, moisture dew point, and relative humidity.
Inspection guidelines for these issues are:
Relative humidity in the work area should not be greater than 90%.
Surface temperature must be at least 5F (3C) above the dew point of the air
in the work area.
Level the fan to the foundation using shim stocks; grout and anchor.
4.4.5.11 Electrical
Completion of the electrical and grounding system is typically one of the last
items to be completed on a new chimney installation. Before demobilization of
4-38
4-39
or recessed away from the main surface. These edges could act as sites for liquid
re-entrainment back into the gas flow. This is also true for the membrane
between the bricks, which must be struck flush with the liner surface. No issues
related to block misalignment have been reported, indicating the existence of
good quality control during the installation process. Very few issues have been
reported when borosilicate blocks have been used in new plants designed from
the start to use the system. Some retrofit applications have had SLD events
related to specific unit-geometry issues, liquid-collection-system drainage issues,
and of course, higher than recommended liner-gas velocities.
Based on discussions with plant operators, most wet stack liquid-collection
systems appear to be operating well. A common theme linking plants without
problems is that they perform liquid-collection-system maintenance on a regular
basisincluding inspections of the absorber mist eliminators, liquid-collection
gutters and baffles, and ring collectors and drains for buildup, pluggage, and
debris. They also clean and flush the system drains. Photographs should be taken
before any cleaning has been performed to document the location and extent of
any buildup within the system. A documented history of the liquid-collectionsystem performance will be very useful should an SLD issue occur.
5.1.2 International Experience with Wet Stacks
The United States was an early adopter of wet stack operation and, for the most
part, wet stack operation has only recently been accepted for use in other areas of
the world. These areas include the EU, countries hoping to join the EU, and
countries in which the plants are being funded by international agencies. As
experienced in the United States, the conversion of older, previously dry, stacks
to wet operation has resulted in a number of units experiencing SLD, primarily
because of unfavorably high liner-gas velocities. Because of this experience, most
new wet-scrubber installations also include a new properly designed wet stack.
These units are using the latest absorber and wet stack technologies, and if
operated according to prescribed procedures, they appear to be operating well.
SLD is not a regulated emission in the United States. As part of the Wet Stack
Design Guide revision process, efforts were made to determine if SLD was
regulated in other parts of the world. To the best of our knowledge, it is not.
However, as in the United States, it is considered a nuisance, and many
countries/regions recommend the installation of a liquid-collection system to
minimize the potential for SLD.
5.1.1.1 European Standard EN 13084
The European Standard for free-standing chimney design, EN 13084 Parts 1-8,
makes no mention of the need for liquid collection.
5.1.1.2 VGB PowerTech
VGB PowerTech Instruction Sheet VGB-M 643 Ue, Chimneys for Operation
without Flue Gas Reheating after FGD, discusses the formation of a liquid film
5-3
on the liner wall and the need for an effective liquid-collection system, although
no specific design details or guidance are provided. As in the EPRI/CICIND
Revised Wet Stack Design Guide, recommendations for the maximum flue gas
velocities inside the liner as a function of the liner material are presented. These
velocities are similar to those recommended in this guide for FRP, alloys, and
coated surfaces; they are higher for acid-resistant brick and slightly lower for
borosilicate block. The need to minimize horizontal discontinuities, as well as to
control plume downwash, are also discussed in this document.
5.1.1.3 China
In 2006 and 2007, more than 100 GWe power-generation capacity was installed
in China. In 2008 through 2010, an additional 130 GWe of capacity was
installed. Many of these plants incorporated FGD systems, and the dominant
technology is wet scrubbers. Most of these units have wet stacks; however there is
no evidence that these plants have incorporated liquid-collection systems.
Discussions with the China Electric Council have revealed that at present there
are no regulations in China controlling SLD from their plants. It is known that
SLD is an issue at many plants in China, and at present there is significant
interestboth at the governmental and the plant levelin the introduction of
liquid-collection technology to these plants.
5.1.1.4 India
No standards or guidelines related to wet stack design and/or operation in India
could be found.
5.2 Before Unit Startup
Before a wet stack is brought on line, a number of reviews and inspections should
be performed to ensure that the liquid-collection and drainage system operates as
expected, with minimal potential for SLD when the unit comes on-line.
5.2.1 Review of Fabrication/Installation Drawings
The first inspection that should be performed is a review of the liquid-collection
system installation drawings. The liquid-collection system design provided by the
laboratory/flow-model vendor is typically limited to providing the size, shape,
and location of the required collection gutters, ring collectors, flow diverters, flow
controls, and drains. These recommendations are then provided to the stack
vendor or architecture engineering (AE) firm, who develops the actual
fabrication and field-installation drawings. These drawings should be reviewed by
the liquid-collection system designer to ensure that their recommendations have
been properly interpreted and that any modifications required for ease of
fabrication or installation do not adversely impact the efficacy of the liquidcollection system.
5-4
What is the velocity of the flue gas in the stack when SLD occurs?
Have there been any recent problems with the absorber chemistry?
Have the mist eliminators been inspected since the SLD issue started, and if
so, was any plugging or fouling observed?
Have any droplet-size measurements been made in the stack, and if so, what
are the results?
When was the last time the liquid-collection system was inspected? Were
there any issues such as plugging or fouling?
5-6
This information will provide a basis for evaluation of the SLD issue. Hopefully,
it will provide the investigators with sufficient information to begin their
assessment of the cause of the issue and its elimination.
Specific actions can also be taken. The first should be to measure the liquid-flow
rate exiting the stack through the liquid-collection system drains. If these flow
rates are significantly lower than the baseline or calculated values, it is likely that
the drain system is plugged. An inspection of the liquid-collection system and
drains should be made at the earliest opportunity. The entire drainage path
should be inspected, but particular attention should be paid to areas where the
collected liquid is constricted. Such areas include drain ports; locations where
liquid enters a drainage pathway, such as found on internally drained ring
collectors; and (on units with side-entry breaches) the drains located on either
end of the breach-top gutter. Past experience has shown that debris tends to
collect in these areas. At some units experiencing unexpected SLD, these areas
have been found to be plugged with plastic sheeting; pieces of the plastic film
used as a separating media during FRP liner fabrication that was not fully
removed prior to startup; and sometimes even dead birds. These locations are the
first places that should be inspected if SLD is encountered.
During this inspection, the mist eliminators and the mist-eliminator wash system
should also be inspected for pluggage and damage. Plugged mist eliminators can
result in higher than design velocity values through the unit, which could lead to
liquid breakthrough and increased levels of droplets re-entrained in the flue gas
flow. Some droplets could also be re-entrained directly from the edges of the
plugged areas of the mist eliminator. Some of these droplets will not be collected
in the absorber-outlet ducting or liner and will pass through the unit, exiting
from the top of the stack.
5.4.1 Preliminary Stack Droplet Testing
The location of the droplets within the liner can often be related to the source of
their generation. Approximate droplet-size measurements and their spatial
distribution within a few feet of the liner wall can be estimated using a simple
droplet-impact probe made from a 1012-ft long 2 x 4 board painted flat black,
the smoother the surface the better (Figure 5-1).
The probe should be oriented with the test side facing horizontally, then rapidly
inserted through a test port between 2 and 8 ft (0.6 and 2.5 m). The probe is
then quickly rotated 90 to place the test face directly into the upward-rising gas
flow, and it is left in this orientation for 34 seconds. The probe is then rotated
back to its original orientation and rapidly removed from the liner (Figure 5-2).
5-7
~2
~24
12
Impact
Probe
Stack Linier
Figure 5-1
Droplet Probe
Insertion
Orientation
Test
Orientation
Withdrawal
Orientation
Test Surface
Gas Flow
Figure 5-2
Droplet-Probe Orientation
Droplet spots will be observed on the surface of the probe (Figure 5-3). The
probe exposure time may need to be increased or decreased, depending on the
number of droplets captured. The objective is to get enough droplets so that the
droplet spots do not overlap and that enough spots are obtained to have some
statistical significance. Tests should be repeated multiple times. The diameter of
the spots should be measured and their distribution on the probe noted. Large
droplets occurring in the area adjacent to the liner surface indicate that the issue
could be related to droplets being re-entrained from the liner surface. Large
droplets located farther from the liner wall are indicative of droplet carryover
5-8
Figure 5-3
Typical Droplet-Probe Test Results
The use of this probe technique is not intended to replace more sophisticated
evaluation methods. However it is a quick, cost-effective way to get a better
picture of what is going on within the liner during the initial stages of evaluating
the sources leading to SLD.
An additional test that can provide valuable insight into a stack experiencing
undesirable levels of SLD is to collect droplet samples downwind of the stack at
ground level. This test can be performed by placing metal collection pans, such as
a cookie sheet or a square piece of plywood covered in aluminum foil, on the
ground in the area where the droplets are hitting the ground. This approach can
be used to establish the number and size of droplets collected as a function of
distance to stack and unit load.
5-9
Section 6: References
1.
2.
Wet Stacks Design Guide. EPRI, Palo Alto, CA: 1996. TR-107099.
3.
FGD Mist Eliminator System Design and Specification Guide. EPRI, Palo
Alto, CA: 1993. GS-6984.
4.
Guidelines for the Fluid Dynamic Design of Power Plant Ducts. EPRI, Palo
Alto, CA: 1998. TR-109380.
5.
6.
7.
8.
9.
VGB PowerTech Service GmbH, Chimneys for Operation without Flue Gas
Reheating After FGD. Instruction Sheet VGB-M 643-Ue. Essen, Germany,
February 2007.
10. Bernhardt Hertlein, ed., American Society of Civil Engineers, Chimney and
Stack Inspection Guidelines. Reston, VA, 2003.
11. Guidelines for FGD Materials Selection and Corrosion Protection. EPRI, Palo
Alto, CA: TR-100680, Vols. 1 and 2.
12. Acid Deposition on Ductwork. EPRI, Palo Alto, CA: 1983. CS-3240.
6-1
13. American Society for Testing and Materials, Standard Specification for
Design and Fabrication of Flue Gas Desulfurization System Components for
Protective Lining Application. (ASTM D4618-92), Philadelphia, PA, 2010.
14. T. S. Clark. Chimneys Subjected to Acid Gases, Industrial and
Engineering Chemistry. Vol.15, No.3, March 1923, pp. 227230.
15. R. Mongia, A. Reza, et al. Effect of Exhaust Stack Geometry on the
Amount of Liquid Condensate During Plant Start-Up, Proceedings of the
95th Annual Conference of Air and Waste Management Association. Baltimore,
MD (2002).
16. H. S. Rosenberg. Wet Stacks: Friend or Foe?, Power Engineering.
November 1998, pp.7681.
17. American Society for Testing and Materials, Standard Specification for
Inspection of Linings in Operating Flue Gas Desulfurization Systems. (ASTM
D4619-96), Philadelphia, PA, 2004.
18. American Society for Testing and Materials, Standard Guide for Design and
Construction of Brick Liners for Industrial Chimneys. (ASTM C1298-95),
Philadelphia, PA, 2007.
19. N. J. Gardner and I. Owen, The Behavior of Liquid Films and Drops in
Relation to Liquid/Gas Separators, Proc. Institute Mechanical Engineers.
Vol. 211, Part E, 1997, pp. 5359.
20. D. S. Miller, Internal Flow Systems. 2nd Edition, Gulf Publishing, Houston,
TX 1990.
21. Leaning Brick Stack Liners. EPRI, Palo Alto, CA: 1989. GS-6520.
6-2
Appendix A: Glossary
A.1 Definitions
absolute humidity - the weight (or mass) of water vapor in a gas water-vapor
mixture per unit volume of space occupied.
absorber - general term for those gas/liquid contacting devices designed primarily
for the removal of SOx pollutants, i.e., scrubber.
absorption - the process by which gas molecules are transferred to a liquid phase
during scrubbing.
air cubic feet per minute (acfm) - a gas-flow rate expressed with respect to
operating conditions (temperature and pressure).
ambient - pertaining to the conditions (pressure, air quality, temperature, etc.) of
the surrounding environment of a plant or scrubbing system.
annual outage - a scheduled period of time (generally four to six weeks) set aside
by the utility once per year to shut down the boiler and/or FGD system for
inspection and maintenance.
annulus - the space between a chimney liner and a chimney shell.
base load - a generating station that is normally operated to take all or part of the
normal load of a system and that, consequently, operates at a constant output.
breeching - the section of ductwork in an FGD system between the absorberoutlet duct and the stack.
British thermal unit (Btu) - the amount of heat required to raise the temperature
of one pound of water 1F, averaged from 32212F.
bypass gas - flue gas that bypasses a scrubber for the purpose of raising wet flue
gas temperatures above the saturation temperature.
bypass reheat - a system that increases the temperature of the saturated flue gas
leaving an FGD system above dew point by ducting a slipstream of particlecleaned flue gas from the ESP exit duct past the FGD system to the absorberoutlet duct or directly to the stack.
A-1
capacity factor - the ratio of the average load on a boiler for the period of time
considered to the capacity rating of the boiler (actual kWh produced/theoretical
kWh produced x 100).
carryover - entrained solids, slurry droplets, and/or liquid droplets that leave with
the flue gas stream exiting a particular stage of a scrubber or absorber.
chimney - a vertical structure at a power plant that encloses one or more flues
which exhaust combustion gases. A chimney is typically constructed out of
reinforced concrete.
chloride - a compound of chlorine with another element or radical.
choke - the constricted upper section of a steel stack or chimney liner.
cladding - a thin sheet of corrosion-resistant alloy (usually nickel alloy or
titanium) that is either resistance-welded or roll-bonded to carbon-steel plate.
closed water loop - the water loop of an FGD system is closed when the fresh
makeup water added exactly equals the evaporative water loss leaving via the
stack and the water chemically or physically bonded to the sludge product.
column - a reinforced-concrete chimney shell. Usually encloses one or more
chimney liners. The purpose of the column is to protect the liner from weather
and to act as a wind shield.
computational fluid dynamics (CFD) - the use of finite elementanalysis
methods to simulate heat transfer, temperature profiles, and fluid and particle
movement in boilers and air-pollution- control equipment.
corrosion - the deterioration of a metallic material by electrochemical attack.
damper - a plate or set of plates or louvers in a duct used to stop or regulate gas
flow.
dew point - the temperature at which vapor contained in saturated flue gas begins
to condense.
efficiency - ratio of the amount of a pollutant removed to the total amount
introduced to the normal operation.
entrainment - the suspension of solids, liquid droplets, or mist in a gas stream.
erosion - the action or process of wearing away of a material by physical means
(friction).
electrostatic precipitator (ESP) - an air-pollution device used to remove particles
from an exhaust stream by initially charging them with electrodes and then
collecting them on oppositely charged plates.
A-2
A-3
A-4
pressure drop - the difference in force per unit area between two points in a fluid
stream as a result of resistive losses in the stream.
rain hood - the component at the top of a stack that covers the annular space.
reheat - the process of increasing the flue gas temperature downstream of a wet
scrubber. Reheat can be supplied by in-line indirect hot air, direct combustion, or
by partial bypass of unscrubbed flue gas.
reheater - device used to raise the temperature of the scrubbed gas stream to
prevent condensation and corrosion of downstream equipment, avoid visible
plume, and/or enhance plume rise and dispersion.
relative humidity (also relative saturation) - the ratio of the weight (or mass) of
water vapor present in a unit volume of gas to the maximum possible weight (or
mass) of water vapor in unit volume of the same gas at the same temperature and
pressure. The term "saturation" refers to any gas-vapor combination, whereas
"humidity" specifically refers to an air-water system.
removal efficiency:
-
SO2 - the actual percentage of SO2 removed from the flue gas by the
FGD system.
retrofit - the FGD unit will be/was added to an existing boiler not specifically
designed to accommodate an FGD system.
saturated - the situation in which a gas or liquid is filled to capacity with a certain
substance. No additional amount of the same substance can be added under the
given conditions.
saturation temperature - the temperature to which flue gas drops when it is
saturated by scrubbing in a wet FGD system.
scale - deposits of slurry solids (calcium sulfite or calcium sulfate) that adhere to
the surfaces of FGD equipment, particularly absorber/scrubber internals and
mist-eliminator surfaces.
scheduled outage - a planned period of time periodically set aside for inspection
and maintenance of the boiler and/or FGD system.
A-5
scrubber - a device that promotes the removal of pollutant particles and/or gases
from exhaust streams of combustion or industrial processes by the injection of an
aqueous solution or slurry into the gas stream, i.e., absorber.
sludge - the material containing high concentrations of precipitated reaction
byproducts and solid matter collected and/or formed by the FGD process
(composed primarily of calcium-based reaction byproducts, excess scrubbing
reagent, fly ash, and scrubber liquor).
slurry - a watery mixture of insoluble matter (usually lime or limestone).
SOx - a symbol meaning oxides of sulfur (e.g., SO2 and SO3).
stack - a vertical structure at a power plant that encloses one or more flues which
exhaust combustion gases.
stack flue - the inner duct or liner in a stack through which the flue gas is
conveyed.
stack-exit velocity - the exiting velocity of the flue gas out the top of the stack.
stack-liner velocity - area average gas velocity inside the liner.
stack liquid discharge (SLD) - liquid that is discharged from a stack and falls to
the ground prior to evaporating.
standard conditions - a set of physical constants for the comparison of different
gas volume flow rates (68F, 29.92 in Hg, barometric pressure).
standard cubic feet per minute (scfm) - units of gas-flow rate at standard
conditions.
steel stack - a vertical structure at a power plant that exhausts combustion gases.
The primary supporting shell is made of steel.
superficial gas velocity - the area average flue gas velocity through a mist
eliminator or other component of an FGD system.
temperature, dry bulb (DB) - the temperature of a gas or mixture of gases
indicated by a thermometer after correction for radiation.
temperature, wet bulb (WB) - a measure of the moisture content of air (gas)
indicated by a wet bulb psychrometer.
total controlled capacity (TCC) - the gross rating (MW) of a unit brought into
compliance with FGD, regardless of the percent of flue gas treated at the facility.
turning vanes (i.e., vanes) - devices used in ductwork or chimney liners to control
gas-flow direction. Usually fabricated from flat or curved plates.
A-6
unit rating:
-
net - gross unit rating less the energy required to operate ancillary station
equipment, inclusive of emission-control systems.
video droplet analyzer (VDA) - uses on-the-fly video image analysis to detect
and measure the diameters of all in-focus droplets that are entirely within the
view of the camera in each video frame
wallpaper - thin sheets of corrosion-resistant alloy material welded to new or
existing carbon-steel plate.
water loop - all aqueous mass flows from inlet (e.g., seal water, quench water,
scrubber liquor) to outlet of an FGD system (e.g., evaporation via stack, pond
evaporation, waste disposal).
wet stack - a chimney, stack, or flue that exhausts saturated, completely scrubbed
flue gas. Wet stacks are located downstream from a wet FGD system. Wet stack
operation does not utilize any flue gas reheat system or partial bypass. Wet stacks
are equipped with corrosion-resistant liners for handling the wet, acidic flue gas
exiting the FGD system.
zero discharge - a pollution regulation requiring that no effluent waste stream be
discharged back into the environment, with the exception of evaporation via
ponds and stacks (e.g., pond runoff or direct piping of spent slurry or waste into
nearby waterways or tributaries would be prohibited).
A-7
A.2
To Obtain
Multiply
By
Atmospheres
Feet of water @ 4C
0.0295
Atmospheres
Inches of mercury @ 0C
0.03314
Atmospheres
0.068
Cubic meters
Cubic feet
0.02831685
Cubic meters
Gallons
0.00378541
Cubic meters
per second
0.0004719
Inches of mercury
@ 0C
2.036
Kilograms per
square meter
4.882
Kilowatts
0.01757
Millimeter
Inches
25.4
Kilograms
Pounds
0.4565924
Kilogram per
cubic meter
16.01846
Liters
Gallons
3.785
Meters
Feet
0.3048
0.3048
Pascal
6,894.757
Pascal
47.88
Square meters
Square feet
0.09290304
Temperature Conversions:
o
F = 1.8 x C + 32
C = ( F - 32)/1.8
A-8
A-9
Access to and use of EPRI Intellectual Property is granted with the spe-
compliance with all applicable U.S. and foreign export laws and regu-
under applicable U.S. and foreign export laws and regulations. In the
event you are uncertain whether you or your company may lawfully
mine whether this access is lawful. Although EPRI may make available
in Palo Alto, Calif.; Charlotte, N.C.; Knoxville, Tenn.; and Lenox, Mass.
export classification for specific EPRI Intellectual Property, you and your
company acknowledge that this assessment is solely for informational
purposes and not for reliance purposes. You and your company acknowledge that it is still the obligation of you and your company to make
your own assessment of the applicable U.S. export classification and
ensure compliance accordingly. You and your company understand and
acknowledge your obligations to make a prompt report to EPRI and the
appropriate authorities regarding any access to or use of EPRI Intellectual Property hereunder that may be in violation of applicable U.S. or
foreign export laws or regulations.
Programs:
Integrated Environmental Controls
2012 Electric Power Research Institute (EPRI), Inc. All rights reserved. Electric Power
Research Institute, EPRI, and TOGETHER...SHAPING THE FUTURE OF ELECTRICITY are
registered service marks of the Electric Power Research Institute, Inc.
1026742