United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Washington DC 20460
EPA 34071-86-021
August 1987
Stationary Source Compliance Series
Technical Assistance
Document:
Recommended
Recordkeeping
Systems for Air
Pollution Control
Equipment
Part I — Paniculate
Matter Controls
-------
EPA 340/1-86-021
Technical Assistance Document:
Recommended Recordkeeping
Systems for Air Pollution
Control Equipment
Part I — Paniculate Matter Controls
by
PEI Associates Inc.
11499 Chester Road
Post Office Box 46100
Cincinnati, Ohio 45246-0100
Contract No. 68-02-3963
Work Assignment No. 112
EPA Project Officer: Louis R. Paley
EPA Work Assignment Manager: Kirk Foster
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air Quality Planning and Standards
Stationary Source Compliance Division
Washington DC 20460
August 1987
-------
DISCLAIMER
This report was prepared for the U.S. Environmental Protection Agency by
PEI Associates, Inc., Cincinnati, Ohio, under Contract No. 68-20-3963, Work
Assignment No. 112. The contents of this report are reproduced herein as
received from the contractor. The opinions, findings, and conclusions
expressed are those of the authors and not necessarily of the U.S. Environ-
mental Protection Agency.
ii
-------
CONTENTS
Figures iv
Acknowledgment v
1. Introduction 1
2. Purpose of Recordkeeping 3
Regulatory requirements and recordkeeping 4
Recordkeeping and maintenance programs 5
Elements of a recordkeeping program 9
3. Multicyclones 17
Operating records 18
Maintenance records 21
4. Fabric Filters 23
Operating records 24
Maintenance records 30
5. Venturi Scrubbers 34
Operating records 35
Maintenance records 39
6. Electrostatic Precipitators 41
Operating records 43
Maintenance records 49
Summary 53
Appendix A. Example work order form 55
iii
-------
FIGURES
Number Page
1 Example of catalog system used by a large pulp and paper
mill 12
2 Bag failure location chart 31
iv
-------
ACKNOWLEDGMENT
This report was prepared for the U.S. Environmental Protection Agency by
PEI Associates, Inc., Cincinnati, Ohio. Mr. Louis Paley was the Project
Officer and Mr. Kirk Foster was the Work Assignment Manager. Mr. David
Dunbar served as the Project Director and Mr. Gary Saunders was the Project
Manager. The senior technical advisor was Mr. Ron Hawks. The principal
author was Mr. Gary Saunders. PEI Associates, Inc., wishes to thank Mr. Kirk
Foster for his guidance and direction on this work assignment.
-------
SECTION 1
INTRODUCTION
One of the many requirements often specified by control agencies for
regulated air pollution sources are recordkeeping and reporting of selected
operating data, excess emissions, and certain maintenance activities. Many
times the recordkeeping and reporting requirements are vague and the specific
format of requirements are not specified. This reflects, in part, the site-
specific nature of recordkeeping activities and uncertainty associated with
knowing what data are useful. As a result, there is a need to establish what
minimum data are needed to evaluate the overall performance of a source and
correspondingly its compliance status. It is the purpose of this manual to
provide minimum recordkeeping considerations for several control equipment
categories. The minimum recordkeeping guidelines presented in the manual
will contain several "mandatory" parameters that should be monitored along
with additional "optional" parameters that may provide useful information.
There will, however, still be situations where flexibility is needed with
respect to what parameters should be recorded and reported to the air pollu-
tion control agency.
Operating parameters and maintenance activity records should be a part
of any preventive maintenance program. They should not be an afterthought or
just kept to fulfill some legal or regulatory requirements. Good records are
an important part of a feedback loop in determining whether preventive main-
tenance is being performed in a timely manner or if operation and maintenance
practices need to be changed. Operating records can also be used to diagnose
performance or changes in performance affecting the compliance status of a
source that require some action or intervention. Records may also be used to
establish representative conditions during stack testing or to establish
baseline conditions when changes in operating practices are planned.
The major problems with recordkeeping are deciding what records should
be kept, in what format, how often the data should be gathered, and how it
1
-------
should be analyzed. The amount of paperwork associated with recordkeeping
activities can, in some cases, be substantial, which can be a serious draw-
back to the effective implementation of a recordkeeping and reporting pro-
gram. There are many approaches to recordkeeping ranging from a simple
logbook to computerized work order/spare parts reporting systems. The type
of recordkeeping system chosen depends on the type and complexity of the
equipment as well as resources available to manage and operate the system.
This manual discusses the types of records that should be maintained for
major particulate matter control systems—multicyclones, fabric filters, wet
(venturi) scrubbers, and electrostatic precipitators (ESP). In addition,
this manual also discusses (through examples) the types of parameters that
should be monitored for selected processes such as asphalt concrete plants,
industrial boilers, and municipal incinerators. In many cases, process
variables have a significant impact on the overall performance of the control
equipment. Therefore, any recordkeeping system that overlooks process
related data may be incomplete or ineffective in diagnosing equipment per-
formance. Examples are provided that make use of combined process and con-
trol equipment data to illustrate the relationship between process and con-
trol equipment parameters. The difficulty in specifying process related
parameters is that there are usually far too many parameters to be monitored
and that specifying all of the potential monitoring areas may prove too
burdensome and yet still not provide the degree of protection desired.
Although some recordkeeping activities are part of a statutory require-
ment, the emphasis of this manual is on recordkeeping as part of a preventive
maintenance program. With an increased emphasis on maintaining long-term
compliance with all regulatory emission limits, there is generally a need for
a preventive maintenance program that uses an effective inspection and
recordkeeping system. With greater interests on toxic air pollutants and
continuous compliance, the role of recordkeeping is expected to be a signif-
icant part of the overall air quality management program in the future.
-------
SECTION 2
PURPOSE OF RECORDKEEPING
The purpose of recordkeeping and any associated reporting requirements
is often perceived differently by various individuals depending on their
circumstances. For some, recordkeeping is merely a regulatory requirement
that must be fulfilled. For others, recordkeeping is just a nuisance with no
real purpose. On the other end of the scale there are those who want to
maintain records on "everything11 just in case they are ever needed in the
future. Although the above observations represent extreme points of view,
they are routinely encountered, sometimes within the same source.
The primary purpose of recordkeeping is not to merely fulfill regulatory
obligations or to make what appears to be excessive demands on personnel and
resources. Instead, a recordkeeping program should be part of a larger
preventive maintenance program that enhances the long-term performance of the
associated equipment (process and/or control equipment). Unfortunately, like
most preventive maintenance programs, the recordkeeping program that must be
established is very site-specific and considers a number of factors such as
the adequacy of the design (redundancy), instrumentation, access for mainte-
nance, and personnel requirements and availability. The same types of deci-
sions that make the development of a preventive maintenance plan difficult
are present in designing a recordkeeping program: how much data are needed,
how often should the data be recorded and summarized, who will be responsible
for gathering the data, who will analyze it, and how will it be analyzed.
Personnel responsible for development of recordkeeping programs as part of
the preventive maintenance program are usually trying to establish a balance
between having enough versus having too much data. In this situation, the
recordkeeping program serves as a tracking mechanism and helps provide the
necessary data to evaluate the effectiveness of the overall preventive main-
tenance program.
-------
2.1 REGULATORY REQUIREMENTS AND RECORDKEEPING
As mentioned previously, recordkeeping practices may be dictated by
certain regulatory requirements. Some agencies (not related to the control
of air pollution) have very specific information that is required on produc-
tion levels and personnel. In the air pollution control field, very few
specific data are required by regulations. In general, recordkeeping and
reporting requirements typically emphasize the collection of continuous
monitoring data, some production data, and excess emission reports. The
Clean Air Act and State Implementation Plan (SIP) regulations codified into
40 CFR 51 give the State agencies the authority to require monitoring and
recordkeeping as part of their program operation, but those powers have not
been used extensively. Recent emphasis on upgrading the content of State
construction and operating permits as well as those permits issued under
prevention of significant deterioration (PSD) regulations have increased the
overall recordkeeping and reporting requirements for certain sources. Many
of these requirements have been imposed in an effort to maintain continuous
compliance.
Although these permit requirements are generally more specific than
found in the general regulations, they quite often do not require reporting
of the data that are to be monitored (e.g., pressure drop) unless the param-
eter that is to be monitored exceeds some threshold value. At that time, the
parameter would provide an indication that some operational change or mainte-
nance is necessary. It should be noted that all values should be recorded
and analyzed not just those that exceed the threshold value. An analysis of
all the values enables one to establish a trend or pattern in the data, which
can be very useful in diagnosing the problem rather that just indicating that
a problem exists.
Care should be taken, however, to ensure that the data are not being
collected because "its nice to have." The data required should be useful to
both the agency and source in analyzing equipment operation and performance.
Operating data that cannot be used for this or any associated purposes should
not be required (although it may be monitored and even reported for other
purposes not related to air pollution regulations).
In addition to operating data and performance related parameters, the
regulations often allow recordkeeping practices to be specified as needed.
-------
Again, some regulations require that specific maintenance activities be
documented (e.g., continuous emission monitor zero and span checks, cleaning
downtime, etc.).. Most regulations, however, are not specific in this regard
and agencies are accorded some discretion as to what maintenance records
should be kept.
Regulatory recordkeeping requirements cannot ensure that compliance with
the applicable emission limitations will occur. Equipment design and opera-
tion and maintenance practices will have the greatest effect on continuous
compliance with the emission limitations. Recordkeeping is merely a tool
that can aid in evaluating the overall performance of the source.
2.2 RECORDKEEPING AND MAINTENANCE PROGRAMS
Maintenance programs can be broadly divided into either preventive
maintenance or breakdown (as needed) maintenance. Preventive maintenance is
geared towards anticipating and preventing failure of operating components to
provide "continuous" operation. Breakdown maintenance responds to equipment
failure when it occurs and demands response when performance is adversely
affected. Recordkeeping is normally thought of as being associated with
preventive maintenance and less often with breakdown maintenance. Most
effective maintenance programs, however, use elements of both preventive and
breakdown maintenance to adequately respond to the day-to-day situations that
occur in the plant environment. A key element in these programs is the
method of recordkeeping and records analysis. By definition, preventive
maintenance seeks to identify avoidable failures and prevent them from happen-
ing by routinely maintaining the equipment. The preventive maintenance
program can either be directed by a schedule that does not rely on analysis
of operating data (but may be dictated by operating history) or by interpret-
ing operating data to determine if and when maintenance may be necessary
prior to catastrophic failure of equipment or components. In this case,
records help determine what action is needed and when.
In situations where preventive maintenance is impractical and the pre-
dominant form is breakdown maintenance, recordkeeping is not without its
uses. For example, routine inspection and operating data may indicate that
problems are occuring or that a failure may be imminent and that maintenance
will be necessary. This may provide lead time to assemble the necessary
-------
spare parts and schedule personnel so that when a failure occurs (or a shut-
down occurs for some other reason), the maintenance will be quickly and
efficiently performed. It is not unusual to see a mixture of preventive and
breakdown maintenance at a facility because not all equipment warrants the
same type of maintenance procedures.
The elements of preventive maintenance programs are often justified on
the basis of reduced operating and production costs (cost/unit). This in-
cludes the cost of gathering data and reducing it as well as the cost for
components and personnel to implement the various phases. In general, where
the cost of preventive maintenance and breakdown maintenance are nearly
identical, preventive maintenance may be difficult to justify although in-
spection and recordkeeping should be retained. Where costs are unexpected
but preventable and malfunctions are severe, the use of preventive mainte-
nance will be favored and, as mentioned previously, recordkeeping is an
important part of the preventive maintenance program.
Unfortunately, for air pollution control, the cost of emitting pollu-
tants short of a major equipment failure does not easily figure into the
"cost" justification for a preventive maintenance program. This is particu-
larly true with toxic pollutants where the cost of an accidential release may
be substantial in terms of human health effects, but not easily quantifiable
in terms of monetary value. Even in the absence of a formal preventive
maintenance program, however, recordkeeping can and should be used to assist
plant personnel in evaluating equipment performance and to establish good
maintenance procedures.
The following two examples 1) show how good recordkeeping practices may
be used to diagnose the existence of problems by presenting the changes that
occurred over a period of time, 2) assist in the selection of maintenance to
be performed, and 3) aid in the selection of the most cost-effective method
of maintenance. These two examples, however, serve to illustrate a very
important point: although better than adequate recordkeeping existed at both
plants, compliance problems existed because plant personnel did not take the
necessary time to examine the data available to them and to correlate the
data with the observed results. The existence of a recordkeeping program
does not guarantee continuous compliance with the emission limits, just as
the lack of recordkeeping does not mean that noncompliance will occur. For a
-------
recordkeeping program to be effective, however, it must be integrated into a
program where essential data are gathered and interpreted to optimize the
preventive maintenance efforts.
Example 1: An asphalt plant near a residential area has been the source
of complaints for several months because of excess emissions from the
stack. Stack opacity is very high and the scrubber is unable to comply
with emission limits established for the source. There is no routine
preventive maintenance program at the source and all work performed on
the system has been breakdown maintenance. The plant does routinely,
however, monitor certain parameters such as scrubber pressure drop and
occasionally measures water flow from the pump. In addition, invoices
for major purchases are available to provide some indication of the type
of maintenance that has been performed.
Examination of the records available indicated that the problem had
actually begun some 2 years prior when pressure drop across the scrubber
began to drop despite closing the scrubber throat and increasing the
scrubber water flow from 250 gal/min to 450 gal/min. The scrubber had
been able to achieve a 26 in. H?0 pressure drop but 2 years later could
only achieve 17 inches. In addition, repair records indicated that the
fan shaft had been replaced three times because of damage from fan in-
balance ($4000 each). The fan also showed considerable erosion and
would require the replacement of wear plates built on the fan blades.
Access to the scrubber throat was provided by an access hatch above the
variable throat held to the scrubber by 12 bolts. The access hatch had
not been opened in the 5 years that the scrubber had been operating.
With the plant shutdown and cool, the hatch was opened and the variable
throat examined. It was found that the plate used to control throat
area and, hence, pressure drop had eroded severely. Examination of the
scrubber arrangement indicated that gas flow and particulate matter
loading was biased heavily towards one side of the scrubber because of a
design deficiency in the duct work. The problem could be corrected by
redesigning and reconstructing the duct work (at a cost of several
thousand dollars) at the scrubber inlet. A more suitable solution was
to replace the scrubber's variable throat, which cost approximately $75,
and to routinely visually check the plate at least once per year for
wear and replace the plate as necessary. This restored scrubber per-
formance had also reduced the problems the plant was experiencing with
the fan. In this case, the data were adequate to suggest the source of
the problem and ultimately led to a reasonable, low-cost solution, if
only time was taken to review the available information.
Example 2: In this example, cost became a compelling factor but some
costs were ignored because they could not be easily quantified. A
manufacturing facility operated three coal-fired boilers to produce
steam. Typically, only two of the three boilers operated (each was
rated for 100,000 Ib steam/h or approximately 120 x 10° Btu/h heat
-------
input) to produce an average 85,000 Ib steam/h each. All three stoker-
fired boilers were ducted to a large four-compartment pulse-jet fabric
filter containing 210 bags/compartment. The bags were felted fiberglass
that cost $65/bag.
This facility had a long history of bag problems. Typically, bags were
changed approximately every 90 days because the boiler opacity was above
the 40 percent limit allowed by the State agency. In fact, it was not
unusual for the boilers to be routinely violating the emission standard
during the previous 30 to 45 days before the bags were changed. After
the bags were changed, pressure drop across the fabric filter would
gradually increase from about 4.5 inches to 14.5 inches over a 30-day
period and then dropped to approximately 10 to 12 inches as the bags
failed. Unfortunately at this pressure drop, the fan had reached the
fan limit and the boilers were unable to produce additional steam if
there was a demand above these operating levels (more common in winter).
Thus, production and/or heating would suffer because of the inability to
produce more steam when needed.
Continuous recorders were available for all parameters needed such as
steam flow, pressure, temperature, boiler oxygen after the economizer
and after the air heater, boiler CO levels, flue gas temperature, pres-
sure drop, and opacity. These same parameters and others were also
recorded hourly by the boiler operator on the boiler log sheets. Plant
personnel were aware that combustion problems probably contributed to
fabric filter failure. Boiler oxygen was exceptionally high at 14.0
percent 02 at a temperature of 385°F. The energy loss from this opera-
tion was substantial as lower 02 and temperatures were possible.
However, the major evidence concerning combustion was the high CO levels
in the flue gas and the quantity of unburned carbon in the fly ash. The
heat content of the flyash was 7,800 Btu/lb (the coal heat content was
only 13,000 Btu/lb) and indicated that substantial quantities of un-
burned coal were entering the fabric filter.
The symptoms exhibited indicated that poor combustion was a major con-
tributing factor to the failures experienced with the fabric filter. In
addition, reduced efficiency because of carbon loss meant more coal had
to be burned, which generated more particulate matter and S02 than
origially estimated. Although poor air distribution could contribute to
poor combustion, the problem was discovered after examining the coal
supply. Coal properties were such that very fine coal was being fed to
the spreaders. Unfortunately, spreader stokers are very sensitive to
fuel size. As a result, there was poor combustion and the carbon carry-
over, as a result of the poor combustion, blinded the bags.
In this situation, the most obvious and effective solution was to change
coal size characteristics from the very fine size noted in the pile to
the i x li distribution needed for proper operation. Based on operating
data, the stack loss included a 12.4 percent stack loss (14.0 percent 0?
at 385°F) and a combined combustible loss of 11.91 percent (2.78 for CO;
9.13 for carbon in the fly ash). By changing coal size, the boiler
efficiency was expected to improve substantially by burning CO and
-------
carbon in the boiler rather than losing them to the atmosphere. There
would also be an opportunity to reduce both oxygen content and flue gas
temperature to further improve efficiency and to reduce the uncontrolled
emission rate of particulate matter and S0? from the boiler. An esti-
mate of the improvement would raise the boiler efficiency from 73.5
percent to 84.6 percent at the improved combustion levels.
However, the change in coal quality caused the cost per ton of coal to
increase. The poor quality coal was delivered at a cost of $30/ton and
the double-screened coal cost $37/ton and could quite possibly increase
to $40/ton if the plant burned this coal instead of the coal being
supplied. The plant realized that this would increase the operating
costs for the boiler even with the improvement in boiler efficiency. A
complete analysis, however, had not been performed to quantify the cost
of changing coal. A more complete analysis indicated that at $37/ton,
the boiler would cost more to operate but when the cost of fabric filter
bag changes were included ($60,000), the more expensive coal yielded a
savings of $16,900 over each 3-month period. At $40/ton, however, the
cost resulted in a net increase of $36,375 per 3-month period giving the
appearance that it was more desirable to burn the cheaper coal and
replace the bags quarterly as the plant had done previously. Although
other factors were overlooked such as increased wear on other components
(fans, conveyors), increased cost of moving waste materials, and loss in
production due to steam limitations, the most obvious item missing was
the economic beneift derived by operating 30 to 45 days out-of-com-
pliance with emission limitations. If the plant had been forced to
replace the bags once every 30 to 45 days when bag failures began to
significantly affect emissions, the economics would have, without ques-
tion, been more compelling for the source to change coal.
2.3 ELEMENTS OF A RECORDKEEPIN6 PROGRAM
A recordkeeping system should not be established in a haphazard manner.
To the extent possible, a recordkeeping program should be designed to provide
the necessary level of useful information with a minimum amount of personnel
resources to complete the necessary checks and paperwork. In a preventive
maintenance program, it is often the paperwork that receives the least atten-
tion because it is the most cumbersome daily activity to complete.
A recordkeeping program should contain five basic items which, in some
cases, may add to the existing paperwork load. These items are an equipment
record, a checklist, an inspection schedule, inspection report, and equipment/
maintenance cost record. Some of these items can be combined to fulfill
several needs. For example, operating logs can be used for scheduling, and
reporting if certain key parameters are selected for routine monitoring. The
need to obtain complete and accurate information must be balanced by the
-------
necessity to keep the recordkeeping system simple to maintain. This will
require careful evaluation by responsible personnel regarding what data are
necessary and important and what is secondary or "nice to know" data. If
there is already a functioning recordkeeping system used by the plant, it may
be possible to modify it to supply the necessary information. If not, then
the opportunity to design a workable system exists.
2.3.1 Equipment Record
An equipment record, although considered essential to the establishment
of a preventive maintenance program, may be of limited use on a day-to-day
basis as the equipment record usually takes the form of specifications and
design parameters rather than recording a set of hourly values. A system
should be devised to catalog this information because even a small industrial
plant can have hundreds of subcomponents where the design information is
critical. Very often, this critical information becomes misplaced and when
the information is needed to determine the initial design conditions param-
eters, the data are unavailable.
Generally, the best method of keeping these data are to have a central
filing and retrieval system. In smaller operations, this may be an office
with a bookshelf or two devoted to the storage of the operations manuals
accompanied by drawings and blueprints kept in a separate but nearby file.
As plants get larger, however, the number of manuals and specifications that
need to be maintained becomes very large and a more sophisticated storage and
retrieval system is needed. In these situations, a "catalog" system works
well where documents are given a file number for later retrieval. The
catalog can consist of cards kept in a filing system according to process or
control equipment grouping. The catalog may be further subdivided into major
subassembly groups. To locate the necessary data, one must locate the major
grouping and subgroup, obtain the file number, and then go to that file
location to obtain the necessary data.
A variation in this procedure is to list all major components and sub-
assemblies under a category either on paper or by computer. Again, all that
is necessary is to look up the major heading and then the subcategory to find
the appropriate file number. In fact, either system allows the addition of
basic design information in the catalog that might save time and effort in
10
-------
finding the appropriate file. Figure 1 is an example of a catalog system
that is used by a large pulp and paper mill.
The equipment record should be kept up-to-date. If modifications to a
system are made (e.g., changing the number and size of tubes in a multi-
cyclone) this information should be reflected in the equipment record. Old
or out-of-date information should be removed from the system.
2.3.2 Inspections and Monitoring
As part of a preventive maintenance program, one of the first decisions
that must be made is what items will be inspected and/or monitored? Items
that will be inspected need some type of form or method of reporting what was
found whether it is numerical value or merely an acknowledgment of proper
operation. The basis for deciding what will be routinely inspected and
monitored is usually determined by its relative importance to the plant
operation. For example, items (and parameters) that are essential to the
operation of the process or the plant and whose failure would cause sub-
stantial damage or loss of production capability would naturally receive more
attention than those items that are easily replaced or of have minor con-
sequences in terms of the overall plant operation. Other items that may
receive more frequent inspection or continuous monitoring are expensive
components or items that can lead to more serious failures if problems are
not corrected in a reasonable time period.
The emphasis on cost may appear to ignore regulatory requirements that
are necessary in order to comply with the emission limitations. There is,
however, a cost associated with continuous compliance when a malfunction
occurs. This cost may be in the form of lost production or noncompliance
penalties or both. The routine inspection and monitoring of control equip-
ment operating characteristics and key process parameters helps ensure con-
tinuous compliance with the emission limits by providing information to plant
personnel to evaluate the situation and to respond as needed.
2.3.3 Inspection Reports
The inspection and monitoring program generates the need to report or
record data, to verify that proper items have been checked, and that any
needed maintenance is scheduled. The report may be only a checklist to
indicate that certain items were inspected or it could include important
11
-------
IS - LIME 4 SALT CAKE STORAGE, and /5 KILN 4 CAUSTICIZING SYSTEM
supersedes 8-1-84 February 1, 1985
9-1-04
9-1-E1
7-48
7-1-84
7-1-8S
7-1-C1
7-1-85
7-1-C2
7-1-C3
7-1-El
7-1-E1
7-1-F1
7-1-F1
1567-63
1567-64
1567-70
1567-71
1567-80
1567-81
1567-82
1567-90
1567-91
1567-92
1567-93
1568-00
1569-00
1570-00
1571-00
1572-00
1572-01
1572-02
1572-03
1572-05
1572-10
1572-11
1572-12
1572-13
1572-14
BURNER, tS kiln, flame safeguard system
n PUMP, o<1 for tS kiln, Roper size 1H15, S/N F3033,
20 hp, 1200 rp« :
FAN, primary air, /S kiln, Lanson Slower model 515, 5-stage,
" 550 scfa 3 5 pslg, Stoddard silencer model F64-5, 30 hp,
CrauB 286T, 3600 rpn, rated 36.3A 3 460V
FAN, strong NCS to IS kiln burner
FAN, Induced draft, tS kiln, Champion, 76,000 cfn 3 9' S.P.
DRIVE, 1.0. fan, /S kiln, 200 hp, frame 5008, 900 rpm, rated
246A 3 460V, Emerson variable frequency drive model A.S.4470
DRAFT CONTROL, 1.0. fan, /S kiln
CONVEYOR, belt, /6 4 /7 filters to cross screw conveyor,
20* belt, 350 fpm, 5 hp, 1800 rpm, Falk S2-50FZ2M, 20.52 ratio
CONVEYOR, screw, lime mud cross to feeder, 22' dia x 9'-5' long,
29.0 rpm, 50 hp, frame 326T, 1800 rpm, rated 59. SA 3 460V,
Falk 1080FZ2A, 25.63 ratio
CONVEYOR, screw, lime mud feeder to /S kiln, 22' dia x 11'-2' long,
29.0 rpm, 50 hp, frame 326T, 1800 rpn, rated 59.5A Q 460V,
Falk 1080F22A, 25.63 ratio
VIBRATOR, 11me mud, tS kiln
ENCLOSURE, IS kiln precipitator, F.L.Smidth, 2 units (north/south),
3/16* steel walls, each unit consists of two sections 22' high x
16-1/2' wide x 20' long, design flow 76,000 cfra 9 9' H20,
efficiency 99.88*
PLATES, JSklln precipitator, 19' 8-1/4' nominal height, l'-9" wide,
16 ga mild steel, 35,120 sq Ft collecting surface area
STACK 4 BREECHING, /S kiln, 4'-6" O.d., 3/16" 3161 S.S. stacks,
one each precipitator, 3/16" mild steel breeching
DAMPER, precipitator inlet, south side, F.L.Smidth disc valve C,
2 hp, 1200 rpm, rated 3.7A 9 440V
DAMPER, precipitator \nlet, north side, F.l.S-.idth disc valve C,
2 hp, 1200 rpm, rated 3.7A 9 440V
HIGH VOLTAGE LINES, SWITCHES S INSULATORS, .'5 kiln precipitstor
tt TRANSFORMER/RECTIFIER, *S kiln precipitator, south side
outlet, 33.6 KVA, 440 VAC input, 86,000 '/DC output,
NWL transformer model 28319 •
t2 TRANSFORMER/RECTIFIER, *5 kiln precipitator, south inlet,
33.6 KVA, 440 VAC input, 88,000 VOC output,
MM. transformer model 28319
/3 TRANSFORMER/RECTIFIER, /5 kiln precipitator, north outlet,
33.6 KVA, 440 VAC input, 88,000 VOC output,
NUL transformer model 28319
/4 TRANSFORMER/RECTIFIER, tS kiln precipitator, north inlet,
33.6 KVA, 440 VAC Input, 88,000 VDC output,
MM. transformer model 28319
Figure 1. Example of catalog system used by a large pulp and paper mill
12
-------
numerical information along with a narrative so that the findings are made
clear. It is usually at this point that many recordkeeping systems begin to
experience difficulty because the paperwork becomes the most demanding part
of the preventive maintenance program. The problem is not so much of one of
being unwilling or unable to perform the routine monitoring and inspection of
key parameters or equipment as it is in analyzing what the information means,
what actions are required, and reporting the findings to the appropriate
individuals within the organization. Strip charts can record key parameters
and people can follow a checklist and fill in forms where data are requested
but these are of little use if they are not in a format that may be easily
analyzed or if they are not examined at all. That is why plant personnel
should carefully examine what parameters and checks are most useful and
carefully select those items to be monitored.
Inspections and reports may also be performed in conjunction with
scheduled or requested maintenance as a follow-up to verify correction of a
noted problem. The decision to include this information with the "normal"
inspection reports or to include it only in the equipment/maintenance record
will depend in part on how retrievable both elements of the system are and
the degree of detail provided. It is sometimes better to duplicate this
effort than to have the possibility of pertinent information being lost. For
example, if a problem is noted and maintenance is requested, it does per-
sonnel tracking the preventive maintenance program little good if there is no
way to report back the findings and actions taken. As a result, key informa-
tion that is needed to fine tune both the recordkeeping and preventive mainte-
nance program may be lost.
As with the inspection scheduling, the inspection reports and records
should be reviewed periodically (quarterly to semi-annually) to determine if
the information gathered is adequate. If there are areas in which data are
being gathered that are of little use, these should be eliminated. Any time
the level of paperwork can be reduced without affecting the overall quality
of the program it will, most likely, be appreciated by the people who have to
take the time and effort to provide the information. Reducing the "clutter"
of unneeded and unused recorded data may also allow or encourage personnel to
focus on those recordkeeping items that are of greater importance.
13
-------
2.3.4 Equipment Maintenance/Cost Records
The most workable system for a wide variety of facility sizes and types
appears to be a system based on job or work orders. For the smallest facili-
ties, a simple maintenance log with pertinent information (such as descrip-
tion of work performed, cost, and time required) may be adequate. The work
order system, however, has several advantages over a simple log book. First,
the work order provides written notification that maintenance has been re-
quested. Second, it can aid in the scheduling of maintenance personnel
particularly when the work is prioritized. Third, the work order should
provide a written record of what was found, what was done, and if any action
was taken to correct to problem, or whether further action is required (and
scheduled). For example, in an ESP, if a broken wire was suspected of caus-
ing the deenergization of a power supply, a work order might specify that the
cause of the ground should be found and removed. Then during a short boiler
outage, maintenance personnel would be scheduled to repair the unit. If they
found the cause was a random wire break, they would remove the wire, possibly
replace it, note its location, and document on the work order. If some other
cause was found (e.g., a warped plate that required removal of extra wires to
prevent breakage and scheduling of straightening for a longer outage), these
would be recorded as well as any additional maintenance performed or sched-
uled for a later date. Lastly, work orders can be sorted several ways for
accounting purposes. They can be sorted by labor craft, equipment type,
process, and by type of problem. Whether done by hand or by computer, the
work orders can summarize the results of a preventive maintenance plan by
showing if failures are occurring "randomly" or if the maintenance and cor-
rective maintenance are not addressing the proper causes of problems. The
cost of such problems would also be available to estimate the cost of chang-
ing the preventive maintenance program. Appendix A contains an example of a
work order form.
Example 3: A pulse-jet fabric filter on a small source works well, when
it is working. But every 45 to 60 days the bags have to be changed
because they become blinded and the pressure drop has gradually in-
creased from 4 inches when the bags are new to 15 inches just before the
bags are changed. The process dust caused by the restricted ventilation
rate causes a nuisance and many citizen complaints are received because
of the fugitive dust escaping from the plant. A complete bag change
14
-------
costs between $2,500 and $2,800. The plant routinely checks pressure
drop and cleaning system operation but is unable to stop the deteriora-
tion. A change in strategy and new options need to be considered.
These options include a modification to the process, a change of control
equipment, or a change in maintenance. Lets examine the problems more
carefully.
Plant personnel were aware that some changes were needed and at current
performance levels that excessive costs were being incurred. No process
changes seemed possible to alleviate the problem as the dust should not
have been causing the observed blinding problems. A change to a new
process might lower the uncontrolled emission rate without substantially
changing the particulate matter characteristics. A change from a fabric
filter to a scrubber was possible but considered undesirable because of
the water discharge regulations. The most obvious solution appeared to
be a continuation of the process of changing all the bags but at shorter
intervals to prevent the fugitive emissions and citizen complaints.
Unfortunately, the plant did not use the information available to dis-
cover the source of the problem and was only responding to a symptom
because the cause of the bag blinding problem was water and oil carry-
over from a pulse-jet air supply compressor that needed to be rebuilt or
replaced. For the cost of three complete bag changes, the compressor
could be replaced with a new one with an air dryer and the bag blinding
problems "solved." Plant personnel had the cost and performance data
available and were ready to make changes. But the key was to use the
data collected to find the cause of the problem rather than treating the
symptom. The plant's own data indicated they were treating the
symptoms.
Larger facilities that use or are familiar with a work order system will
find that the inclusion of air pollution control equipment into the system
will present very few difficulties. A few additional records and analysis
may be needed but the function of the system will be the same as for process
equipment. There will be the same need for a priority system and a method to
ensure that the findings are sent to appropriate departments and personnel.
For those just beginning a system, careful planning and adequate time and
training will be needed to ensure that the system is effectively implemented.
2.3.5 Summary
Recordkeeping can be a monumental task whether it is part of a preven-
tive maintenance program or not. The ability to retrieve and analyze the
data can present a significant logistics problem and requires careful plan-
ning to make the recordkeeping program effective. The differentiation be-
tween required data and "nice to know" data is critical in attempting to keep
the paperwork at a minimum. The following sections for the various types of
15
-------
control equipment provide examples of the type of information that should be
considered and, indeed, may be required in a permit or to operate an effec-
tive preventive maintenance program. The list is not all inclusive and
special conditions and exceptions can always be found but, hopefully, most
situations are addressed.
16
-------
SECTION 3
MULTICYCLONES
Of all the participate matter control equipment groupings discussed in
this manual, multicyclones probably represent the least complicated control
devices. Generally equipped with few moving components, these devices have
the potential for reliable service if it were not for two limitations.
First, multicyclones have a tendency to plug. Although some units see years
of operating service without problems, others need to be taken out of service
almost weekly for cleaning. Second, of all the control equipment types, the
multicyclone generally does not perform well when the gas volume through the
device decreases. The performance of a multicyclone is closely tied to the
volume of gas passing through it. Although there are designs that try to
reduce or overcome these limitations, most do not address'these problems.
The simplicity in design means that there are relatively few parameters
to be monitored for multicyclones. Unfortunately, the parameters that can be
monitored do not provide the best indication of control equipment perform-
ance. There is no direct measure of the amount of inertial forces applied
within the collector although this might be inferred from the gas volume.
The parameters available to be monitored include:
Operating
Pressure drop
Temperature
Fan motor current
Dust discharge operation checklist
Opacity
Maintenance
Gasket replacements
Tube replacement
Vane replacement
17
-------
Smoke-bomb tests
Tube pluggage
It should be noted that although opacity is included on the multicyclone
opearting parameter list, it is not an effective parameter to be monitored as
it is for other types of control devices discussed later. In general,
opacity is of little use for multicyclones because they are usually not
capable of capturing small size particulate matter in the light scattering
range. Thus, substantial changes in particulate matter emissions may occur
without any apparent change in opacity. Conversely, a change in opacity may
mean that a process change has occurred, which changes the particle size
distribution entering the collector. Notwithstanding this limitation,
opacity may be worthwhile to record in that it will indicate a change in fuel
or process conditions that tend to generate excessive quantities of fine
particulate matter, which readily pass through the multicyclone leading to
opacity violations and possible mass emission standard violations.
3.1 OPERATING RECORDS
Parameter Measured: Pressure drop
Applicability: All sources
Methods of Measurement: Manometer, magnehelic, and continuous strip
chart recorders
Limitations: Pluggage of pressure taps, will not necessarily indicate
tube pluggage problems
The most common monitoring parameter for multicyclones is pressure drop.
Pressure drop provides an indication of gas flow through the multicyclone
because what is being measured is the resistance to flow through the turning
vanes and tubes, which provides an indication of the amount of force being
applied to remove particles from the gas stream. Within limits, a higher
pressure drop means more removal force and higher collection efficiency.
Pressure drop will vary with gas flow as increases or decreases in gas volume
through the collector will correspond to increases or decreases in pressure
drop.
Pressure drop may also be a useful indicator of pluggage of the turning
vanes. In theory there should be an equal gas flow through each tube and
18
-------
blockage of the tube inlet turning vanes will force gas to travel through
other tubes and increase the gas volume (and velocity) through each tube.
This will, inturn, increase the pressure drop. It takes, however, a number
of tubes to become blocked before there is a noticeable increase in pressure
drop, which would have to be coordinated with some other parameter such as
gas volume and temperature to determine if pluggage is occurring. Normal
variations in gas volume will produce larger shifts in pressure drop than
isolated tube pluggage.
Pluggage of the bottom of a multicyclone tube, which is common, or the
outlet tube, which occurs less frequently, will rarely be noticed as a result
of a change in pressure drop because 80 to 85 percent of the pressure drop is
generated across the turning vanes. Therefore, it is possible to have most
of the multicyclone discharge tubes plugged and not notice a discernible
change in pressure drop.
Some operations monitor pressure drop on a continuous basis (e.g., some
industrial boilers and other boiler applications of multieyeTones). Most,
however, only periodically monitor the values. Minimum time periods should
be once per day. If existing process parameter monitoring allows for more
frequent monitoring and recording than once a day, more frequent monitoring
should be conducted.
Parameter: Temperature
Methods of Measurement: Thermocouple
Applicability: All high temperature sources
Limitations: None
Temperature problems are usually not associated with multicyclones. Two
areas of concern, however, are excessive temperature that affects some
gaskets used to seal the tubes and tubesheets and the possibilities of low
temperatures causing sticky particles that lead to or aggrevate pluggage
problems. High temperatures may come from process operation or from hopper
fires (combustion sources). Hopper fires may also result in other problems
such as tubesheet warpage and annealing of the steel in the tubes and turning
vanes resulting in exceptionally fast abrasive wear.
Temperature monitoring should generally be continuous at the multi-
cyclone outlet. The outlet should be equipped with temperature alarms to
19
-------
alert operator of excessively high or low temperatures in the collector. In
addition, any periods of high or low temperatures should be noted in a log
book or on the operating records. Temperature will also affect pressure drop
and fan motor current.
Parameter: Fan motor current
Method of Measurement: Ammeter (inductance- type)
Applicability: Moderate to large sources with substantial variations
in gas flow
Limitations: Electric drives only. Steam driven fans not measureable.
Fan motor current provides an indirect measurement of the gas flow rate
through the system. Higher gas volumes usually mean high amperage drawn by
the fan motor. Even with different damper arrangements, this is qualita-
tively true. This is useful in determining if an observed change in pressure
drop is caused by a change in gas flow or from pluggage within the multi-
cyclone. It should be noted that gas temperature affects the gas density,
which will affect both pressure drop and fan motor amperage. An increase in
gas temperature will decrease gas density and increase actual gas volume,
which may reduce pressure drop and the amperage drawn by motor. Thus, read-
ings should be normalized to a constant temperature [i.e., Amp (70°F) =
measured amps x — TO — where T = temperature, °F)].
Fan motor current may be measured and recorded constantly by a strip
chart recorder. The minimum recordkeeping requirement, however, should be
once per day whenever pressure drop is monitored (or at every pressure drop
reading).
Parameter: Dust discharge operation checklist
Method of Measurement: Visual
Applicability: All sources
Limitation: Quantitative measurement is often not practiced
The proof of operation of any multicyclone is that material is being
collected and discharged from the device. Although a precise measurement of
the amount of material captured is not required, periodic checks of the dust
discharge should be conducted. This should include checking of the rotary
airlock, double gate valves, or other sealing devices for proper operation
and maintenance of all associated conveying systems. Minimum frequency is
20
-------
usually once per shift, although multicyclones with very heavy dust loadings
should probably be checked more frequently to avoid serious hopper pluggage
problems. If hopper vibrators are used to help remove dust from the hoppers,
they should also be checked during routine inspections.
3.2 MAINTENANCE RECORDS
Work order records should provide adequate information on maintenance
work performed on multicyclones. The following items should be included if
maintenance in these areas is performed. Of course, records from each in-
ternal check should always be maintained.
Maintenance Parameter: Gasket replacement
Applicability: All sources
Limitations: None
Gaskets fail because of age, excessive heat, or improper installation.
The gaskets between the dirty side and the tubes are particularly susceptible
to failure if not properly installed or maintained. The location of all
failed and replaced gaskets should be maintained. Erosion of the tubesheet
where a gasket has failed is possible because of the pressure differential
between the two sides of the tubesheet. This may make adequate installation
of replacement gaskets difficult or impossible and may eventually require
replacement of the tubesheet or the complete "blanking off" of a tube from
the gas flow.
Maintenance Record: Tube replacement
Applicability: All sources
Limitations: None
Some sources are capable of operating for 20 years or more without
having to change tubes while others may only be capable of operating from 3
to 6 months before the tubes are severely worn. The wear characteristics are
very site-specific. Isolated worn tubes, however, may indicate a flow or
particulate imbalance problem in the collector, that may require modification
to extend collector life and to improve long-term performance. The location
and extent of tube wear as well as any tube replacement should be noted.
21
-------
Maintenance Record: Vane replacement
Applicability: All sources
Limitations: MuHicyclones with replaceable vanes
The same considerations that apply to tube replacement apply here as
well. A few multicyclone designs are equipped with replaceable vanes instead
of vanes that are cast into place.
Maintenance Record: Smoke-bomb tests
Applicability: All sources
Limitations: Cost
A smoke bomb test of a multicyclone is both expensive and time consum-
ing. It may also be the only way to document the presence or lack of tube-
sheet/hopper leaks or dirty-side/clean-side leaks. The location of any leaks
noted during such tests should be recorded. In addition, such tests will
also document the presence of tube pluggage. Again, pluggage problems should
be recorded and efforts made to eliminate the pluggage as soon as possible.
A smoke bomb test basically involves the closing (stoppering) of all
outlet tubes and dust discharge tubes. Smoke bombs are lit while the multi-
cyclone is pressurized to draw the smoke through any openings. Smoke is very
visible and a well sealed collector will show no leaks.
Maintenance Record: Tube pluggage
Applicability: All sources
Limitations: None
Periodic checks (when the collector is off-line) should be performed for
indications of tube wear and pluggage. Any pluggage should be noted in the
records giving the tube location and the pluggage location (turning vane,
dust discharge, outlet). Most often tubes near the wall see the greatest
pluggage. The frequency of such checks will depend on a number of factors.
Those exhibiting few problems may only need periodic checking once or twice
per year while others may need to be checked once per week. A good starting
point is monthly checks that are adjusted on an as needed basis.
22
-------
SECTION 4
FABRIC FILTERS
Of all the particulate matter air pollution control devices, the fabric
filter probably receives the widest application in terms of the range of
source types and sizes. Fabric filters range in size from two to four bag
units to greater than 2,000 bag units and with proper fabric selection they
can be operated at temperatures ranging from ambient to 550°F. The amount of
available cleaning energy supplied as well as inlet particulate matter load-
ing will help determine the physical size requirements of the fabric filter.
All fabric filters rely on the same method of operation to remove
particulate matter from the gas stream. The fabric provides the support
material for the establishment of a dust layer or dust cake that performs
most of the filtration. During operation, this dust layer increases in
thickness and, thus, increases pressure drop across the fabric filter.
Periodically the dust cake must be removed by the cleaning system, which may
be categorized as either pulse cleaning, reverse air, or shaker. The low
energy systems (shaker and reverse-air) require lower air-to-cloth ratios
(acfm/ft2 fabric) typically in the range of 1 to 2 acfm/ft . The pulse
cleaning systems provide a greater amount of cleaning energy and they usually
2
have higher air-to-cloth ratios (4 to 12 acfm/ft ). All fabric filters are
sensitive to the process operation and dust characteristics. As such, the
records obtained must be coordinated with appropriate process data.
The list of operating and maintenance related data and records that may
be used is limited. Some of the monitored information would be useful only
on larger or higher temperature sources but there are enough of these in-
stallations to warrant their discussion. Although the data are limited, the
information provided is very useful in evaluating performance and maintenance
considerations. These data include:
23
-------
Operating
Pressure drop
Temperature
Opacity
Fan motor current
Inspection Checklist
Cleaning system operation
Dust removal system operation
Maintenance Records
Bag replacement location
Cleaning system components
Dust discharge system components
Other miscellaneous components
There may also be other data that may be of use such as process data, bag-
failure analysis, etc., that may influence the long-term performance of the
unit, however, the data to be gathered must be determined on a case-by-case
basis.
4.1 OPERATING RECORDS
This subsection summarizes the key operatiing data that should be ob-
tained when a fabric filter is in operation. There may be several methods of
obtaining these data and site-specific factors will govern the method
selected.
Parameter: Pressure drop
Applicability: All installations
Method of Measurement: Manometers, magnehelic/photohelic gauges,
pressure/electronic converters for continuous
readings
Limitations: Pluggage of pressure taps, leakage in pressure lines
Pressure drop is one of the more useful parameters that can be monitored
on a fabric filter. When taps are provided for the inlet and outlet static
pressure, the pressure drop across the bags provides an indication of the
resistance to gas flow and a relative indication of bag cleaness. Under
normal circumstances, pressure drop will range from 2 to 8 inches with 3 to 6
24
-------
inches H^O being most common. Most often, a substantial change in pressure
drop will indicate the existence of a problem. If the pressure drop de-
creases, it may indicate loose bags, holes in the bags, or problems with a
tubesheet. If the pressure drop increases, it may indicate a cleaning system
failure, blinding of the bags because of changes in the dust properties, and
even a dust removal system failure. Failures such as these are usually noted
in periods ranging from a few minutes to several days. Much more gradual
changes may be noted over weeks or months if bags are blinded due to exces-
sively high A/C ratios, fine or sticky particulate matter, or acid and
moisture dewpoint problems. The higher pressure drops observed may be used
to schedule bag replacement because the bag is reaching the end of its useful
life.
Example: A fabric filter with pulse cleaning was used to remove sander
and planes dust from the gas stream at a furniture plant. The A/C ratio
varied between 8.5 and 11.0 depending on the number of machines operat-
ing and the bag conditions. When the bags were new (filtering dust
layer established) the pressure drop was typically 3.5 to 4.0 inches.
Over a 6-month period, the pressure drop gradually increased to approxi-
mately 8.5 in. H20. Above 10 in. H20, flow was restricted to the point
where the hood systems would not efficiently pick up the dust (causing
fugitive emissions) and bag failures would increase from virtually none
to one to two per day. Thus, the plant scheduled bag changes every 6
months to minimize the energy required to overcome the pressure drop and
to reduce periods of noncompliance. Failure of the cleaning system
caused an increase in the pressure drop to be noted within several
hours. Holes in the bags caused both decreased pressure drop and high
opacity.
One of the problems associated with measuring pressure drop is pluggage
of the pressure taps, particularly the inlet tap. The taps should be cleaned
or periodically back-flushed with air so that valid data can be obtained.
Although there is no flow of gas and particulate matter in the static
pressure lines, there may be some deposition of moisture and particulate
matter in the instruments, which may require periodic changing of the
manometer fluid or cleaning the instruments.
Larger, multicompartmented fabric filters equipped with shaker, reverse
air, or plenum pulse cleaning may be equipped with continuous strip chart
recorders for overall recording of pressure drop. These charts are useful
for diagnostic confirmation that each compartment is isolated for cleaning.
25
-------
As each compartment is isolated, the pressure drop should increase and as it
is returned to service the pressure drop should decrease.
The use of individual manometers or gauges on each compartment is less
useful than an overall pressure drop reading. Each gauge will read nearly
the same pressure drop as the overall value because static pressure will tend
to balance over the system although flows may vary in each compartment. The
individual pressure readings are useful, however, for determining if the
compartment is properly isolated during cleaning (or for on-line maintenance)
because a no-flow condition should yield a "0" pressure drop reading.
Parameter: Temperature
Applicability: High temperature conditions (non-ambient)
Method of Measurement: Thermometer, thermocouple or thermistor for
continuous readout, temperature alarm and con-
trol
Limitations: Improper location due to stratification, slow response to
temperature changes, wrong thermocouple type for
electronic conversion
There are a wide variety of bag fabrics with a number of characteristics
that are important to the operation of the fabric filter. One of the most
important characteristics is the temperature limitation of the fabric. The
potential to operate the fabric filter above the maximum allowable fabric
temperature can cause concern about fabric life. The loss of fabric
integrity may result in pinholes, tears, or destruction of part or all of the
fabric in the bag(s).
Temperature indicator/recorders are used most often on "hot" sourses
(e.g., combustion sources or metal processing facilities). A temperature
monitoring device, however, may be recommended on any process that has the
potential to operate at 50°F above ambient temperature and within 50°F of the
permissible temperature limit of the fabric. The temperature monitoring
equipment should be located to measure gas temperature on the inlet of the
fabric filter. Temperature losses as great as 100°F have been noted between
the inlet and outlet readings in some situations because of inleakage and
heat loss from the fabric filter enclosure. Measurement downstream of the
collector may result in a false sense of security because bags may be sub-
jected to excessive temperature although there may be no indication of prob-
lems.
26
-------
The use of electronic temperature measuring devices with output displays
in the control room are most useful too as they provide the operator with
continuous temperature data. In addition, the temperature output may be
connected to an alarm circuit to provide warning of high temperatures and
potential damage to the bag. The alarm setpoint should be set high enough
not to be a nuisance but low enough to provide the operator with time to
react to the situation to prevent fabric damage. Generally, the temperature
setpoint will be 25 to 75 degrees below the maximum allowable fabric tempera-
ture. More sophisticated temperature monitoring systems can evaluate not
only the temperature but the rate of temperature change to govern alarm
conditions.
In addition to high temperature conditions, temperature monitoring
equipment may also be useful in situations where acid or moisture dewpoints
are of concern. In some applications, operation of the fabric filter below
dewpoint conditions may result in bag blinding and/or acid attack of the
fabric resulting in unacceptable performance.
Temperature monitoring is of little use on fabric filters that operate
at or near ambient temperature and collect a flammable dust. The monitor is
usually unable to detect a fire condition unless measurement is within or
downstream of the collector. Even then, as fire protection the thermocouple
would react too slowly to be of use.
Not all sources that can be defined as a "hot" source will elect to
continuously record inlet temperature. This may be suitable for sources that
experience little variation in operating temperature such as product dryers
that operate on a constant feed. The temperature readout, however, should be
prominently displayed for the operator and the temperature data should be
recorded each time other operating data are recorded.
Parameter: Opacity
Applicability: All fabric filter applications except some positive
pressure units
Method of Measurement: Reference Method 9, opacity monitor
Limitations: Reference Method 9 reading requirements may miss "puffs";
opacity monitor impractical for small sources. Must have
particles less than 2 microns to scatter light. Conden-
sation may cause false indications of problems.
27
-------
In a large number of applications, opacity observations or measurements
are useful indicators of fabric filter performance. As long as there are
particles up to the 5 to 8 micron range to scatter light, the gas stream
will be capable of exhibiting opacity if there are problems within the fabric
filter. The opacity measurement can be used to determine the presence of bag
leaks and, if analyzed carefully, the measurements can also identify the rows
or compartments where the leaks may be occurring.
In most fabric filter applications, the opacity that is visually ob-
served or measured with a transmissometer should be extremely low. The
measurements with a transmissometer, however, will be much more sensitive to
slight variations, particularly at low opacity levels. Observation of opaci-
ties above 10 percent generally indicate that some problems exist within the
fabric filter and that maintenance is required. Continuous elevated opacity
indicates that the leak could be caused by a bag tear, tubesheet leak, or
fallen bag whereas puffs associated with the operation of the cleaning system
may indicate the presence of pinhole leaks that reseal with the dust layer
between cleaning.
Opacity should be visually observed at least once per day. For plant
personnel involved in maintenance or preventive maintenance programs, it may
not be necessary to make observations in strict accordance with Reference
Method 9. What is important is observing the base opacity and any puffs
associated with the cleaning system operation. For sources equipped with
opacity monitors, the monitor should be allowed to record data on a "real-
time" basis and not a 6-minute average for at least one complete cleaning
cycle once per day. The presence of spikes should be correlated with com-
partments or rows of bags if they are present on the strip chart recorder.
For sources that do not have enough small particles to provide a
reliable opacity indication, other methods for evaluating performance are
necessary. This may include daily checks of the fabric filter outlet or
clean-side for excessive deposits. These checks should be made daily so that
loss of performance and damage to other bags can be minimized.
Parameter: Fan motor current (amps)
Applicability: All fabric filter systems with electric drives
Method of Measurement: Ammeter
28
-------
Limitations: Precise relationship between motor current and gas
flow requires fan curve. Electric drives only. .
The use of fan motor current to provide an indication of the gas volume
being moved through the exhaust system is becoming more common in facilities
where energy use and monitoring is of concern. Because there is a relation-
ship between fan motor current (and horsepower computed from the fan motor
conditions) where more current means more energy and (usually) more gas
volume through the system, it is a simple procedure to obtain a relative
indication of gas flow.
The measurement of fan motor current is better suited to sources where
temperature variations are minimal or occur over a long period of time.
Density changes (gas temperature changes) will influence the fan motor horse-
power use. Cooler gas stream density causes an increase in the required fan
motor horsepower to move a cubic foot of gas. Thus, comparisons of fan motor
current must be normalized to some reference temperature and wide variation
in temperature can make visual "quick checks" difficult.
Fan motor parameters provide confirming data when combined with fabric
filter pressure drop. Data on fan motor current should be obtained whenever
pressure drop is recorded. Plant personnel can use the combined data to
determine the degree of bag blinding or the presence of pinhole leaks, tears,
or excessive gas flow.
Parameter: Dust removal system operating checklist
Applicability: All systems except venting bins or silos
Method of Measurement: Visual
Limitations: None
Fabric filters not returning captured dust directly to a bin or silo are
usually equipped with hoppers that need some method of removing the captured
dust from the fabric filter. Unless equipped with manual dust removal, the
dust discharge system should be checked periodically to confirm the operation
of the system. Because it is generally recommended that dust discharge
systems be operated on a continuous basis, finding items such as stopped
conveyors or airlocks on the dust discharge will usually indicate that there
is some sort of problem. Some plants avoid the problem of not noticing a
failure of the dust discharge system by interlocking it with fan operation.
29
-------
In most cases, a visual check every 1 to 2 hours is all that is needed (mini-
mum of once per shift). A periodic check of the quantity of dust being
removed from the fabric filter, however, is helpful as any gross variation
from normal quantitites would indicate problems. Alternative methods of
evaluating performance would include weighing material discharged or measur-
ing the conveyor drive motor current.
4.2 MAINTENANCE RECORDS
Type of Record: Bag replacement
Applicability: All sources
One of the most telling indicators of proper operation and maintenance
of a fabric filter is in the number of bags replaced and the average bag
life. Most bags are designed to give 2 to 3 years of service before replace-
ment is contemplated. The actual life will vary depending on service condi-
tions. For each complete bag change, however, there should be a graphical
chart to show the location and type of bag failure as shown in Figure 2.
There are various approaches to keeping bag failure records depending on
the size of the fabric filter and the number of failures that occur. For
larger multicompartmented fabric filters, the record may be kept by compart-
ment to keep the diagrams manageable and/or by month if there are a signifi-
cant number of failures. For most installations, however, a diagram for the
entire fabric filter is practical and can, at a glance, help identify if a
design or operating problem is occurring. Although a narrative can explain
which bag was replaced, it is usually easier to show graphically where the
bag was damaged and the date of the failure.
Parameter: Cleaning system operating checklist
Applicability: All installations
Method of Measurement: Pressure gauge (pulse-jet), magnehelic or
manometer (shaker and reverse-air)
Limitations: None
The operation of the fabric filter cleaning system is one of the most
overlooked items observed today. Although the bags may be in place and
physically intact, many times the cleaning system will not be operating.
Generally, a failure in the cleaning system will result in an excessive
30
-------
BAG FAILURE LOCATION RECORD
A B C DBF 0 I I
* ooo ooo ooo *
:888 888 888:
»000 000 000 -
- OOO OOO OOO -
, OOO OOO OOO.
. ooo ooo ooo.
T ooo ooo ooo T
. ooo ooo ooo.
, ooo ooo ooo.
> OOO OOO OOO-
, OOO OOO OOO 3
.ooo ooo.
.000
ACCESS DOOR
MODOLZ MO.
RZPUCK - R CAP-CfT - C
PATCH - P MOTSICB - T
Figure 2. Bag failure location chart.
31
-------
pressure drop across the fabric filter. A source that fails to monitor
pressure drop, however, may not be aware of a failure in the cleaning system
unless its operations are routinely checked or plant personnel notice that
the exhaust gas stream is not being adequately removed because of the in-
creased resistance.
Pulse-jet fabric filters should be checked every 24 hours at a minimum
to ensure proper operation. Items to be checked include proper solenoid
firing and pulse air pressure. Misfires, weak pulses, and improper pressure
should be noted on the checklist for scheduling of maintenance. For shaker
or reverse air fabric filters, there should be a confirmation that dampers
are activating and that the cleaning system is operating. Because the A/C
ratio for reverse air and shaker fabric filters is considerably lower than
most pulse-jets, the need to check the cleaning system operation need not be
as frequent. Once per shift is usually adequate for most applications.
In addition to external checks of the cleaning system operation, in-
ternal checks of shaker and reverse-air fabric filters on a weekly basis
should be considered. Check points include the presence of fallen bags,
proper bag tension, and adequate shaking from shaking mechanisms.
An additional note for those fabric filters using compressed air to
activate dampers on shaker and reverse air or to clean the bags on pulse-jet
type fabric filters; water and/or oil in the compressed air lines can cause
problems. In winter, sufficiently low temperatures can cause the water to
freeze or an oil/water emulsion to become sticky causing a failure in the
cleaning system. For pulse-jet fabric filters, the situation is more damag-
ing because water/oil can cause bag blinding when blown into the bags. The
water/oil separators and airline dryers should be checked and serviced
periodically along with the compressor(s) as they may be physically separated
at some distance from the fabric filter. In addition, the oil that may be in
the compressed airline can attack the rubber components such as seals and
diaphragms, which cause them to fail as well.
Type of Record: Cleaning system and dust removal components
Applicability: All sources
In addition to the bags, the maintenance on the cleaning and dust re-
moval system is an important indicator of performance. The information to be
32
-------
recorded includes components replaced, reason for replacement, and date.
This may be recorded in a maintenance log or by using a work order system.
Excessive repairs may indicate the need for changing the design, operating
mode, or inspection and maintenance schedule.
33
-------
SECTION 5
VENTURI SCRUBBERS
There are numerous scrubber designs for removal of particulate matter
and gaseous emissions. The most commonly employed scrubber for the control
of particulate matter emissions is the venturi scrubber. Even within the
classification of venturi scrubber there are several different design types:
circular throats, rectangular throats, and fixed and variable throat designs.
Although there are a number of different designs, the basic operating princi-
ples remain the same.
The goal of the venturi scrubber is to capture the particulate matter on
water droplets that are easier to collect and remove. The particulate matter
is usually much smaller than the water droplets. The mean particle size for
particulate matter might be 5 to 10 microns, whereas water droplets may be 25
to 200 microns. The water droplets are usually easily separated in a cy-
clonic separator following the scrubber. The gas stream approaches the
scrubber throat and accelerates to maximum speed (5,000 to 15,000 cm/s) as
the cross-sectional area is reduced. Water is introduced at the scrubber
throat where it is atomized by the gas stream (water can be introduced by
spray nozzles, which form droplets that can be further atomized). The high
relative velocity of the particulate matter compared to the water droplets
causes the particulate matter to impact the droplets. This mechanism works
well for particles above 2 microns. For particles less than 0.1 micron a
diffusion mechanism predominates to remove very fine particles. Between 0.1
and 2 microns neither collection mechanism works well and collection effi-
ciency is worst in this range. In general, more energy means greater collec-
tion efficiency as more water is accelerated to higher speeds, which results
in greater pressure drops.
Venturi scrubbers can also collect gaseous pollutants simultaneously
because intimate mixing between the gas and scrubber liquor occurs at the
throat. There are other designs more suitable for collection of gaseous
34
-------
compounds alone, but in some circumstances where particulate matter can
present pluggage and fouling problems in packed-bed scrubbers, a venturi
scrubber with its nonfouling characteristics may be more satisfactory. One
example of a venturi scrubber used for gaseous emission control is in the
sulfite pulping industry where recovery boilers burning magnesium-based,
spent sulfite liquor use three or four venturi scrubbers in series to capture
and reuse sulfur dioxide (S02) from the gas stream. Through careful control
of pH, these scrubber trains can be quite effective in removing S02 from the
gas stream.
A number of parameters are available to monitor scrubber performance and
some are even used to control scrubber operation. The records associated
with these operating and maintenance parameters include:
0 Operating
Pressure Drop
Water Flow Rates (Recirculation, Make-Up, and Slowdown)
PH
Temperature
Solids Content of Recirculated Scrubber Water
Solids Removal from SettlingJanks or Ponds
Fan Motor Current
0 Maintenance
Nozzle Replacement
Throat Replacement or Adjustments
Pump Impeller Wear
Opacity is not an operating parameter to be routinely monitored because
wet plume characteristics typically interfere with proper transmissometer
operation. Although some work has been done to develop a method of
monitoring opacity in wet plumes, its application is not currently widespread
and testing is still being performed to determine the acceptability of the
data.
5.1 OPERATING RECORDS
Parameter: Pressure drop
Method of Measurement: Manometer, magnehelic, continuous chart recorder
Applicability: All sources
Limitations: Pluggage of taps and water in pressure sensing lines
35
-------
One of the most useful operating parameters to be monitored in a venturi
scrubber is pressure drop. The pressure drop should be monitored across the
throat of the venturi scrubber and not across the entire scrubber train
(presaturator, scrubber, and separator) for the most useful information.
Pressure drop is the parameter monitored and used to control the operation of
variable throat scrubbers (opening and closing the throat as needed with
constant water flow rate). It is also one of the first parameters available
to indicate problems in scrubber operation.
Smaller sources using low pressure drop scrubbers probably only need to
have pressure drop indicated in a control room where operators can periodi-
cally record the values in the daily logs (i.e., once every two or three
hours as a minimum). Larger sources should consider continuous chart re-
corders. Sources with little fluctuation in operating charactristics and/or
variable throat scrubbers with automatic controls should see only small
changes in pressure drop. Sources with variable process rates and fixed
throat scrubbers will likely observe greater fluctuation in pressure drop.
Parameter: Water flow rates
Method of Measurement: Orifice or venturi meters, rotameters,
ultrasonic meters, pump motor current
Applicability: All sources
Limitations: Pluggage or miscalibration of meters, high solids can
cause problems in all but noncontact devices
The second most useful parameter in monitoring scrubber operation is
•water flow rate. Unfortunately, it is rarely monitored except at the largest
sources. The difficulty usually lies in the equipment used to monitor flow.
Orifice and venturi meters are subject to wear and buildup from suspended
solids in the water that changes meter calibration. The same problem can
occur in rotameters as well as turbidity problems making reading of the meter
very difficult because one must read a floating bob position against a cali-
brated sight glass. Ultrasonic meters are usually noncontact devices and are
not subject to the wear problems associated with the other meters mentioned.
The ultrasonic meters, however, are very expensive and generally do not
handle shocks well, which means that these devices must be handled carefully
during installation. In addition, special maintenance is required. Another
alternative is to monitor pump motor current to indirectly monitor flow. But
36
-------
on all but the largest pumps the horsepower requirements are low and subtle
differences between power input may be difficult to distinguish.
Given the problems and costs associated with the monitoring of scrubber
water flow rates, sources will often opt to measure or monitor water supply
pressure to the scrubber. Although this indicates the header is pressurized,
it does not provide an indication of water flow rate through the pipe. Water
pressure is not an adequate substitute for water flow rate.
In addition to scrubber flow rate, the makeup to and/or blowdown rates
of water from the scrubber system should also be monitored. This is
important to the solids buildup rate and amount of evaporative losses in the
scrubber. Flow monitor outputs should be available to operators and if not
measured and recorded continuously then data should be recorded at a minimum
of once every 2 hours. Water flow rate to the scrubber is typically not
varied (the exception is some fixed throat scrubbers that vary flow rate to
control pressure drop) although makeup or blowdown may vary. A QA program
including routine calibration of flow meters should be required (usually once
per month).
Parameter: pH
Method of Measurement: pH meter or monitor
Applicability: Limited
Limitations: pH sensors prone to pluggage and failure
Monitoring scrubber water pH is rarely needed in many of todays scrubber
applications. The use of pH monitoring is typically limited to carbon steel
scrubbers that are susceptible to acid attack and to applications where
gaseous emission control is part of the scrubbers function. The pH monitor
helps operators maintain the proper pH to limit corrosion or to operate in
the most effective absorption range for the scrubber.
Generally, pH monitors have required a substantial amount of maintenance
to remain operational. Most successful applications of pH meters for con-
tinuous pH monitoring employ sidestream monitors where only a small of the
flow is monitored rather than the total water flow through the scrubber.
Parameter: Gas temperature
Method of Measurement: Thermocouple
Applicability: All high temperature sources
37
-------
Limitations: None
Temperature of the gas stream both before and after the scrubber are
important indicators for scrubber operation. Saturated gas conditions repre-
sented by gas temperatures in the 150° to 160°F range should be expected as a
result of evaporative cooling by the scrubber. Water loss by evaporative
cooling may account for 5 to 10 percent of the total water flow in high
temperature application. This may cause an increase in emission rate if the
solids levels are high in the scrubber water.
The inlet temperature is useful for indicating the need for a cooler/
presaturator ahead of the scrubber. High temperature gas streams may contain
components that are gaseous until cooled in the scrubber. These materials
may form in particle size ranges that are difficult to capture. In addition,
the outlet temperature can be used to determine if saturated conditions are
being achieved. Temperatures higher than saturation usually indicate maldis-
tribution of gas and/or water within the scrubber.
Temperatures should be monitored periodically (once every 2 hours mini-
mum) for relatively steady-state source. Sources that vary considerably in
operation should consider continuous chart recorders to record temperature
data.
Parameter: Solids content of scrubber water
Method of Measurement: Water sample, balance
Applicability: High temperature sources
Limitations: None
A frequent problem observed is the resuspension and regeneration of
particulate matter into the gas stream because the solids levels are too high
for the particular application. Resuspension generally refers to suspended
solids in the scrubber water that are suspended once the water is evaporated
away from the particles. Regeneration generally refers to the particle
formation that occurs from dissolved solids and salts when water is evapo-
rated. Typically regenerated particulate forms a very fine, very difficult
to collect fume.
Periodic samples should be collected, particularly at sources where
little blowdown of scrubber water occurs. Buildups of 5 to 10 percent dis-
solved solids can occur and cause significant problems with plume opacity and
38
-------
in some cases, erosion of scrubber components. Weekly grab samples should be
considered minimum for sampling from the scrubber sump return. If water with
high dissolved and suspended solids are used in a presaturator, high residual
opacity may occur. The level of dissolved and suspended solids that can be
tolerated will be very site specific.
Parameter: Solids removal
Applicability: All scrubbers with settling ponds
Method of Measurement: None
Limitation: None
Part of the operation of a scrubber involves the settling and removal of
solids captured by the water droplets. Although automatic clarifier systems
do exist, most scrubbers use settling ponds that have to be emptied manually
(usually draining and removing sludge with a front-end loader). Each time
the pond is cleaned it should be noted in the operating records to establish
representative operation. In this way the representativeness of a stack test
may be determined if the scrubber was operated with clean water and a clean
settling pond.
5.2 MAINTENANCE RECORDS
With few moving parts, the required maintenance records for a scrubber
are not extensive.
Maintenance Parameter: Nozzle Replacement
Applicability: All scrubbers equipped with nozzles
Limitations: None
New scrubber designs consider nozzle accessibility and ease of replace-
ment for maintenance personnel. Consequently in many applications, nozzle
replacement can be accomplished while the scrubber is on-line. Nozzle re-
placements should be part of the maintenance records including the reason the
nozzle was replaced (i.e., erosion, pluggage, preventive maintenance, etc.).
In addition, a routine check of the nozzle operation should be included as
part of a preventive maintenance plan. A minimum of once per week is neces-
sary although once per day would be preferable.
Maintenance Parameter: Throat adjustments or replacement
39
-------
Applicability: All scrubbers
Limitations: None
Adjustments to the scrubber throat or to the controls of a variable
throat scrubber should be maintained in the maintenance records. Without
this information sudden changes in performance are not easily explained. An
example of this was found at a grey-iron foundry when extreme erosion of the
scrubber throat forced maintenance personnel to patch the scrubber with a
steel plate. After applying several layers of steel patches, the configura-
tion of the scrubber (inside) was considerably different, which severely
reduced the draft on the blast furnace causing excessive fugitive emissions.
As a result, maintenance personnel would go to repair another patch and
decide to replace all the previous repairs and duct again. This repair would
always return the scrubber to near initial design and performance would be
restored until the next major repair 3 to 4 months later after performance
progressively got worse. No records of maintenance were kept so that it was
merely by accident that the cause of the variable performance was discovered.
Pressure drop and water flow rate seemed normal during the time that perform-
ance grew worse. However, if records of what had been done had been kept,
the solution would have been easier to identify.
Maintenance parameters: Pump impeller wear
Applicability: All scrubbers
Limitations: None
Pump impellers should be checked periodically for wear. The greater the
suspended solids content of the scrubber solution, the greater the chance for
pump impeller wear. As the pump wears, the ability to pump a given quantity
of water at a fixed pressure decreases. This affects both water volume
capacity to the scrubber and delivery pressure to the scrubber (nozzle de-
signs may generate different drop sizes with different pressure characteris-
tics). Each time the pump housing is opened and checked or the impeller is
replaced it should be noted in the maintenance record.
40
-------
SECTION 6
ELECTROSTATIC PRECIPITATORS
The ESP is probably the most complicated of the control equipment types
discussed in this manual. Although the operating principles are relatively
straightforward there are a large number of parameters to be specified and
most designs are a compromise between the various factors to arrive at an
economical yet reliable control device. To adequately monitor ESP perform-
ance, however, there are a large number of parameters that should be moni-
tored. Unfortunately this can require a substantial investment in personnel
to obtain and/or correlate the information recorded. Thus, on a generously
designed and reliable ESP there may be justification for reducing the amount
of data obtained or, at least, the amount of daily data reduction.
Unlike other types of control equipment discussed in this manual, the
energy expended in collecting particulate matter is applied directly to the
particulate matter and not to the entire gas stream. Consequently, pressure
drop through an ESP is not an indicator or even a factor to be considered
during normal operation. The electrical energy used goes to creating a
corona discharge, charging of particles, moving the particles across the gas
stream to the collection plates, and partially discharging the particles
while retaining them on the plates until they are ultimately removed from the
ESP. The operation of the ESP may seem very complex and, indeed, it can be.
But the main indicators of ESP performance are closely related to the trends
of the voltage and current characteristics within the ESP. In fact if more
time was taken evaluating the voltage/current readings of the ESP, many
sources could probably function quite well taking and analyzing only the
electrical readings and opacity values to determine if an ESP is functioning
properly. There are, however, many other parameters that can be calculated
for ESP performance as well as many items that should be periodically
checked. Briefly, these items are:
41
-------
Operating Records
Voltage and current readings
Opacity
Gas temperature
Gas composition
Gas volume
Conditioning agents addition
Gas load V-I curves (supplemental)
Resistivity tests (supplemental)
Carbon loss or loss-on-ignition (LOI) tests
Rapper operation checklist
Transformer-rectifer (T-R) set checklist
Hopper heater and dust discharge checklist
Insulator heater checklist
Maintenance Records
Wire breakage, removal, and replacement diagram
Rapper replacement diagram
T-R repair/replacement
Insulator/heater replacement
Dust discharge system repair
Misalignment
Air load V-I curves
Gas flow distribution tests (supplemental)
As can be seen the list is very extensive and can extend even further
when one considers the process data that can also be recorded. Because ESP's
are generally sensitive to variation in process operating conditions or other
factors (e.g., sulfur content in coal burned at a boiler and its effect on
ash resistivity) process data and evaluation of their effects on ESP opera-
tion are particularly important. Arguably each of these records can be
viewed as necessary to diagnosing the performance of an ESP or troubleshoot-
ing if problems are noted. There are sources that integrate each and every
one of these elements into their overall preventive maintenance program. The
key, however, is not the amount of data created but what is done with it
42
-------
particularly in fine tuning operations or minimizing problems once dis-
covered. Some items will be listed as "optional" although the arguments for
including that element into the recordkeeping program will be strong. Only
site-specific consideration will help determine what elements are truly
needed.
6.1 OPERATING RECORDS
Parameter: Voltage and current readings
Applicability: All sources
Measurement method: Voltage and current meters for each T-R control
Limitations: Meters may provide an indirect measurement of the
individual parameter to be recorded. Require periodic
recall'bration. Small positive corona two-stage ESP's
may not be equipped with meters
The most important parameters that should be recorded for an operating
ESP are the voltage and current values for each T-R set. The individual
values for each T-R voltage and current are important but more important are
the trends noted within the ESP. The most notable trend is the increase in
current density and power density in the various fields as the gas moves from
inlet to outlet.
There can be a significant variation in the operating values for an ESP
and changes sometimes occur within minutes. Variations caused by gas volume
and temperature, particulate matter characteristics and even the depth of
buildup of dust on the plates can all play an important role in changing the
values of voltage and current. But these readings also provide the first
warnings of problems within the ESP.
The readings should be taken at least once every 8 hours during normal
operation although once every 4 hours may be preferable for giving more
advanced warning of problems. The values for primary voltage and current as
well as secondary voltage and current should be recorded where available.
For small ESP's this produces only a small quantity of data. For large
systems, application of digital based data loggers may be required. In all
situations, the values should be compared against a baseline or "normal"
value to determine if significant variations are occurring and if further
action should be taken.
43
-------
Parameter: Opacity
Applicability: Most larger and moderate sized sources
Measurement Method: Double-pass transmissometer
Limitations: Variations in particulate characteristics (particle size)
may limit usefulness. Six-minute integration smooths out
spikes in data that may be useful in optimizing ESP per-
formance. Siting criteria must be met
In many situations, ESP operation is usually evaluated in terms of the
opacity observed by a transmissometer on a real-time basis. Under optimum
conditions the ESP should be capable of operating at some "constant" or base
opacity with a minimum amount of spiking from rapper reentrainment. These
data can also be supplied to an integrator to provide 6-minute averages to
fulfill regulatory requirements. Some sources are capable of providing both
data. Provisions should be available to provide the real-time data for plant
or agency personnel to evaluate the degree of spiking that occurs and any
significant changes in operation. (Note: Some methods of integrating can
potential miss the problems associated with spiking although they will show a
higher than "base" opacity.)
Many sources rely on the opacity meter as being the first indicator of
problems and this is useful as long as the opacity versus mass relationship
does not vary significantly with operating conditions. When used with
electrical readings the opacity measure can be very reliable. In those
cases, however, where extreme changes in particle characteristics may occur
(e.g., poor combustion of fuel in a boiler, some kraft recovery furnaces,
alkalies in cement production), these mass concentrations may change by an
order of magnitude without changing opacity. Sometimes opacity can change
without a corresponding change in emission rate (concentration). These,
however, tend to be the exception although both agency and plant personnel
need to be aware that such changes can occur.
Some sources have tried to correlate ESP power readings with emission
rate and/or opacity. At the midpoint of.the range of power levels there may
be good correlation with the relationship deteriorating at very high and very
low power readings. The opacity reading used is usually the 6-minute average
at the time of the readings. Over a long period of time there may be strong
correlation, but for any given set of readings there may be no apparent
relationship between opacity and power input.
44
-------
Parameter: Gas temperature
Applicability: All sources
Measurement Method: Thermocouples, thermometers
Limitations: Stratification in duct and representative values
Changes in gas temperature can have profound effects on ESP performance.
The temperature variation can be very small and yet cause a significant
change in ESP power levels and opacity (in some cases as little as 15°F).
Although gas temperature variations may have some effect on corona discharge
characteristics and physical characteristics of the ESP (corrosion, expan-
sion/contraction), its most important effect is on particle resistivity. For
sources with the potential for high resistivity, the change in performance
can be dramatic as a result of temperature changes when all other parameters
seem to be the same. The gas temperature should be checked once per shift
for smaller sources and measured continuously on larger sources and on those
sources with temperature sensitive performance.
Temperature measurement can also be a useful tool in finding excessive
inleakage or an unequal flow through the ESP. These can affect localized
velocity patterns within the ESP which, while the average value seems
adequate, can cause reduced performance within the ESP. Most ESPs are de-
signed with a minimum of two chambers. The gas temperature for each chamber
should be measured at both the inlet and outlet, if possible. Significant
temperature changes between the inlet and outlet values may indicate in-
leakage problems that should be confirmed by measurement of gas composition.
A single point temperature probe may miss stratification within the duct and
the duct should be checked periodically to assess the degree of stratifica-
tion and determine the representativeness of the selected measurement loca-
tion.
Parameter: Gas composition
Applicability: Source dependent, optional
Measurement Method: Continuous monitor for appropriate pollutant(s) or
gases
Limitations: Some monitors used for protection from explosions give
limited protection. Stratification considerations are
important
45
-------
Continuous monitors are often required to fulfill regulatory require-
ments. In some cases, however, monitors are placed to prevent or reduce the
chances of explosions within the ESP. The most common example is the use of
hydrocarbon and/or CO monitors on cement plant applications.
A useful monitor on combustion sources is a continuous oxygen analyzer
usually placed at the inlet of the ESP but may be colocated with another
monitor. The oxygen analyzer can be useful in determining the amount of
excess air that is brought into the ESP along with other combustion products,
although not particular useful for process control. If continuous monitors
are not available, usually a weekly grab measurement of oxygen with ORSAT or
Fyrite analyses are sufficient. It should be noted that stratification in
the ductwork can present measurement problems.
Parameter: Gas volume
Applicability: Optional, sources with "marginal" design parameters
Measurement Method: Pi tot tube, F-factor, fan current and speed
Limitations: Can be time consuming to obtain and reduce data
Gas volume is important to ESP operation because most of the key design
p
and operating parameters such as specific collection area (ft /1000 acfm),
superficial velocity, treatment time, and specific corona power (watts/1000
acfm) are dependent on gas volume. Routine gas volume calculations are
recommended for those designs and operating practices that do not provide an
adequate margin for operation. For example, significant variations in oxygen
may indicate large swings in the gas volume that may decrease ESP performance
and indicate the need to routinely determine ESP gas volume. Low SCA values,
high superficial velocities (6 ft/s), short gas treatment times (5 s or
less), and much higher oxygen levels at near full load conditions are indi-
cators that would suggest that excess gas volume may be decreasing ESP per-
formance.
Parameter: Conditioning agent addition
Applicability: Units equipped with gas conditioning equipment
Measurement Method: Various
Limitations: Periodic calibration
Conditioning agents are usually added to the gas stream to modify re-
sistivity of the particulate matter entering the ESP. Conditioning agents
46
-------
may include water (dry process cement plants), SCL and H^SO^ (coal-fired
boilers with low-sulfur coal), and even the addition of sodium compounds to
the coal for some boilers. Feed rates should be monitored for changes rela-
tive to the ash production rate. In addition, access to nozzles for periodic
maintenance checks should be provided for water and sulfur based conditioning
systems to prevent pluggage problems.
Parameter: Gas load V-I curves
Applicability: As needed
Method of Measurement: Voltage/current meters on T-R controls
Limitations: May vary due to internal ESP factors
Gas load voltage-current curves (also known as V-I curves) are usually
performed on fields where problems are suspected although in rare cases the
entire ESP may be checked. The curves can be useful in diagnosing alignment
problems, dust buildup problems, changes in resistivity, insulator tracking
problems, and some T-R controller problems. When more than one field is
being checked the outlet fields should be checked by first moving forward
towards the inlet. This minimizes interferences within the ESP while the
field is deenergized to establish the corona initiation voltage. It may be
difficult to compare V-I curves from day-to-day or even with air load condi-
tions. The gas load V-I curves, however, should become part of voltage/
current records for diagnostic purposes.
There are some cases where the V-I curves are used to establish the
optimum level for collection of high resistivity dust because the controllers
are unable to "recognize" high resistivity. In this situation, the controls
are usually left in the manual mode and it is generally unnecessary to show
these curves developed for each period before readjustment.
Parameter: Resistivity
Applicability: As needed
Measurement Method: In-situ resistivity probe, bulk sample (external)
Limitations: Difficulties in measurement and differences in two methods
may not show agreement in resistivity
Particulate matter resistivity tests are usually performed when other
possible reasons for poor performance have been eliminated. Two methods are
47
-------
generally used: in-situ and external bulk sample. The methods differ
significantly and can produce drastically different results.
. The in-situ method measures voltage and current under clean probe condi-
tions and the electrostatically collected particles on a sample plate where
the "dirty" voltage/current values are generated. The test is usually con-
ducted at the inlet of the ESP. The advantage of this system is that the
sample is measured under action operating conditions for temperature,
moisture, gas composition, and particle characteristics. The disadvantage of
the method is that it is often difficult to measure the thickness of the dust
layer and the resistivity is measured only under actual gas conditions with
no real ability to modify them.
The bulk external method usually aspirates a sample out of the ESP
inlet. In the lab it is fused in a crucible at approximately 800°C and then
placed on a plate and subjected to various gas conditions to generate a
resistivity profile. The advantage of this system is that sample thickness
can be more easily controlled and gas conditions can be varied both for
temperature and for gas components. The disadvantage is that the action of
fusing the sample may volatize or oxidize resistivity modifiers that under
actual operating conditions assist ESP performance. Generally, the two
methods will not agree precisely although each has its advantages and dis-
advantages. The results of such analyses should be kept as part of the
operating record.
Parameter: Carbon loss or loss on ignition (LOI)
Applicability: Fossil fuel fired utilities, or any sources where carbon
in the ash can be substantial. Optional
Measurement Method: ASTM method
Limitations: None
The LOI tests are routinely conducted at some utilities as a method of
tracking combustion performance. This periodic measurement is usually
carried out daily or weekly. Carbon in the sample can act as an effective
resistivity modifier in some cases. Generally, the improvement in re-
sistivity can be seen in levels up to 10 to 12 percent carbon in the ash.
The economics of carbon loss, however, may force these levels to remain low.
Below 1.5 to 2.0 percent carbon in the ash, the carbon generally does not
modify resistivity. Extremely high levels of carbon in the fly ash can lead
48
-------
to extremely poor ESP performance as well as increase the potential for fires
and explosions. High levels usually indicate the need for burner adjustment.
Parameter: Operation checklists for rappers, T-R's, hopper heaters,
dust discharge, and insulator heaters
Applicability: All sources
Method of Measurement: Various
Limitations: Access to some items
In addition to the daily monitoring of data some operation checkpoints
should be included. These include a once per shift check of rapper opera-
tion, dust discharge system, and T-R set operation (indicating the ones in
the "off" position). Rappers not functioning should be scheduled for mainte-
nance particularly if large sections of rappers are out of service. Dust
discharge systems should have highest priority for repair as dust should not
accumulate in the bottom of the ESP for long periods of time because of the
potential for causing severe plate misalignment problems. Hopper heaters can
usually be repaired with little difficulty after removing weather protection
and insulation. Insulator heaters may be difficult to repair except during
short outages. These heaters are important to keep condensation on the
insulators to a minimum although hopper heaters generally help keep the dust
warm and free flowing. As each shift data sheet and checklist are completed,
they should be evaluated for maintenance needs.
6.2 MAINTENANCE RECORDS
The work order system for prioritizing, dispatching, and providing
written information on what conditions were found and what repairs were made
seems particularly well adopted to an ESP where a large number of records may
be necessary for retrieval. There may have to be supplemental files,
however, for such items as internal inspection reports, air-load V-I curves,
and wire replacement charts for ESP maintenance. Although the log book
concept may still be used, on large ESP's the "routine" maintenance reporting
is much more important.
Type of Record: Wire replacement chart
Applicability: All ESP's equipped with wires
49
-------
Although an ESP can operate effectively with up to 10 percent of its
wires removed, care must be taken that no more than 5 to 10 wires in any one
gas lane are removed. The loss of wires down any one lane can result in a
substantial increase in emissions. The only way to adequately track where
wires have failed and been clipped out of the ESP is with a wire replacement
chart. In general, a few wire breaks randomly distributed can be expected
after installation, with few during normal operation and then more frequent
occurrences of breakage at the end of wire life. Many factors influence wire
performance including sparking, alignment, corrosion, and even construction
and installation practices.
It is not unusual for wires to be simply removed and not replaced if
they break. The wire location and the location of the break (top, middle,
bottom) should be noted. If wires are replaced, this should also be noted.
Wire breakage can keep a substantial portion of the ESP deenergized even
if there is only one wire break in each section deenergized. The wire
breakage chart can help maintenance personnel determine if patterns are
random or more concentrated in a given area.
Type of Record: Rapper replacement/repair
Applicability: ESP's equipped with rappers
With the exception of wet ESP's, all other ESP's use rappers to remove
dust buildup on the collection plates and the discharge electrodes. For the
rigid-frame and rigid-electrode type ESP's, checks and replacement of the
rapping system components is usually limited to periods when the ESP is
deenergized and open for access. Only the drive and the gear reduction
mechanisms are available outside the ESP for maintenance. On weighted-wire
ESP's, however, the rappers are usually external and more easily checked.
Rappers should be checked each shift for operation. This check may only
be cursory to determine if the rapper system is functioning. Once each week,
however, the rappers should be more carefully checked to determine which, if
any, rappers need repair or replacement. This check also extends to the
rapper control system. The failure of an individual rapper or two will
probably have little effect on overall ESP performance. But the loss of an
entire section of rappers or the rappers on an entire ESP can be more severe.
The type of repair should be noted to see if a pattern exists. For example,
50
-------
on one ESP the number of coil burn-outs and failure of rapper control cards
suggested high voltage was somehow feeding back through the system. Ground-
ing the rappers and supplying the system with diodes eliminated the problem
but it took several years to accumulate enough data to realize a repeating
pattern was occurring.
In addition, any adjustment to the rapping frequency and intensity
should be noted. Very often, this is an easy item to be changed when things
seem not to be working and when the values are changed without any notifica-
tion, it is extremely difficult to restore the rapping intensity and/or
frequency back to their original, perhaps more optional, levels.
Type of Record: T-R problems
Applicability: All ESP's
Generally, the number of T-R problems is small. Failure of controllers,
changes in allowable spark rate, current limit, and replacement of meters,
however, should be noted in this file. Also, changes in the normal wiring
arrangement of the ESP should also be noted. In some applications it is not
uncommon to rewire or "jumper" the ESP so that additional sections of the ESP
may be energized when a T-R failure occurs. It is not always possible to
determine this from T-R meter readings and if this practice occurs, diagrams
or listings of what T-R energizes what section should be maintained so that
future analyses of the operating data can be adequately conducted.
Type of Record: Insulator heater/hopper heater replacement
Applicability: Optional
Insulator heater failures and replacement may correlate with the need to
replace insulators because of excessive tracking. For most sources, however,
the problem is not severe and separate records are not necessary. Hopper
heaters may only be needed at certain times in some applications (startup,
cold weather) and sometimes plugged hoppers can be traced in part to failure
of hopper heaters. Again, this is a relatively minor problem and separate
records are not necessary.
Type of Record: Dust discharge system repairs
Applicability: All ESP's
Each time there is a failure in the dust discharge system there is the
potential for dust buildup to cause temporary problems (dust buildup between
51
-------
plates, reduced power at increased sparking, grounding of portions of the
ESP) as well as permanent problems (misalignment). Each time a failure
occurs and is repaired, the cause of the problem (if found) should be noted
and the electrical characteristics of the fields near the problem area should
be checked after the repair to see if they return to normal levels using gas
load V-I curves. The main concern is misalignment within the ESP. It is
possible in some cases to trace misalignment within an ESP back to a specific
incidence of hopper pluggage or failure of the dust removal system. This is
one of the most common ways to misalign an ESP.
Type of Record: Misalignment and repairs
Applicability: All ESP's
The presence of misalignment of the ESP internals can often be detected
through the electrical readings. The precise location of misalignment,
however, can usually be determined only by an internal inspection of the ESP
during a shutdown period. In some cases, the cause of the misalignment can
be repaired and in others it may be impossible or impractical to repair.
Misalignment can be caused by warped plates, shifting of support frames,
broken insulators, and even structural failure of the ESP. The analysis of
wire breakage records may indicate localized misalignment because those areas
have greater occurrences of wire breakage. Misalignment most often causes
localized areas of high sparking that may result in wire breakage. If mis-
alignment can be repaired it will help reduce wire failures and increase
operating voltages. If it cannot be repaired, other steps such as cutting
out wires in the affected areas to increase operating voltages, reduce the
effect of wire breakage, and keep the T-R's in service may be an acceptable
alternative to reduced or no energization.
Type of Record: Air load V-I curves
Applicability: All ESP's
Initially, before startup, the ESP should be energized to establish if
it is ready for operation. The voltage/current characteristics established
under these conditions should become part of the permanent record for the
ESP. After each shutdown when maintenance is performed, an air-load V-I
curve should be produced that verifies that the ESP is ready for operation.
In fact, before and after V-I curves (if possible) are a good idea to provide
52
-------
evidence that problems were found and corrected. The patterns in a well
aligned, well maintained ESP should be similar to the original V-I curves.
If problems were not corrected, the air-load V-I curve provides evidence that
further maintenance is required to find and isolate the problem(s).
Type of Record: Gas Flow Distribution
Applicability: Optional
Gas flow distribution problems are often blamed for a number of ESP
performance problems but it is often difficult to prove because the testing
methodology is cumbersome. A hot-wire anemometer test is usually conducted
during shut down to demonstrate the presence of unsuitable gas flow distribu-
tion. This can be considered a special test for most ESP's as most flow
orientation devices are installed within the ESP duct work on the basis of
experience and flow model studies. If severe maldistribution is noted, then
corrective actions should be scheduled. Repairs and modifications to the
original design should be recorded. Hot-wire anemometer tests taken after
the repairs or modifications have been made should also be kept to demon-
strate problems have been corrected. Other ESP functions such as T-R set-
points, rapper operation, and dust discharge system functions may have to be
optimized to reflect the new operating conditions within the ESP. Because
most ESP's are very sensitive to gas flow distribution problems, correction
will usually result in improved performance.
6.3 SUMMARY
The listing of items to be recorded and tracked for ESP's is extensive
even with the exclusion of items such as lubrication, routine filter changes,
and process parameters. This is due, in part, to the sensitivity of ESP
performance to a number of operating parameters and the operating principles
of applying energy directly to remove particulate matter. There is also,
however, more instrumentation available on most ESP's when compared to other
control equipment that provides useful information about performance and the
effects of changes of various operating parameters. As stated earlier, many
sources find that a work order system works well with ESP maintenance because
there are many items that must be addressed. On the operating data side, the
manipulation of electrical data can become manpower intensive and typically
53
-------
requires at least one person to become familiar with the operating charac-
teristics to quickly analyze and evaluate performance characteristics.
Properly designed, operated, and maintained, the ESP can provide high particu-
late matter removal efficiency to maintain continuous compliance with the
applicable emission limits.
54
-------
APPENDIX A
EXAMPLE WORK ORDER FORM
55
-------
WORK ORDER
STATION
WOMIWOUIST
LOCATION .
AVACLA»H • OATt
OAT(
. I : !
UNIT i MIIOMTV
NO.
MQ-0. COM*
OATf
; mo TAP
11 vtsil NO
APMOVCO •» OATf
COUIMfNT NAMC OM M» TITU
CMAWU TO:
COUI^. NO.
PIANNCR
COOI
oiscntmoN of M*
i i i i i
TOOkUST
ESTIMATIO LAMM
*TCHA^T»OUJ
HO.
TOTAt
tSTIMATIO
MAN/MOUM9
COWIMMII
MIMAMKS:
UArCMIAI. HCOUMfMCNT
OATC
OCLivCN TO:
OCSCMTION
STOCK NO. OTV. *WA1U UMO
IOO<»
•CCOTCO IV
! OATC
WAMTCNANCC SU*C«VIS
(Copyright©April 1983, EPRI Report CS-2908, "Proceedings: Conference on
Electrostatic Precipitation Technology for Coal Fired Plants". Reprinted
with permission. )
56
-------
U«T
SYSTB1
SU8SYSTB1
conpONGir
SUBCOWMEJIT
nftlKTEMNCE REQUEST r-QRfl
000000
DATS:
AMI HMD TO;
PmoniTr:
UNIT STATUS:
PIIOIUH
OATI:
Jot STATUS:
CAUSC <3t PwoiUfl:
USBJ:
TOTAL IANHOUDS
"UTMIAL COST
(Copyright April 1983, EPRI Report CS-2908, "Proceedings: Conference on
Electrostatic Precipitation Technology for Coal
with permission. )
Fired Plants". Reprinted
57
------- |