Beruflich Dokumente
Kultur Dokumente
Workshop Proceedings
Technical Report
9862704
Fourth Intelligent Sootblowing
Workshop Proceedings
1004112
EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com
9862704
DISCLAIMER OF WARRANTIES AND LIMITATION OF LIABILITIES
THIS DOCUMENT WAS PREPARED BY THE ORGANIZATION(S) NAMED BELOW AS AN
ACCOUNT OF WORK SPONSORED OR COSPONSORED BY THE ELECTRIC POWER RESEARCH
INSTITUTE, INC. (EPRI). NEITHER EPRI, ANY MEMBER OF EPRI, ANY COSPONSOR, THE
ORGANIZATION(S) BELOW, NOR ANY PERSON ACTING ON BEHALF OF ANY OF THEM:
EPRI
ORDERING INFORMATION
Requests for copies of this report should be directed to EPRI Orders and Conferences, 1355 Willow
Way, Suite 278, Concord, CA 94520, (800) 313-3774, press 2 or internally x5379, (925) 609-9169,
(925) 609-1310 (fax).
Electric Power Research Institute and EPRI are registered service marks of the Electric Power
Research Institute, Inc. EPRI. ELECTRIFY THE WORLD is a service mark of the Electric Power
Research Institute, Inc.
Copyright © 2002 Electric Power Research Institute, Inc. All rights reserved.
9862704
CITATIONS
EPRI
3412 Hillview Avenue
Palo Alto, CA 94304
Principal Investigator
J. Stallings
The report is a corporate document that should be cited in the literature in the following manner:
Fourth Intelligent Sootblowing Workshop Proceedings, EPRI, Palo Alto, CA: 2002. 1004112.
iii
9862704
9862704
REPORT SUMMARY
This document describes the proceedings of the Fourth Intelligent Sootblowing Workshop and
Exposition held on March 19–21, 2002, in Houston, Texas.
Background
To become more competitive, power generators have been purchasing more lower cost and/or
environmentally benign fuels to reduce the cost of electricity. One potential impact of firing
these varied fuels is a dramatic increase in slagging and soot deposition. Although savings from
alternative fuel costs are often substantial, operators have to deal with continual changes in the
combustion process. Intelligent sootblowing (ISB) is an effort to optimize removal of ash
slagging and fouling for the least impact on boiler performance and operability.
Objective
• To bring together all major players in the sootblowing arena—equipment manufacturers,
software developers, control vendors, and utility operators—to address the current state of
development for ISB and its concomitant research needs
Approach
Because there are still many unanswered questions regarding the most effective ISB solutions,
the workshop focused on panel discussions with leading developers in the field and
knowledgeable utility engineers. Numerous presentations described the different potential
solutions, and utility experience to date was highlighted by a site visit to Reliant Energy’s W.A.
Parish Plant to see available equipment in operation.
Results
The workshop incorporated sessions on utility experience, hardware and sootblowing media, ISB
optimization and control, and utility issues. Vendor panels addressed issues surrounding different
hardware approaches, including the potential for eventually optimizing the process. A utility
panel summarized industry needs and future R&D requirements.
EPRI Perspective
Two philosophical approaches toward intelligent sootblowing are currently being applied in the
industry. One incorporates heat flux monitors to gather real-time heat transfer data to determine
which areas of the furnace need cleaning. The other—a less intrusive approach—uses indirect
temperature and pressure data to infer locations where sootblowing is needed. Each concept has
its strengths and weaknesses. Only by demonstrating both approaches at multiple plants can the
industry determine which of these works best in specific situations. Proceedings of EPRI’s first
three ISB workshops—held in 1998, 1999, and 2000—were not published. However, EPRI
v
9862704
report 1000410, Guidelines for Intelligent Sootblowing Control, published in December 2000,
serves as a useful state-of-the-art description for the technology.
Keywords
Intelligent sootblowing
Boiler optimization
Boiler efficiency
Instrumentation and control
Slagging
Fossil plant operation and maintenance
vi
9862704
CONTENTS
vii
9862704
16 RELIANT ENERGY AND SARGENT AND LUNDY: APPROACH TO
ALTERNATIVE CLEANING SYSTEMS USED ON SCR APPLICATIONS............................16-1
viii
9862704
1
W.A. PARISH UNITS 7 & 8 ISB PROJECT
Presented by:
1-1
9862704
W.A. Parish Units 7 & 8 ISB Project
Presented by:
• Background
• Project Objectives
• Scope of the ISB Systems
• Unexpected Problems and Issues
• Test Plans
• Preliminary Results
• Preliminary Conclusions
– Unit 7 Overview
• 590MW T-Fired
• PRB Coal
• New DCS in Fall 2001 Outage
• Overfire Air / Low NOx Burners
• Economizer / LTSH surface added
• 19 new Retracts installed
• Lower Slopes replaced
• SCR being installed
– Unit 8 Overview
• 590MW T-Fired, Uprated to 650MW
• PRB Coal
• New DCS in Spring 2001 Outage
• Overfire Air / Low NOx Burners
• Economizer / LTSH surface added
• 19 new Retracts installed
• Lower Slopes replaced
• Turbine Uprated (HP/IP rotor)
• SCR being installed
• Welding Problems
– 58 out of 60 heat flux sensors installed with no problems - one had
to be replaced, one had two damaged thermocouples (still usable)
• Wiring Problems
– wrong size cables used in some cases
– heat flux sensors with remote leads wired incorrectly
– inadequate wiring checkouts performed on Unit 7 prior to
commissioning activities
2-1
9862704
W.A. Parish Unit 5 ISB Project
presented by:
H. S. (Steve) Blinka, P.E.
Engineer, W.A. Parish Technical Support
2-2
9862704
W.A. Parish Unit 5
• 690 MW Gross
• B&W Opposed Wall
Fired Drum Unit
• PRB Fuel
• In Service 24 Years
• Diamond Power BOS
Sootblowing System
with 43 wallblowers, 80
convection pass steam
lances
• 22 waterlances added
in 1995
2-3
9862704
Furnace Wall Quenching
9862704
2-5
9862704
Quench Cracks
2-6
9862704
After Repair
2-7
9862704
Improvements Needed
2-8
9862704
Sootblowing Optimization Plan
2-9
9862704
ASI System For WAP 5
9862704
Schedule for WA Parish Unit 5
2-11
– Plan to install water cannons in a future outage
9862704
Existing BOS Terminal
2-12
9862704
New ISB Terminal
2-13
9862704
Cleanliness Graphics
2-14
9862704
Remote CPU Location
2-15
9862704
Heat Flux Sensor
2-16
9862704
ACOUSTIC FEGT TECHNOLOGY
9862704
combine equations and express in degrees C:
T c = ( d B τ ) x 1 0 6 − 2 7 3 .1 6
2
w h e re :
T c = g a s te m p e ra tu re , C
B = a c o u stic c o n sta n t = γ R M
τ = flig h t tim e
Yields AVERAGE gas temperature along path
2-18
9862704
Acoustic Furnace Exit Gas
Temperature Instrumentation
Components
2-19
9862704
Configuration
2-20
9862704
Send/Receive Horn
2-21
9862704
Accumulator & Air System
2-22
9862704
Preamp
2-23
9862704
Processor Cabinet
2-24
9862704
Advisory Mode Experience
9862704
2-26
9862704
3
THERMAL FATIGUE CRACKING OF WATERWALL
TUBES FROM WATERLANCES AND WATER
CANNONS
R. E. Kessler
Reliant Energy
12301 Kurland Drive
Houston, Texas 77034
3-1
9862704
THERMAL FATIGUE CRACKING
OF WATERWALL TUBES
FROM WATERLANCES AND WATER CANNONS
R. E. Kessler
Reliant Energy
12301 Kurland Drive
Houston, Texas 77034
Abstract
Water blowing of fireside deposits in coal-fired utility boilers is an effective method of
removing slag deposits. Due to the concern of utilities that repeated thermal shocks from
water spraying damages boiler tubes, EPRI evaluated the rate of thermal fatigue cracking
of the outer surface of waterwall tubes from water blowing. EPRI developed, and
subsequently publicized in Report CS-4914, a model that enables the prediction of the
remaining life of tubes subjected to on-line water cleaning.
These predictions of thermal cracking as a function of thermal cycles made by EPRI are
in line with actual experiences at Reliant Energy. These predictions indicate through-wall
circumferential cracking can occur in less than 50,000 cycles (2 to 3 years operation).
The OEM’s, Diamond Power and Clyde Bergemann, say a circumferential crack will be
16 % deep after 40 years of operation.
This report explains the reasons for the differences found between EPRI and Reliant
Energy experiences versus OEM waterwall tube remaining life predictions.
Conclusions:
1. The thermally induced axial stress associated with the production water lance/
water cannon is approximately 1/5 that of the stress developed by EPRI simulations.
2. The EPRI axial stresses placed on test panels are over fourteen times the stress
of the tube supporting 200 feet of its own weight.
Discussion
Thermal fatigue cracking of waterwall tubes result from water spraying from water lances
or water cannons. The cracks are typically multiple circumferential cracks located very
close to each other as shown in the photograph below:
9862704 3-2
The factors that influence the rate of formation of thermal fatigue cracks in waterwall
tubes are both the number of cycles and the severity of quench. Factors that influence the
severity of quench are:
The greater the rate of cooling, the higher the thermal stress. In the EPRI Report CS-
4914, a water lance simulator was built. Two panels of eight tubes, one of SA-210 Grade
A1 and one SA-213 Grade T2, were sprayed without an insulating slag.
9862704 3-3
The temperature of the cooling water used in the cooling trials by Reliant Energy, EPRI,
Diamond Power and Clyde Bergemann are ambient.
The water spray velocity or lance projection velocity relates to the speed that the water
spray moves across the tube surface. This equates to the duration of each specific quench
cycle. The following table compares the conditions utilized by ERPI, Reliant Energy and
Diamond Power water lances:
The tube temperature after quench is the lowest temperature reached after the water
spraying cycle.
The cooling rate is the length of time it takes for the tube temperature to drop from the
tube temperature before water spraying to the lowest tube temperature after spraying. In
the EPRI simulation, the tube cooled from 750 F to 200 F in approximately ¼ second.
Data from the field reported a lower temperature above 600 F. Water shut-off
immediately after slag removal accounts for the lower temperature being above 600 F.
The difference between the EPRI 550 degrees F drop in temperature versus the water
lance/ water cannon’s 100 degrees F drop equates to a significantly lower thermal
contraction in the field for the water lance/ water cannon. The thermally induced axial
stress associated with the production water lance/ water cannon is approximately
1/5 that of the stress developed by EPRI simulations.
9862704 3-4
When comparing the cooling curves obtained on waterwall tubes by EPRI and field
experiences, two significant differences are apparent. The following figure illustrates the
EPRI cooling curve. Note that the tube was cooled to approximately 200 F. Note that the
temperature of the tube after cooling recovered to only 650 F. The temperature after
cooling is lower than the temperature before cooling. The temperature of the tube wall
was reduced by 100 F suggesting an overcooling condition (ie.excessive cooling).
A typical cooling curve is shown on the next page. This curve illustrates the tube was
cooled to approximately 605 F after water spraying. The temperature of the tube after
cooling, 730 F, is higher than the temperature before cooling, 715 F. The slag insulated
the waterwall tube from radiant heating causing it’s metal temperature to reduce. This is a
quantitative measurement of the reduced heat absorption of the waterwalls due to
slagging. Also the fact that the tube is cooled to only 605 F indicates better control of
excessive cooling (ie, removing heat content from the tube wall).
9862704 3-5
The presence of axial stresses of waterwall tubes can contribute to cracking. The load on
the tube is added to thermally induced stresses to produce thermal fatigue cracks.
EPRI applied an axial load of 59,000 pounds per panel. This is equivalent to an axial
stress, in each tube, of approximately 3700 lbs./sq. in.
A bend stress of 89,000 inch-pounds was placed on each panel during the EPRI
simulation – which must be added to the axial load. The bend loading on the 77-inch long
panels, when added to the tensile stress, develops a total surface tensile stress of 9656 psi
at 650 F.
A tube supporting 200 feet of its own weight, the axial stress would be approximately
680 lbs./sq. in. The EPRI axial stresses are over fourteen times the stress of the tube
supporting 200 feet of its own weight.
9862704 3-6
4
RELIANT W.A. PARISH FIVE-YEAR ISB PLAN
Rabon Johnson
EPRI I&C Center
4-1
9862704
Reliant W.A. Parish Five-Year
ISB Plan
zBackground
zScope
zMember Interest and Potential Benefits
zFunding
z Background
» 1999/2000: Technology Assessment conducted
– Process reflected in ISB Guidelines Report
» 2000/2001: ISB Project initiated on Unit 8,
expanded to Units 5 and 7
» Mid-2001: ISB Five-Year Plan
– ISB and related systems
– Novel instrumentation technologies
– Test and demonstration bed
z Scope
» Includes all four units
– Two 590-MW T-Fired Units, one upgraded to 650 MW
– Two 690-MW Wall-Fired Units
» Near Term Performance Benefits
– Sootblower Usage
– Boiler Performance
– SCR Impacts
– NOX Emissions
z Scope (Continued)
» Long Term Benefits
– Sootblower Maintenance
– Boiler Availability and Life
– SCR and Emissions
z Funding
» Estimated Cost - $7M, with over $3M in place.
Supported by Reliant and above utilities.
» Reliant is committed to providing the additional
matching funds needed.
» Additional TC credits are being sought.
» In return, funders will be invited to site meetings
for technical updates.
Don Sands
Synergetic Engineering Pty Ltd
GPO Box 783
Brisbane QLD 4001, Australia
Dennis Linkins
Tennessee Valley Authority
1101 Market Street
Chattanooga, TN 37402
5-1
9862704
Intelligent Sootblowing within a Global Intelligent Operations
Management System
Don Sands
Synergetic Engineering Pty Ltd
GPO Box 783
Brisbane QLD 4001, Australia
Dennis Linkins
Tennessee Valley Authority
1101 Market Street
Chattanooga, TN 37402
Abstract
When optimizing a sootblowing regime all of these influences need to be catered for
and weighted so that the optimization is for all influences and not just one aspect.
9862704
5-2
There is little point in having the most efficient sootblowing regime for energy loss, if
that sootblowing regime reduces reliability and increases maintenance.
This paper will discuss how these sometimes conflicted influences can be combined in
an auditable way, so that the appropriate sootblowing regime can be established for
efficiency, reliability, maintenance, life expenditure and emissions using an Intelligent
Operations Management System. Results and current status from a project currently
underway at TVA’s Bull Run Plant to implement Intelligent Sootblowing with economic
optimization will be described.
This paper will also outline how material tracking within an Intelligent Operations
Management system can be added to the Intelligent Sootblowing aspect to achieve
further gains in optimization and capacity gains and provide a sootblowing system that
is proactive to known coal slagging/fouling properties.
During the operation of a coal fired boiler one of the by products of the combustion
process is ash. Typically the ash comprises of silicates, clays, metals and other non
combustible products with its exact characteristics being dependant on the type of coal
and the combustion process. One of the many affects of the ash being part of the coal
fired combustion process is that the ash deposits on the boiler surfaces which will result
in loss of efficiency and if left unchecked can create a loss of reliability by requiring the
unit to be shut down to effect removal of monster clinker. The magnitude of this
problem is dependant on the load profile of the unit, the ash characteristics and the
boiler design. However, the problem is of such a magnitude for all coal fired boilers that
it must be addressed by some cleaning regime during the operation of the boiler.
The major control mechanism for controlling the negative affect of ash build up on the
boiler tubes is the injection of steam to remove the ash and this is commonly known as
Sootblowing (SB). Other cleaning mechanisms other than sootblowing exist such as self
9862704
5-3
shedding, water cannons for furnace cleaning, and ultrasonics for back-end and
airheaters cleaning. We include all cleaning mechanisms that are available in the
cleaning part of the fouling and cleaning cycle. While the cleaning cycle of the program
is where currently the greatest emphasis is the fouling process is also another area
where significant opportunities exist for managing bolier performance. Fouling is a
variable dependent on many influences such as load profile, ash properties and
combustion process. Although any control of the fouling part of the cycle is much
more difficult there are opportunities available in the control of the fouling part of the
cycle.
The removal of ash by SB improves boiler efficiency at the cost of lost steam, increased
maintenance, and life expenditure through fatigue cracking of boiler tubes. There is also
a potential loss of reliability as a result of over SB caused by tube erosion and cutting of
tubes by the steam. The risk of loss of plant also increases if the boiler elements are
getting close to the end of reliable life as a result of fatigue cracking. It is therefore
essential that SB operation needs to be carefully considered to ensure that the boilers is
run efficiently as possible but not at the cost of lost steam, maintenance, life and plant
reliability.
The factors therefore that need to be taken into consideration for optimised sootblowing
are the steam usage, maintenance for repair of erosion of tubes; stress imposed on the
boiler tubes and this is to be balanced against the efficiency gains made by SB within a
load profile. The barriers for determining the optimal SB program are that many of
these factors are varying in real time such as the fouling rate or that the variable is not
measured such as the stress imposed on the boiler tubes or in the case of the efficiency
gain per SB cycle that it is not directly measured and changes in real time.
Due to these limitations SB in the pass has been carried out by adding safety factors and
setting policy for the worst coal type. Variables such as fatigue cracking of the boiler
tubes, or for the time variation of the fouling properties are generally not considered.
Sometimes boiler efficiency is considered through a periodic review of the efficiency
gains and a resetting of the regime, however even this is generally an exception. This
approach results in a far less than optimal SB program. Studies by independent
engineering consulting companies have estimated that a 1% increase in efficiency could
be achieved by improved SB. This equates to a saving of over $1,000,000 Pa for a station
of 1400MW capacity and an improvement in greenhouse emissions of 120,000 tonnes
per year. While efficiency is important to all coal fired power stations, the cost to plant
experiencing a of loss of reliability of plant as a result of the fouling and cleaning cycle
(clinkers forming or tube failures through fatigue) could run into $10s of million per
year in lost opportunity plus the additional expense for replacement of tubes as the
tubes reach the end of their life. The opportunity to increase reliability and life through
optimized sootblowing could be many time more that the opportunity for efficiency
gains.
9862704
5-4
It is therefore evident that the optimisation of SB is a worthwhile objective and this is
where Intelligent Sootblowing (ISB) can make a significant contribution. ISB is the next
evolution of SB where instrumentation measures a number of process variables and
provides feedback on the boiler status and then this information is used to determine a
new SB program. The difficulty in describing a SB system as being Intelligent is as with
all generic labels is that there is a wide range of systems with varying degrees of
performance. The success of the ISB system giving a truly optimal SB program being
largely determined by the systems ability to accurately determine the critical process
variables needed to determine the optimal SB program and adjusting for the real time
behaviour of the boiler.
The removal of ash from the boiler via steam injection sootblowing is essential for
efficient operation of the boiler. The current methods for determining the SB program
result in sub optimal boiler operation which results in an estimated opportunity of 1%
efficiency gain. To address this gap ISB systems are required which take account of the
boiler state and additional parameters. It is critical that for truly optimal SB the ISB can
account for additional parameters, accurately determine these critical parameters and
account for real time changes in the boiler.
Background
TVA and EPRI have entered into a collaborative agreement to implement ISB at Bull
Run. This demonstration project will establish a basis for duplicating similar
technology in other fossil units within TVA that can benefit from ISB technology.
The Bull Run Fossil Plant is located on the East side of Melton Hill Lake in Anderson
County, Tennessee, approximately 12 miles west of Knoxville, Tennessee and 5 miles
east of Oak Ridge, Tennessee.
This unit is a significant baseload asset and contributes a significant amount of energy
to the grid. Due it it loading and contribution, efficiency and reliability are paramount
to such a unit.
9862704
5-5
In the life of the boiler a major source of failure and maintenance expense was the
failure of boiler tubes. Analysis of the problem showed that fatigue cracking also called
circumferential cracking was the major cause of this failure and that the
fouling/cleaning cycle was initiating the fatigue cycle.
The impact of the fouling cleaning cycle has on circumferential cracking is described
below:
Crack Mechanism
The cause of circumferential cracking is usually one or more of the following factors:
• Corrosive environment;
• Cyclic stress;
• Crack initiation.
Through a combination of one or more of the above items circumferential cracking may
occur with catastrophic results.
3. Make thermal cycles smaller (smaller temperature changes) and/or have less
cycles; and
Goal 1 has been addressed at Bull Run by using oxygen treatment. The remianing goals,
2, 3 and 4, are part of the SB program and therefore needs to be addressed by Intelligent
Sootblowing.
9862704
5-6
As a result TVA Bull Run team initated a project to develop a ISB system that would
address objectives 2 to 4 and be completed before the boiler tube replace program
occurring in 2003.
To overcome these significant challenges the URS&S solution utilises a unique patent
pending framework. The framework is referred to as Sentient System which allows for
multiple inputs and correlates the information to ensure that all the inputs provide the
same interpretation. This interpretation is enhanced by processing the inputs through
multiple models each with separate philosophies which adds to the robustness of the
solution. If the interpretation is different it allows reasoning to take place and discard
the input that is conflicting or can enter into a learning mode where models are fine
tuned to ensure a greater alignment of the outputs and therefore a higher confidence in
the solution.
A simple example may best illustrate the URS &S Sentient System. Human beings are
sentient creatures that is the ability to use a range of senses and models to determine a
possible result. An example of this would be a temperature probe in a kettle. If the
reading was 100 deg C and there was steam coming off the top and you could feel the
heat from the kettle then all inputs correlate in the correct manner, however if the same
circumstance occurs with a reading of 40 deg C then you would discard the
9862704
5-7
temperature reading, as from your experience if there is steam coming off the kettle
then the water temperature will be close to 100 deg C. Further information and models
can increase both the accuracy and the confidence in the decision such as including
atmospheric pressure etc. or the person can learn that the radiant heat from an insulated
and uninsulated kettle differ and therefore not all radiant heat levels relate directly to
water temperature but that the insulation must also be taken into account.
We also believe that any automated system needs to understood by a large cross section
of personnel rather than being a black box that provides outputs based on inputs. This
is the reason that we put all our operations management systems into a common
framework that is easily understood by a great number of operational, engineering and
management personnel
A framework of allowing the tube life expenditure to be incorporated with energy loss
so that total optimization is available is paramount in achieving the primary (increase
tube life) and secondary (maintain or improve efficiency) objectives in a defined
manner. It is the unique patent pending framework within URS & S’s ISB that allows
for optimization of conflicting requirements. This Sentient framework also allows other
aspects to be included at any time in the optimization such as emission, maintenance
and reliability.
Primary Measures
To measure the impact of ISB controls over objectives 2 and 3 (see above) we are using
the following:
9862704
5-8
• Chordal thermocouple temperatures which directly measure metal temperature
for objectives 2 and 3.
There is the option of direct and indirect temperature measurement. The use of non-
obtrusive indirect measurement allows the calculation of the fireside metal conditions
as an alternative to direct metal temperature from chordal thermocouples. Initial
results at Bull Run however have proved our initial proposal of being able to soft sense
metal temperature through advanced modelling and existing instrumentation have
been correct eliminating the need for expensive sensors and the associated maintenance.
Secondary Measures
The other measurements being used to determine the change that is occurring in the
boiler and furnaces or to optimize on some other objective are:
• Gas exit temperatures and flow rate to allow performance optimization within
objectives 2 and 3 to prevent circumferential cracking;
There are a number of identifiable inputs that affect the optimization of objectives 2 and
3. These inputs can be classified as either controllable or uncontrollable. Some of these
inputs are:
9862704
5-9
at what pressure and for Hydrodynamics and cooling flows.
how long).
Boiler proportional heat pickup.
Water blasting (intentional
and unintentional). Internal tube fouling.
Fuel and ash constituents. Combustion (air/fuel ratio and combustion intensity).
(if there is scope)
Burner tilt.
Load.
Control mechanism
The main control available for the controllable inputs is the sootblower sequence. URS
& S’s ISB optimizes the sootblower sequence to achieve objectives 2 and 3. URS & S’s
ISB also monitors the sootblowing system to ensure unintentional water blasting is
prevented.
Another control being used to control the intensity of the cleaning cycle is the insertion
depth of the furnace wall blowers and the blowing pressure. The blow pressure and the
insertion depth is a manual adjustment and hence is a once off optimization process
with feedback coming from the ISB system. This along with the optimized sootblower
sequence assists in achieveing objective 2, 3 and 4.
A secondary control available is controlling the fuel supply. This indirectly controls the
fouling cycle and is only available if there is variation in the fuel supply and suitable
facilities to manage different fuel qualities.
• Plant capacity
• Plant reliability
9862704
5-10
• Maintenance costs
• Plant life
• Particulate emission
• NOX emissions
We can expand to another level by including the emission influences such as NOx,
Particulate and CO2 into the solution space which adds to the complexity of the
solution space.
We can increase the complexity of ISB and expand to another level by considering SB
influence on maintenance and reliability.
Yet another level is introduced with different cleaning mechanism such as sonic horns
and water cannons.
If we expand the solution space to now include the quality of coal in determining the
optimum SB regime a new solution space will appear. The quality of the coal is not
fixed nor is it totally variable but also lies within in a solution space through
negotiations with the supplier or in the case of multiple coal sources the blending of
different coal qualities or both. Therefore by tracking coal supplies and controlling
blending the fouling/cleaning cycle may be further optimised.
In the case of a sole supplier of coal the quality of the coal may be varied through the
benefitiation plant and blending grades from various sources within the mine. The
savings from SB alone may not be sufficient to renegotiate the coal supply contract but
if we also consider boiler performance, capacity constraints and emission constraints
etc then further savings may be realised which would make this option more viable.
9862704
5-11
If we could also model the mine performance then a different solution will also occur
for coal quality now taking into account the mine cost, boiler performance and SB.
Sentient system technology and methodology allows the seemless expansion from 3
variables steam, efficiency, tube failure to include all other variables such as boiler
performance, coal quality and mine performance.
The expansion of the variables being used in the development of optimal SB can,
through the URS & S Sentient System, be developed as point solutions and then
integrated as the point solutions interconnect. This has significant benefits as the
implementation of optimal SB can be tailored for each plant, the most effective areas
addressed first and staged so that careful evaluation of improvements can be made
before proceeding to the next stage.
In the case of Bull Run the primary objective is not to minimize energy loss but to
minimize tube failures while maintaining or improving efficiency. As a result the
solution model is different from traditional Intelligent Sootblowing applications.
Overview
As has been shown it is only by expanding the variables which affect the outcome can a
truly optimal solution be found. While intuitively the further removed the variable is
from the process the less affect it has the interrelationship with other processes within
the plant then become more apparent in finding the optimal plant solution.
It is only by applying the technology and methodology of URS & S Sentient System that
the entire wide range of variables can be monitored, modelled and optimal solutions
found.
A further difficulty is found when expanding the range of the variables influencing
plant operation is that accuracy of real time performance is essential as is the
9862704
5-12
measurement of critical process variables many of which are not measure via
instrumentation. The technology and methodology employed by URS & S Sentient
System overcomes this issue as they are in real time altering the models to reflect
current plant performance and use a range of modelling techniques and current
instrumentation to determine the critical but often not instrumented process variables.
The process can also be used to evaluate off line equipment modifications or selection
and different operating regimes using the current real plant performance.
• Schedules sootblowing into zones and times that have maximum benefit to the
overall plant performance with minimum disruption to operations (emissions,
minimum load etc).
It also:
• Predicts where the next optimum time to sootblow is within constraints; and
URS & S’s ISB has been designed to be a stand alone application or imbedded as a point
solution within URS & S’s Intelligent Operations Management System, to be part of a
global Power Plant monitoring and optimization system.
9862704
5-13
to the specific application and address specific problems. The system is flexible enough
to optimize for performance, life expenditure or maintenance and reliability – or a
combination of these in predefined weighting.
Tools:
The system uses installed instrumentation, which means there are no expensive heat
flux monitors to install and maintain. Alternatively, heat flux monitors, temperature
probes or other instrumentation can be incorporated into the system if required.
Our system is self-learning to the extent that it learns the behavior of your plant as a
result of sootblowing events. Based on the learning action, the system is able to predict
the fouling rates, metal temperatures on the boiler elements, reduced efficiency and
hence the estimated change of these parameters as a result of the sootblowing activity.
• Metal temperatures;
• Burner tilt;
• NOx emissions;
• Life expenditure;
• Maintenance; and
• Energy.
This list is not exhaustive and the addition of other parameters is a very simple process.
Within URS & S’s ISB system, it is the prediction of the changes of key parameters as a
result of sootblowing activities and the relationship between these changes allows
optimization of the sootblowing sequence.
9862704
5-14
In the case of the Bull Run project where the primary requirement is to maximize tube
life, it is the ability of the system to predict the changes in metal temperature as a result
of sootblowing activities that allow the fatigue life expenditure to be calculated and
included in the optimization objective function. The secondary requirement of
maintaining efficiency is included in the objective function by determining the change
in efficiency as a result of sootblowing activities.
There are a number of levels of predicting the change of the parameters as a result of
the sootblowing.
These methods can be used independently or at the same time within URS & S’s ISB
system to provide a check and balance on the estimated change.When an actual
recovery as a result of sootblowing does not map to the expected result then the root
cause of why the predicted result did not map to the actual is possible.
ISB allows for a prediction of the unit load is conducted to allow an estimation of the
next sootblow activity. This prediction can be done as a replication of the previous
week’s load profile or as manually entered profile that can be configured to reflect a
known operating regime. This can be made more sophisticated using neural network
models to provide a generation forecast if required although we have not found this to
be necessary.
Learning Mechanism.
The learning mechanism of URS & S’s ISB involves the following:
Predicting Mechanism
The predicting mechanism of URS & S’s ISB involves the following:
9862704
5-15
Determination of the load band the system is operating in over the evaluation period.
Calculating the rate of change of the key parameters (particularly metal temperature in
the case of Bull Run) over the period.
Integration of the absolute change as a result of a sootblow from the last sootblow
event.
Tracking the overall cost structure as a result of the deviation in the key parameters.
The costs incurred as a result of changes in the metal temperature is related through a
fatigue life assessment.
Predicting the next sootblow event based on minimizing the overall tube life
expenditure and costs.
Action Mechanism
The actions as a result of the learning and the predicting mechanism of URS & S’s ISB
allow the system to provide advice when a sootblow is required and can take automatic
action when operating in closed loop control. Action can be a sootblow sequence
developed by URS & S’s ISB, a real-time signal to initiate a sootblow, launch a dialog or
procedure to the operator, send out a pager or email message to a designated person.
The types of actions can be continuous control actions or exception actions as a result of
some criteria being met. The criteria may be set with hard and soft limits or derived
from historical data analysis. In the case of the Bull Run project URS & S’s ISB provides
the optimized sequence to the sootblower control system for the optimum sootblow
regime (closed loop control to achieve waterwall protection) and to provide exception
reports when a deviation from the expected behavior is encountered.
When the actual sequence deviates from the ISB sequence (a possible communication
problem or a sootblower is unavailable);
When the times allowed for each sootblower in the sequence deviate. (This will prevent
sootblowers being left in or jammed);
When the change of key parameters such as waterwall metal temperatures, back-end
temperatures, mass flow, burner tilt or NOx levels deviates from the expected change.
(This will indicate a damaged sootblower, or a change in the fouling properties of the
fuel.)
9862704
5-16
Benefits of URS & S’s ISB system
• Minimizes total life costs associated with sootblowing (life expenditure, energy
loss, emissions, reliability, maintenance);
• Provides early warning when conditions change (eg. sootblow not effective,
change in monitored parameters);
• Self learning: The system automatically changes with changes to operation, coal
type and sootblower availability, minimizing any maintenance of the system;
• Can be turned on and off without affecting the normal sootblower control.
Additional benefits arising from the adoption of URS & S’s Intelligent Operations
Management systems are:
URS & S’s Intelligent Operations URS & S’s Intelligent Operations
Management Expert Analysis Management Analysis Structure
Provides the ability to capture and Standard way to bring point analysis
deploy in real time expert knowledge into a global1 analysis system
1
Global Analysis meaning analysis and optimization across multiple components, point solutions, units and power stations. At its highest
level the global analysis and optimization would cover an entire organization and its suppliers and consumers.
9862704
5-17
Ability to bring in third party analysis to
increase the resolution of the process
URS & S’s ISB uses raw instrumentation data and performance data. Typical
information used is:
• NOx;
• Opacity;
• Spray flows;
• Metal temperatures;
• Nozzle temperatures;
• Chordal thermocouples.
Acknowledgments
We wish to thank the following personnel from EPRI and TVA for their valuable
contributions to this project:
9862704
5-18
Appendix A
Fouling / Cleaning Cycle –Influences
Within a fouling cleaning cycle there are numerous influences on the operations. The
main influences are as follows:
Fatigue life and creep is related to the base metal temperature, the temperature cycles
and the corrosive environment. As the change in temperature increases a higher loss of
life is experienced due to the cycle. The higher the number of cycles the quicker the loss
of life.
Emissions
The effect that sootblowing has on emissions is related to change in NOx emissions due
to changes in burner tilt, combustion intensity and excess air. These are variables in the
fouling cleaning cycle and typically follow the cleaning cycle. By monitoring the
influence that sootblowing has on NOx emissions, NOx emissions can be introduced into
the objective function. CO2 emission are also influence by sootblowing as a result of
changes in the efficiency of the plant. This is described below.
Efficiency
The effect of fouling / cleaning on the cycle efficiency is due to energy loss from the
back-end of the boiler as a result of fouling and a reduction in the heat transfer that
occurs in the boiler. The other efficiency loss from the cycle due to the cleaning is the
energy loss as a result of using steam to sootblow the fouling from the tubes.
Performance and efficiency can be included in the objective function.
Maintenance
The effect of sootblowing on the boiler maintenance can be included in the optimization
provided that sufficient maintenance cost information is available and can be related to
sootblowing activities
Boiler Weights
The effect of clinkering in the pendants and fouling in the furnace can be monitored
using their weight on the boiler hangers. This patent applied for use of weights to
determine fouling within a boiler can be used to determine the relative positions of
fouling in the furnace area and the weight of the pendants can be used to provide an
indication of where clinker has built up for overhaul preparation. It can also be used
within the ISB system to provide information on the effectiveness of a sootblow and
9862704
5-19
included in the optimization provided that strain gauges are installed on the boiler
hangers.
The effect of the fouling / cleaning cycle on boiler reliability can be included in the
optimization if sufficient information is available regarding loss of availability of plant
due to the fouling and cleaning cycle. SynEngCo’s Intelligent Operations Management
system can provide a mechanism for capturing reliability issues and associating this
loss of reliability to plant and operation on the plant.
Material Tracking
By using the system in association with the material tracking there is the ability to relate
sootblowing activities and the effectiveness of the sootblowing activities with the coal
being received from various mines.
• ISB to optimise the sootblowing regime and minimise the energy loss and
maintenance requirements.
With the combined functionality the following additional benefits are available:
• Relationship between the fouling properties and where the material comes from
provides an increased knowledge of changes to the quality of the inbound coal
deliveries.
Other Influences
Any other influence can be incorporated into URS & S’s ISB system. The weighting
given to the influence provides the weighting in the optimization function.
9862704
5-20
Optimization
As fuels and the slagging properties change, the optimum fouling / cleaning cycle may
change. This corresponds to a changed base metal temperature target and a change in
metal temperature due to the cycle. Material tracking can be used to provide early
warning of a difference in fuel type and provide a feedforward to be proactive in
establishing an optimum cleaning cycle.
This optimization methodology automatically caters for all the influences associated
with the fouling and cleaning cycle and the preservation of tube life (base metal
temperature, cyclic metal temperature and number of cycles). It also automatically
caters for the conflicting requirements of sootblowing. For example, achieving
maximum tube life may result in lower performance.
URS & S’s ISB system uses a number of optimization techniques including genetic
algorithm-based optimizer, which can find optimal solutions to problems that cannot be
solved with standard linear and non-linear optimizers (discrete or discontinuous
domains). Genetic algorithms (GA) are based on the theory of evolution with natural
selection and survival of the fittest. ISB optimization is achieved by breeding
sootblower cycle solutions and selecting the solutions with the greatest fitness then
continually breeding from these. Mutations are also allowed to check that a better
solution cannot be found in another domain. Genetic algorithm optimization
automates “what-if” scenarios in a fast, systematic manner.
URS & S’s ISB system and methodology provides a systematic approach to sootblow
optimization, with influences involved for the objective function easily identified and
weighted. URS & S’s ISB provides a clear audit trail for optimization and the ability to
add other influences very easily.
URS & S’s Intelligent Operations Management system also provides the framework for global
plant optimization and has knowledge capture and deployment functionality that has application
for sootblower control and other process areas including the entire plant cycle.
9862704
5-21
Appendix B
Sentient System Features
• Over 160 I/O devices supported to Databases, Historians, PLC, DCS, SCADA
and Dataloggers
Can alter models from monitoring real time plant performance such that the model
aligns with actual performance.
• Specific knowledge
Knowledge captured at all levels to allow high level information to be drilled into.
• Management
9862704
5-22
• Engineering
• Maintenance
• Trading
• Operations
• Dedicated clients
• Email clients
• Pager
• PDA
Modular and integrated in its application. i.e. from any point solution to enterprise
solution.
Can model off line using current plant performance to evaluate equipment and
operational modifications.
9862704
5-23
6
FEGT SYSTEM TESTS AND OVERVIEW OF ISB AT JP
MADGETT STATION
Duane Hill
Manager, Performance Admin.
Dairyland Power Cooperative
6-1
9862704
FEGT System Tests and
Overview of ISB at
JP Madgett Station
Duane Hill
Manager, Performance Admin.
Dairyland Power Cooperative
6-2
9862704
SUPERHEATER
SHTR
PENDANTS
RHT
WINGWALLS
R
PRIM SH
RHTR
RHTR PRIM SH
ECON. ECON.
JPM Boiler
6-3
9862704
Gas Temperatures at JPM
6-4
9862704
Current Gas Temperature
Measurement Status
6-5
9862704
Test Purpose
6-6
9862704
Types of Instruments Tested
6-7
9862704
Test Criteria
6-8
9862704
Test Exceptions
6-9
9862704
JPM Gas Traverse 9th Elv
1900
1800
Degrees F
1700
1600
2 ft 4 ft 6 ft 8 ft 10 ft 12 ft 14 ft 16 ft 18 ft 20 ft
6-10
9862704
JPM Gas Traverse 8th Elv
1800
Temperature, Deg F
1700
1600
1500
1400
1300
2 ft 4 ft 6 ft 8 ft 10 ft 12 ft 14 ft 16 ft 18 ft 20 ft
6-11
9862704
JPM Gas Traverse, 7th Elv
1450
Temperature, Deg F
1400
1350
1300
1250
1200
1150
1100
2 ft 4 ft 6 ft 8 ft 10 ft 12 ft 14 ft 16 ft 18 ft 20 ft
6-12
9862704
JPM Gas Traverse, 6.5 Elv
1100
Temperature, Deg F
1000
900
800
700
600
2 ft 4 ft 6 ft 8 ft 10 ft 12 ft 14 ft 16 ft 18 ft 20 ft
6-13
9862704
Test Results
6-15
9862704
Conclusions
6-16
9862704
7
ACOUSTIC CLEANING OF REGENERATIVE
AIRHEATERS SUBJECTED TO PRB ASH
Mel Freund
Advanced Acoustic Technologies, LLC
Monkton, MD
Randy Vann
Xcel Energy, Roy Tolk Station
Muleshoe, TX
7-1
9862704
ACOUSTIC CLEANING OF REGENERATIVE
AIRHEATERS SUBJECTED TO PRB ASH
Mel Freund
Advanced Acoustic Technologies, LLC
Monkton, MD
Randy Vann
Xcel Energy, Roy Tolk Station
Muleshoe, TX
Mel Freund
Advanced Acoustic Technologies, LLC
3022 Shepperd Road
Monkton, MD 21111
Randy Vann
Xcel Energy, Roy Tolk Station
Muleshoe, TX
Abstract
Acoustic cleaning of fireside deposits has been done for over twenty years. It has been
used as a remedial tool for most applications. Infrasound has been shown to be effective
in removing ash deposits in applications where large heat exchanger elements are used.
It has been particularly effective in removing the aggressive deposits formed by PRB
ash from regenerative air heaters. Xcel Energy’s Roy Tolk Station has been using
infrasound to clean Ljungstrom air heaters subjected to PRB deposits since 1996. Data is
presented, showing the effectiveness of infrasound versus sootblowing as measured by
pitot tube tests of relative flow rates in two side-by-side air heaters. Criteria for
equipment justification, cleaning mechanism, hardware, performance optimization and
cleaning results are all discussed.
Conclusion
Background
Xcel’s Roy Tolk Station lies in the panhandle of Texas, about 80 miles northwest of
Lubbock. There are two CE 565MWe units firing low sulfur coal from the Powder River
Basin of Wyoming (PRB). Unit 1 went on line in 1982, with Unit 2 following in 1985.
The effect of the infrasonic cleaners was unknown since they were used in tandem with
the sootblowers. Pressure drop across the air heaters remained relatively constant. The
original infrasonic cleaners were de-commissioned in 1989.
Within a couple of years the pressure drop across the air heaters started to climb.
Because this is typical of general industry experience with PRB ash in air heaters, it is
not possible to know whether the removal of the sonic cleaners from service contributed
to the degradation in cleanliness.
Summary
A newer, more powerful infrasonic cleaner was installed on the West air heater of Unit
1, and started service in May of 1996. This cleaner was installed on the gas outlet,
operated at 22 Hz, and was capable of outputs up to 5000 W acoustic power. Based on
experience from the latter 1980s with infrasound on PRB ash deposits in regenerative air
heaters, it was felt this would be a much-improved combination.
Because the air heaters are in parallel on the gas side, pressure drop data is not capable
of determining the difference in cleanliness between the two. Pitot tube tests were
performed to evaluate the performance of the infrasonic cleaner by showing the relative
difference in gas flow between the air heaters. Each test comprised a 54-point traverse
of the inlet gas duct of both the East and West air heaters.
Figure 1 is a plot of the flow differential between the air heaters. It can be seen that the
relative cleanliness of the West air heater steadily increased when compared to the East.
In November of 1996, new hot and intermediate layer elements were put in the Unit 1
air heaters. At that time it was also decided to switch the West air heater sootblower off
to get a feel for how much of the cleaning was being done by the infrasonic cleaner. The
trend continued as evidenced by Figure 1.
The highest measured difference in flow was experienced in August of 1998, when the
West air heater was passing over 22 percent more flow than the East one.
During this period a decision was made to install more infrasonic cleaners to take
advantage of the cleaning improvements. New infrasonic cleaners were installed on
both air heaters of Unit 2 in the spring of 1999, along with replacement of all three
The pressure differential was starting to climb across both air heaters of Unit 1 (because
the gas flues are in parallel, the pressure drop across both air heaters is necessarily the
same), so a high-pressure wash was done in February of 2000. By December of 2000 the
flow difference was back up to almost 10 percent.
To see if any measurable change would register, the sootblowers on Unit 2 air heaters
were operated. Since a reduction in pressure drop of about ½ inch resulted, the
sootblowers were put back in service along with the infrasonic cleaners. That is how
things remain on Unit 2 to the time of this writing. There has been no measurable
increase in pressure drop since.
Because of the results on Unit 2, the West air heater sootblower on Unit 1 was placed
back into service as well. In addition, a fourth infrasonic cleaner was purchased in 2001,
and is scheduled for commissioning in February of 2002. This will complete the
complement of cleaning equipment.
Evaluation
Pressure Drop
The cost of increased pressure drop through a regenerative air heater is substantial. Fan
power requirements increase through both the ID and FD circuits.
Depending on the criteria used for cost evaluations, the actual cost figures can vary
substantially. Fuel costs have been used historically for such evaluations, but the cost of
replacement power has been used by some in recent years due to the competitive nature
of the market
At the point where 17.3 percent more gas flow was going through the West air heater on
Unit 1, this corresponded to about 1½ inches of increased pressure drop through the air
heaters if neither air heater had infrasonic cleaning, versus both having it. On the Tolk
boilers that corresponded to a 700 kW increase in fan power consumption. In the worst
case this could be an annual loss of margin of $ 140,000, based on a gross margin of 3.5
ct/kWh.
PRB ash contains a high percentage of CaO, which makes gypsum after reacting with
sulfur in the flu gas, mixing with water and then being allowed to dry in place. This
happens to a certain degree every time a regenerative air heater is washed that has PRB
ash residing in it. No water washing that is done with the air heater elements in place
will remove all ash, but all ash that is not removed will be hydrated and will harden
into a deposit that is all but impossible to remove.
Experience with inadequate washings has led to practices where extremely conservative
washing schedules are used. Unfortunately even these practices cannot remove all
deposits. Channeling occurs that takes water in the intermediate elements through open
paths and leaves deposits behind. Without being able to force water through all points
of all passages, this cannot be avoided. This is espescially true for some of the high
thermal efficiency element designs currently in use.
Even when water washes are avoided at all costs, if there is PRB ash in the elements, it
will eventually turn hard during an outage. As the air heater cools, water condenses out
that mixes with the ash. This dries and hardens into the same cement.
The only way to avoid these deposits, which eventually force replacement of the
elements, is to keep the deposits from forming in the first place. If the PRB ash is not
allowed to accumulate, the cement cannot form.
One thing that can be said of the experience at Tolk is that there has never been a need
to wash the air heaters when infrasonic cleaning and sootblowers were used in tandem.
Whether you consider the first 7 years or so of operation, or Unit 2 since installation of
the infrasonic cleaners in 1999, there has not been a degradation of air heater pressure
drop that required water washing.
With fuels other than PRB, air heater element replacement has generally come when the
metal has been eroded by sootblowing, or when the temperature cycles on the elements
are numerous enough to cause failure. With PRB ash, however, the elements become
unusable when they have accumulated enough “cement”. This phenomenae is clearly
demonstrated by plant data, see Figure 2 for details.
Elements have been redesigned to allow the plates to shift, sometimes referred to as
“loose pack”, to allow scale to break up and be removed. These improve the situation,
but not allowing the ash to form in the first place is a preferred solution.
Application of Infrasound
Infrasound has properties that separate it from higher frequencies. The most significant
of these is its ability to fill large enclosures with resonant sound. This allows one
infrasonic generator to provide enough sound intensity to clean large areas. The
technology has matured and application is now based on three-dimensional acoustic
modeling to determine operating frequency and optimum location of the infrasonic
cleaner.
The acoustic modeling has not only provided more consistent results of applied
infrasonic cleaners, it has also explained why past installations did not work. In some
instances, infrasonic cleaners were changed to produce different frequencies, and/or
moved to a more preferred location.
Figure 3 shows a three-dimensional mesh of the inlet and outlet duct areas of Unit 1 air
heaters at Tolk Station. The model uses this mesh to calculate the sound levels within
the boundaries. Physical obstructions (such as the heating elements) and structures
within the volume are considered, along with temperature, gas velocity, and the
location of the sound sources.
Figure 4 shows the sound levels for different frequencies of injected sound energy. The
sound frequency that provides the most efficient utilization of energy to produce the
required cleaning is the one that is chosen. Because of these results, the original location
and frequency chosen for the first three infrasonic cleaners were retained for the fourth
and final application.
Figure 5 shows the sound levels at the boundaries of the different element layers within
the air heaters. One of the benefits of Ljungstrom type air heaters is that small areas of
low sound levels will not adversely affect cleaning. Because the elements are constantly
moving, if the sound level is low one moment, it will increase the next moment.
Infrasonic Hardware
To generate lower tones efficiently, the equipment must get longer, like musical wind
instruments. Infrasonic generators are about 14 feet long and can have outlet openings
about 40 inches in diameter. They typically operate in the 15 – 35 Hz range. Only one
device is required per air heater, and since low-pressure air is adequate, a positive
displacement blower can supply the air. This significantly reduces the electric energy
required, eliminates the need for a receiver tank, and reduces the length of piping runs.
0.25
22%
0.2
Flow Difference (%)
17%
14%
0.1 10%
HP Waterwash
0.05
4%
2%
0
Nov '95 May '96 Sep '96 Nov '96 Apr '97 Nov '97 Aug '98 Feb '00 Dec '00
12
APH Pressure Drop (IN H2O)
11
10
Nov-89
Apr-93
9
Dec-94
Jun-94
8
6
0 5 10 15 20 25
months after APH wash
Gary M. Gilmartin
ASI
8-1
9862704
Evaluation of a Non-Intrusive
Heat Transfer Sensor
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Introduction
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Presentation Topics
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Current ISB - System
Closed Loop
Feedback System
Sootblowing
Measurement of FEGT
Heat Transfer
Furnace Cleanliness EGOT
Module Software
Adaptive Controls to
Furnace
Direct Cleaning Cleanliness
Module
Efforts
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Current ISB - Standard Heat Flux
Sensor
Tube Mounted Device
Over 1200 in Service
Rugged and Reliable
Including Spare
Thermocouples
Calibrated to Measure
Crown Heat Flux
Recognized as the
Standard for
Continuous
Measurement of
Furnace Heat Transfer
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Potential Benefits of a Non-Intrusive
Sensor
Potential Cost Savings
In-situ Installation, No
Tube Removal
Required
Potential for More
Detailed Furnace
Mapping
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Heat Sensor – Theory
Use the Clarke Chapman Membrane Method of
Heat Transfer Measurement
Relate Delta-T Between Back of Tube and
Membrane to Heat Flux
Standard
Sensor
Type K TC’s-Non-Intrusive
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Design
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Evaluation
Evaluation Standards:
– Can the non-intrusive sensor be related to crown heat
flux?
– Is the relationship consistent for all sensors?
– Can the sensors be used to determine fouling condition of
the tube to drive sootblowers?
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Evaluation
Temperature (F)
880.7 to 887.7
873.7 to 880.7
866.7 to 873.7
859.7 to 866.7
852.7 to 859.7
Calculated
845.6 to 852.7
838.6 to 845.6
831.6 to 838.6
Results
824.6 to 831.6
817.6 to 824.6
810.5 to 817.6
803.5 to 810.5
Assumed No
Boundary
Between Tube
and Membrane
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
50 80
Trend Appears to 40 60
deltaT 6239
– Arrows indicate
Flux
30 40
sootblower operations
– Graph over 7 hours 20 20
10 0
02:56:30 AM 03:45:30 AM 04:37:02 AM 05:25:30 AM 06:13:30 AM 07:01:02 AM 07:49:02 AM 08:37:30 AM
03:20:02 AM 04:10:30 AM 05:02:00 AM 05:50:00 AM 06:37:30 AM 07:25:00 AM 08:13:30 AM 09:01:02 AM
deltaT Flux
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
6239 VS deltaT state 22nd
90 40
Same 80 35
Location, 70
Different Day
30
60
25
Sootblower 50
Flux
20
Operations 40
15
Affect Crown 30
10
not Delta-T
10
0 0
0:00:00 2:24:00 4:48:00 7:12:00 9:36:00 12:00:00 14:24:00 16:48:00 19:12:00 21:36:00 0:00:00
deltaT
flux deltaT
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
100
Correlation Over 80
Several Days 70
The Relationship 60
of Delta-T to Flux
50
Heat Flux is 40
Moving Target 30
Results 20
Predicted by a 10
Study in 1976 0
0 5 10 15 20 25 30 35 40 45
deltaT
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
Technologies at -20%
-30%
EuroCorr2000
-40%
Conference. -50%
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Results
chance to determine 8
clean as dirty 6
– Non-intrusive sensor 4
other techniques of 0
19th 20th 21st 22nd 23rd 24th
clean and dirty Day
determination
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Heat Transfer Sensor –
Results 50
Heat Transfer v s. Position
10 0%
45 90 %
Important on Crown
30 60 %
25 50 %
20 40 %
NOT Membrane 15
10
30 %
20 %
5 10 %
0 0%
0 0.2 0 .4 0.6 0.8 1
D istance from Tube C L
Heat Flux
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Application
Must be Membrane
Tube Design with
Integral Fin
Measures Slagging of
Membrane not Tube
Not Useful for
Intelligent Sootblowing
Applications in Sub-
Critical Boiler
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Non-Intrusive Sensor – Summary
Three Sensors
Constructed Under Ideal
Conditions
Extended Testing of Heat
Transfer Measured by the
Difference in Temperature
Between Membrane and
Rear of Tube
Method Unreliable for
Advisory or Direct
Sootblower Controls
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
Questions
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
9862704
Applied Synergistics Proprietary Information ©2002
Printed with permission of the copyright holder.
9
WATER CANNON APPLICATION AT ALABAMA
POWER’S PLANT MILLER
Mike Carlisle
John Sorge
Brian Mead
Southern Company
9-1
9862704
Water Cannon Application at
Alabama Power’s Plant Miller
Mike Carlisle
John Sorge
Brian Mead
Southern Company
• 660 MW
− Commercial operation - October 1978
• Boiler
− B&W opposed wall-fired furnace
− 56 burners B&W DRB low NOX burners
− Compartmentalized windbox
9-3
9862704
Plant Miller
9-4
9862704
The Problem
9-5
9862704
Quench Cracking
Refs:
Water Blowing of Fireside Deposits in Coal-Fired Utility Boilers (EPRI CS-4914)
9-6
9862704
Quench Cracking
crazing pattern
on the surface
Circumferential
crack leading to
steam leak cracks in
progress
tube at nose arch
2.75" OD x 0.290" MWT, SA210 A-1
cleaned by water lance 9-7
9862704
Primary Goals
9-8
9862704
Miller 1 Installation
Clyde Bergemann SmartCannon
9-9
9862704
New for Miller 1 Installation
9-10
9862704
Major Components
Existing Sootblowers
Water Cannon
Heat Flux Sensor
9-12
9862704
Testing
9-13
9862704
Testing
9-14
9862704
Program Schedule
9-15
9862704
10
IMPULSE DEPOSIT REMOVAL SYSTEM
10-1
9862704
Impulse Deposit
Removal System
EPRI Project 1759\C721
• Definitions:
– Pulse combustion represents a
spontaneous combustion process
where sound, i.e., a standing
acoustic wave is created due to
specific design of a pulse
combustor.
– Relatively low flame propagation
speed and pressure
Impulse Deposit
Removal System
9862704 HTI (770) 804-9309 10-10
EPRI (650)855-2895
Impulse Deposit
Removal System
• Design:
– A. Boiler Simulator
– B. Control Panel
– C. Safety System
– D. Instrumentation System
fuel
air
HMI
PP
Air CV Fuel CV
10.00
0.00
1
5
9
13
17
21
25
29
33
37
41
45
49
53
57
61
65
69
73
77
81
85
89
93
97
101
105
109
-10.00 Series1
-20.00
-30.00
-40.00
Next Steps:
• Transfer the project to Phase
Three:
– Find a suitable host utility site
– Modify the IDRS to comply with
requirements of a system operating
in conjunction with a utility boiler
– Test the IDRS at various locations of
a utility boiler where the technology
could be of assistance
Impulse Deposit
Removal System
9862704 HTI (770) 804-9309 10-31
EPRI (650)855-2895
11
AMEREN SOOTBLOWING OPTIMIZATION UPDATE
Mark Litzinger
Ameren
No paper submitted.
11-1
9862704
12
SOOTBLOWING OPTIMIZATION AND INTELLIGENT
SOOTBLOWING
Nenad Sarunac
Carlos E. Romero
Energy Research Center
Lehigh University
Bethlehem, PA 18015
12-1
9862704
Sootblowing
Optimization and
Intelligent Sootblowing
Nenad Sarunac
Carlos E. Romero
Energy Research Center
Lehigh University
Bethlehem, PA 18015
12-2
9862704
Introduction
• Ash and slag deposition on boiler HT
surfaces is a primary cause of reduced
operating efficiency in fossil-fired boilers.
• Represents a major concern for coal-fired
units.
• Sootblowing is used to control the level of
ash and slag deposits on boiler HT
sections.
• While ash and slag deposition is a gradual
process, sootblowing results in an abrupt
change in local heat transfer.
transfer
12-3
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Performance Gradually Deteriorates
1,030 70
60
1,020
50
Steam Temperature [F]
30
1,000
20
0.85
0.80
0.75
0.70
0 1 2 3 4 5 6 7 8 9
Elapsed Time [hr]
12-5
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Activation of IRs Causes Abrupt Change
in Heat Transfer to Furnace Walls
2,820
Sootblowing
2,810
2,800
2,790
FEGT [F]
2,780
2,770
2,760
2,750
13:00 14:00 15:00 16:00
Time
12-6
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Activation of IKs Causes Abrupt Change
in Convective Pass Heat Transfer
0.33
IK 4
0.32
CF Pendant SHT
0.31
0.30
14:00 15:00 16:00 17:00 18:00
Time
12-7
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Furnace Sootblowing Reduces NOx
Emissions – Results are Site-Specific
0.42
Waterwalls Sootblowing
0.41
NOx Emission Rate [lb/MBtu]
0.40
0.39
0.38
0.37
0.36
11:15 11:45 12:15 12:45 13:15 13:45 14:15 14:45
Time
12-8
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Thermal NOx is Affected by FEGT
0.44
NOx Emission Rate [lb/MBtu]
0.42
0.40 O2 = 2.3%
0.38
O2 = 2 %
0.36
2,740 2,750 2,760 2,770 2,780 2,790 2,800 2,810 2,820
o
Furnace Exit Temperature [ F]
12-9
9862704
Sootblowing Affects Steam Temperatures
and Desuperheating Sprays
550 SB Test 6
Baseline Upper elevation SB Test 7 SB Te s t 8
Up p er furna ce SB Test 10
Test furnace blowers Mid d le fu rna ce
ele va tion, Clean RHT:
(repeat of ele va tion:
m os tly cold IKs 59 A and B
SB Test 5) All b lowers
corners
545
Main Steam Temp. [C]
540
SB Test 9
535 Lower
furnace
elevation:
All blowers
530
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00
Time (Wednesday, December 19, 2001)
12-10
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Location of Furnace Wallblowers
CAMERAS
FEGT B 8 Sou th
51 54
Hot Corne r Sootblowe rs
50 53
Ea s t 52
49
4 Top 3 9 ,4 0 ,4 5 ,4 6 ,5 1 ,5 4
FEGT A 48 7 B Mid 2 1 ,2 2 ,2 7 ,2 8 ,3 3 ,3 6
47 6 Bottom 3 ,4 ,9 ,1 0 ,1 5 ,1 8
42 EL 6 9 0 ft
46 41
3 Cold Corne r Sootblowe rs
5 A 2 4 0 Wes t
39
38 36 Top 3 7 ,4 2 ,4 3 ,4 8 ,4 9 ,5 2
45
37 35 Mid 1 9 ,2 4 ,2 5 ,3 0 ,3 1 ,3 4
44
1 33 Bottom 1 ,6 ,7 ,1 2 ,1 3 ,1 6
North 43 34
32
30 31 Removed for SOFAs: 2 0 ,2 3 ,2 9 ,3 2
24 Locked Sootblowers: 1 0 ,3 5 ,1 1
29 23
28 22 EL 6 7 9 ft
21 COAL BURNERS
20
27
19 18
26 15
25 14 17
13 16
12
11 6
EL 6 3 3 ft
10 5
4
3
9
8 2
7 1
12-11
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing Affects Steam Temperatures
and Desuperheating Sprays
50
Baseline SB Test 7 SB Te s t 8
Test Up p er fu rna ce Mid d le fu rna ce
eleva tion, eleva tion:
m os tly cold All b lowers
40 corners
SHT Spray Valve Opening
SB Test 6
30 Upper elevation
furnace blowers
(repeat of
SB Test 5)
SB Test 9 SB Test 10
20 Lower Clean RHT:
furnace IKs 59 A and B
elevation:
All blowers
10
0
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00
Time (Wednesday, December 19, 2001)
12-12
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Controllable Range for Steam Temperature
555
Valve A
550 Valve B
MST Temperature [F]
545
540
535
530
525
0 5 10 15 20 25 30 35 40 45 50
SHT Attemperating Spray Valve Opening
12-13
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing Affects Steam Temperatures
and Desuperheating Sprays
545
Baseline
SB Test 6 SB Te s t 8
Test
Upper elevation Mid d le fu rna ce
furnace blowers eleva tion: SB Test 9
540 (repeat of All b lowers Lower
Hot Reheat Steam Temp. (Deg. C)
SB Test 5) furnace
SB Test 7 elevation:
Up p er fu rna ce All blowers
eleva tion,
535
m os tly cold
corners
530
525
SB Test 10
Clean RHT:
IKs 59 A and B
520
515
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00
Time (Wednesday, December 19, 2001)
12-14
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Current Sootblowing Practice
• Typically the operator is provided with
little or no guidance concerning
sootblowing.
• Sootblowing is typically done on a rigid
time schedule (sootblow because it is time
to do it, not because it is needed).
• Large variations between the operators.
12-15
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing
Optimization
12-16
9862704
Sootblowing Optimization Goals
12-17
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing Optimization: Maintain
Optimal Surface Cleanliness
250 1,040
Change in Unit Heat Rate [Btu/kWh]
980
150 960
940
100 920
900
50 880
860
0 840
70 75 80 85 90 95 100
12-18
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing Optimization:
Reduce Emissions and Minimize Impact on Heat Rate
150 0.68
125
0.66
Heat Rate Penalty [Btu/kWh]
0.64
75
50
0.62
25
0.60
0
-25 0.58
70 75 80 85 90 95 100
Waterwall Cleanliness Factor [%]
12-19
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sootblowing Optimization Methodology
• Step 1:
1 Instrumentation and
Calculations Setup
• Step 2:
2 Sootblower Characterization
• Step 3:
3 Data Analysis
• Step 4:
4 Development of Sootblowing
Strategy
• Step 5:
5 Implementation and Evaluation
of the Strategy
12-20
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Step 1: Instrumentation and
Calculations Setup
• Install necessary instrumentation on
the steam side and fireside.
• Setup calculations for the convective
pass.
12-21
9862704
Install Instrumentation for FEGT Measurement
Why is Measurement
of FEGT Needed?
• Uncertainty in calculated
R e t ra c t a b le
S o ot b lo we r s value of FEGT is of the
same order of magnitude
as the difference in FEGT
FEGT corresponding to the clean
and dirty furnace
Fur n a c e
Wa llb lo we r s conditions.
• Calculated values of FEGT
do not always follow the
same trend as
measurements.
12-22
9862704
Setup Convective Pass Calculations
Mmst Mrht
Tsht,out Trht,out
Psht,out Prht,out
Tww,out
Pww,out
FEGT
SHT RHT(1)
(Pendant) Pendant
FURNACE
(WW)
Tpsht,out , Ppsht,out PSHT
Mmst,spray
Trht,hor,out
RHT Prht,hor,out
Horizontal
Teco,go
Mrht,spray
12-23
9862704
Step 2: Sootblower Characterization
• With operator’s help, group sootblowers
into groups.
• Perform characterization tests for each
sootblower group:
– Activate one sootblower group at a time.
– Record plant response, i.e., effect on
steam temperatures, attemperating sprays,
NOx emission rate and opacity.
12-24
9862704
Typical Response to Sootblower
Group Activation
Boiler C: Wall Blowers - Hot Corner
1,120 85
Main Steam
Reheat Steam
WWCF
1,100
1,080 80
1,060
1,040 75
1,020
1,000 70
0.00 0.25 0.50 0.75 1.00 1.25
Time [hr]
12-25
9862704
Step 3: Data Analysis
• Use sootblower characterization test data
to create a database (SBC DBase)
DBase on the
effect of sootblower groups on:
– Boiler section cleanliness status
– Steam Temperatures
– Attemperating Sprays
– Gas Temperatures
– NOx Emissions
– Opacity
12-26
9862704
Effect of Sootblower Groups on Steam
Temperature
1,050
Hot Reheat Steam Temperature [F]
SIDE WALL: IR 17
1,040
1,030
IK 1&2
1,020
HOT CORNER: IR 11-20
1,010
1,000
0.00 0.25 0.50 0.75 1.00 1.25
Time [hr]
12-27
9862704
Effect of Sootblower Groups on Emissions
1.02
1.00
Normalized NOx [NOx/NOx,0]
0.98
0.96
0.92
70 75 80 85 90 95 100
Water Wall Cleanliness Factor, CF-WW [%]
12-28
9862704
Steps 4 and 5: Development and
Implementation of Sootblowing Strategy
• Use SBC DBase to develop sootblowing strategy
that satisfies optimization objective and operating
constraints.
• The resulting strategy could be schedule- or
event-driven.
driven
• Implementation Options:
– Manual (Written schedule provided to the operator)
– Semi-Automated or Automated (Plant control system
generates alarms to prompt the operator or automatically
activates sootblowers at appropriate times.)
– Intelligent Software (Provides expert advice to the
operator or activates sootblowers.)
12-29
9862704
Evaluation of Sootblowing Strategy – Boiler A
Optimized Sootblowing Evaluation, Boiler A
40 100
WWCF
HR Penalty 90
35
80
70
30
60
25 50
40
20
30
20
15
10
10 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
Number of Days
12-30
9862704
Evaluation of Sootblowing Strategy – Boiler B
Evaluation of Sooblowing Schedule, Boiler D
20 280
Reheat Temperature
15 270
260
10
NOx @ 3% O2 [ppm],
250
5
NOx
240
0
230
-5
220
-10
210
Burner Tilt
-15 200
-20 190
0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00
12-31
9862704
Intelligent
Sootblowing
(ISB)
12-32
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
What is Intelligent Sootblowing ?
ERC Definition:
An adaptive process which uses
Artificial Intelligence techniques to
maintain optimal boiler surface
cleanliness (determine what boiler
when subject to the
sections to clean and when),
optimization goal and operating
constraints.
constraints
12-33
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sample Optimization Goals
• Lowest emissions
• Target value of FEGT to control slagging
• Minimal performance penalty
• Lowest emissions with predetermined
performance penalty
• Maximum generation
12-34
9862704
Sample Operating Constraints
12-35
9862704
Why is Intelligent Sootblowing Needed?
• Maintain optimal boiler surface cleanliness.
• Remove operator variability.
• Deal with changes in fuel quality and
maintenance condition of the equipment.
• Relieve operators of additional duties.
Operators are already pressed by:
- Emission constraints
- More complex equipment (low NOx burners, SNCR, SCR,
FGD Systems)
- Downsizing (loss of experienced personnel)
- Firing off-design fuels, fuel blends, spec fuels, etc.
• Handle changing optimization objectives (ozone
vs. non-ozone season operation).
12-36
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
What is a Right Tool for ISB?
Considerations:
• Data volume
• Data quality (typically moderate to low)
• Additional instrumentation requirements
• ISB should mimic decision process of the
human expert (operator)
– Intuitive in nature
12-37
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sample Fuzzy Rule for Determining
Furnace Cleanliness Status
0.41
0.40
2,790
FEGT [F]
0.39
2,780
0.38
2,770
0.37
2,760
0.36
2,750 11:15 11:45 12:15 12:45 13:15 13:45 14:15 14:45
13:00 14:00 15:00 16:00 Time
Time
555
550 SB Test 6
Baseline Upper elevation SB Test 7 S B Te s t 8
Up p e r fu rna ce SB Test 10 Valve A
Test furnace blowers Mid d le furna ce
Clean RHT:
(repeat of e leva tion, e le va tion: 550 Valve B
m os tly cold IKs 59 A and B
SB Test 5) All b lowers
corne rs
545 MST Temperature [F]
Main Steam Temp. [C]
545
540 540
535
SB Test 9
535 Lower
furnace
elevation:
All blowers 530
530 525
8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 0 5 10 15 20 25 30 35 40 45 50
Time (Wednesday, December 19, 2001) SHT Attemperating Spray Valve Opening
12-40
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Fuel Quality Affects Sootblowing
Effectiveness
2880 1
FEGT
2860 SB Indication
2840
2820
SB Indication
FEGT [F]
2800
2780
2760
2740
2720 0
0 100 200 300 400 500 600 700 800 900 1,000
Data Sample No.
12-41
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sooblower Maintenance Condition
Affects Sooblowing Effectiveness
0.33
IK 4
0.32
CF Pendant SHT
0.31
0.30
14:00 15:00 16:00 17:00 18:00
Time
12-42
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Implementation: Sample Screens
12-43
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Sample ISA Operator Screen
Gross Unit Load 583 MW Avg. O2 3.35 % Opacity 7.6 %
o
NOx Emission Rate 0.452 lb/MBTU Avg. Burner Tilt -1 ISA Goal NOx
ISA Recommendation
Sootblower Groups Reason
Furnace Sootblowers : IR Group(s) 1 Details Reduce NOx
Convective Pass Sootblowers: IK Group(s) 5, 9 Details Recover SHT Steam Temperature
12-44
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
Conclusions
• Sootblowing optimization is important for
continuous efficient operation of boilers.
• Intelligent sootblowing is needed to maintain
optimal boiler cleanliness and deal with changes
in fuel quality, operation, and maintenance
condition of sootblowers and firing equipment.
• With funding from the electric utility industry ERC
is developing an intelligent sootblowing advisor
(ISA).
• ISA will be deployed this year at two large coal-
fired power plants (tangentially and wall-fired)
12-45
4th Intelligent Sootblowing Workshop, Houston Tx, March 2002
9862704
13
SOOTBLOWING ADVISOR AT PLANT HAMMOND
UNIT 4
David Turner
PowerGen
John Sorge
Southern Company
13-1
9862704
Sootblowing Advisor at
Plant Hammond Unit 4
David Turner
PowerGen
John Sorge
Southern Company
13-3
9862704
Unit description
13-4
9862704
Flow paths
SOOTBLOWER
STEAM IP/LP FEEDWATER
HP
UPPER SB4
SPRAY
R/H
ECON
LOWER
SPRAY
13-5
9862704
Sootblower data
• Blower types
− Retractable blowers, Copes-Vulcan rotating retractable type
− Wall deslaggers, Copes-Vulcan rotating retractable type (not
used)
• Operating data
− Header pressure - 600 psig
− Retract blower operating pressure - 450 psig
− Wall deslagger operating pressure - 200 psig
− Supply from radiant SH outlet header - (2400 psig)
13-6
9862704
Sootblower Group
1 38L
Sootblower
37L, 37R, 38L, 38R, 39L, 39R, 40L, 40R
groupings
2 41R 41L, 41R, 42L, 42R, 43L, 43R, 44L, 44R
3 45R
45L, 45R, 46L, 46R, 47L, 47R, 48L, 48R
4 49R
49L, 49R, 50L, 50R, 51L, 51R, 52L, 52R, 53L, 53R
5 W54
W54, W55, W56, W57
6 W13
W2-W15
7 W17
W16-W51
Furnace Wall
42L
8th Floor 39R 39L
43R 43L
W3 W4 47R 47L W1 W2
Economizer
Secondary
Air Heaters 6th Floor 53R 53L
48L
W9 W10 37R 48R 37L
W11 W12
40' - 0"
Overfire Air Gas Out
Windbox 5.5 Floor W17 W18 W19 W20 W13 W14 W15 W16
EL.569'-6" Primary
Pulverizers(6)
Air Heaters(2)
W50 W51 W48 W49
EL.553'-0"
2nd Floor
13-7
9862704
Existing sootblowing
operator interface
13-8
9862704
Sootblowing procedure
13-9
9862704
Sootblower advisor
13-10
9862704
The sootblowing problem
13-11
9862704
Plant data show sootblowers
have no effect on temperatures
600
1000
500
Temperature F
990
400
Load MW
300 980
200 970
100 960
0 950
0.000 20.000 40.000 60.000 80.000 100.000
Time (hrs)
13-12
9862704
Plant data showing the effect of
Group 1 sootblowers
600 140000
500 120000
400 100000
80000
300
60000
200 40000
100 20000
0 0
0.000 20.000 40.000 60.000 80.000 100.000
load group1 upper spray
13-13
9862704
Backpass damper movement
following sootblower use
500 120
Damper position %
400 100
LOad MW
80
300
60
200
40
100 20
0 0
0.000 20.000 40.000 60.000 80.000 100.000
Time (hrs)
13-14
9862704
Cleanliness factors
13-15
9862704
Evaluation of R/H cleanliness
Damper Position
Dirty
Clean
S/H Spray
13-16
9862704
User interface
• Test interface
• Resides on PC
• Migration to DCS dependent
on benefit evaluation and
plant acceptance
13-17
9862704
Implementation status as of
January 2002
13-18
9862704
Performance metrics
13-19
9862704
Summary of
January 2002 testing
13-20
9862704
Future Plans as of January 2002
13-21
9862704
Summary
13-22
9862704
14
TAMPA ELECTRIC’S NEURAL NETWORK BASED
INTELLIGENT SOOTBLOWING
14-1
9862704
TAMPA ELECTRIC’S NEURAL NETWORK BASED
INTELLIGENT SOOTBLOWING
Mark A. Rhode, PE, Sr. Consulting Engineer, Tampa Electric
Neel Parikh, Director of Development, Pegasus Technologies, Inc.
Abstract
Boiler combustion dynamics change continuously due to several factors including coal
quality, boiler loading, ambient conditions, changes in slag/soot deposits and the
condition of plant equipment. NOx formation, Particulate Matter (PM) emissions, and
boiler thermal performance are directly affected by the sootblowing practices on a unit.
As part of its Power Plant Improvement Initiative program, the US DOE is providing co-
funding and NETL is the managing agency for this project at Tampa Electric’s Big Bend
Station. This program serves to co-fund projects that have the potential to increase
thermal efficiency and reduce emissions from coal-fired utility boilers. A review of the
Big Bend units helped identify intelligent sootblowing as a suitable application to achieve
the desired objectives. The existing sootblower control philosophy uses sequential
schemes, whose frequency is either dictated by the control room operator or is timed
based.
The intent of this project is to implement a neural network based intelligent sootblowing
system, in conjunction with state-of-the-art controls and instrumentation, to optimize the
operation of a utility boiler and systematically control boiler fouling. Utilizing unique,
on-line, adaptive technology, operation of the sootblowers can be dynamically controlled
based on real-time events and conditions within the boiler. This could be an extremely
cost-effective technology, which has the ability to be readily and easily adapted to
virtually any pulverized coal fired boiler.
Fouling of the boiler leads to poor efficiencies due to the fact that heat which could
normally be transferred to the working fluid remains in the flue gas stream and exits to
the environment without beneficial use. This loss in efficiency translates to higher
consumption of fuel for equivalent levels of electric generation, hence more gaseous
emissions are also produced. Another less obvious problem exists with fouling of various
sections of the boiler relating to the intensity of peak temperatures within and around the
combustion zone. Total NOx generation is primarily a function of both fuel and thermal
NOx production. Fuel NOx, which generally comprises 20%-40% of the total NOx
generated, is predominately influenced by the levels of oxygen. Thermal NOx, which
comprises approximately 20% - 50%1 of the total NOx, is a function of temperature. As
the fouling of the boiler increases and the rate of heat transfer decreases, peak
temperature increases and so does the thermal NOx production.
Traditionally, utility boilers are equipped with sootblowers, which are lances that use,
steam, water or air to dislodge and clean the surfaces within the boiler. The number of
lances on a given unit ranges from several to over a hundred. Traditional sootblowing
schemes involve fixed schedules for activating the blowers or the experience of the
operators who manually activate various fixed sequences. Time based sequencing of
sootblowers has been a traditional method employed by power plants, both domestically
and abroad, to improve cleanliness within boilers. These systems are generally automated
and are initiated by a master control device. In some cases, operators activate the systems
manually on the basis of established protocols or generic procedures. These methods
result in indiscriminate cleaning of the entire boiler or sections thereof, regardless of
whether portions are already clean. Hence, traditional methods of sootblowing may be
effective in assuring that a boiler is clean, but they fail to optimize the heat transfer rates
1
EPRI – TR-102906 Retrofit NOX Controls for Coal-Fired Utility Boilers
Simultaneously optimizing the objectives of NOx, PM and heat rate is difficult and
unrealistic for a control room operator, even more so when that operator is also required
to maintain control of the balance of the unit(s) equipment. The industry has recently
been introduced to a number of “Intelligent” Rule-Based systems that derive their
knowledge base from operator experiences, static plant design data, and general thermal
principles. Whereas these systems are better than the traditional methods, they also fail to
fully respond to the dynamic operation and condition of boilers. Rule-based systems are
not readily adaptable to transitional operation of present day boilers, which, as a result of
deregulation, are subject to volatile changes in operation and fuel types or blends.
Furthermore, time or rule based systems are not the answer due to the complexity of the
individual components, combinations thereof and the desire to satisfy multiple objectives
in a dynamic real-time environment. Additionally, rule-based systems are only as good as
the rules that drive them and established rules cannot accommodate the diverse set of
operating conditions that may be encountered on a daily basis.
Intelligent Sootblowing
The goal of the project will be to develop a Neural Network driven Intelligent
Sootblowing (NN-ISB) system module that proactively modifies the sequence of
sootblowing in response to real-time events or conditions within the boiler, in lieu of time
or general rule-based protocols. To date, the ability to intelligently blow soot while
satisfying multiple and specific user defined objectives has not been integrated with an
on-line, automatic and adaptive neural network driven sootblowing system. The NN-ISB
module will provide an asynchronous, event-driven technology that is adaptable to
changing boiler conditions.
Some of the basic technology components proposed for the project are commercially
proven. However, the project also incorporates the use and application of several new or
newly applied components and/or systems in conjunction with the NN-ISB system. The
objective will be to reduce emissions and provide improvements in efficiency and
reliability by employing synergistic approaches, which have not been possible with prior
technologies. Some of the salient technologies planned for implementation during this
Technology advancements in the past few years have resulted in the introduction of
several diverse systems that could change the basic process of sootblowing. Specifically,
robust temperature measurement products have emerged that allow localized
measurement of fireside temperatures and heat transfer rates in both the furnace zone as
well as the convection and backpass regions. The combination of these advanced
measurement techniques coupled with today’s high speed numerical processing allows
for real time determination of tube fouling and levels of boiler tube cleanliness. Albeit
limited in nature and scope, utilization of some of these technologies have seen some
successes in their ability to improve the efficiency of the sootblowing process for US
utility boilers.
Although prior testing and limited demonstrations have yielded some benefits in regard to
NOx, PM and heat rate, these efforts have not been fully exploited in the development of
a system that has the ability to understand, evaluate and optimize the process with
multiple real-time objectives. The advantages of the knowledge capture and adaptive,
counter-intuitive interactions with the NN-ISB system provides, the opportunity for a
modular sootblowing optimization subsystem capable of significant operational benefits.
Furthermore, since all utility boilers that fire pulverized coal and oils generate varying
levels of soot and slag, the commercialization and benefits of this innovative technology
has the potential to be readily and easily applied to a large population of power plants.
Independent manual sequencing of specific sootblowers has shown benefits in the area of
heat rate efficiency improvement, NOx reduction and other areas relevant to efficiency
and reliability. It is expected that additional, hard to quantify, gains will be realized in the
areas of: tube erosion (minimized), auxiliary power consumption (minimized),
perturbations in extraction steam flow (made more level), and particulate generation
(managed to minimize impact on ESP). Of particular note, traditional sootblowers are
high cost O&M devices. Steam consumption rates of 30,000#/hr are not uncommon and
create substantial heat rate penalties. The maintenance costs are also very high
considering the high pressures and temperatures, well in excess of 1000 degrees F, that
exist in many cases.
The NN-ISB system proposed herein will utilize the Neural Network technology from
Pegasus Technologies, which has been implemented successfully for combustion
optimization applications. This project shall use Neural Network based optimization, and
state-of-the-art sensing and sootblowing equipment to direct the operation of the
sootblowing systems in such a manner as to reduce NOx & PM emissions, while
concurrently improving the heat rate. Neural networks have not yet been fully
implemented for ISB applications within the utility industry. Through these development
activities, a NN-ISB will react to and take into account the heat distribution within the
boiler, equipment life, emissions, and the overall cost of generating power. The objective
is to develop a system to automatically determine the need for sootblowing in specific
Sootblowing Neural Network – This is the artificial intelligence system that obtains
real-time data from several sub-systems, adapts to changing plant conditions and
optimizes the operation of the sootblowers. The recommendation for the blower or set of
blowers to activate is then sent to the sootblowing control system. Tampa Electric has
already installed a neural network based combustion optimization system on the target
unit and it will be allowed to communicate directly with the Sootblowing NN system, as
necessary.
Sootblower and Water Cannon Control – State-of-the-art controls are included in the
supply of equipment.
Heat Flux and Slag Sensors – Approximately 16-24 heat flux sensors and 8 slag sensors
are scheduled for installation. The information collected shall be provided to the fouling
module, which in turn shall process the data and provide it to the NN system for further
evaluation.
Water Cannons –Water cannon technology has evolved over the years to become a
highly effective and safe means to clean water tubes. The water cannon technology
involves a pump skid with flow and pressure control which regulate the flow of water to
articulating spray nozzles mounted on the boiler walls. These nozzles direct a jet of
water on the opposing wall in pre-programmed areas. The heat flux sensors and slag
sensors are required to provide real-time information to the water cannon control system.
Conclusion
The intent of this project is to implement a neural network based intelligent sootblowing
system in conjunction with state-of-the-art controls and instrumentation, to optimize the
operation of a utility boiler, and systematically control boiler fouling. State-of-the-art
heat flux and slag sensors, dual plane acoustic pyrometers, directional water cannons, and
integration of boiler cleanliness and performance models with a neural network are some
of the prominent components of this project. Operation of the sootblowers can be
dynamically controlled based on real-time events and changing conditions within the
boiler using on-line, adaptive technology. A new generation of cost-effective sensoring
equipment has the potential to provide sufficient measurable inputs to a NN-ISB
sootblowing process to meet one or more of the objectives, which may include:
NOx Reduction through more stable control of furnace exit temperatures, and more even
distribution of temperature across the furnace exit and convection zones.
Particulate Matter Reduction through reduced excess carbon, uniform ESP inlet
temperatures, and coordination of sootblowing execution with ESP rapping execution.
Heat Rate Improvement through improved localized temperature consistency and better
control of furnace and subsequent heat transfer zone temperatures.
This could be an extremely cost-effective technology, which has the ability to be readily
and easily adapted to virtually any pulverized coal-fired boiler. The net impact to the
industry will be the demonstration of a commercially viable system that improves overall
plant reliability and operations by reducing production cost, while also minimizing
emissions.
Disclaimer
“This report was prepared as an account of work sponsored by an agency of the United States Government.
Neither the United States Government nor any agency thereof, nor any of their employees, makes warranty,
express, or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process disclosed. Or represents that its use would not
infringe privately owned rights. Reference herein to any specific commercial product, process, or service
by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States Government or any agency thereof. The
Randy Carter
ASI
15-1
9862704
EPRI Fourth Intelligent
Sootblowing Workshop
Intelligent Sootblowing –
Boiler Cleaning Management System
Operator interface
When and where to clean
Sootblower effectiveness
Wayne S. Jones
Reliant Energy
Frank Massi
Sargent and Lundy
16-1
9862704
Reliant Energy and Sargent and Lundy:
Approach to
Alternative Cleaning Systems Used on SCR Applications
By:
Wayne S. Jones
Reliant Energy
Frank Massi
Sargent and Lundy
With recent regulatory tightening of NOX emission limits by the Texas Natural
Resource Conservation Commission (TNRCC), Reliant Energy has made the
commitment to add selective catalytic reduction (SCR) systems to most of its
larger generating units. These new systems have brought with them new
challenges with respect to the cleaning of fly ash from the catalyst and air
preheater surfaces. An additional concern associated with SCR systems is the
build up of a sticky substance on the air preheater know as ammonium bisulfate
(ABS). Steam sootblowing systems have been used for many years as the
principal method of removing unwanted ash accumulation on boiler and air
preheater surfaces. This paper will explore other cleaning systems that are being
used with SCR applications both in the US and in Europe. 16-2
9862704
W. A. Parish Site Parameters
• The W. A. Parish site consists of 4 coal fired and 4 natural gas fired units.
SCRs are being added to all 4 coal fired units and 1 natural gas fired unit.
Cleaning systems are being added to the 4 coal fired units.
• Unit 5& 6
– Maximum Unit Load - 690 MW
– B&W Boilers
– Fuel Type - 100% Powder River Basin
– Flue Gas Temperature Range - 780 - 620°F
– Hot Side SCR (high dust)
– Enamel Coated Air Preheater Baskets
• Units 7 & 8
– Maximum Unit Load - U7-590 MW; U8-652 MW
– ABB/CE “T” Fired Boilers
– Fuel Type - 100% Powder River Basin
– Flue Gas Temperature Range - 800 - 620°F
– Hot Side SCR (high dust)
– Enamel Coated Air Preheater Baskets 16-3
9862704
Need for Cleaning SCR Catalyst and Air Preheater
Baskets
SCR Catalyst
16-4
9862704
Air Preheater Baskets
ABS will condense from the gas stream and form a sticky deposit on the heat
transfer surface of the air heater at a temperature of 380 - 450°F. Fly ash
particles will tend to stick to the ABS resulting in the gradual pluggage of the
APH. This in result in an increase in APH pressure drop (ID fan capacity) as
well as loss in thermal efficiency for the plant.
ABS is also corrosive (acidic in nature) and will corrode the mild steel or low
alloy steel surfaces of the APH. The rough surface of corroded material
further enhances the deposition of ammonia bisulfate and accelerates the
plugging mechanism.
16-5
9862704
Catalyst Supplier Requirements
16-6
9862704
Catalyst Cleaning Equipment
• Historically, steam sootblowers have been used to clean catalyst.
• The aim of sootblowing is to remove the particulates while
avoiding condensation of water vapor on the catalyst and nay
physical damage.
• The following specification covers most cases:
– Linear velocity at catalyst surface: 45 ft./sec.
– Maximum temperature of steam after nozzles: 590°F
– Minimum temperature of steam after nozzles: 520°F
– Maximum steam pressure at sootblower lance inlet: 145 psig
– Approximate distance from catalyst surface to nozzles: 20 in.
• Recently, the trend in new SCR installation is to use sonic horns.
• Sonic horns is a general term used to describe audible horns (75 -
250 hz) and infrasonic horns (< 25 hz). 16-7
9862704
Cleaning Philosophy
• Sootblowers - sootblowers are generally operated on a
limited basis (once/shift, three times/day, etc.). Fly ash is
allowed to accumulate on the catalyst. The build-up is
blown with steam which forces the fly ash through the
catalyst channels. A portion of the fly ash will be
reintrained into the flue gas steam and removed by the gas
flow.
• Sonic Horns - sonic horns operate on a continuous basis
usually for about 10 seconds each minute. The operation
is staggered thus reducing plant air requirements. Fly ash
is not allowed to accumulate on the catalyst. The fly ash is
excited by the sound waves and causes them to jump off
the catalyst surface and become reintrained into the flue
gas. Removal of ash is by reintrainment into the flue gas
16-8
and by gravity.
9862704
Sootblower Installation
16-9
9862704
Audible Horn Installation
16-10
9862704
Infrasonic Horn Installation
16-11
9862704
Advantages and Disadvantages to Each SCR Cleaning
Method
16-12
9862704
Sootblowers
• Advantages
– No significant auxiliary power loads due to infrequent
blowing schedule
– Adequate cleaning of catalyst
– Large number of SCR installations
• Disadvantages
– High initial cost for equipment and installation.
– High operational & maintenance costs.
– Potentially erosive to catalyst due to steam jet impingement.
– Risk of water introduction into SCR
• Possible acceleration of catalyst deactivation
• Possible hardening of PRB to cement like material can
lead to deactivation of the catalyst due to masking of the
pore structure. 16-13
9862704
Sootblowers - Disadvantages (Cont.)
16-14
9862704
Sonic Horns - Advantages
• Audible Horns
– Minimum maintenance requirements (Only diaphragm
needs maintenance)
– Minimum spare parts (Diaphragm is only moving part)
– No risk of water induction.
– Minimum piping requirements - only air piping.
– No risk of structural side effects
– Can operate on existing service air
– Low initial installation cost
– Low operational cost
16-15
9862704
Sonic Horns - Advantages
• Infrasonic Horns
– No noise considerations (Horns are not audible)
– Minimum maintenance requirements (Only sound
generator, diaphragm, and air compressor needs
maintenance)
– Large uniform cleaning of catalyst and entire reactor
vessel.
– Horn location is determined by computer 3-D
modeling.
– No risk of water induction.
– Minimum piping requirements - only air piping.
– Easier installation - Resulting in reduced installation
time. 16-16
9862704
Sonic Horns - Disadvantages
• Infrasonic Horns
– Minimal experience in SCRs
– Risk of structural side effects in ducting
– High initial installation costs
– High operational costs
• Audible Horns
– Gaining experience in new SCRs
– Noise considerations
– Ash accumulations then removal
16-17
9862704
Sonic Horn Installations Visited
16-18
9862704
SCR Sonic Horn Installation List
• Existing Installations
– PG&E - Indiantown Station (Fall 1997)
– TVA - Paradise Station (Spring 2000)
– Constellation Power - Brandon Shore (Spring 2001)
– Kansas City Power & Light - Hawthorne (Spring 2001)
– Carolina Power & Light - Roxboro (Summer 2001)
– American Electric Power - Gavin Station (Spring 2001)
– Mirant Americas - Birchwood (Fall 2000)
– Stadtwerke Bremen AG, Germany
– Tiefstack Power Station, Hamburg, Germany
16-19
9862704
SCR Sonic Horn Installation List
• Installations in Process or Planned
– First Energy - Bruce Mansfield
– TVA - Cumberland
– TVA - Widows Creek
– Constellation Power - Wagner Unit 3
– Dominion Energy - Kincaid
– Black Hills Power & Light - Wygen
– Carolina Power & Light - Mayo
– Santee Cooper - Cross Unit 1 & 2
– Cinergy - East Bend Unit 2 (S&L)
– Cinergy - Miami Fort Units 7 & 8 (S&L)
– Cinergy - Gibson Units 1 - 5 (S&L)
– Cinergy - Cayuga Unit 1 (S&L)
– Southern Illinois Power Cooperative - Marion Unit 4 (S&L)
– Reliant Energy - W. A. Parish Units 5 - 8 (S&L)
16-20
9862704
Air Heater Cleaning
16-21
9862704
17
SOOTBLOWER EROSION/CORROSION MITIGATION
VIA THE APPLICATION OF WELD METAL OVERLAY
Doug Vail
Welding Services, Inc.
17-1
9862704
Abstract
In an effort to improve steam generation efficiencies throughout the fossil fuel power
generation industry, soot blowing has become a widely used technique. Wall blowers,
water lances and water cannons are being utilized to remove slag and increase thermal
conductivity and heat transfer on water walls and in the convection passes.
These methods have led to rapid thinning of boiler tubes due to increased
erosion/corrosion. Increased maintenance cost and reduced unit reliability are a direct
consequence. The practices of boiler tube replacement, shielding, and metal sprays have
been employed to try to reduce the thinning phenomenon, but with limited, if any, long
term success.
This paper reviews boiler tube damage mechanisms under conditions of soot blowing,
with or without Low NOX conditions. The long-term success of utilities that have chosen
automatic weld overlay is reviewed along with the basis for evaluating the cost of
alternative methods and the return on investment.
Introduction
All fossil fired boilers contend with slag deposits on their furnace walls affecting heat
transfer and thermal conductivity. A number of factors determine the extent of this
problem for a particular boiler; however, the most prominent differentiators are boiler
design and fuel composition.
Due to the erosion induced corrosion mechanism of this failure, it is our experience that
reliably solving and removing the corrosion aspect from the equation has eliminated the
rapid thinning and solved the problem. This paper will exhibit the reliability, feasibility,
and cost effectiveness of Unifuse Automated Weld Metal Overlay in eliminating this
issue.
Examples used are for 500MW Coal utility looking at refurbishment of 54 soot blower
regions of 50 s.f. each.
9862704 17-2
Discussion/Evaluation
Feasibility of Options
In a review of feasibility there are a number of key areas of review. Primary to these are
critical path outage time and boiler maintenance availability during the outage. Since in
most outages the boiler and boiler related activities are critical path, a restoration option
that will both minimize the critical path impact and have minimal impact on other
activities carries a large advantage.
Reliability of Options
Prevention of down time resulting from tube failures and rework is the next area of
improvement. In the case of replacement this is successful as the weak areas are
eliminated and a known quantity of time is thereby purchased. In the case of Unifuse this
is also achieved through a full fusion bond to the substrate. This bond not only
substantially increases and restores tube thickness and strength, it creates a very tight
dendritic structure. This structure protects the substrate from attack and eliminates the
opportunity for separation of the overlay from the base material. In the case of Metal
Spray, a mechanical bond is created with the base material. Not only does this bond add
no material strength, the material structure is very porous and ultimately allows
sulfidation of the substrate and material failing in the form of spalling. While “sealing”
the spray has made some gains, this failure over time must be reworked during outage
critical path or tube failures will again result.
9862704 17-3
Cost Effectiveness of Options
In reviewing cost effectiveness of options, total cost must be considered. Since the focus
of discussion is in the elimination of forced downtime and outage critical path, it is
imperative that these costs be considered when evaluating options.
The following costs are extracted from a 500MW boiler in the Eastern U.S. restoring 54
soot blower regions. Comparison is with tubing replacement, Metal Spray, and Unifuse
in a soot blower atmosphere seeing 5-8 Years service life.
9862704 17-4
Technical Analysis
Technical Summary
17-5
9862704
Economic Analysis
Key Factors
•Installation Cost
•Installation Schedule
•Scope Flexibility
•Future Inspection Requirements
•Future Maintenance Requirements
•Life Expectancy 17-6
9862704
Economic Analysis
Installation Cost
•Abrasive Blast
•Assessment/Scope Definition
•Grinding
•Unifuse(R) Application
•Final QA/AC
17-7
9862704
Economic Analysis
Installation Schedule
• No Critical Path
• All Work Performed in Parallel with
Exception of Abrasive Blast
17-8
9862704
Economic Analysis
Scope Flexibility
• Assess As-found Conditions
• 200 sq. ft. to 10,000 sq. ft. during Outage
Window
17-9
9862704
Economic Analysis
17-10
9862704
Economic Analysis
17-11
9862704
Economic Analysis
Life Expectancy
• Soot Blower Erosion less than .003/year=25+
years
• Thermal Fatigue Resistance = 30% - 50%
stronger than SA210 and SA213T2 materials
at 800 deg F
• Significant life expectancy improvements
over current methods
17-12
9862704
Technical & Economic Conclusions
17-13
9862704
18
FOULING MONITOR AND ALARM SOFTWARE -
DEVELOPMENT OF A USER INTERFACE
R. Thompson
Fossil Energy Research Corporation
23342-C South Pointe
Laguna Hills, California 92653
A. Hickinbotham
TransAlta Utilities
110 12th Avenue, S.W.
Box 1900, Station ‘M’
Calgary, Alberta, Canada T2P 2M1
T. Holden
S. Frostad
AMEC, Incorporated
35 Boulder Boulevard
Stony Plain, Alberta, Canada T7Z 1V6
18-1
9862704
FOULING MONITOR AND ALARM SOFTWARE -
DEVELOPMENT OF A USER INTERFACE
R. Thompson
Fossil Energy Research Corporation
23342-C South Pointe
Laguna Hills, California 92653
A. Hickinbotham
TransAlta Utilities
110 12th Avenue, S.W.
Box 1900, Station ‘M’
Calgary, Alberta, Canada T2P 2M1
T. Holden
S. Frostad
AMEC, Incorporated
35 Boulder Boulevard
Stony Plain, Alberta, Canada T7Z 1V6
9862704
18-2
FOULING MONITOR AND ALARM SOFTWARE -
DEVELOPMENT OF A USER INTERFACE
R. Thompson
Fossil Energy Research Corporation
23342-C South Pointe
Laguna Hills, California 92653
A. Hickinbotham
TransAlta Utilities
110 12th Avenue, S.W.
Box 1900, Station ‘M’
Calgary, Alberta, Canada T2P 2M1
T. Holden
S. Frostad
AMEC, Incorporated
35 Boulder Boulevard
Stony Plain, Alberta, Canada T7Z 1V6
Abstract
TransAlta Utilities has installed specialized software to monitor boiler fouling on three
units that have had a history of outages caused by plugging in the convective section.
The software was developed to provide the boiler operators with adequate warning of a
fouling episode so that preemptive action could be taken (e.g., sootblowing, a change in
coal blend, etc.) to potentially avoid a costly outage. The software utilizes a unique
combination of combustion diagnostic techniques and convective section heat
adsorption analyses to identify boiler operating conditions where ash deposition rates
may be high and conducive to triggering a fouling episode. Existing plant
instrumentation is used to provide input data to the software.
The paper includes examples of operating experience with the Fouling Episode Monitor
Alarm (FEMA) software but the primary focus is on the development of a user friendly
interface. The current user interface represents a major improvement over the original
Beta version. One of the unique features of the second generation software is the ability
of engineers and supervisors to view the boiler operator screens at any time directly
over the corporate LAN. This is accomplished using a standard web browser such as
9862704
18-3
Microsoft Internet Explorer. The FEMA software allows the user to quickly assess the
fouling status of individual boiler sections from a single overview graphic. The fouling
history of the unit is indicated by a graphical record of the fouling index over the most
recent 12-hour period. Additional graphical displays can be assessed showing fouling
history and heat absorption patterns over a 24-hour period. A database of boiler fouling
and operating parameters is maintained that can also be accessed remotely to download
historical files of past performance for more detailed engineering analysis.
Introduction
Many utilities have increased their use of low sulfur Western coals because of the
significant NOx and SO2 emission benefits. More specifically, conversion to a 100%
Western Powder River Basin (PRB) coal, combined with a low-NOx firing system (such
as the TFS2000), has allowed some utilities to completely avoid the cost of installing a
Selective Catalytic Reduction (SCR) NOx Control System. This can result in a substantial
reduction in power generation costs in a competitive deregulated market. However,
conversion of older units designed for bituminous coals to 100% PRB can result in
increased slagging and fouling, load derates, and possible forced outages if adequate
precautions are not taken.
The reflective nature of PRB ash can lead to a substantial increase in furnace exit gas
temperature and accelerated ash deposition. These adverse impacts can often be offset
by more frequent sootblowing, installation of additional sootblowers, and more careful
attention to sootblower maintenance and burner zone combustion conditions. Some
utilities, acustomed to firing Eastern coals, have been somewhat slow to recognize the
need for very high sootblower availability and uniform combustion in the burner zone.
The economic penalties of a load derate or a forced outage can be substantial compared
to a more proactive approach to slagging and fouling prevention.
Low-NOx firing, particularly under deeply staged combustion conditions with high
overfire air (OFA) flows, can lead to local regions of increased ash deposition in the
burner zone. More specifically, furnace ash deposits can build in local fuel-rich regions
to the point where ash fusion temperatures are exceeded and the ash becomes “sticky”
or semi-molten. This can be aggravated by non-uniform combustion in the burner zone
caused by an uneven coal flow distribution to the burners. Under these circumstances,
localized delayed combustion may result in burning high in the furnace and into the
convective section. Sticky or semi-molten ash can lead to rapidly growing deposits that
bridge the entrance to the superheater pendants. A rapid build-up of ash deposits can
also occur in the reheat section, particularly on units with narrow reheat tube spacings
(i.e., designed for eastern bituminous coals). Although bridging of the superheat
entrance is usually quite visible and is often monitored by TV cameras, a rapid build-up
and complete pluggage of the reheat section can occur in areas where visibility is poor
9862704
18-4
and/or boiler operators become busy with other duties and, thus, monitor deposits
infrequently.
For the reasons outlined above, a cost-effective strategy to achieve NOx emissions
compliance without SCR by low-NOx firing of PRB coals can backfire, resulting in
frequent unit derates and forced outages due to heavy ash deposition and fouling of the
convective passes (i.e., superheat, reheat, and primary superheat sections). Many
fouling episodes can be avoided by religious sootblower maintenance, optimized
sootblowing cycles, frequent monitoring of coal properties, direct measurement of
FEGT (Furnace Exit Gas Temperature), maintaining uniform combustion, and frequent
boiler inspections. However, the reality is that this ideal world seldom exists at most
commercial power plants. Therefore, a need has developed for a fouling episode
monitor and alarm system that can alert boiler operators to a potential boiler plugging
and prevent a forced outage. Ideally, this monitoring system would not only detect the
onset of a fouling episode, but also provide enough advanced warning to alert boiler
operators to take evasive action to arrest the fouling and avoid a forced outage.
This paper outlines the development of software intended to monitor boiler fouling and
provide an advanced warning to the control operators. With adequate warning,
preemptive action could be taken (e.g., aggressive sootblowing, a change in combustion
conditions, etc.) to potentially avoid a costly forced outage.
Project Objective
The primary objective of this project was to prepare a custom software module that
would:
• Identify adverse boiler firing practices that increase the risk of fouling;
9862704
18-5
The fouling episode software would initially monitor full-load operating parameters
that directly affect the rate of ash deposition. More importantly, it would be custom
calibrated using unit design and coal-specific test data. The Fouling Episode
Monitor/Alarm (FEMA) should not be confused with commercial soot blowing
advisors or boiler performance monitor software. FEMA® utilizes a unique combination
of combustion diagnostic techniques and convective section heat absorption analyses to
identify boiler operating conditions where ash deposition rates may be high and
conducive to triggering a fouling episode. Each FEMA is custom-configured and tuned
for a specific unit based on its unique design, fouling history, and data acquisition
system.
A detailed analysis of combustion conditions and boiler firing practices that occurred
just prior to past fouling episodes suggested that a combination of events, rather than
just the fuel properties alone, were the primary cause of the fouling episodes. In some
respects, each fouling episode appeared to have one (or more) typical “signatures” of
related events and parameter excursions. Potentially, they could be monitored with
custom PC software to detect an impending fouling episode before it had reached its
avalanche-type threshold.
Acceptable ranges and alarm tolerances were established for a total of 20 key operating
parameters, available from the plant historian. A logic module was developed to
process the various parameter alarms resulting in a “FEMA® Fouling Index” which
indicates the relative fouling potential for the current unit operating conditions. A high
FEMA® Fouling Index indicates that a fouling event is imminent and the boiler control
operator should take immediate action to reduce the potential for fouling. These steps
may include:
9862704
18-6
It should be noted that the FEMA® Fouling Index is computed using an algorithm
developed from unit-specific historical fouling data and it bears no relationship to
fouling indices that have been developed in the past based on coal properties. The
FEMA® Fouling Index is based on boiler operating parameters and calculated values
and currently does not contain any coal property data.
Several candidate FEMA® Fouling Index algorithms were back tested using historical
fouling data to assess the beta software performance. These simulations provided
warnings of potential fouling episodes of from one to three hours before the boiler
control operators took evasive action to try to prevent the boiler from plugging up and
going “off-line”.
The Fouling Episode Monitor/Alarm (FEMA®) was subsequently installed at two other
units in the TransAlta Utilities system. Each FEMA® was custom configured for the
characteristics of the individual unit. The software and FEMA® Fouling Index were
initially designed for operator convenience so that a yellow warning occurred at a
FEMA® Fouling Index of 5 and it turned red at 7.5. An example of a serious fouling
episode that occurred on the south side of the superheat section of Sundance 1 on
October 6, 2000 is shown in Figure 1. The FEMA® Fouling Index and the normalized
secondary superheat heat absorption are shown for comparison purposes over the 24-
hour period 03:00 on October 6th to 03:00 on October 7th.
The FEMA® Fouling Index issued its first yellow warning at 14:47 and a red alert at
15:37 (50 minutes later). The FEMA® Fouling Index peaked at a value of 10 at 15:52. The
FEMA® Fouling Index dropped with aggressive soot blowing, but returned to yellow at
16:42 and subsequently shot up to 10.7 at 18:37, at which time the operators dropped
load to 250 MW to clean the boiler. The heat absorption graphs continued to indicate
that the south side was not cleaning well and a third red alert was issued at 01:47 on
October 7th. We believe the FEMA® Fouling Index warnings and red alerts at three
different time periods (including one more than three hours before the operators
dropped load to clear the furnace) illustrates the value of the FEMA® software in
alerting operators to a fouling event. Other examples of the response of the FEMA®
software to fouling events were presented in previous papers on the subject (1) (2).
The FEMA® software was originally installed on a separate stand alone computer as a
Microsoft Excel Visual Basic Application (VBA) running under a Microsoft Windows 95
operating system. Evaluation of the software, hardware, data exchange, and operator
interface identified a number of issues that were addressed in a new version of the
software installed at the Sundance station in the spring of 2001. Some of the lessons
learned from the use of the first version of the software were:
9862704
18-7
N SSH S SSH FI
1.20 16
1.10 14
Normalized Heat Absorption
1.00 12
0.90 10
Fouling Index
0.80 8
0.70 6
0.60 4
0.50 2
0.40 0
10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/6/00 10/7/00 10/7/00 10/7/00
1:00 3:00 5:00 7:00 9:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 1:00 3:00 5:00
Date/Time
Figure 1. Sundance Unit 1 – FEMA® Fouling Index and Normalized Heat Absorption
• Boiler control operators preferred to have the FEMA® screen displays on the
same console as other process control screens, once the software was operating
reliably.
• Data retrieval and validation of input data from the historian was unreliable for
real-time operation due to continual upgrades and changes in Tag addresses.
The initial Beta version of the FEMA® software was installed on a separate stand-alone
PC for cost reasons and also to provide some isolation from the plant process control
system. Once the proof-of-concept testing had been completed, discussions with the
control operators clearly indicated that FEMA® needed to be installed on the operators
console for convenient access. AMEC prepared Bailey Process Control View (PCV)
screens for the operators console that include a tabular summary of key FEMA®
parameters as well as a graphical trend of the fouling index over the last 24 hours. In
addition, a horizontal status bar to the right of the current fouling index value was
9862704
18-8
programmed to change length and color in correspondence with the fouling index.
Submenus allow the operator to quickly examine FEMA® parameter trends that may be
of concern. The main control operator screen interface is shown in Figure 2. The format
and graphical display characteristics of this screen were dictated primarily by the Bailey
PCV and requirements to be consistent with other process control screens at TransAlta
Utilities.
Figure 2. FEMA® Main Boiler Control Operator Interface Screen – PCV Version
Significant changes were also made in the data acquisition, communication, and storage
system. The input data stream to FEMA® now comes directly from the unit process
control computer instead of being routed into and out of the plant historian. Input data
and calculated results are now archived in a manner that they can be directly
downloaded over the TransAlta Utilities Local Area Network (LAN). Data can be
analyzed “off-line” in a batch mode ensuring that detailed data analyses give results
9862704
18-9
that are consistent with the real-time data provided to the operator. Data validation
routines have been enhanced and substitute default parameters defined to allow
FEMA® to continue operating if an input parameter is missing.
9862704
18-10
The current fouling index is shown in the right center of the screen. It has been rescaled
so that it turns yellow at a value of 15 and red at a value of 20. The bottom of the
overview screen includes a historical trend of the fouling index over the last 12 hours.
This provides the user with valuable insight regarding the frequency and history of
exclusions above the warning level. This is particularly valuable in assessing whether
sootblowing and other remedial efforts have succeeded in reducing the fouling index
once an excursion to unsafe operating levels has occurred. Since some users prefer a
tabular display, an alternate summary screen display is shown in Figure 4.
For those occasions where additional parameter trends are desired to provide insight
into the history of a fouling excursion, trend graphs for 14 key parameters and
parameter groups are available. These trend graphs cover the most recent 24-hour
period and are full screen displays. An example of the fouling index over a 24-hour
period where difficulty was experienced in burning old yard coal is shown in Figure 5.
9862704
18-11
Figure 5. Example of 24-Hour Display of Key FEMA® Parameters
Even more user flexibility is provided by the ability to design custom trends where up
to six individual tag names can be displayed in the same trend (see Figure 6). The
maximum and minimum limits can be manually entered, left as default values of 0 and
100, or can be auto-scale based on the high and low limits of the tags.
The second generation FEMA® also includes a report screen that gives a summary of
FEMA® information for a 12-hour shift. A sample report screen is shown in Figure 7.
The intention is to print the report at the end of each shift to record the key events for
the period. An hourly summary of the fouling index includes the average value for the
period, the maximum, and the number of minutes the index was in the “Red” above the
20% alarm limit. Data access and index calculation errors are also tabulated for
diagnostic purposes. If an essential tag was out of range during the reporting period,
the tag name and the number of times it was out of range is noted. An area is provided
for the operator to include notes about the FEMA® software performance or unit
9862704
18-12
Figure 6. Ability to Define Custom Screen Trends for Diagnostics
Included in FEMA® Capability
operating conditions that may be important. If high fouling indexes were experienced
during the shift for reasons that were known to the operator (e.g., sootblowers out-of-
service for maintenance), convenient check boxes are provided to identify the reason. A
trend graph of the fouling index for the shift is included at the bottom of the report page
for reference purposes.
The latest version of the FEMA® software was installed in April 2001 and has proven to
be very stable and crash resistant. (At the time this paper was written, January 2002, the
program had not crashed once.) A boiler control operator training program on the use
of FEMA® was conducted in late fall of 2001 and the fouling monitor is now fully
integrated into daily plant operating practices at Sundance Units 1 and 2.
9862704
18-13
Figure 7. Sample FEMA® Report Summary Screen
9862704
18-14
Conclusions
• Fouling episodes leading to a unit derate (or possible forced outage) may become
more common as units designed for bituminous coals are converted to 100% PRB
coals operating in a deeply staged low-NOx mode.
• High furnace gas exit temperatures and convective section fouling are often
related to poor sootblower maintenance, inadequate or infrequent sootblowing,
and non-uniform combustion (often caused by uneven coal flow distribution to
the burners).
• A FEMA® Fouling Index Algorithm has been defined that typically provides
sufficient advanced warning of a fouling episode to allow boiler control
operators to implement preventative measures to clean the boiler and reduce the
potential of a forced outage with modified boiler firing practices.
• A second generation FEMA® was installed in the Spring of 2001 featuring the
following enhancements:
- The FEMA® Fouling Index is now calculated using input data obtained
directly from the process control computer interface instead of the plant
historian.
• The latest FEMA® software is very stable and has not crashed since its
installation in April 2001.
9862704
18-15
• The most recent FEMA® software and FEMA® Fouling Index algorithm shows
considerable promise in providing an advance warning of an impending fouling
episode.
Acknowledgements
This project was sponsored by the TransAlta Utilities Fuel Supply Department, but
would not have been possible without the cooperation and support of the Wabamun
and Sundance Station staffs. The VBA version of the FEMA® software was developed
internally by Fossil Energy Research Corp. (FERCo) but the programming of the second
generation VB version was provided by the engineers at the Stony Plain, Alberta office
of AMEC Energy, Inc. (formerly AGRA-Monenco).
References
9862704
18-16
Target: About EPRI
Coal Boiler Performance Optimization EPRI creates science and technology solutions for
and Combustion NOx Control
the global energy and energy services industry. U.S.
electric utilities established the Electric Power
Research Institute in 1973 as a nonprofit research
consortium for the benefit of utility members, their
customers, and society. Now known simply as EPRI,
the company provides a wide range of innovative
products and services to more than 1000 energy-
related organizations in 40 countries. EPRI’s
multidisciplinary team of scientists and engineers
draws on a worldwide network of technical and
business expertise to help solve today’s toughest
energy and environmental problems.
EPRI. Electrify the World
1004112
EPRI • 3412 Hillview Avenue, Palo Alto, California 94304 • PO Box 10412, Palo Alto, California 94303 • USA
800.313.3774 • 650.855.2121 • askepri@epri.com • www.epri.com
9862704