EPA 340/1-87-002
Municipal Waste Combustion Systems
Operation and Maintenance Study
Prepared by
Allen Consulting & Engineering
317 Howland Avenue
Gary NC 27513
Prepared for
EPA Project Officer: Louis Paley
EPA Work Assignment Manager: Pamela Saunders
U.S. ENVIRONMENTAL PROTECTION AGENCY
Stationary Source Compliance Division
Office of Air Quality Planning and Standards
Washington DC 20460
June 1987
-------
DISCLAIMER
This report was prepared by Allen Consulting and Engineering for the Stationary
Source Compliance Division of the U.S. Environmental Protection Agency in fulfillment
of Purchase Order No. 7W-6610-NASX. The contents of this report are reproduced
herein as received by the contractor. The opinions, findings, and conclusions expressed
are those of the authors and not necessarily those of the U.S. Environmental Protection
Agency. The mention of product names does not constitute endorsement by the U.S.
Environmental Protection Agency.
11
-------
ACKNOWLEDGEMENT
Appreciation is expressed to the staff of the Bureau of Sanitation, City of
Stamford, and particularly to Mr. Marvin E. Serra, Superintendent for providing the
months of operating data and many hours of assistance in helping us compile the field
study presented in Chapter 3. In addition, appreciation is expressed for the assistance
provided by Peer Consultants, Inc. in preparing the statistical analyses performed on the
data generated by the field study.
in
-------
-------
TABLE OF CONTENTS
LIST OF FIGURES. „ .v
LIST OF TABLES v
CHAPTER 1 INTRODUCTION 1
1.1 Objectives and Scope of this Project 1
1.2 Overview of MWC Sites Visited 1
1.3 Report Organization 3
CHAPTER 2 GENERAL OPERATING CONCERNS 4
2.1 Introduction ... 4
2.2 Fuel Preparation and Handling , 4
2.3 Furnace Operation 7
2.4 Ash Handling 15
2.5 Air Pollution Control Equipment 18
2.6 Fans, Ducts, and Stacks .. 24
2.7 Startup and Shutdown 29
2.8 Equipment Inspection Frequency 32
2.9 Recordkeeping 34
2.10 References 40
CHAPTER 3 SITE VISITS AND OPERATING VARIABLES STUDY 42
3.1 Introduction „. 42
3.2 Site Visits - Case Histories 42
3.3 Operating Variables vs. Opacity Study 46
3.4 References 59
APPENDIX A AUDIT: OPACITY METER -STAMFORD MUNICIPAL INCINERATOR
IV
-------
LIST OF FIGURES
Number
2-1
2-2
2-3
3-1
3-2
3-3
Page
Continuous Monitor Quarterly Report • 35
Incinerator Shift Operations Record 37
Combustion Condition Observation Form 38
Schematic of Stamford Incinerator 48
Conceptual Relationship Between
Underfire/Overfire/Temperature -51
Sample Data Sheet for Stamford Incinerator «55
LIST OF TABLES
Number
1-1
2-1
2-2
2-3
2-4
3-1
Page
Types of MWC's Reviewed 2
Typical Composition of Municipal Solid Waste 5
Electrostatic Precipitator Log 22
Typical Fan Inspection and Preventive
Maintenance Schedule •
Example Incinerator Equipment
Inspection Frequencies
ESP Parameters
.26
.33
.53
v
-------
CHAPTER 1
INTRODUCTION
Municipal Waste Combustors (MWC's) have traditionally had problems in
complying with participate matter and visible emission (VE) limitations and are
often the subject of citizen's complaints. Due to stricter waste disposal regulations
and decreasing availability of space for landfills, an increase in incineration is
anticipated. It is believed that, in many cases, particulate matter and visible
emissions of MWC's can be reduced by improving the incinerator and control
equipment operation and maintenance (O&M) practices.
1.1 OBJECTIVES AND SCOPE OF THIS PROJECT
The objectives of this project were twofold. The first objective was to
determine the significant operation and maintenance considerations that bear
directly on the day in/day out air pollution compliance status of MWC's. Rather
than performing theoretical analyses of furnace and air pollution control device
performance, achievement of this objective was sought by visiting several MWC
facilities and discussing the issues on-site with operators and management
personnel.
The second objective was to determine the effect of operating variables on
visible emissions (opacity) at one facility. Any correlation found would provide
insight into the relationship between operating variables and opacity.
1.2 OVERVIEW OF MWC SITES VISITED
An attempt was made to select several facilities which, when taken
together, would represent a reasonable cross-section of the types of equipment in
service at the time of the visits (1985 and 1986). To keep travel costs to a
minimum for the EPA headquarters and contractor's staff, site visits were limited
to the eastern United States. Seven sites were visited with a total of 19 furnaces
in operation. Table 1-1 summarizes the types of MWC's reviewed. Recent permit
applications for construction of MWC's indicate that the fraction of operating
MWC's that burn refuse derived fuel (RDF) and/or utilize boilers to recover heat
will substantially increase in the future. Processing municipal waste into RDF is an
operation that is often performed off-site and then trucked to the MWC facility.
RDF processing facilities are outside the scope of this report.
-------
TABLE 1-1
TYPES OF MWC's REVIEWED
Type
No. of of
Furnaces Furnace
4
2
3
1
4
2
1
2
Rotary
Combustor
Rectangular
Rectangular
Rectangular
Rotary
Combustor
Rectangular
Rectangular
Traveling
Grate
Continuous
or
Batch
Feed
Continuous
Batch
Continuous
Continuous
Continuous
Batch
Continuous
Continuous
Fuela
MB
MB
MB
MB
MB
MB
MB
RDF
Boiler
None
None
None
None
None
None
None
Water
Wall
Air Throughput
Pollution per
Control Furnace
Device (tons/day)
ESP
ESP
ESP
ESP
Venturi
Scrubbers
Baffle Wall
Scrubber
ESP
ESP
300
100
300
240
250
215
200
300
aMB = Mass burn (unprocessed municipal waste)
RDF = Refuse Derived Fuel
-------
1.3 REPORT ORGANIZATION
Chapter 2 of this report contains a discussion of general O&M problems,
concerns, and procedures described by MWC operators during the site visits. This
discussion covers the whole range of operations at MWC's from furnace charging
and combustion control to ash handling and air pollution control equipment.
Chapter 3 contains two major sections. The first section describes eight
brief case histories of specific O&M related air pollution problems and the solutions
implemented by the MWC operators interviewed for this study. The second section
describes the results of the study on the effect of operating variables on opacity.
This discussion includes a description of the facility, the operating variables
studied, and the analytical techniques employed. The chapter ends with some
conclusions that can be drawn of the study's results. FinaUy, Appendix A contains a
report of the audit performed on the opacity meter at the facility.
-------
-------
CHAPTER 2
GENERAL OPERATING CONCERNS
2.1 INTRODUCTION
This chapter discusses MWC operating concerns as described by MWC
operators and their supervisors at seven facilities visited over the course of this
study. The operating concerns discussed herein include operating practices,
preventive maintenance requirements, and recordkeeping necessary to operate
MWC's reliably and in compliance with air pollution emission limitations. The
operating concerns are discussed on a unit operation or equipment basis starting
with fuel preparation and ending with the stack.
2.2 FUEL PREPARATION AND HANDLING
Mass burn MWC's are those facilities that charge municipal waste to the
furnace on an as-received basis. The alternative to mass burning municipal waste
is to process the waste into refuse-derived fuel (RDF). The extent of processing
waste into RDF varies from plant to plant, but usually it entails any one or
combination of the following processes: magnetic separation, shredding, screening,
and air classifying. Most RDF facilities produce a fuel that can pass through a 10
mesh screen (0.034 in. square), or coarser fuels that can pass through screens up to
six inches square. The 10 mesh screen produces a fuel that can be fired in a
tangentially fired furnace that is similar to units that burn pulverized coal. The
coarser fuel is fired in various types of stokers, most commonly those with a
traveling grate.
The objective of processing municipal waste into RDF is to create a
uniform fuel by improving the fuel's important combustion parameters, e.g. size
distribution and combustible content. Indeed, one of the major challenges faced by
operators of mass burn furnaces is the wide swings in fuel quality that are inherent
in unprocessed municipal waste. This subsection discusses operating procedures
employed by MWC operators to deal with the difficulties imposed by burning
unprocessed refuse in mass-burn furnaces.
2.2.1 Trash Composition
Table 2-1 lists a "typical" composition of municipal solid waste. Trash
composition will vary among facilities depending on factors such as region of the
4
-------
country, season of the year, and the extent a given facility accepts non-hazardous
commercial and industrial waste. Design values for heat and moisture content of
the waste will vary among facilities. Factors that effect the design valves include
the nature of the waste anticipated for the specific facility and whether (and to
what extent) the waste will be processed into RDF.
TABLE 2-1
TYPICAL COMPOSITION OF
MUNICIPAL SOLID WASTE
Component
Weight Percent
Paper, Other than Newspaper
Newspaper
Garbage
Yard Wastes
Glass, Ceramic, Stone
Metal
Cardboard
Wood
Textiles
Plastic Film
Leather, Rubber, Molded Plastics
Total
Source: Reference 1
25
14
12
10
10
8
7
7
3
2
2,
100%
The combustion characteristics of any given truckload of waste may be
substantially different than the design value or general composition guideline. For
example, municipal trucks collecting household garbage and trash will deliver waste
that has substantially different characteristics than the trucks for a specific
commercial operation that deliver nothing but shredded paper or process (non-
hazardous) waste from a plastic film or corrugated box manufacturer. Regardless
of the source of waste, mass burn MWC operators are often challenged by extended
periods of wet weather, when much of the material they are attempting to burn has
been soaked by several days of rain.
-------
2.2.2 Pit Operation
The fuel preparation and handling procedures discussed here apply directly
to those continuous and batch furnaces that utilize gravity feed chutes that are
loaded with an overhead crane from a sub-elevation trash pit. A properly sized pit
and one overhead crane can serve up to four furnaces with a total capacity of over
1000 tons of refuse per day. Pit sizing is important to the efficient movement of
raw trash through the facility. Considerations include not only the furnace
capacity, but also the anticipated operating schedules for the furnace and the trash
receiving operation. Undersized pits require that municipal waste be stored for
significant periods of time in a pile either inside the receiving room if space is
available, or outside. In either case, odor problems are likely to ensue.
Municipal waste arrives at the incineration facility via truck. The trucks
are weighed prior to tipping their trash into the pits. Weigh records provide
.reasonably accurate data regarding the amount of trash processed over a week or
more and such data is useful in establishing long term trends about the production
of municipal waste for the area serviced by the incinerator facility. The scale data
can also be used to estimate hourly or daily furnace throughputs, but these
estimates are crude at best since they do not account for changes in the amount of
trash in the pit.
The crane operator can make or break a successful MWC operation. The
crane operator should constantly mix wet and dry refuse in order to produce
"uniform" refuse; although he or she should keep some "wet" refuse in one area of
the pit and some "dry" refuse in another area for emergency use when required by
the furnace operator. Should it become necessary to go from feeding normal refuse
to feeding either wet or dry refuse, the crane operator should do it gradually and
give the furnace operator time to change the combustion conditions. Constant
observation of incoming loads by the crane operator will give that person an idea of
the condition of the storage pit at all times and the location of the wet and dry
refuse.
The feed hopper serves as a funnel to direct refuse into the feed chute. To
prevent bridging in many hopper designs, the crane operator must take care not to
fill the hopper more than approximately one foot above the entrance to the feed
chute.
When not feeding the incinerator units, the crane operator should utilize all
his time mixing the wet and dry refuse in the pit. Large non-combustibles such as
refrigerators, drive trains, etc. should be carefully removed from the pit. Some
-------
2.3
facilities set such materials aside for a metal recovery contractor, others ship such
materials to a landfill. If allowed to pass through the furnace, large non-
combustibles may damage the grate system and/or the bottom ash handling
equipment.
Similarly, certain excessively combustible items such as tires and large
foam rubber items should be set aside for disposal by other means, as should any
excessively large items such as chairs and couches.
The crane operator should also attempt to clean the bottom of the pit
occasionally and should not just take the refuse from the top of the pit. The best
method is to remove everything down to the bottom of the pit in one area, then
another, and so forth. If possible this program should be used weekly. The purpose
of periodically cleaning the pit to the bottom is twofold. First, the longer trash
stays in the pit, the larger its potential for creating odor problems. Second, decay
of municipal trash produces methane gas that becomes trapped in the waste pile.
Thus, periodic turnover of the pile minimizes fire and explosion hazards.
The trash in the feed chute provides the seal between the combustion
chamber and the pit room for a continuous feed furnace. If this seal is broken, the
combustion chamber will draw large volumes of cold ambient air, thus upsetting the
combustion process and creating significant air pollution until the seal can be re-
established and until the furnace can return to normal operating temperatures.
Therefore, the crane operator should never "get behind" a continuous furnace.
For a batch furnace, when the furnace operator determines that it is
necessary to charge the furnace, a guillotine gate located at the bottom of the
charging chute is briefly opened to allow the trash to faU to the furnace. It is
important that this procedure be coordinated with the crane operator to ensure
that trash is available in the chute and that the gate is open for the minimum time
necessary so that as little cold, ambient air as possible is drawn into the furnace.
Many facilities maintain a backup crane so that the facility can operate
when the primary crane is down for maintenance. In a one crane operation, even as
little as one hour of crane downtime may require initiation of shutdown procedures.
FURNACE OPERATION
The objective of MWC operation is to reduce the volume of municipal
waste that eventually will go to landfill to the maximum extent practicable. This
objective is achieved only when the MWC is operated in a manner that provides
good combustion. Good combustion requires maximizing the classic three Ts of
-------
combustion; time, temperature, and turbulence. As discussed below, the MWC
operator has meaningful control of only temperature; residence time and
turbulence having been maximized through the design process.
This section discusses operating practices commonly utilized to provide
good combustion at MWC facilities. Common upset conditions are described, as are
operating procedures to minimize the impact of such upsets. The reader should
keep in mind that providing good combustion minimizes air pollution by minimizing
the emission of unburned particles and maximizing the combustion of volatile
organic compounds (VOC) that evolve from the municipal waste in the combustion
chamber.
2.3.1 Combustion Air
Oxygen is necessary to complete the combustion reaction. Oxygen is
introduced to the furnace as preheated or unpreheated ambient air. In order to
completely burn combustibles in the furnace, it is necessary to provide combustion
air in excess of the theoretical quantity required by stoichiometry to completely
burn (i.e., oxidize) the waste. A minimum of 50 percent excess air is necessary to
provide adequate turbulance of the volatile gases generated from heating the solid
waste. In general, waterwall furnaces require excess air in the range of 50 to 100
percent excess air, while refractory furnaces require 150 to 200 percent excess air.
An insufficient quantity of combustion air cannot provide adequate oxygen
and turbulence for good combustion. Such a condition produces increased emissions
of carbon monoxide, unburned organic gases and soot. Curiously, too much
combustion air also produces increases in soot, CO and organic emissions. The
reason is that the increased mass flow of cool combustion air tends to lower the
furnace temperature. In addition, this increased mass flow reduces the residence
time of combustible gases in the furnace and has a tendency to entrain significant
quantities of ash out of the furnace.
All the furnaces observed in this study provide combustion air through
underfire systems and most also utilize overfire systems. A brief description of
these systems follow.
Underfire Air - Underfire air is provided by a fan with airflow controlled by
a fan inlet damper. The air is manifolded to several areas under the grates, where
it is directed upward so that the air passes through the grates, refuse and ash. In
some MWC's the operator can control the amount of air independently to various
sections of the furnace.
-------
Underfire air systems are designed to provide sufficient velocity (energy)
to the underfire air so that the air penetrates the fuel bed. This provides intimate
contact of combustion air with the solid, combustible material in the fuel, thus
enhancing the combustion process. However, in addition to lowering the furnace
temperature, excessive flow of underfire air has the velocity necessary to carry ash
and unburned solid material out the furnace in sufficient quantities to create
visible emissions and particle fall out problems.
Some incinerator systems utilize an air preheater to heat the underfire air
prior to introduction into the furnace. A preheater is an air-to-air exchanger that
utilizes the waste heat in the furnace exhaust gases. Preheated underfire air
increases the flame temperature, thus improving the combustion and energy
efficiencies of the system.
Overfire Air - Overfire air is introduced above the fuel bed via high
velocity nozzles through the furnace sidewalls, perpendicular to the flow of fuel.
Optimum combustion usually occurs when the overfire air is somewhere in the
range of 40 to 60 percent of the total combustion air introduced into the system.
However, operating experience (i.e., trial and error) sometimes dictates that the
fraction of overfire necessary for optimum combustion will fall outside of this
range in a given furnace. As for design purposes, overfire air is any air introduced
that is not underfire air. Therefore, overfire air also includes air introduced for
sidewall cooling (if any) and air in-leakage from sight ports, furnace cracks, etc.
The primary reason for introducing air over the fuel bed is to provide
turbulance to the volatile gases that have evolved from the solid refuse fuel bed,
thus improving the combustion of those gases. A secondary benefit results to the
extent that lower underfire air velocities reduce the entrainment of fly ash and
other solid material in the exhaust stream.
2.3.2 Temperature
Maximizing time, temperature and turbulence is essential for good
combustion. Of these parameters, temperature is the only one that can be directly
measured and also the only one over which the incinerator operator has reasonably
direct control.
The temperature not only varies with position within the combustion
chamber, but also varies with time at a given position. Most systems monitor the
temperature (using a shielded thermocouple) near the roof at the exit of the
combustion chamber. Operating temperatures at this location for most MWC's are
-------
in the range of 1200° to 1850° F. MWC's with waterwalls tend to operate in the
lower end of this range, whereas refractory wall MWC's operate in the upper end of
the range.
Furnace geometry may indicate a need to monitor temperature in several
locations. For example, in rotary kiln designs, the combustion chamber divides
itself into a drying chamber, an ignition chamber, the rotary kiln, and finally a
mixing chamber where the furnace gases undergo final combustion. Temperatures
in such a case are often monitored at the exit of the ignition chamber and mixing
chamber. The common operating ranges for these locations are 1600° to 1800° F
and 1800° to 2000° F, respectively.
Furnace temperature is controlled by adjusting the combustion air supply.
The details as to how this is done regarding the split between underfire and overfire
air will vary among facilities. In general, high temperatures are reduced by
increasing the air supply, i.e., quenching the furnace with cool air. Conversely, low
temperatures are increased by reducing the combustion air.
Maintaining an adequate furnace temperature is absolutely essential to
attaining complete combustion and minimizing air pollutants. "Adequate furnace
temperature", in effect, is the maximum achievable temperature that is consistent
with the long term mechanical reliability of the system.
2.3.3 Furnace Pressure and Draft
MWC's are always operated at a modest negative static pressure,
approximately 0.2 inches of water. There are several reasons for this. First,
operation under negative pressure eliminates emission of toxic combustion products
through site ports, furnace cracks and so forth into the workplace and general
environment. All gas leakage related directly to the furnace leaks into, rather than
out of, the furnace. Second, positive pressure exacerbates a phenomenon called
torching. Torching occurs when the flame propagates from the fuel bed downward
through the grates. Localized conditions can cause occasional torching in any
furnace, but operating under negative pressure minimizes the phenomenon and the
resulting problems with heat damage to the grate system and the underfire air
system.
In order to remove the gaseous combustion products from the combustion
chamber, it is necessary to establish the proper furnace draft, which is done by
balancing the airflows among the over and underfire air systems and the induced
draft (ID) fan. The ID fan is usually located donwstream of the air pollution control
10
-------
system. (Some MWC systems utilizing wet scrubbers have located the ID fan
upstream of the scrubbers to minimize fan corrosion.) If the furnace generates
more combustion gases than the fan can handle the furnace pressure goes positive
and puffing of combustion products from the furnace enclosure can usually be
observed. This condition often is caused by an excessive burn rate which can
originate from an excessive feed rate or feed that has an unusually high
combustible content.
2.3.4 Feed Rate
In continuous, gravity feed systems, refuse is piled up in the feed chute.
Refuse in the chute provides an air-tight seal while the furnace is in operation.
Refuse is drawn from the chute by the action of the grates, either rocking or
reciprocating grates. Thus, the feed rate is governed by the grate speed. Traveling
grate systems do not employ a gravity feed arrangement. With traveling grates,
the feed chute dumps on to an auger that moves the refuse to the traveling grate.
The refuse is transferred from the auger to the traveling grate by an air blast
distributor. The feed rate depends on the speed of the auger. Augers and traveling
grates are used in systems utilizing RDF.
Improper feed rates create a variety of problems. If the feed rate is too
low, insufficient energy is available to maintain an adequate furnace temperature.
If the feed rate is too high the furnace temperature can reach unacceptably high
levels. In either case, the bottom ash will contain a significant quantity of
unburned material.
The feed rate also affects the depth of the fuel bed on the grates. A fuel
bed that is too thick will not allow proper penetration of underfire air. On the
other hand, a fuel bed that is too thin allows underfire air to penetrate the fuel bed
with velocities that are sufficient to entrain significant quantities of ash (and
unburned material) thus creating opacity and downwind fallout problems. Excessive
furnace temperatures also cause premature wear and structural problems (buckling,
warping, etc.) of the furnace refractory and grate system components, as well as
slagging. Slagging is discussed in Section 2.3.7.
2.3.5 Dealing With Fuel Problems
This section discusses procedures for controlling the combustion process
when especially low or high combustible fuel, respectively, is introduced to the
MWC. Despite the efforts of the crane operator, it becomes necessary from time
to time to burn fuel that has less than optimal characteristics.
11
-------
All the facilities visited during this study utilize automatic combustion
control (trim) systems that employ the furnace exit temperature as the operating
parameter that controls the combustion process. A signal from the temperature
sensor is processed by a central controller that operates the dampers on the
overfire and underfire air fans. The details of how such a system operates will vary
from plant to plant, but basicaUy, each system responds to high temperature
excursions by increasing the volume of combustion air in the furnace. The
additional air cools the furnace. Conversely, the response to low furnace
temperature excursions is to decrease the volume of combustion air in the
furnace. The MWC visited that utilizes heat recovery also automatically controls
the feed rate based on steam demand. There is a trend for newer MWC's with heat
recovery to employ oxygen trim systems similar to those utilized in coal-fired
utility boilers.
MWC's trim systems will compensate for moderate swings in fuel quality.
However, if especiaUy poor quality fuel is fed to the furnace, it may be necessary
for the furnace operator to operate the furnace manually until the combustion
process has stabilized. Whether manual operation is required or not, the
adjustments to poor fuel quality depicted below are appropriate responses to the
fuel conditions described.
2.3.5.1 Wet Refuse and Low Combustibles
Burning wet refuse or low combustibles requires running a shallow fuel bed
on the ignition grates and maintaining high underfire air pressures. This is often
accompanied by a slight reduction in the feed rate in the case of wet fuel. The
wet, or low combustible content, of the refuse will effect a reduction in furnace
temperature; thus, increasing the underfire air may appear contradictory to the
discussion in Section 2.3.2 where it was stated that a decrease in furnace
temperature should be countered with a decrease in combustion air. A couple of
points need to be noted.
First, especially with wet fuel, the refuse simply will not burn until it is
dry. Running a shallow fuel bed with high underfire air ensures good contact and
mixing of the fuel with the air, enhancing the drying process. Second, cutting back
on the feed rate reduces the amount of heat consumed per unit time utilized to
evaporate the moisture. If the low combustible fuel is not wet, the forge effect of
the higher underfire air rate assists in oxidizing such refuse. However, it is not
usually desirable to cut back on the feed rate as long as the dry refuse is reasonably
porous.
12
-------
Automated combustion control systems that control both the underfire and
overfire air will throttle back on the overfire air as the underfire air is increased,
thus minimizing the cooling effect of the increased underfire air.
2.3.5.2 Dry Refuse and High Combustibles
To burn dry, high combustible content refuse, it is necessary to maintain a
high fuel bed on the grates and set a low underfire airflow while maintaining a high
overfire airflow and reducing the feed rate slightly. Highly combustible refuse
creates excess turbulence in the furnace. If this turbulance (which is normally
good) is exacerbated by a high underfire airflow and a shallow fuel bed, significant
quantities of bottom ash and unburned waste will be carried out of the furnace and
ultimately emitted to the atmosphere. The potential for excess furnace
temperatures is eliminated by maintaining high overfire airflow and reducing the
feed rate slightly.
2.3.5.3 Undesirable Waste
The best solution for dealing with combustion problems resulting from
undesirable waste is to prevent such waste from ever reaching the furnace. The
kind of waste that is usually amenable to this solution is certain kinds of highly
combustible waste such as rubber tires, mattresses and certain types of commercial
and industrial waste.
Some municipalities do not allow neighborhood trash collection crews to
pick-up tires. Such prohibitions often include items that can cause mechanical
damage to the MWC such as large appliances, axles, engine blocks, etc. Other
municipalities pick-up all trash and leave it up to the crane operator to pick out
undesirable refuse.
2.3.6 Grates
The purpose of the grate system is to transport the refuse and residual ash
through the combustion chamber. The grates are designed to allow the passage of
underfire air from underneath the grates up through the grates and fuel bed.
Certain grate designs (rocking grates and reciprocating grates) provide a significant
amount of agitation of the refuse as it tumbles from one grate section to the
next. Such agitation helps promote complete combustion. Traveling grates provide
less agitation and are used primarily for RDF. In the United States, most stoker
fired MWC's utilize reciprocating, rocking, or traveling grates.
13
-------
Regardless of which of the three grate types are employed, the grates can
be qualitatively divided into three sections. The first section is referred to as the
drying or charging grate. Most of the drying in this section is a result of radiant
heat from the furnace arches and walls. The second section is referred to as the
ignition grates. The refuse is ignited in this section and a significant amount
(approximately 50 percent) of the burning takes place in this section. The third and
final section is referred to as the burnout or combustion grates. Combustion is
completed in this section. The remaining residue (bottom ash and non-
combustibles) falls off the end of the burnout grate, is quenched, and conveyed out
for final disposal. (Ash handling is discussed in Section 2.4.) If a rotary kiln is
employed, it normally is preceded by drying and ignition grates of the reciprocating
type. The kiln serves the same purpose as the burnout grates.
From an operating point of view, the main concern about grates is
preventing mechanical and thermal damage to the grate system. Mechanical
problems include damage caused by allowing excessively heavy items to get into
the furnace. Engine blocks, automobile transmissions and even refrigerators have
been known to find their way into MWC's. As such heavy items tumble through the
furnace they can bend, break, or otherwise severely damage the grates, sometimes
requiring the plant to shutdown. Smaller material such as chains, wire, small pipes,
etc. fall through the grate system and if not cleaned out periodically can jam the
grates, requiring a shutdown.
Excess heat can warp or buckle the grates and the support system
underneath the grates. This problem can be caused by high furnace temperatures
or, probably more commonly, by excessive torching.
2.3.7 Slagging
Excessive furnace temperatures can cause slagging. Slagging occurs when
temperatures in the furnace (or some portion thereof) are sufficient to fuse the
ash. This causes the ash to run and to evaporate, but it will solidify or condense at
the first opportunity. Different types of refuse produce different types of slag that
have different melting and running temperatures. Slagging creates havoc with the
grate systems by filling underfire air slots, thus creating poor air distribution. Slag
can also interfere with the movement of the grates themselves. If the grates
become completely frozen due to solidified slag, an immediate shutdown is
required.
14
-------
Slag formation on the furnace walls win cause premature deterioration of
the refractory. Many MWC operators monitor for slag formation by regularly
observing the appearance of the furnace refractory in the lower portion of the
furnace. Fused ash sticking to the furnace refractory gives the walls a wet or
glassy appearance.
2.4 ASH HANDLING
Many MWC operators report that the ash handling equipment creates more
O & M problems on a day in - day out basis than any other system in the facility.
As discussed below, some of these O & M problems result in air pollution
compliance difficulties.
2.4.1 Ash and Residue Composition
If the residue produced by MWCs contained only bottom ash and fly ash the
problem of ash handling would be much simpler than it is. Unfortunately, however,
the residue contains a substantial fraction of non-combustible material, including
glass fragments, metal material (cans, wire, axles, springs, sheet metal, motors,
etc.), clinkers, rocks, and unburned organic material. This collection of variously
sized material is abrasive and difficult to handle. Ash can be acidic or basic,
depending on its source.
2.4.2 Ash Collection Points
The number and location of ash collection points will vary somewhat among
MWC's, even if the facilities are of rather similar design. The following five
collection points are common to many facilities in the United States.
1. Grates - Siftings are the ash and residue that fall through and are collected
underneath the grates.
2. Quench Pit - The quench pit collects the bottom ash that faUs off the end
of the combustion grate. The quench pit contains water so that the ash is
cooled for further handling.
3. Mixing Chamber - The mixing chamber holds the volatile furnace gases at
the proper temperature long enough to allow complete combustion to
occur. Heavy flyash falls out of the gas stream at this point. In some
designs, this fallout is collected in the quench pit, i.e., the mixing chamber
is directly above the quench pit. If the mixing chamber is slightly
downstream of the quench pit, it is necessary to collect the ash at the
bottom of the mixing chamber and evacuate the ash through a rotary air
lock valve.
15
-------
4. Cooling Tower - Cooling towers lower the temperature of the furnace
exhaust gases via evaporative cooling prior to the gases entry into the air
pollution device. The cooling towers often function as ash collection points
because the tower vessel also acts as a gravity collector. To a limited
extent, they also perform as prescrubbers. Sometimes cooling towers are
referred to as quench towers or chambers, and if they precede a wet
scrubber they may be referred to as a presaturator, if saturation of the gas
stream is their function. Most MWC's with heat recovery do not utilize
cooling towers.
5. Air Pollution Control Equipment - Electrostatic precipitators (ESP's),
baghouses and scrubbers^ of course, collect the fly ash produced by MWC's.
Rectangular batch furnaces typically have ash collection and storage
hoppers directly underneath the grates. Periodically, ash and residue are removed,
quenched, and accumulated in a residue hopper. Discharge from the residue hopper
normally is done through a watertight gate valve into trucks or other containers
and shipped to a disposal area, usually a landfill.
2.4.3 Ash Conveying Systems and O & M Considerations
The means of moving the ash and residue produced at MWC facilities can
be broadly classified as either manual or continuous, with some plants using a
combination of these techniques. This section focuses on ash handling problems
that will adversely effect atmospheric emissions.
Quench Pit - Continuous ash removal operations employ a variety of means
to move ash through the plant. The bottom ash moving off the grates is usually
dumped into a quench pit. The water in the pit actually provides an airtight seal
for the furnace. If the water level in the pit is not maintained at the proper level,
the seal is broken and it becomes impossible to maintain proper furnace pressure
and draft. A drag conveyor, submerged at its lower end in the quench pit, pulls ash
from the pit to a holding hopper or in some cases directly into trucks for transport
to a landfill. Since the drag conveyor is inclined, it allows some of the entrained
quench water to drain back to the quench pit. Often, these drag conveyors become
hung up with the odd sized material that accumulates in the quench pit. If these
hang-ups occur above the water line they can usually (but not always) be cleared in
a short time without interrupting operation of the furnace. If they occur below the
waterline, it is usually necessary to drain the quench pit to gain access to the
problem source. Draining the pit will break the airseal and result in higher
emissions until the problem is corrected.
16
-------
Grates - Siftings are removed from under the grates by vibrating or screw
type conveyors, or by water sluices. The collected material is sent to the quench
pit. Most facilities will observe the siftings conveyors area on an hourly basis to
ensure that excessive amounts of material are not catching on beams and the grate
drive system or jamming the conveyors. Sometimes such problems can be resolved
by opening inspection doors and knocking the material off with long poles. This
draws a lot of cold air into the furnace and thus usually creates an opacity
problem. If solving the problem requires shutting down the furnace, even greater
air emissions result.
ESP Hoppers - Fly ash from ESP or baghouse hoppers may be dumped
through an airlock valve directly to a sluicing system, or it may be moved by a
screw conveyor system prior to being dropped into a quench pit. Condensation in
the hopper area can cause bridging of the ash, and can even cause the ash to create
clinkers that adhere to the hopper walls. When the ash backs up into the ESP's
electrical field it causes excessive sparking and eventually the ESP will short out.
Most commonly, bridging and clinkering are caused by air in-leakage (ESP's are
operated under negative pressure) in or near the hopper area. Cool ambient air
leaking into the hopper condenses water vapor and acid gases. Another cause is
excessive cooling at the cooling tower, or cooling tower spray nozzles that have
deteriorated to the extent that they produce large droplets that cannot completely
evaporate prior to entering the ESP. Yet another potential cause is leaking boiler
tubes in MWC's equipped with heat recovery.
Sometimes the subsequent bridging can be rodded out while the MWC is in
operation. If this fails, it is necessary to shutdown the facilitly to gain internal
access to the hoppers.
Some operators report that bridging is sometimes caused by sticky material
such as plastic film reaching the hoppers without being burned. This may be
indicative of excessive airflows, especially underfire air, in the furnace.
Prevention of hopper problems entails aggressive maintenance procedures.
The ESP's, hoppers and air lock valves need to be inspected for leaks at every
opportunity. The air lock valves should be externally inspected and lubricated
weekly when the MWC is in service.
From a design perspective, ash bridging and flow problems can be
minimized through the installation of vibrators on the hoppers. Condensation can
be minimized by insulating the hoppers or installing hopper heaters.
17
-------
Mixing Chamber - The rotary air lock valve should be inspected for leakage
internally at each opportunity and inspected and lubricated weekly when the MWC
is in operation.
Summary - Basically, the important components of the ash handling system
should be monitored frequently to ensure that residue is not backing up and thus
creating a potential shutdown condition. These observations include checking
water levels, flow rates, pumps, motors for rotary valves, and free movement of
conveyors and residue.
The details of specific systems will vary from the description above. For
example, facilities that recycle process water may send sluice residue directly to a
clarifier, rather than to the quench pit. Also, recycle systems sometimes control
pH of the process water by use of chemical additives. Regardless of the details of
the methods and equipment employed at a given facility, clean and efficient
operation of the MWC is tied directly to the reliable operation of the ash handling
system.
2.5 AIR POLLUTION CONTROL EQUIPMENT
This section discusses the control devices encountered during the site
visits: electrostatic precipitators (ESP's) and high energy wet scrubbers. The
discussion is limited to the specific concerns brought to our attention by MWC
operators. Regardless of the type of control device, a common concern expressed
was corrosion control in the control equipment and ductwork. (Broader, in-depth
discussions of control equipment design, O & M, and inspection, recordkeeping, and
performance evaluation are discussed in the literature. For examples, refer to
References 2, 3, 4, 5, 6, 7 and 8.)
2.5.1 Electrostatic Precipitators
ESP's are the most common particulate matter control device applied to
MWC's today. This section will briefly discuss corrosion problems and useful
monitoring and recordkeeping techniques for ESP's.
18
-------
2.5.1.1 Corrosion: Causes and Prevention
A successful corrosion prevention program must first recognize that, as a
result of the combustion reaction, MWC exhaust gases contain a significant water
content. Also, water is added to the exhaust gases at many facilities either by
design or accident. The acid content (mostly HC1) of the furnace exhaust gases
further complicate the problem.
The effect of excess moisture on the ESP's ash hoppers was discussed in
Section 2.4. The corrosion problem discussed here manifests itself primarily as ESP
plate corrision. Within a very few years holes of several feet in area can develop.
This requires municipal authorities to decide whether to engage in costly repairs or
limp along with reduced plate area. Indeed, some MWC operators have expressed
the opinion that a major ESP rebuild is required within seven years for units in
MWC service. As operating experience has accumulated over the years, however,
it becomes apparent that corrosion in MWC ESP's can be controlled to levels
similar to those in other industries where rebuilding of the unit is not necessary for
at least 15 to 20 years.
Cooling Tower - The purpose of the cooling tower is to lower the
temperature of the exhaust gases to a point where the ESP will not suffer thermal
damage. (Plate warping and buckling are the most common heat related problem.
In more severe temperature excursions, the ESP's structural members can also be
damaged.) The cooling is accomplished by evaporating a fine mist introduced into
the gas stream by water sprays in the cooling tower. If too much water is
introduced into the gas stream, all of the water cannot evaporate before entering
the ESP, thus creating a potential corrosion problem.
If the spray nozzles in the cooling tower have eroded substantially, they
may produce a course spray (large droplets) that cannot completely evaporate prior
to entering the ESP. Again, moisture in the ESP chamber establishes a corrosion
hazard.
• A thermocouple is placed near the entrance to an ESP. The signal is sent
to the control panel so that the MWC operator can monitor the inlet temperature.
The signal is also sent to a controller that governs the quantity of water supplied to
the cooling tower. It is imperative that the operator monitor this system so that
he can take appropriate action if the performance of the cooling tower deteriorates
for any reason. The cooling tower spray nozzles and the thermocouple should be
inspected at every scheduled outage for MWC's that operate for at least a month
continuously.
19
-------
Air In-Leakage - Since ESP's operate under negative pressure any leakage
through cracks or pinholes on the enclosure will allow air to move from the outside
to the inside. Cool air coming into the unit is likely to cause condensation of
moisture and acid gases. If plate corrosion is confined to a limited area, air in-
leakage should be suspected since airflow in the ESP is not turbulent; air streams
from holes and cracks tend to maintain their integrity, do not diffuse very much,
and thus tend to attack rather small areas.
The insidious characteristic of air in-leakage, however, is that it tends to
feed on itself. If an air stream from a leak flows in a manner that causes it to
contact the shell of the ESP, the ensuing condensation will cause corrosion at that
location, ultimately creating another route for leakage. Another common source of
leaks are deteriorating seals on access doors.
The ESP should be thoroughly inspected for leaks at least on a quarterly
basis if the operating schedule allows it. It is important to correct any problems
found during the inspection before the system is brought back on-line. Otherwise,
the situation will likely be much more severe at the next shutdown.
As unlikely as it may seem, it is advisable to inspect a new ESP prior to the
initial startup to ensure that the field erection crews have welded all appropriate
seams in the ESP shell and hopper areas. MWC and other ESP operators have
learned this technique through the school of hard knocks.
Shutdown - Every time an MWC is shutdown, the hot, water vapor laden
exhaust gases in the ductwork and ESP must cool to below the dewpoint. In- effect,
the ducts and air pollution control device take an acid bath every time the system
is shutdown. Obviously, the corrosion this process inevitably causes is more severe
if shutdowns are frequent. Facilities that regularly shutdown on a weekly schedule
probably have the most severe problem.
Two steps can be taken to alleviate condensation during shutdowns. The
first is to minimize the number of unscheduled shutdowns by implementing an
aggressive O & M plan, particularly on those parts of the system whose failure can
cause a shutdown situation, e.g., the ash handling equipment, the grate drives, and
the air pollution control system itself.
Second, if the normal operating schedule calls for a weekly shutdown, it is
then feasible to install an ESP heater. An ESP heater is basically an oil or gas
burner installed in the duct on the inlet side of the ESP. The heater is fired up for
short term (weekend) shutdowns, preventing the exhaust gases in the ESP from
condensing during the shutdown period.
20
-------
Performance Monitoring - Virtually all MWC operators who employ ESP's
do some performance monitoring. The commonly monitored electrical data include
primary current, primary voltage, secondary current, secondary voltage, and spark
rate. Many older units do not monitor secondary voltage and some of the newer
systems do not include a spark rate guage. Regardless of the details as to precisely
what data is available for a given unit, most operators record the electrical data on
an hourly or bi-hourly schedule. Table 2-2 is a typical log for recording the
secondary electrical data.
The ESP electrical data is a powerful tool for evaluating the overall
performance of the unit and in diagnosing performance problems. The reader is
referred to the EPA publication in Reference 2 for an excellent discussion of the
utility of these parameters.
Another useful parameter for evaluating ESP performance is opacity.
Federal New Source Performance Standards (NSPS) do not require continuous
monitoring of opacity at MWC's. Indeed, some MWC's normally operate with
condensation occurring downstream of the control device, making continuous
opacity monitoring unfeasible. Yet many other units do not create such steam
plumes. Thus, many states have incorporated continuous opacity monitoring permit
requirements on a case by case basis. Where opacity data is available it can be
very useful in evaluating performance and diagnosing problems. Again, Reference
2 provides an excellent discussion on how to utilize opacity data as a diagnostic
tool.
2.5.2 Wet Scrubbers
High efficiency scrubbers for MWC's include variable and fixed throat
venturi scrubbers and flooded disc type venturi scrubbers. One of the major
problems with wet scrubber performance as applied to MWC's is control of the
condensible, partially oxidized organic matter which forms when incinerator
operating temperatures are too low. The organic material condenses while passing
through the scrubber, creating an aerosol with a particle size too small for
effective impaction in the scrubber. Few scrubbers have been able to achieve the
NSPS emission limits because of this problem (Reference 3).
Corrosion is a major concern in wet scrubber operations, as it is with
ESP's. For scrubbers, the solution is centered around control of the scrubber's
liquor quality. The paragraphs below briefly discuss the impact of liquor quality on
system performance.
21
-------
TABLE 2-2
ELECTROSTATIC PRECIPITATOR LOG
Incinerator No.
Date
Field No. 1
Field No. 2
Secondary
Voltage
(KV)
Secondary
Current
(Ma)
Spark Secondary
Rate Voltage
(per min.) (KV)
Secondary
Current
(Ma)
Spark
Rate
(per min.)
1 am
2 am
3 am
4 am
5 am
6 am
7 am
8 am
9 am
10 am
11 am
Noon
1 pm
2 pm
3 pm
4 pm
5 pm
6 pm
7 pm
8 pm
9 pm
10 pm
11 pm
Midnight
22
-------
Liquor Quality, Corrosion, and Performance - From both a performance and
maintenance perspective, liquor quality is critical to the scrubber's reliable operation.
The critical parameters are pH, and suspended and dissolved solids. The rigor with which
a source operator monitors these parameters may not need to be as severe for a scrubber
utilizing a once through system, but most operations recycle a significant fraction of the
spent liquor; some recycle all of it with fresh water introduced only to make up for
evaporative losses. The concern for liquor quality applies to both the liquor supplied to
any cooling tower or presaturator and to liquor supplied to the scrubber itself.
In municipal incinerators, scrubber liquor tends to become acidic due to the
absorption of HCl'and organic acids in the liquor. Acidic conditions can quickly lead to
serious corrosion problems everyplace downstream of where the acid liquor is
introduced. Thus, the spray chamber, ducts, scrubber, entrainment separator, ID fan and
stack can suffer corrosion problems. In areas of the system that are under positive
pressure, (i.e., the scrubber is preceded by the ID fan) the holes and cracks from
corrosion in the system will allow untreated combustion gases to escape to the
environment. For those portions of the system under negative pressure, holes and cracks
in the system can substantially reduce the ID fan's ability to control draft in the
furnace. In addition, in-leakage of ambient air causes localized condensation of moisture
and acid gases, the resulting liquid exacerbating the corrosion problem. Thus, you have
something similar to a monster feeding on itself.
Liquor pH and water make-up rate can be monitored and controlled
automatically at the clarifier just before the recycled liquor is reintrodueed to the
system. The pH of the liquor should be maintained above the levels at which carbon steel
is attacked. A pH of 6 or greater is usually satisfactory. Chemicals utilized to reduce
pH are usually sodium hydroxide, ammonia, or lime.
The dissolved and suspended solids content of recycled liquor varies from source
to source, but can be as high as 15% by weight. Since a major portion of the liquor
injected into the quench tower and/or presaturator evaporates, some of the solids
entering as suspended and dissolved solids in the liquor are released as particulate matter
(Reference 11). High solids content also is often responsible for excessive erosion of
pipes and pumps, and pluggage of spray nozzles.
Samples of the liquor should be analyzed periodically, even if this requires
sending the samples to an outside laboratory. High solids content in the liquor can be
deviated by several methods including more frequent blowdown of the scrubber liquid
system and a lower recycle rate.
23
-------
Proper control of other scrubber variables such as pressure drop and liquid-to-gas
ratio are also critical to maximizing scrubber performance. For a more detailed
discussion of these parameters, the reader is referred to Reference 3.
2.6 FANS, DUCTS, AND STACKS
As auxilliary systems the importance of fans, ducts and stacks to the
overall operation and performance of an MWC is easy to underestimate. This
section briefly describes some of the salient O & M concerns for these auxilliary
systems.
2.6.1 Fans
Operation and maintenance considerations for fans depend on several
factors. These factors include the type of service (and locations) the fan is in, fan
size, fan type, and whether the fan is a direct drive or belt driven unit. Regardless
of any general "rules of thumb" provided here or elsewhere, source operators should
follow the specific fan manufacturer's recommendations for the nature and
frequency of fan maintenance procedures as minimum requirements for reliable fan
performance.
The fans of principle concern at municipal incinerators are the combustion
air fans (underfire, overfire, and if applicable, sidewall cooling fans) and the
induced draft (ID) fan. Generally, the combustion air fans are constant speed
devices with airflow controlled by dampers. The ID fans may also be constant
speed units, but variable speed units are probably more common.
Most ID fans are located between the air pollution control system and the
stack. If the ID fan precedes the air pollution control equipment, problems related
to abrasion, erosion, corrosion, and build-up of solids on the fan will be more
severe. The only "solution" to this is frequent inspection and cleaning.
ID fans serving scrubber systems operate under substantially more severe
conditions than those serving dry systems. This is due to the humid conditions that
enhance the possibility of condensation of acids on the fan and fan housing. Even if
the ID fan preceeds the scrubber, the fan is usually placed after the cooling tower,
since the cost of a fan that can handle the substantially higher gas volumes and
temperatures that occur upstream of the cooling tower would be prohibitive.
Regardless of its location, the ID fan is usually equipped with water cooled
bearings and a vibration sensor that will sound an alarm in the plant and illuminate
emergency lights in the control room if vibration becomes excessive. In some
24
-------
systems, the sensor controls will automatically shutdown the fan and open an
emergency stack in preparation for an emergency shutdown. Aside from the fact
that the furnace cannot operate properly without the ID fan, everyone working
around industrial fans should be aware that vibration is a preliminary indication
that the fan might shake apart at anytime, possibly throwing fan parts considerable
distances at substantial risk to life and property.
Fan current for all fans is often monitored in the control room. At a
minimum, this information tells the incinerator operator whether fans are
operating or not and can provide rough estimates of airflows. These meters may
also include audible alarms or emergency lights that indicate no current. If the ID
fan loses current, many systems will, again, automatically begin shutdown
procedures. \
Table 2-3 illustrates a typical fan inspection and preventative maintenance
schedule. Operating experience or the manufacturer's recommendations may
dictate a more frequent inspection and maintenance than indicated here.
2.6.2 Duets
Air ducts are a passive part of the incinerator system. Nevertheless,
proper maintenance of the ducts is critical to the overall performance of the
system. Two problem areas are addressed in this section, leakage and deposition.
Leakage - Leaks in air ducts result from any of several causes, including
corrosion, abrasion/erosion, flex failure and failure of expansion joints. Leaks can
occur as holes, cracks, and fissures in the duct system. The causes and effects of
such leakage will vary somewhat, depending on whether they occur in ducts that
are under positive or negative pressure. :
Positive pressure ducts occur downstream of the ID fan. In most systems,
this is only a small fraction of the total duct system. A few incineration systems
however, locate the ID fan upstream of the air pollution control device, thus
creating substantial duct lengths between the fan and control device and continuing
on to the stack. In a positive pressure environment, leaks occur from the duct to
the ambient atmosphere. Between the fan and the control device, this means that
untreated furnace gases are exhausted directly to the atmosphere.
Air in-leakage occurs in negative pressure ducts. Negative pressure ducts
occur upstream of the ID fan. In most systems, the ID fan is the last unit operation
prior to the stack, so most of the ductwork, in fact, most of the entire system,
including the furnace, the air pollution control device and cooling towers, operate
under negative pressure.
25
-------
TABLE 2-3
TYPICAL FAN INSPECTION AND PREVENTATIVE MAINTENANCE SCHEDULE
INSPECTION/MAINTENANCE ITEM
DAILY
WEEKLY
MONTHLY
SEMI-
ANNUAL
Vibration Check
Oil Level
Oil Color
Oil Temperature
Lubricate
Bearings Check
Noise
Leaks
Cracks
Loose Fittings
Inspection (Internal)
(Clearances, Wear,
Pitting, Scaring)
Clean and Inspect Blades and
Internal Housing
Fan Belt
Noise Check
Belt Tension and Wear
X
X
X
X
X
X
X
X
X
X
XJ
X
Whenever the fan is out of service
LFans in scrubber service may require cleaning on a monthly or quarterly schedule.
26
-------
Air in-leakage creates several problems. The first is that it tends to feed
on itself. Leakage of cold ambient air into the ducts encourages localized
condensation of acidic vapors that tends to create more holes and cracks. As this
situation propagates, the ID fan attempts to pull more and more air. It becomes
difficult for the furnace operator to maintain the proper furnace pressure and
draft. I
Ducts should be inspected externally for leaks on a monthly basis, more
often if the facility has a history of leaky ductwork. Small leaks are difficult to
locate by an external inspection. Internal inspections performed during daylight
hours have the advantage of highlighting small holes and cracks due to contrast
between the dark duct interior and bright daylight. Internal inspections should be
performed on a quarterly schedule.
Particle Deposition - Dust build-up is an inevitable difficulty in many
horizontal duct sections upstream of the air pollution control device due to the
impractical economics of maintaining air velocities that are sufficient to carry the
very largest particles in the gas stream. Consequently, many designers provide
conveniently located clean out ports so that the build-up can be easily removed.
Excessive build-up in MWC systems result from oversized ductwork, or, more
commonly, excessive underfire air and/or furnace draft that tends to carry
significant quantities of unburned material out of the furnace and into the air
handling system.
The consequences of excessive dust build-up are several. First, by reducing
the cross-sectioned area of the duct, build-up increases air velocity. This
consequence alone can deteriorate the performance of the air pollution control
device. Second, by moving the entire air stream over to one portion of the duct,
maldistribution problems in ESP's are likely to occur, despite turning vanes and gas
distribution plates, which were never designed to correct for excessive dust build-
up. Finally, duct systems are not usually designed as load bearing structures. The
build-up of excessive quantities of material in air ducts can (and sometimes does)
lead to collapse of the duct support system or the duct itself. Aside from requiring
an immediate shutdown and rather expensive repair, collapse of ductwork is
potentiaEy dangerous to plant workers.
27
-------
2.6.3 Stacks
Stacks are another passive system from an operational procedures
perspective. Air pollution agencies are concerned about only a few specific aspects
of the initial design. First, of course, is the stack height. The function of stacks is
to lift incinerator effluents to a suitable elevation above ground level so that the
air pollutants remaining in the effluent can be adequately dispersed and diluted
prior to being returned to ground level. Second, if the stack is to be the sight of
stack sampling ports and/or pollutant monitoring instruments (transmissometers,
S(>2 gas analyzers, carbon dioxide monitors, etc.) then the design must provide for
safe access to the ports and instruments, including appropriate scaffolding.
Twenty or more years ago it was common for municipal incinerators to
operate without air pollution control equipment. In those days, MWC's operated by
natural draft, rather than with ID fans. The installation of air pollution control
equipment required cooling the incinerator effluent to provide temperatures in a
range that would not damage scrubbers, baghouses, or ESP's. A second important
aspect of the gas cooling was the reduction in gas volume that followed. The
cooler, slower gas stream now required an induced draft fan so that the incinerator
operator could maintain proper furnace draft and pressure. The complications this
created for stack design relates to the increased potential for condensation of acid
containing vapors on the stack liner. Most incinerator operators schedule a
thorough stack inspection on a yearly basis. The inspection addresses both the
structural integrity of the stack and chemical attack of the liner. Often these
inspections are conducted by outside firms who specialize in stack maintenance.
Observation of the physical stack can sometimes assist in identifying or
diagnosing problems elsewhere in the incinerator system. Stacks in scrubber
service, for example, sometimes develop a mud lip, that is, the stack exit becomes
covered with wet particulate matter. This is generally the result of a
malfunctioning entrainment separator.
Some operators have been known to cut "windows" in the stack near the
base. The purpose for doing this has never been clear. Aspirating ambient air into
the stack lowers the stack gas temperature, thus increasing problems related to
condensation of acids in the stack. In addition, the lower gas temperatures
decrease plume bouyancy, thus decreasing the effective stack height and resulting
dispersion at ground level. Finally, in most jurisdictions, dilution of the stack gases
in this manner is considered a circumvention of the opacity emissions limitation.
28
-------
2.7 STARTUP AND SHUTDOWN ;
Startup and shutdown are routine activities at MWC's that can cause
periods of excess emissions. The excess emission problem is much more severe for
startup than it is for shutdown. This section describes operating procedures that
can minimize excess emissions during startup and shutdown.
2.7.1 Startup
It is common for newer MWC's to use auxiliary burners as a means of
achieving a prescribed minimum furnace temperature before waste is charged to
the system. However, for MWC's which do not utilize auxiliary burners, the typical
startup is a procedure in which the refuse loading and the combustion air are
gradually increased, and during which time the furnace temperatures are
intentionally controlled below the levels required for optimum combustion. The
initial charging rate normally varies from one-third to one-half of the design rate
and is increased stepwise over the duration of the startup period.
During startup, particulate matter and opacity emissions can be extremely
high. This is due to the low furnace temperatures that exist for the first two to
four hours. The result is incomplete combustion and the generation of large
quantities of soot. The reason for the long startup interval is the necessity to
slowly heat the MWC refractory material and downstream components and thereby
prevent damage from rapid expansion. Rapid temperature increases can result in
spalling and premature failure of the refractory and similar heat stress problems
for the grate system. For units equipped with boilers, all of the pressure parts of
the boiler must be heated and gently expanded into their position at full
operation. Typically, the rate of temperature increase at the boiler outlet is
limited to 100°F increase per hour as a uniform gradation in temperature, i.e., a
25°F increase in boiler outlet temperature every 15 minutes. Since the boiler
outlet temperature at full load operation conditions is approximately 650°F, this
would result in a startup period of approximately 6 hours. Excessive opacity will be
generated for the first 2 to 4 hours. On the more sophisticated MWC systems, it is
possible to reduce the startup time somewhat by bypassing the economizer or ,air
preheater at the tail end of the boiler so that the flue gas temperature is increased
as rapidly as possible to the air pollution control equipment.
Electrostatic Precipitators - During the initial 4 hours of startup, the flue
gas products going to an ESP can be wet and sticky due to 16w flue gas
temperatures. If the ESP is placed in operation at the time the furnace is ignited,
29
-------
it can be difficult to remove the sticky particles from the plates. As the flue gas
temperature increases these particles can bake onto the plate. This situation can
reduce the efficiency of the ESP when the MWC is fully charged. This is why many
manufacturers of electrostatic precipitators do not recommend power be turned on
until the acid dew point temperature has been exceeded at the ESP outlet. If the
MWC is equipped with ESP heaters as described in Section 2.5, the heater can be
used to preheat the ESP prior to furnace charging, thus allowing the control device
to be energized much earlier in the startup process.
Economizers - In order to exceed dew point temperature in the control
equipment as quickly as possible, a bypass of the water around the economizer is
required so that the economizer will not reduce the flue gas temperature during
startup. Most economizers are equipped with bypass piping so that the feedwater
can be bypassed and not permitted to pass through the economizer. This can
decrease the time necessary to achieve an adequate flue gas temperature by as
much as 2 hours. The problem with this procedure occurs in putting the economizer
into service once the boiler has been brought up to pressure and placed on line. At
this time the economizer will generate steam when feedwater is introduced into it
and this steam must be vented out through the vent valve at its top until"the
steambound economizer can begin to circulate water to the boiler without being
injurious to the boiler. This requires a high degree of operating skill "on the part of
the plant personnel.
Air Heaters - Air heaters can also be bypassed. This requires a duct
between the discharge of the forced draft fan and the windbox area of the stoker.
This prevents the forced draft air from being preheated by the flue gases, therefore
allowing a more rapid increase of the flue gas temperature at the air polltition
control equipment.
Multiple Furnace Facilities - An effective way to eliminate opacity
problems during startup at multiple furnace facilities is to provide inter-
connections among the various furnaces and the air pollution control equipment.
With such a setup, the air pollution control equipment (ESP or baghouse) associated
with a boiler coming on line can be preheated to a point above the acid dew point
temperature by a slip stream(s) from the flue gas of other fully operating
furnaces. Thus, baghouses and ESPs can be brought on-line much earlier in the
startup process. Maintenance of damper seals is very important in an
interconnected setup so that the potential for air infiltration and out-leakage is
eliminated.
30
-------
Superheaters - During startup of a unit, the superheater section and
reheater sections could be flooded with feedwater if hydrostatic testing has been
conducted. If so, the refuse feed rate into the furnace must be held low enough to
allow the water that was in the superheater (or reheater section) to be converted to
steam and vented to the atmosphere. During the initial startup of the boiler,
thermocouples are normally attached to determine the time it takes to boil away
all of the water in each and every superheater or reheater element. The refuse
feed rate must be limited to ensure that the superheater and reheater elements are
heated and cleared slowly so that saturated steam may pass through them and not
cause overheating of superheater metal. This time/temperature curve is drawn up
by the boiler manufacturer for each and every boiler at initial startup. Time must
be allowed, especiaUy at the very low firing rates where excessive opacities can be
generated, to clear these units. After clearing the liquid water from the
superheater, heating the superheater metal is usually limited to 50°F per hour.
Thus, clearing liquid water and gentle heating of the superheater usually requirfes
about ten hours. After this ten hour period, the stack gas is hot enough to allow
energizing the ESP or bringing the baghouse on-line.
Short-term Shutdowns - The MWC startup time can also be greatly reduced
if the MWC is to be shut down for a short time over a weekend, by converting the
MWC and air pollution equipment into a "thermos bottle". This requires (1) the
complete stoppage of the forced draft fan air flowing through the furnace after the
MWC has been brought off line and steam is no longer being generated in the unit
and (2) not allowing any air leakage into the setting and not allowing natural stack
effect of the high temperature flue gases (i.e., gases at 300°F or 400°F should riot
be allowed to pass through the MWC to heat the tailend economizer, air heater, or
ID fan, and then pass out the stack). By making a "thermos bottle", the MWC will
minimize the loss of heat and fast startup (3 to 4 hours) can be accomplished.
When startup is initiated the following Monday morning, the internal MWC
temperature is already up to 250°F to 300°F instead of the ambient air
temperature of 70°F to 80°F.
Most operators do not make the effort to stop the natural draft through the
MWC, thereby cooling the furnace and increasing the length of time for system
startup. It is most desirable to prevent a MWC system from cooling down if the
unit is to be off line for a short period of one to three days. Such cooling is
undesirable because it takes more time to startup a cold furnace. Also, a furnace
that is allowed to cool down from operating temperature to ambient air
31
-------
temperatures (e.g. 70°-80°F) requires more maintenance than a unit that cools to
only 250°-300°F. The greater and more frequent the expansion and contraction of
the pressure parts and refractories, the more maintenance that will be required.
2.7.2 Shutdown
The shutdown of an MWC does not generally cause significantly increased
emissions. Charging is discontinued, but grate movement, airflow, and operation of
the air pollution control system continue as appropriate until all the refuse in the
furnace has been burned. Heat retained by the refractory and other interior parts
keeps temperatures at acceptable levels for the 1 to 2 hours required to completely
burn out the refuse after charging has been discontinued.
2.8 EQUIPMENT INSPECTION FREQUENCY
Clean, efficient MWC operation requires an aggressive preventive
maintenance program. Such a program includes getting out of the control room or
office to periodically inspect, evaluate and repair equipment.
Table 2-4 presents an example of a schedule for inspection of incinerator
equipment by the incinerator staff. It must be understood that, as an example, the
inspection frequency tabulated here should not be interpreted as the only
acceptable frequency to ensure reliable operation of the system. For example,
many of the hourly and daily inspection checks itemized here may be inverted at a
specific facility or may be performed on a "once per shift" basis. Similarly, in
some facilities the monthly and semi-annual inspections may be performed on a
quarterly basis, especially if the facility routinely shuts down on a quarterly basis
and the nature of the inspection requires internal access to the furnace, ductwork,
control equipment, etc.
The inspection schedule implies a certain amount of routine maintenance
must be performed as needed. For example, if an hourly check of the conveyors
indicates a jam, obviously it must be corrected immediately. The fact that
different inspection frequencies are recommended for specific pieces of equipment
implies that inspections of different levels of detail are required. For example, the
cooling tower exit temperature is the only means to measure the performance of
the cooling water sprays on an hourly basis. However, many plants will'check the
spray pumps for excessive bearing noise on a per shift basis, and lubricate the
bearings on a weekly basis. Stand-by pumps are often run through their transfer
cycle on a weekly basis. The spray nozzles are usually inspected (and cleaned or
replaced as needed) during every scheduled shutdown.
32
-------
TABLE 2-4
EXAMPLE INCINERATOR
EQUIPMENT INSPECTION FREQUENCIES
Equipment
Cranes
Feed Hoppers
Grates
Hourly Daily
X
X
X
Weekly
Lube &
Inspect
X
X
X
Semi-
Monthly Annual
X X
X
X
Bi-
Annual
X
Air Box
Conveyors
Refractory
Fans
Rotary Kiln
Diversion Gate
Dump Trucks
Rotary Valve
Air Locks
Cooling Water
Sprays
Cooling Chamber
By-Pass Stack
ESP
Stack
Controls &
Sensors
Quench Pit
Water Level
x
x
x
x
See Table 2-3
x x
X X
X X
X X
X
X
X
X
X
X
33
-------
Differences in technology, design detail and operating experience
ultimately dictate required inspection and maintenance frequencies for specific
incinerator components. More can be inferred about system reliability from
whether or not a facility has a written inspection and maintenance schedule that
they adhere to than can be inferred from the details of such a schedule.
2.9 RECORDKEEPING
Management and technical personnel at MWC facilities require certain
information about the plant's operation so that they can monitor the efficiency,
production and reliability of the operation and take corrective action when
necessary. This information includes data about throughput, operating parameters
for the furnace, combustion, and air pollution control processes. Fortunately,
efficient and reliable MWC operation is totally consistent with good air pollution
practice. Therefore, the records and reports maintained by the MWC staff are
normally more than adequate to satisfy recordkeeping and reporting requirements
that may be desirable as stipulations in air pollution permits. An exception, excess
emission reporting, is discussed later in this chapter.
2.9.1 Furnace and Combustion Data
Section 3.3 contains an example furnace and combustion data log. Such
logs are maintained by the furnace operator. Typically, these logs are manually
compiled with readings recorded hourly or bi-hourly. However, there is a clear
trend on newer installations to compile this data on computer with current and
historical data available to the operator via CRT and/or paper.
Precisely what data is available and recorded at a given facility will .depend
on the basic MWC configuration, the sophistication of the automatic combustion
control system and the experience/opinions of the MWC management.
Monitoring the instruments is essential to maintaining a clean, efficient
operation and preventing minor process excursions from becoming substantial .and
costly equipment problems and environmental compliance difficulties. Creating a
historical record of the process operation provides MWC management with a tool to
evaluate trends in process operation and malfunctions.
2.9.2 Stack Gas Monitoring and Excess Emission Reporting
Figure 2-1 is an illustration of a typical Continuous Monitor Quarterly
Report. These reports can be rather laborious for plant staff to prepare jf. there is
34
-------
CJ
i
o
o
§ I
115
E 1 S
CO
«= a
UJ
£ Ul
A 1
^ O
CO
« 5 g
=
c. i_ ®
a; o 2
*" . "= "55
=5 a -s
§. £ §
^ ox §
o. w es
o O w
v? CO O
t 7T ^
— c "5
•a 2 a
0) tn O.
~ to o
i. to o
2- i «
es «
m »^ >?• 51
— • CD
•
5= o
= 6
35
C/3
-------
a large quantity of excess opacity excursions during the quarter. If the facility is
maintaining continuous compliance adequately, the reports can be compiled in a
very few minutes. It is the former case that presents a compelling argument for
permit conditions that require the preparation and submittal of such reports.
In addition to documenting the time, duration, and severity (percent
opacity) the report requires the identification of the nature and jcause of the
incident and the corrective action taken. These last two data columns provide the
most important information in the report. Most often, those facilities that
experience difficulty in achieving continuous compliance do so because they have
one or two operating problems that create the bulk of the violations (Reference
12). If the data are summarized to indicate which problems are causing the most
violations, the plant management and, if necessary, the state or local air pollution
control agency, can focus on the solution to the problem.
To date, EER's have not been required at MWC's utilizing
transmissometers. In view of the wealth of useful information that can be derived
from EER's, it is desirable to re-evaluate this approach.
2.9.3 Miscellaneous Records
Additional recordkeeping at MWC facilities will vary widely from plant to
plant and, in general, will be of marginal use to air pollution authorities. It is
helpful, however, to be aware of the broader recordkeeping activities since some
facilities will have useful environmental data tucked away on a record that
otherwise may seem rather obscure.
Figures 2-2 and 2-3 illustrate two forms utilized by the Montgomery
County (Ohio) Solid Waste Management Division in operating two MWC's at each of
two sites in the County. Figure 2-2 is the Incinerator Shift Operations Record
filled out at the end of each shift by the Shift Foreman. The record provides
summary information for the shift regarding operating time and throughput
(grapples fed and loads of residue hauled). Ammonia is used to control pH of the
process water, so the amount of ammonia used and the pH at each clarifier is
recorded once per shift at this facility. The ESP is equipped with an oil heater that
is fired during brief shutdowns to prevent acid condensation. Thus, the amount of
diesel fuel used per shift is also recorded.
Figure 2-3 is the Combustion Condition Observation Form. This form is
used as part of the Division's training program and is used on a random basis to
document operating observations. The value of the form is to encourage operators
to make appropriate observations of the combustion conditions and, just as
important, make appropriate corrections or explain why not.
36
-------
FIGURE 2-2
INCINERATOR SHIFT OPERATIONS RECORD
Plant
Date
Shift No.
Furnace 1
Furnace 2
Shift Foreman
Ooerating Hours
From
To
Truck// Gals
Trucks? Gals
Downtime Hours
From
To
-
Diesel Gals
PH 111 PH #2
Grapples
Fed
Loads of
Residue
Hauled ,
Ammonia Reading
Amount Used
Remarks :
•
37
-------
FIGURE 2-3
DATE
COMBUSTION CONDITION OBSERVATION FORM
Extremely Dry , Very Wet
Condition of Refuse
Condition of Ash #1
Condition of Ash #2
Condition of Plume
Poor Burnout
Difty
Drafts: All Balanced & Approximately Correct:
If not, Explain:
Good Burnout
Clear
yes
no
Grate & Kiln Speeds: All Balanced & Appropriate Conditions: yes no
If not, Explain:
Combustion Air Fans: All running, quiet and clean intake:
If not, Explain:
yes
no
Combustion Air Seals:
filled with water:
If not, Explain:
All doors shut, rotary locks working, and ash quench
yes no.
Division Chief Engineer
Response of Operator:
Signatures:
Incinerator Operator
In Charge
Shift Supervisor
Chief Operation
Supervisor
Comments Plant Manager
(Forward to Superintendent After Reviewed & Signed by All.)
38
-------
2.9.4 Charts and Computers
Many facilities continuously record furnace and cooling tower temperatures
and opacity on strip or circular charts. Even in such cases the operators are usually
required by management to manually log these parameters.
The extent to which computers are used at MWC's varies greatly from
plant to plant. Any automatic combustion control system is a computer, if
somewhat rudimentary by today's standards. Some facilities also utilize the
computer as a data recorder and logger. Operating parameters such as furnace
temperatures and oxygen concentration can be monitored via a CRT. In addition to
the instantaneous values, the computer is usually programmed to provide historical
summaries and hourly averages of the operating parameters. Precisely what data
are routinely printed will, again, vary from plant to plant. Facilities involved with
steam production usually have the most sophisticated control systems.
2.9.5 Summary
MWC facilities utilize rather extensive recordkeeping practices to operate
and control the combustion process and the movement of municipal waste and
combustion residue through the plant. These records are also useful to the plant
staff in diagnosing and resolving operational problems, whether those problems
result in excess air poUution emissions or not. The air poUution control official will
generally not need records beyond those maintained at responsible MWC
operations. He should be aware of the normal (baseline) operating ranges of
critical parameters such as furnace temperature, oxygen concentration, ESP
secondary power levels, and scrubber pressure drop that are specific for the
faculties in his jurisdiction. He should also be aware that operation outside the
baseline ranges for these critical parameters is an indication of a potential air
pollution problem.
Transmisso meters have proven successful in a number of MWC
applications. Facilities should be encouraged to install and maintain
transmissometers that meet the 40 CFR 60 Appendix B specifications in any new or
existing facility where water vapor in the stack or duct is not a problem, and, in
the ease of existing units, where the remaining useful life of the unit can justify
the expense. It must not be forgotten, however, that the opacity data generated by
such instruments is totally useless if it is simply stored in a box at the plant or the
air agency's offices.
39
-------
2.10 REFERENCES
1. Brunner, C.R. (1985). Hazardous Air Emissions from Incineration.
Chapman & Hall, Ltd., New York, New York.
2. Air Compliance Inspection Manual. U.S. EPA Publication No. EPA-340/1-
85-010, September 1985.
3. Richards, John R. and Robin Segal. Wet Scrubber Performance
Evaluation. U.S. EPA Publication No. EPA-340/1-83-022. September 1983.
4. Theodore, Louis and A. Buonicore. (1982). Air Pollution Control
Equipment: Selection, Design, Operation and Maintenance. Prentice-Hall,
Englewood Cliffs, New Jersey.
5. Azabo, M.F., and Y.M. Shah. Inspection Manual for Evaluation of
"Electrostatic Precrpitator Performance. U.S. EPA Publication No. EPA
340/1-79-007. March 1981.
6. McDonald, Jack R., and Alan H. Dean. A Manual for the Use of
Electrostatic Precipitators to Collect Fly Ash Particles. U.S. EPA
Publication No. EPA-600/8-80-025. May 1980.
7. Control Techniques for Particulate Emissions from Stationary Sources -
Volumes 1 and 2. U.S. EPA Publication Nos. EPA-450/3-81-005 a and b.
September 1982.
8. Yung, Shui-Chow, et al. Venturi Scrubber Performance Model. U.S. EPA
Publication No. EPA-600/2-77/172. August 1977.
9. Proceedings: Operation & Maintenance of Electrostatic Precipitators. Air
Pollution Control Association, Pittsburgh, PA. April 1978.
10. Proceedings: Operation & Maintenance Procedures for Gas Cleaning
Equipment - A Specialty Conference. Air Pollution Control Association,
Pittsburgh, PA. April 1980.
40
-------
11. Kalika, P.W. How Water Recirculation and Steam Plumes Influence
Scrubber Design. Chemical Engineering. July 28, 1969. pp. 133-138.
12. Schmidt, Charles M., and R. D. Allen, "Reducing Opacity Emissions from
Industrial and Utility Boilers," Paper No. 84-110.3, presented at the 77th
Annual Meeting of APCA, San Francisco, CA, June 29, 1984.
41
-------
CHAPTER 3
SITE VISITS AND OPERATING VARIABLES STUDY
3.1 INTRODUCTION
This chapter discusses several operating and air pollution compliance
problems described by MWC operators during the MWC site visits performed under
this study. Each "case history" describes the nature of the problem or concern
faced by the operator and the action or solution implemented to resolve the issue.
The kinds of problems addressed in the case histories are fairly typical in the MWC
industry. Thus, it is hoped that these presentations may be useful to other MWC
operators.
Also discussed in this chapter are the results of a field study conducted at
the Stamford, Connecticut MWC. The purpose of this field study was to determine
the relationship of operating variables routinely monitored at the facility to
opacity emissions.
3.2 SITE VISITS - CASE HISTORIES
Case History No. 1 - This facility operates two mass burn rectangular batch
furnaces rated at 100 tons per day each. The two cell furnaces are equipped with
rocking grates. The furnace exhaust gases are treated by a single two field
electrostatic precipitator. The facility operates on a five day per week schedule.
The ESP was installed in 1980 as part of a refurbishing and upgrading
program for the overall system. The operating staff and engineering contractor
were aware of the corrosion difficulties ESP's suffer in MWC service. Further, the
facility's five day per week operating schedule was expected to exacerbate the
corrosion potential since each shutdown would cool gases in the ESP to below the
acid dewpoint.
To eliminate this problem, an oil heater was installed in the inlet duct of
the ESP. The heater is fired during weekend shutdowns to prevent acid
condensation. The oil burner consumes 200 to 300 gallons of fuel oil during the
weekend to maintain ESP internal temperatures around 275°F. Often it is not
necessary to fire the burner in the summer.
The ESP's are inspected internally once per month. Corrosion has been
negligible over the seven years of operation. The facility superintendent feels that
the cost of the fuel oil is insignificant compared to the benefits derived from the
42
-------
burner. However, in view of current fuel oil costs, he would recommend a burner
that utilizes natural gas.
Case History No. 2 - This facility was experiencing fugitive dust problems
resulting from tipping floors that were not fully enclosed and unpaved roadways on
the plant premises. In 1982 the tipping areas were completely enclosed with sheet
metal siding and large overhead garage doors. Further, a program to install asphalt
paving on unpaved roads was initiated in 1981 and completed within three years.
These newly paved roads are swept regularly with automatic road sweeping
equipment. This program has brought the facility into compliance with local
fugitive dust control'requirements and has resulted in a substantial reduction in
citizen complaints.
Case History No. 3 - Exhaust gases from four rotary kilns are treated by
each of four venturi (cone-type) scrubbers. The scrubber liquor is a 100 percent
recycle: make-up is for evaporative losses only. Liquor treatment is provided by a
clarifier/still well system. No provision is made for pH control.
The system was suffering from persistant and advanced corrosion in the
scrubbers, mist eliminators, ducts, and ID fan housing. Observations at the site of
stack gas emissions confirms opacity measurements varying from 20 percent to 80
percent during normal operations. Furthermore, little difference in the plume
appearance could be noted with and without the venturi scrubbers in service. This
suggests a plume comprised predominately of submicron particles that are not
removed by the scrubber. Such particles are typical of the soot and cracked
hydrocarbon tars which characterize opacity due to incomplete combustion.
At the time of the site visit, the MWC management had installed an
ammonia system to control the pH of the scrubbing liquor. In addition, they had
begun a program to refurbish each of the four furnaces and control systems one at
a time. The refurbishment of the first system was near completion at the time of
the visit. The project had included rebuilding the ducts, scrubber and mist
eliminator to replace corroded parts and repair holes and cracks as necessary.
Additional water capacity was added to the presaturator. To improve combustion
conditions, overfire air was added to the furnace.
Case History No. 4. - The ESP ash hoppers at this MWC were suffering
frequent bridging problems. The original installation of the system did not include
vibrators on the hoppers to aid ash removal. Installation of vibrators solved the
problem.
43
-------
Case History No. 5 - This facility consists of two rotary kiln MWG's. As is
typical with such systems, the kiln is preceded by a furnace equipped with drying
and ignition grates and combustion air supplied by over fire and under fire nozzles.
Air pollution control is achieved by two ESP's with three fields each that vent into
a common stack. The ESP's were installed in 1982, replacing a wet scrubbing
system.
During the first year of operation the ESP's experienced re-entrainment
problems when the third fields were rapped, causing opacity violations. The re-
entrainment problem did not occur everytime the third fields were rapped, but
appeared to occur only when poor combustion occurred in the furnace, specifically
when the furnace was burning highly combustible waste. The resulting incomplete
combustion produces fine, light particles of soot that the ESP's can collect, but due
to the low resistivity of the particles, they become re-entrained during rapping.
To resolve the problem, the frequency of third field rapping was reduced to
once every 24 to 48 hours and interconnect ducts were installed between each MWC
line and ESP. This allows the exhaust gases of both MWC's to be vented to one ESP
while the third field on the idle ESP is rapped. With no airflow in the idle ESP, no
opacity violations can occur during rapping.
After installation of the interconnect system, opacity violations were
reduced to 1.67 percent of the total operating time. Almost half of the remaining
violations were related to startup of the facility. Yet, the larger half still
appeared to be related to some kind of uncontrolled combustion caused by a
specific waste. By coordination of various members of the operating staff the
problem was isolated and identified as a highly combustible waste from a styrofoam
processing plant. Apparently, the resistivity/re-entrainment problem related to
this waste was so severe, the ESP's simply could not handle it, even with the
interconnect procedure. The firm was contacted and informed that the MWC
facility could no longer accept the firm's waste.
Case History No. 6 - In 1983 this facility shutdown two furnaces to perform
an extensive rebuilding of the ESP's serving the furnaces. The rebuilds were
necessary because of extensive corrosion in the ESP's, particularly in the outboard
plates where holes of several square feet in area had developed. Prior to the
rebuilding effort the ESP's had operated reliably for almost nine years. The facility
superintendent felt that this is good performance for ESP's in MWC service.
Nevertheless, the superintendent is taking precautions to ensure that the MWC
system is operated in a manner consistent with long term reliability of the rebuilt
44
-------
unit. This includes maintaining furnace exit temperatures and ESP inlet
temperatures as high as possible to minimize the chance of condensation and
carryover of cooling tower spray in the ESP. For these units, maximum furnace
exit temperatures are 1600°-1650°F. Temperatures over 1700°F tend to cause
slagging. The ESP maximum inlet temperature is set at 650°F. The units have also
had some difficulty with welds breaking in the hopper assembly, allowing air to leak
in. These breaks are repaired during scheduled downtime.
Case History No. 7 - This facility consists of two RDF fired traveling grate
stokers with tail end boilers. Emissions are controlled by ESP's. Local air pollution
officials had documented several opacity violations and had received complaints
about ash fallout in the neighborhood.
Although the combustion air for this system is controlled automatically,
the feed rate, as dictated by the speed of the screw conveyor, was controlled
manually. The MWC staff developed software to provide automatic control of the
feed rate based on the steam demand. Automatic feed control. eliminated
quenching the bed by feeding too much fuel. To gain further control of the excess
air rate, the staff repaired cracks in the furnace system to minimize air in-
leakage. The combination of these measures eliminated the opacity and ash falldut
problems.
Additional system refinements are planned for the future. First, the
operators are planning to install, oxygen and carbon monoxide monitors. This
installation is part of a long term plan to provide combustion control through these
parameters rather than furnace temperature, which is the current practice. The
operators also believe they can provide better combustion control by eliminating
the fly ash reinjection system. Finally, the ESP's pneumatic rappers will be
replaced by hammer rappers. It is believed that the pneumatic rappers do not
provide sufficient energy to properly clean the plates.
Case History No. 8 - Two ESP's control particulate matter emissions from
two travelling grate RDF fired furnaces. Auxiliary fuel is used during startup and
when flame out occurs on the fuel .bed. Originally No. 6 oil was the auxiliary fuel.
The oil created opacity violations (the standard is 20 percent) and citizen's
complaints were registered. These problems were eliminated by switching to
natural gas.
Another problem at this site has been clinker formation in the ESP ash
hoppers. The clinker causes the ash to bridge, thus backing the ash up into the ESP
itself. When the ash approaches the plates the electric field is disrupted, causing
45
-------
the ESP to short out. The MWC staff has investigated the problem and they are
convinced that the bridging is not caused by air in-leakage/condensation. At the
time of our visit they were considering rebuilding the hoppers so that the hopper
walls are almost vertical and the openings to the ash conveyors below are
substantially larger than the 8-inch diameter valves currently in service.
The fuel delivery system (air blast distributors) has suffered some abrasion
problems in the ducts where the RDF is blown onto the traveling grate. The
abrasion apparently is caused by glass contained in the RDF. (Refuse processing at
this facility does not include an air classification system that would eliminate glass
from the fuel. Basically, the fuel processing at this facility entails hand sorting,
magnetic separation, and shredding by a hammer mill.) Although the problem has
not been completely eliminated, some improvement was noted when a local bottle
bill went into effect. Returnable glass bottles resulted in less glass in the RDF and
also resulted in a small but noticeable increase in the heat content per unit mass of
the RDF.
3.3 OPERATING VARIABLES VS. OPACITY STUDY
In an effort to determine how MWC operating practices affect visible
emissions, a field study was initiated in the City of Stamford, Connecticut at their
co-disposal (municipal solid waste and sewage sludge) incineration facility. The
purpose of the study was to attempt to correlate "real world" operating data
routinely compiled by the MWC staff to the visible emissions emitted by the
facility. The operating parameters such as furnace temperature, under fire air
pressure, furnace pressure, etc. recorded hourly over ninety days were statistically
correlated to the hourly opacity readings from the opacity monitor.
The 360 ton per day (tpd) Stamford MWC was chosen because the facility
met several selection criteria established by the EPA. Specifically,
1. The facility has a good overall visible emission compliance record and is in
generally good mechanical condition. Therefore, it was known that good
operating practices can produce good results at this facility.
2. The Stamford staff has conducted several studies on combustion efficiency
and throughput and thus has experience in data collection and analysis.
3. Except for the co-firing of sewage sludge, the equipment configuration at
Stamford is reasonably typical of intermediate to large size MWC's
currently in operation in the United States.
4. The Stamford facility has an opacity monitor.
46
-------
5. An east coast location for the test facility was necessary for economical
travel for EPA Headquarters and contractor staff.
This section provides a process description of the Stamford 360 tpd MWC.
This is followed by a description of the field study protocol. Finally, a brief
discussion of the results and conclusions are presented.
3.3.1 Process Description
The 360 tpd MWC in all ways, except one, is a "typical" mass-burn unit with
continuous feed of unprocessed municipal waste, rocking grates, and an ESP to
treat exhaust gases. The one exception to this unit being typical is the use of
energy to process the solids removed from the sewage treatment plant. The MWC
is designed to process 30 tons per day of dried sewage sludge with a maximum
moisture content of 20 percent from belt filter presses. The sludge comes from the
city's 20 million gallon per day (mgd) wastewater treatment plant located on the
same site. Sludge processing is accomplished by two belt filter presses, a rotary
kiln that uses a slip stream from the furnace exhaust to dry the sludge, and the
MWC. The dried sludge is fed to the MWC continuously from the kiln through
several openings in the furnace ceiling. Figure 3-1 is a schematic of the Stamford
waste processing facility. Flyash laden water from the process water system is
pumped to a separator/clarifier that is part of the MWC. The city also operates a
150 ton per day batch MWC at this site that is not a part of this study.
Waste Handling and Storage - Trucks dump refuse into a receiving bin that
has a capacity of 425 tons level fuU and 725 tons heaped. This design capacity is
based on the assumption of a refuse density of 12.4 pounds per cubic foot. (More
typically, refuse density is 25 lbs/ft^ which means that the receiving bin capacity
may hold twice the design capacity). The refuse consists of residential,
commercial, and industrial waste. Other types of waste such as pathological or
hazardous wastes are not processed at this faciliity. The proximate analysis of the
municipal refuse burned is typically 55 percent combustibles, 30 percent moisture,
15 percent ash and has a heat content of approximately 5400 Btu/lb. The furnace
feed hopper is served by two air conditioned cab cranes each with a 2-cubic yard
grapple. The hopper delivers refuse to a refractory lined continuous feed chute
connected to the furnace. Refuse in the chute provides an air-tight seal for the
furnace when in operation. A hydraulically powered cover seals off the hopper
during shutdown periods.
47
-------
Figure 3-1
Schematic of Stamford Incinerator
PRECIPITATOR
225,000 CFM at 600°
PRECIPITATOR
750,000 CFM at 600'
*rr- — » f
HOPPER^ /SLUDGE INJECTION NOZZLES I
CENTRIFUGALS
EXHAUST SILENCER
II - -
UNDERF1REOUCT
UNOERFIREFAN
"ARY SLUDGE DRYER
OVERFIRE DUCT j QVERFIRE , ,
"/FAN CHEMICAL MIXING TANKS
lENCY GENERATOR
SLUDGE RECEIVING
WELLS
Source: Reference 1
48
-------
The receiving bin is equipped with water sprays that are designed primarily
for fire suppression. When sensors in the bin roof detect abnormally high
temperatures, alarm annunciators in the control room signal for manual operation
of the spray nozzles.
Three sludge wells are located in the building in addition to sludge pumps,
belt filter presses, drying equipment, and chemical conditioning storage/mixing
facilities. With a maximum moisture content of 20 percent, the sludge burns at a
maximum hourly rate of 1.25 tons.
Hot gases from the furnace are used in a sludge drying heat exchanger
(rotary kiln) with heat exchange exhaust ducted back to the combustion chamber.
Dried sludge is introduced into the furnace through openings in the furnace
roof, with burning taking place in suspension prior to reaching the furnace bed.
Typically, the dried sludge fed to the furnace will have a heating value of 3000
Btu/lb and a moisture content less than 10 percent.
Furnace and Grates - The combustion chamber is of wet-bottom design;
i.e., a poor of water is maintained in the bottom into which bottom ash and larger
particles in the exhaust stream settle out and are removed in a slurry. A water
make-up system automatically maintains the desired water level with manually
operated dump valves provided for the draining of the flyash laden slurry.
The lower furnace walls are subjected to abrasion from moving refuse.
Thus, this portion of the furnace is outfitted with air cooled silicon carbide
refractory materials.
Waste is agitated and transported through the furnace by a series of
rocking grates on which the refuse burns. Rocking grates are arranged in rows
across the width of the furnace, at right angles to the solid waste flow. Alternate
rows are mechanically pivoted or rocked to produce an upward and forward motion,
thus advancing and agitating the solid waste. The rocking grate stoker is
hydraulically driven. The same hydraulic system also operates the refuse hopper
cover and a chute at the high end of the ash drag-out conveyor. The furnace grate
area is 355 ft2 with a grate loading of approximately 80 lb/hr/ft2. The furnace
dimensions are 12 ft. in width, 39 ft. in length, and 29 ft. in height.
Combustion Air - Combustion air is provided by underfire air, sidewall
cooling, and overfire air forced draft systems. Underfire air is that portion of the
combustion air which is introduced to the chamber from beneath the grates. The
underfire air is manifolded to several areas under the grates, where it is directed
upward so that the air passes through the grates and refuse.
49
-------
The underfire air is furnished by a 70,000 cfm fan that is modulated by an
inlet damper. The underfire air fans are designed to provide sufficient velocity
(energy) to the underfire air jets so that the combustion air penetrates the fuel bed,
thus providing maximum contact of combustion air with the solid, combustible
material in the refuse.
In the automatic mode, the volume of air delivered by the
underfire/overfire system is regulated through a control" system driven by a
temperature sensor located near the furnace ceiling. The control system increases
the volume of underfire air if the furnace temperature is too low simultaneously
reducing the over fire air. The volume of underfire air is decreased if the furnace
temperature is too high, simultaneously increasing the overfire air. The movement
of the underfire/overfire air dampers is always proportional, i.e., they move in
opposite and approximately equal amounts. By viewing Figure 3-2 one can see the
position that the underfire and overfire are at in respect to each other both inside
and outside the selected burning temperature. The amount of air being delivered to
the furnace by the overfire air system is monitored by a draft gauge on the overfire
air duct just after the overfire air fans, the amount delivered by the underfire air
system is monitored by a draft gauge on the underfire air duct just after the
underfire air fans, and the total air movement through the incinerator is monitored
by a draft gauge on the outlet side of the forced draft fan.
Two 12,500 cfm fans with inlet damper controls supply overfire air. The
reasons for introducing overfire air are to ensure, through added turbulence, the
complete oxidation of combustible gases that have evolved from the solid refuse
fuel bed, and to control high temperatures in the furnace. (As described above, an
increase in the overfire airflow quenches the furnace gases.) Overfire air is
introduced above the fuel bed through the furnace sidewaUs. The overfire air
system was designed for complete penetration of the exhaust gases.
SidewaU cooling is supplied by two 15,000 cfm fans. Although the primary
purpose of sidewall cooling is to protect the refractory, sidewall cooling air
contributes to the overfire combustion air. The sidewall cooling air fan damper is
manually operated and normally set at 40 percent open.
50
-------
CD £
CD
CD
.9-E
JZ 0
O
CD .—
13 ^
D) CD
±o
CC
D
CD
o-t:
CD 0
SI
< \
I <
1
1
-------
Air Pollution Control - The emission control system consists of a spray
chamber followed by an ESP. The water spray chamber reduces the combustion gas
temperature to between 550° and 650°F prior to entering the ESP. The spray
chamber consists of a multiple bank spray system, with cluster type nozzles
providing 100 micron diameter droplets at 125 psi. Spray volume is automatically
controlled in 45 to 50 gpm increments. The spray chamber has high-heat heavy
duty fireclay brick wall§ with a castable refractory similar to that of the
combustion chamber.
The ESP is located on the roof because of site restrictions. The ESP
consists of two electrical fields and is designed to handle 225,000 acfm at 600°F
and a maximum pressure drop of 0.5 inches water column. Motorized dampers on
the inlet duct control flow under abnormal operating conditions. The cross-
sectional area of the ESP was designed to provide a velocity of 3 to 4 feet per
second through the unit in order to allow adequate detention time and limit dust re-
entrainment. ESP design parameters are summarized in Table 3-1. An induced
draft (ID) fan is downstream from the ESP. The fan has a radial'tip blade, a
variable speed drive unit, and is sized for 202 bhp.
Inlet dust loading conditions are 3.5 pounds of particulate matter per 1000
pounds of dry gas, adjusted to 50 percent excess air for product of combustion.
Maximum particulate matter concentration in the ESP discharge is 0.175 pound per
1000 pounds of dry gas was specified; equivalent to 95 percent efficiency.
Sampling ports are installed on the inlet and outlet ducts of the ESP for
performance testing. Sampling ports are also available on the exhaust stack.
Multiple hoppers at the bottom of the ESP collect flyash, with rotary
valves for removal. Slide gates have been installed upstream of these valves to
facilitate maintenance without emptying the hoppers. From the rotary valves, the
collected flyash drops into a sluicing system and is conveyed by suspension in water
to the ash drag-out conveyor trough.
The ash handling system consists of stationary packers receiving the
residue from the ash drag-out conveyor. The residue is then transferred into ash
receiving containers and hauled to disposal sites.
52
-------
TABLE 3-1
ESP PARAMETERS
Equipment Parameter
Control Efficiency (Design)
Plates
No. of Fields
Wires
Plate Area
Distance from wire to plate
Rappers
Hoppers
Instrumentation
Value
95% minimum
18 gauge cold rolled mild steel
2
Weighted Wire Design
39,000 sq ft
4.5 in
Solenoid (Replaced later with pneumatic
rappers)
60° hopper slope, electric heaters and
vibrators. Discharge valves: rotary feeders
Analog meters for primary and secondary
current and voltage for each field. Spark
meter on inlet field only.
Flue Gas Parameters
Air Volume (design)
Gas Velocity
Inlet Temperature
Maximum Temperature
Exhaust Gas Moisture
Inlet Particulate Loading
Outlet Particulate Loading
Value
225,000 acfm at 600°F
3.0 ft/sec minimum
4.0 ft/sec maximum
550°F - 650°F
700°F
1500 Ib/min
3.5 lb/1000 Ib1
0.175 lb/1000 Ib1
'•Pounds particulate per 1,000 pounds dry gas @ 50% excess air.
53
-------
Instrumentation Controls - An air-conditioned control room is located on
the ground floor near the discharge end of the furnace. The control room is
furnished with basic instrumentation and controls for the furnace and ESP.
Automatic controls and panel readouts are provided for air and water supply
pressures and flow rates. Fans, pumps, and control dampers are monitored and
controlled manually when not in the automatic mode. The temperature probe
measuring furnace exit temperature is located near the center of the primary
chamber ceiling. AU the temperature probes are shielded to prevent false high
readings attributable to radiant energy within the combustion chamber. Continuous
gas sampling and analysis for smoke density is conducted by a self contained system
located on the third floor. Operating procedures require daily calibration.
An opacity monitor is installed on the exhaust stack. A chart recorder for
data storage is included with the opacity monitoring system and is located in the
control room. Records were not available to indicate if the unit has passed a
performance specifications test or when the last relative accuracy test was
conducted. The instrument zero and span drift is checked and adjusted daily by a
certified instrumentation technician.
3.3.2 Field Study Protocol
The field study was a cooperative effort between the City of Stamford's
Bureau of Sanitation and the U.S. Environmental Protection Agency's Office of Air
Quality Planning and Standards. After review and acceptance of the field study
protocol by each of the parties, the Bureau of Sanitation was responsible for
coUecting approximately 60 days of data on process and operating conditions, and
visible emissions.
Figure 3-3 illustrates the operating parameters provided by the City of
Stamford. All the data except oxygen, carbon monoxide and carbon dioxide were
part of the original study concept. The gas concentration data were coincidentaUy
being coUected by the MWC staff as part of a preliminary program to look at
certain conditions to assess the feasibility of: 1) if and what should be replaced of
the antiquated nitrogen dioxide, oxygen, carbon monoxide and carbon dioxide
monitoring equipment supplied with the facility in 1973; and 2) the possible addition
into the over fire/under fire/temperature control loop of feed with trim control
within span provided by the carbon monoxide concentration. To minimize expenses
for the preliminary program, the Sanitation Department borrowed a CO/CO2/O2
54
-------
Q*
r
g
H
G.
-H
i
0
«J
u
tl
CA
V
X.
0
hi
to
•a
o
o
CM
Q 4)
E 41
4) U
•U O*
U
a
H
&
U
4>
«*>
10
11
12
13
IS
«r¥j
Vf
u->
17
ia
19
20
21
23
/ •» f;'
?*<>'
«*
A3
*?,.;
ts
tffi:
condition
-dry/vet
Hall
Furnace pre»«ure
;2
- II
\~\f-.
'/
**,
Vf
/O
Smoke »eter rending
Through put/hr
Iv'W&o
'it
•tl
• n
h
in
ta
n
n
-X
7.3
7s
Dioxide
Carbo
S
E
S
S
£35_J_W?/
•/c
M
J*<
Ml
z^a.i
it.?
rf",
/o /y£
/t)
li
ia
'f*.
/?
?>
17*
t.o
•r /
i^v/
' -/
,i/ A » | ,?/
IM
•:/
11..
.w
?'H
,'4^
1.-1
1,3.
*
?fl
.0 V//
;r I «s
f>
l,7«
'-
511,24
3Ll,n
IW
Kl
V t
1) Peed stroke • add strokes/sect ion x 12 •«• s«c delay ti:
[
\
7 feed section setting
30.67
27 x 250 • lb« per hour
ncinerator
T3
U
O
UJ
S
-------
instrument that was designed to compute combustion efficiency for oil, gas or coal
combustion.
During the course of the study, it was desirable to manipulate some of the
operating variables over specific ranges to broaden the data base. Specifically, it
was agreed to attempt the following:
Furnace Temperature: During the course of the study, it was
recommended that the furnace exit temperature be run from 1600°F to
1900°F and back to 1600°F in 50 degree increments. To ensure that the
furnace stabilized at each temperature level, each increment was to be
held for at least four hours. After considerable discussion with the
Superintendent of Sanitation, Mr. Marvin Serra, and the Solid Waste
Division Staff, it was decided that determination of upper and lower bounds
through opacity violation would not be an acceptable means. Further it
was determined that 1) variation of the temperature, because it affected
throughput, could only be accomplished when time permitted, and 2)
operating above 1900 degrees F could not be permitted because of possible
high heat problems with the ESP.
Corona Power: The corona power would be adjusted through
controlling the voltage. The run would be similar to that done for the
furnace temperature. However, some experimentation was required to
determine how low the power could be set without causing opacity
violations. At the upper end, an excessive spark rate could be noted. Once
the upper and lower limits were defined, the test would begin at the lower
power setting and increased incrementally every four hours to the upper
power limit and back to the lower limit.
Manipulation of other operating variables was considered but rejected as
not being practical. For example, fly ash resistivity is dependent on flue gas
temperature (among other things). The range of responsible ESP operating
temperatures is not very large. The upper limit is defined by what the unit can
take without suffering heat damage. The lower limit is defined by how much
water can be injected at the cooling tower without carryover of liquid droplets into
the ESP. Since this temperature range is limited, the fact that the temperature
would not be controlled to better than plus or minus 10 to 20°F became significant
to the overall temperature range available.
3.3.3 Discussion, Results, and Conclusions
The operating parameters were run through a multiple regression analysis
to determine which parameters had a good correlation with opacity. It was
anticipated that several parameters could provide a good correlation, indicating
that monitoring and controlling certain process operating variables would produce
acceptable emission levels.
56
-------
Operating personnel were only partially successful in regulating furnace
temperature and ESP corona power and thus determining their impact on stack
opacity. The reason that the furnace temperature could not be easily regulated can
be attributed to several factors. The two most significant factors are the broad
range of refuse burned and the inability of the system to control the refuse
through-put (the majority of the time it was required to be at maximum).
One of the methods most commonly used to analyze data is a simple linear
regression and correlation, from which a correlation coefficient can be derived.
The correlation coefficient is an index measure of the degree of linear association
between dependent and independent variables. The linear regression equation is the
straight line that best fits the data. The closer the correlation coefficient is to
unity (1.00), the stronger the linear relationship is likely to be. Correspondingly,
for inverse relationships, the strongest correlation coefficient would be -1.00.
Conclusions - The following conclusions were drawn from the analyses.
1. The statistical analyses that were performed emphasized opacity and
carbon monoxide concentration in the exhaust gases as dependent variables,
while other process parameters such as carbon dioxide concentration,
oxygen concentration, furnace throughput, furnace temperature, ESP
temperature, etc., were considered independent variables. The analysis
involved determination of the arithmetic mean and standard deviation of
the variables, the correlation coefficient to measure the strength of
relationship between two variables, and a simple regression equation (which
shows the best aggregate correlation between variables).
2. Statistical analysis of the available data was unable to establish any
reliable correlation between stack opacity and any other independent
variable.
3. Analysis of the data also showed no strong correlation between CO
concentration and any variables such as furnace temperature (-0.12),
carbon dioxide (-0.05), oxygen (0.14), refuse burn rate (0.07), ESP inlet
temperature (0.05), or stack opacity (0.03). The low correlation
coefficients among the gas concentrations may be the result of utilizing a
gas monitoring instrument that was not designed for the use employed by
this study.
4. All of the outputs (meters and strip charts) indicated that the variables
were constantly varying at various rates. This is typical and expected for
the operating variables at an MWC. Recording instantaneous readings at
57
-------
the end of each hourly period may raise the question as to how
representative the data may be of the actual operation for that hour.
Ideally, the pluses and minuses associated with each reading would, over a
90 day period, cancel each other out. The correlation factors resulting
from this analysis indicate strongly that 1) the sample population may not
have been large enough, or 2) the high variability in refuse composition and
operating conditions had a strong impact on the analytical results. If the
latter is true, an automated data logger that can output hourly averages of
the dependent and independent variables, rather than instantaneous
readings at the end of each hour, may have produced higher correlation
coefficients. However, installation of such a device would have violated
the "real world" operating and recordkeeping criteria defined in the first
paragraph of Section 3.3.
Another difficulty in analyzing MWC operating data is the lag time
between cause and effect when adjusting one operating variable to improve
the value of another. For example, if the combustion air is adjusted
because the furnace temperature moves out of the acceptable range, the
furnace temperature will not respond instantaneously, but only after some
period of time - perhaps anywhere from several minutes to one-half hour,
or so.
The opacity monitor was audited on July 24, 1985. The monitor is of a
design common in the 1960's and early 1970's where an automobile spotlight
is used as the light source, and an unfiltered selenium photocell is used as
the detector. The monitor meets the requirements of the operating permit
issued by the Connecticut Department of Environmental Protection. The
U.S. EPA does not require opacity monitors at municipal incinerators.
However, monitors of this design are not capable of meeting the U.S. EPA
performance specifications contained in 40 CFR 60 Appendix B. However,
the audit report (Appendix A) concluded that the unit's performance was
adequate for the purpose of the Stamford pilot study. In addition to
summarizing the audit results, the report provided recommendation for
calibration and upgrading the system. The audit was originally scheduled to
coincide with the start of the pilot study. However, due to process
equipment replacement requirements for the study (ESP control system
replacement which was completed before the data collection), and the
replacement of the opacity monitor (which was not completed before the
58
-------
3.4
1.
data collection) the pilot study was delayed. Startup of the pilot program
actually occurred on March 31, 1986.
REFERENCES
Gates, D.W., Incinerator is Part of Integrated Waste Disposal System,
reprint from Public Works Magazine, May 1974.
59
-------
APPENDIX A
AUDIT
OPACITY METER
STAMFORD MUNICIPAL INCINERATOR
-------
-------
AUDIT
OPACITY METER
STAMFORD MUNICIPAL INCINERATOR
BY
THOMAS H. ROSE
JULY 26, 1985
EASTERN TECHNICAL ASSOCIATES
BOX 58495
RALEIGH, NC 27658
-------
-------
TABLE OF CONTENTS
I. INTRODUCTION
II. TRANSMISSOMETER SYSTEM DESCRIPTION
III. AUDIT PROCEDURES
IV. ' AUDIT RESULTS
V. RECOMMENDATIONS FOR CALIBRATION/OPERATION OF PRESENT SYSTEM
VI. LIGHT SOURCE AND CONTROL SYSTEM
VII. RECOMMENDATION FOR REPLACEMENT OF PRESENT SYSTEM
-------
-------
I. INTRODUCTION
Eastern Technical Associates was contracted to perform a survey audit of the
transmissometer located on the municipal incinerator in Stamford, Ct. The
purpose of this survey was to determine the suitability of the transmissometer
in determining the effects on in-stack opacity by varying the operational
parameters of this incinerator. The incinerator is a well maintained and
operated unit equipped with an electrostatic precipitator. The unit has a
history of operating within the emission limitations of the State and EPA.
The transmissometer is on the incinerator stack for operational controls only.
This source is not required to have an opacity meter. Thus the transmissometer
is not required to meet the EPA specifications for a continuous emission
monitor. The study of the effects of changes in incinerator and collector
operational parameters on in-stack opacity does not require .cull compliance
with SPA GEM specifications.
The management and staff of the municipal incinerator were extremely helpful
in assisting in the conduct of this audit to the point of working overtime to
assure our success. Their assistance and cooperation is sincerely
appreciated.
-------
II. TRANSKISSOMETER SYSTEM DESCRIPTION
A. GENERAL SYSTEM DESCRIPTION
Tlie transmissometer system is very similar in design and operation to those
used in smoke generators used for opacity training and certification in the
late 60's and early 70' s. It is a single pass unit consisting of a automobile
Searchlight source, a selenium photocell and a stripchart recorder.
B. INSTALLATION
The transmissonieter is installed on the stack just above the roof over the 4th
floor of the incinerator building. The stack diameter at that point is
approximately 10 feet. The stripchart recorder and control console is located
remotely in the main incinerator control room.
C. SOURCE
The source is a 6.3 volt automobile spotlight that is controlled via a variac
and stepdown transformer. This allows the adjustment of the light intensity
to match the range of the photocell. As the light intensity goes down the
beam becomes rich in infrared radiation and the system becomes insensitive to
the smaller particles responsible for the scattering of light most visible by
human observers. Thus this type system is almost never photonic and generally
underestimates opacity. The type of light used has two distinct beam widths.
There is a stronger inner beam and a wider weaker beam. This is not really
suitable for a transmissoineter .
D. DETECTOR
The detector is a selenium photocell probably manufactured by International
Rectifier and repackaged. It is unfilterecl and therefore not fully photopic in
response. This type of cell can be linear if operated with an output of over
15 railli volts. At the time of the survey the sensitivity or output of the cell
was attenuated to 20 millivolts with a 1 turn 20 1C Ohms pot. The wiper ^ of the
pot was worn out and it was difficult to adjust and was not stable. With tne
assistance of the staff electrician we replaced the pot with a Burns 10 turn
precision pot of equivalent resistance. This corrected the instability
problem and made the adjustment of the output simpler.
E. RECORDER
The recorder appeared to be in good condition and responded to values across
the scale. Due to the nature of the source and the detector the chart was
difficult to read. The quick response time of the system can be reduced to
alleviate this problem. The recorder is set up to report in Ringleman
•lumbers, from 0 to 5.
-------
III. AUDIT PROCEDURES
A. STANDARDIZATION
The normal standardization procedure (calibration) could not be performed.
This was due to the fact that it is a single pass instrument and the
incinerator could not be turned off. Instead we performed a relative
calibration on the system. Using' the average opacity from visible emissions
observations conducted on the stack the baseline was established. Normal
procedure would be to establish this as zero on a clear stack. The span value
or slope was established by blocking the light source off .This is in
accordance with established practice.
B. LINEARITY
Normal linearity tests are performed on a transmissometer with a dynamic zero
and span. As this is a single pass instrument on an operating stack, this was
not possible. After the preceding standardization procedure the test was
conducted with the stack in operation. This is not proscribed procedure. It
will only give relative results. If the in-stack opacity is reasonably stable
an estimate of linearity is possible but it is not accurate. Linearity was
tested by insertion of National Bureau Of Standards traceable neutral density
filters into the light path. To assure that all of the light entering the
photocell was attenuated, the filter was attached to the face of the
photocell. Three filters were used with nominal opacities of 20,50, and 80%.
To account for the variation in response due to stack opacity, multiple
insertions of each filter were conducted.
C. INFRARED RESPONSE
Infrared response was audited by placing an infrared filter over the photocell
in the system. This filter passes less than 3% of visible light. Thus the
photocell output should be near zero or less than 5% of full output to
indicate no infrared response.
D. ANGLE OF PROJECTION AND VIEW
Due to the inability to bench test the actual angle of projection and view, no
actual figures are reported. However, extensive experience with similar
systems and a careful visual inspection of the installation did give some
information.
E. RESPONSE TIME
Response time was checked by timing the full scale response of the recorder
while alternating turning the source on and off. Both upscale and downscale
response were measured.
-------
IV. . AUDIT RESULTS
A. CALIBRATION
The calibration could not be accurately determined on the instrument
at the time of the audit. Instead a; relative calibration was
performed without the actual zero value.
B. LINEARITY
The following table was developed from equations derived from the following
relationship. It allows the establishment of a baseline within approximately
10% opacity. The average opacity can then be used in the same equation to
estimate the response of thE meter to filter insertion with the baseline
removed.
100-T
100
F Filter value in % opacity
E Estimated opacity of stack in %
T Total measured opacity by meter in %
WHERE
TABLE,1
ESTIMATION OF OPACITY FROM FILTER AND METER DATA
Data Opacity Opacity Trans.
point observed filter observed
Trans. Trans. Opacity
filter estimate estimate
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
44
74
34
52
38
58
42
60
100
30
60
40
28
80
80
60
44
44
71
21
44
21
44
21
44
100
0
44
21
0
71
71
44
21
56
26
66
48
62
42
58
40
0
70
40
60
72
20
20
40
56
56
30
79
56
79
56
79
56
0
100
56
79
100
30
30
56
79
100
88
83
85
78
75
73
71
0
70
71
76
72
68
68
71
71
* Point taken prior to system stability
** Total light path blocked
0 *
12
17
15
22
25
27
29
100 **
30
29
24
28
32
32
29
29
-------
Table 2 was developed from the data on table 1. It utilizes the average
opacity of the stack as the baseline ( E ) in the equation to determine the
effect of the filter and the linearity of the system.
TABLE 2
ESTIMATE OF METER RESPONSE TO FILTER INSERTION
DATA OPACITY TRANS OPACITY
POINT OBSERVED OBSERVED BASE
1
2
3
4
5
6
- 7
8
9
10
11
12
13
14
15
74
34
52
38
58
42
60
30
60
40
28
SO
80
60
44
26
66
48
62
42
58
40
70
40
60
72'
20
20
40
56
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
75
75
75
75
75
75
75
75
75
75
75
75
75 .
75
75
TRANS ESTIMATE FILTER
BASE RESPONSE VALUE
65 71
12 21
36 44
17 21
44 44
23 21
47 44
7 0 *
47 44
20 21
4 0 *
73 71
73 71
47 44
25 21
* No filter inserted
Charts 1 and 2 illustrate the approach that was used. In chart 1 the actual
response of the meter is plotted against the known filter value. The meter is
set up with an upscale zero based upon visible emissions observations. Scatter
is due to changing opacities in the stack. In spite of the scatter, the points
indicate that the transmissometer is linear. In chart 2 the average of the
estimated opacity values in table 1 (25) is used to calculate the estimated
response with the upscale value calculated out. Once again the plot is linear and
this time goes thru zero. This demonstration of the technique was based upon a
quick evaluation of the stack opacity and is offered for demonstration purposes
only.
-------
CHART 1
tt
111
_!
E
a
in
|_
_l
E
80
72
64
56
48
40
32 ,
24 '
16
8
0
RESPONSE
D
a
a
n
" n a
g
i
-
-
i i i i \ § i i i
0 50 100
CHART 2
100
90
80
70
€0
50
40
30
20
RESPONSE
.
-
D-
-
a
-
a
UJ . _
n
n
•MM
D
10 L D -j
. 0 1
1 1 1 1 1 1 II 1 1
i
.0 50 ; 100
-------
C. INFRARED AND PHOTOPIC RESPONSE SURVEY
The selenium detector used in the system is near photopic. That is it matches
the photopic response curve of the human eye on the long wave end but not on
the ultraviolet end. The infrared response test indicated that 5% or less of
the response was due to infrared radiation. Considering the rough setup this
is within acceptable limits.
D. ANGLE OF PROJECTION AND VIEW
The angle of projection and view were not accurately determined. However, the
spotlight used as a source has a wide angle of projection and a more intense
beam with a narrower angle. The sensitivity of the system to movement of the
light indicates that the narrovr beam has the greatest effect on the
transraissometer. The photocell has a wide angle of view. There are no
aperture plates to restrict the view of the photocell and prevent it from
being negatively biased by scattered light.
E. RESPONSE TIME
The response time of the transraissometer is under 1 second for full scale
response from either zero to 100 or 100 to zero. This results in a difficult
to read strip chart. A response time of approximately 4 seconds would be more
useful in this system.
F. SUNLIGHT INTERFERENCE
No appreciable interference from sunlight or ambient light was detected.
V. RECOMMENDATIONS FOR CALIBRATION/OPERATION OF PRESENT SYSTEM
Purchase 4 inch neutral density filters of nominal opacities of 20,50,and 80%.
This is the equivalent of transmissions of 80,50,and 20%. For the purpose of
calibrating this system gelatin filters available from a camera store would be
adequate. Higher quality filters would be xvasted.
PROCEDURE
1. Establish the span by use of a certified smoke reader and adjust the meter
to that value by adjusting the sensitivity or photocell output.
2. Establish the transmission baseline by turning off the light source or
blocking it. Adjust the recorder zero to just on scale.(a live zero)
3. Repeat steps 1 and 2 several times until consistent results are
achieved.(no adjustment necessary)
4. With system on, insert the filters one at a time in the opening between the
photocell and the stack, and record the results. Multiple insertions are
highly recommended.
-------
5. Perform a linear regression of the results but do not force it thru zero.
6. Alternately generate a calibration curve by compensating for the instack
opacity by using equation #1. Results illustrated in table 2.
VI. RECOMMENDATIONS BOR REPAIR/MODIFICATION OF PRESENT SYSTEM
A. LIGHT SOURCE AND CONTROL SYSTEM
••The light should be operated at its rated voltage to assure proper light
wavelength emissions. This can be accomplished by installing a constant
voltage (regulated) power supply for the light and checking that the light
Voltage is correct with a meter at the light with the light on. Selection of
the proper light and its associated voltage must be in conjunction with the
proper output of the photocell.
B. PHOTOCELL
The photocell should be modified by the addition of an 102 filter available
thru any retail camera store. A better solution would be to purchase a
photocell with encapsulated filter. The photocell must be producing at least
25 millivolts with the selected light source to be linear. An aperture plate
should be installed in the photocell housing to restrict the angle of view to
at least 15 degrees preferably 10 degrees. This would result in the detection
of less scattered light.
C. STRIP CHART RECORDER
The strip chart recorder does not need replacement. However it should be
'modified by installation of a 200 mfd 25 volt electrolytic capacitor in
parallel with the input to the recorder. Additionally the recorder should be
set up to measure transmission of light not opacity. Thus zero input would
equal 100% opacity and 20 millivolts input would equal 0% opacity. This may
require the adjustment of the alarm system but should cause little difficulty.
cell output
millivolts
20
15
10
5
0
transmission opacity
100%
35%
50%
25%
0%
0%
25%
50%
75%
100%
VII RECOMMENDATION FOR REPLACEMENT OF PRESENT SYSTEM
It is recommended that when the present unit is replaced, that the replacement
transmissometer be a double pass instrument from one of the several major
manufacturers. Additionally the purchase should include a full EPA
performance test with a condition that payment be conditional to demonstrated
compliance with all EPA requirements.
-------
VISIBLE EMISSION OBSERVATION FORM 2
-------
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA 340/1-87-002
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Municipal Waste Combustor Systems
Operation and Maintenance Study
5. REPORT DATE
June 1987
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Roger D. Allen
8. PERFORMING ORGANIZATION REPORT NO.
3. PERFORMING ORGANIZATION NAME AND ADDRESS
Allen Consulting and Engineering
317 Howl and Avenue
Gary, NC 27513
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
7W-6610-NASX
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Stationary Source Compliance Division, OAQPS
401 M. Street SW
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
SSCD Project Officer: Pamela Saunders, EN-341 (202) 382-2889
16. ABSTRACT
This study was undertaken to determine the significant operation and
maintenance (0 & M) considerations that bear directly on the day in/day out
air pollution compliance status of municipal waste combustors (MWC's). Two
independent tasks were performed to compile and document these 0 & M considerations.
First, seven MWC sites were visited to interview equipment operators and
management personal to determine the nature of the problems they routinely
encounter in operating MWC systems. The information gathered from their visits
is reported herein through a description of 0 & M problems and solutions for
each major subsystem in the MWC process. The subsystems discussed include: fuel
preparation and handling; furnace operation; ash handling; air pollution control
equipment; fans, ducts and stacks; and procedural considerations such as startup/
shutdown, equipment inspection, and recordkeeping.
Second, a field study was conducted to determine the effect of operating
variables on visible emissions (opacity) at one facility. Ninety days of hourly
readings of process and opacity data were collected by the MWC staff. An attempt
was then made to statistically correlate the process and opacity data.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
Air Pollution Emissions
Municipal Waste Incinerators
Control Equipment
Operation and Maintenance
Visible Emissions
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (This Report)
21. NO. Oi- PAGES
2O. SECURITY CLASS (Thispage)
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------
------- |