Performance Tests of Water Cannon Furnace Cleaning Systems

Electric Power 2003, Houston, Texas

Charlie Breeding, P.E.

Clyde Bergemann, Inc.

Atlanta, GA

Table of Contents

INTRODUCTION 3

FURNACE SLAGGING 3

FURNACE CLEANING 4

INTELLIGENT SOOT BLOWING 5

TEST PROGRAM 5

TUCSON ELECTRIC SPRINGERVILLE 6

ALABAMA POWER, MILLER 9

BELLE RIVER, DETROIT EDISON 9

SMARTGAUGE TESTING 11

CONCLUSION 14

REFERENCES 14

Introduction

Clyde Bergemann has installed water cannon furnace cleaning systems in over 60 utility power plants.This paper presents the results of performance and start up test from a number of those plants. These testsquantified the changes in furnace exit gas temperature, unit output, effect on steam flow, boiler efficiency,NOx production, and other plant parameters due to furnace cleaning with water cannons. Special testing ofa new technology utilizing strain gages to identify fouling of convection pass elements is also presented.Data from special instrumentation such as HVT probes and emissivity instruments is included in theanalysis along with data from permanent plant instruments and heat flux monitors. A special technique ofutilizing the heat flux monitor as an indicator of thermal impact is described and results are presented.Most of the water cannon installations use heat flux sensors as a feedback mechanism to determine thetiming and frequency of cleaning. These instruments can also be configured to infer the thermal impact ofthe water spray on the furnace wall tubes.Most coal fired power plants are no longer burning the fuel that the boiler was originally designed for. Inmany cases the current fuel is vastly different than the design fuel. This is especially the case with boilersthat have converted to Powder River Basin (PRB) coal. Often the change in coal is accompanied withincreased slag formation and deposits in the furnace. Cleaning the furnace with water cannons can improvethe heat absorption of this section of the boiler and return it to near the design conditions of cleanliness.

Furnace slagging

The ash component of coal can leave significant deposits on the heat transfer surfaces of the furnace. In thehigh temperature range of the furnace slagging can be particularly tenacious and cannot always be removedby steam and air blowers. Manual lances using water had been used to clean deposits off furnace walls.The success of this technique lead to the mechanization of the lance into the water cannon. Water was alsothe medium used in water lances, which are a adaptation of the steam wall blowers.

In a typical coal fired power plant burning 800 lbs of coal per Mwh a 500 Mw unit would combust ~ 200tons of coal per hour. Ash contents of the various coal types range from 6 to 40 precent. If we use 20% asan average this unit would handle 40 tons of ash per hour.

In this 500 Mw example 20% to 30% of the ash is typically bottom ash, and the remaining 70% to 80% (or28 tons per hour) is fly ash entrained in the gas stream.

Deposits are formed when the ash in the flue gas is at a temperature above its melting point. The laboratoryterm for the initial melting point is the initial deformation temperature. Typical values of initialdeformation temperature for PRB coal are in the 2100F to 2200F range. With typical gas temperatures of2500F in the furnace, the ash is in a semi-molten condition and when it comes in contact with furnace waterwalls that are at a relatively cooler temperature, a deposit is formed.

The majority of the ash passes through the boiler and is collected in the ash removal equipment(electrostatic precipitator or bag house). However, even a small portion collecting on the heat exchangesurfaces can have a dramatic impact on plant performance. Not only is there an insulating effect from thedeposit, but there is also a change in the emissivity of the surface. Surface emissivity is critical to heattransfer in the furnace area since most of the heat transfer is from radiant heat. The general equation forradiant heat transfer is:

surfaceofetemperaturT

gasofetemperaturTftsqareaS

emissivityThrftsqBtu

hrBtuQwhere

TTSQ

=

===

×=

=

−=

−

2

1

49

42

41

..,

,,/1071.1

/:

)(

εσ

εσ

Some coal such as PRB contains high levels of calcium oxide and magnesium oxide, which are majorsources of the reflective property of PRB ash deposits. Calcium oxide can be in the 24% range for PRBcoal compared to 1 to 2 % for eastern coals. Also, magnesium oxide is in the 5% range for PRB coalcompared to about 1% for eastern coals. The normal emissivity of a boiler tube with a coating of ironoxide is from 0.85 to 0.89. However, a slag deposit can cut this value to 0.5, and thus have a significanteffect on the amount of heat absorbed by the furnace. From the radiant heat transfer equation it can be seenthat this reduction in emissiviity (or increase in reflectivity) would reduce heat transfer by almost half.This reduction of heat absorbed in the furnace results in higher furnace exit gas temperatures. Less heat isabsorbed by the water walls and therefore more heat must be absorbed by the superheater and reheatersections in the convection passes to maintain full load steam conditions. It is often the case that this upsetin boiler balance is enough to reduce the capacity of boilers. The higher furnace exit gas temperaturesusually result in higher stack gas temperatures and therefore lower boiler efficiency. A reduction in boilerefficiency increases heat rate and therefore increases the cost of generation. The elevated temperatures ofthe gas leaving the furnace often results in ash entering the convection passes containing ash that is aboveits fusion temperature. Thus this ash will deposit on the superheat and reheat surfaces a further reduce theability of the furnace to make steam conditions. High temperatures that exist in the furnace for longer timeperiods also result in more production of thermal NOx.

Furnace cleaning

The traditional method of furnace cleaning is the retractable wall blower. This device is inserted into thefurnace wall through an opening. The blower tube rotates while a jet of steam or air is sprayed on thefurnace wall to clean off deposits. A blower typically can cover a 10 foot diameter area.

Wall blowers are often not adequate to clean boilers that are burning PRB, lignite, or other coals that formfurnace deposits. Water lances were developed to clean larger areas and use a water jet as the cleaningmedium. This device is a modification of the retractable sootblower. It sprays a jet of water back on to thewall it is mounted on in a spiral pattern as it is inserted into the furnace.

Water lances cover about a 20-foot diameter area. Thus, a 500 Mw size boiler my be outfitted with 40 ormore in an attempt to keep the furnace clean. They have small nozzle area and require high purity water.Lances can be bent if they are hit by falling slag while inserted into the boiler.

Water cannons spray a jet of water from an opening in the furnace wall to the opposite wall. In manyinstallations a cannon can clean from the nose arch to the bottom slope tubes and from one side to the other.Thus 4 cannons, one on each wall, can clean an entire furnace. A concentrated water jet, which isproduced by a special nozzle, crosses the boiler inside and impacts on the slagged wall surface. Thecleaning effect is based on the fact that the impacting water which penetrates the topmost layer and expandsinto steam. In this way the slag is broken up and removed from the surface.

Water cannons offer a number of advantages over conventional wall blowers and water lances. The largenumber of wall blowers and/or water lances requires an extraordinarily high maintenance expense. The useof steam, air, or high purity water as cleaning agents results in high operating costs. Water cannons can usefiltered water with a clear reduction in cost. The water cannons also allow areas to be cleaned that cannotbe outfitted with wall blowers and water lances. Division walls, nose arches, and lower slope tubes areexamples of areas routinely cleaned by water cannons.

In a typical furnace the area cleaned can often be doubled. This increase in cleaning has the effect ofreducing furnace exit gas temperature (FEGT) by increasing the amount of heat absorbed in the furnace.An additional benefit is the reduction in thermal NOx production. Reductions of 100 F in FEGT arecommon and NOx reductions of 10% or more have been documented.

Intelligent soot blowing

The water cannon mechanism allows for 90 degrees of movement in both the vertical and horizontal planes.Also, variable speed motors drive the mechanism providing for constant speed of the water sweepingacross the furnace wall. The speed of the water jet moving across the wall is referred to as Jet ProgressionVelocity (JPV). EPRI studies and actual experience indicates that the JPV is directly related to the thermalimpact on water wall tubes. A JPV of 300 ft/min. results in about 100 deg. F metal temperature excursion.Computer control of the cannon operation allows for any area of the boiler to be bypassed by the cannonspray. Thus, over fire air ports, manholes, burners, and any other special area can be excluded from thewater spray. Also, special zones can be created, to clean above burners, or bottom slope tubes, or any othertroublesome surface.

Traditional boiler cleaning was done on a time or condition basis. Operators monitored steam temperaturesand operated the cleaning devices to control drops in steam temperature. Blowers were set up to run in aset sequence. Often blowing was set on a time basis, every 12 hours, every shift, once per night, etc. Thistype of operation often resulted in the blowing operation following a build up of deposits , it did not act toclean deposits at they occurred. When computer control is applied to boiler cleaning, we often find thecleaning performance, measured by heat flux, can double.

Water cannon systems are combined with heat flux sensors in the furnace walls to detect deposits andoperate the cleaning sequence as needed to keep the furnace in a clean condition at all times. This type ofcomputer controlled automatic system is often called Intelligent Soot Blowing. Heat flux sensors operatewith embedded thermocouples in selected water wall tubes. Paired thermocouples detect the heat fluxgradient in the tube. This is monitored and any reduction in heat flux at constant load is associated withdeposits forming on the furnace surface. The sensor is also used to determine when an area is clean. Withthis type of system the furnace cleanliness can be optimized and steady predictable operation is assured. Inaddition to monitoring the heat flux, the thermal impact on boiler tubes can be monitored. With this systemof monitoring and control precise cleaning can be accomplished with minimal impact on tubes.

Test Program

Most of the water cannon systems are provided with start up testing and fine tuning. Some systems haveadditional performance tests that determine specific accomplishments and compliance with contractguarantees. The data presented in the following sections of this paper are taken from papers that have beenpresented previously or from recent tests. Clyde Bergemann is continually conducting research intoimproved cleaning methods and methods of controlling cleaning equipment. Data from field tests is usedto improve equipment and control systems.

Tucson Electric Springerville

Unit 1 at the Springerville plant is equipped with 2 water cannons. The graph below is typical of thecleaning events at the plant. Heat flux can be seen to improve from the 50 kBTU/hr/ft2 to almost 100kBTU/hr/ft2 when the heat flux sensor is cleaned. The graph also displays the thermal impact to thesurface of the tube. This is done by monitoring the thermocouple in the heat flux sensor that is closest tothe fire side surface.

The installation of cannons at Springerville also resulted in the reduction in variation in main steam andreheat steam temperatures. This is probably due to the cannons ability to control furnace exit gastemperatures to a relatively constant level and the reduction in spray flow necessary to control the steamtemperatures. The charts below indicate the reduction in temperature fluctuation.

The reduction in spray flow is shown on the next chart. This data is over a 12 month period. Data wascollected when the unit was at full load. Most of the data points represent a week or more of data. Thereduction in average spray flow was from 69.3 kLbs/hr to 47.6 kLbs/hr. It is estimated that this reductionin spray flow decreased (improved) heat rate by almost 15 BTU.

Springerville unit 1 Reheat spray flow

30

40

50

60

70

80

90

11/5/2001 12/25/2001 2/13/2002 4/4/2002 5/24/2002 7/13/2002 9/1/2002 10/21/2002 12/10/2002

Date

Reh

eat

Sp

ray

flo

w (

klb

/hr)

Water Cannon Operation Start

69.3

47.6

Alabama Power, Miller

Unit 1 at the Miller plant is equipped with 4 Clyde Bergemann SmartCannons. These cannons incorporatethe latest technology available in furnace water cleaning. The following graph indicates the results of HVTtraverse tests at the plant during the initial performance tests of the cannons. Before cannon installationthere were large differences in temperatures. After the cannons were installed the profile at the nose of theboiler was much more even with lower peak temperatures. Peak temperatures were reduced from levelsabove 2300 oF to 2150 oF. Such a reduction reduces the tendency for slag accumulations in the convectionpass of boilers.

10 20 30 40 50 60 70

5

10

15

20

14001450150015501600165017001750

18001850190019502000205021002150220022502300

Alabama Power Company Miller Steam Plant Unit 1 HVTTest #1 February 18, 2002

10 20 30 40 50 6 0 70

5

10

15

20

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

2100

2200

Alabama Power CompanyMiller Steam Plant Unit 1 HVTTest #7 February 21, 2002Without Water Cannons With Water Cannons

Belle River, Detroit Edison

Detroit Edison has reported on a long term study of water cannon operation at the Belle River unit 1installation. The unit is equipped with 4 water cannons. It is a 640 MW unit with a B&W wall fire boiler.The unit is sub-critical and designed for western sub-bituminous coal with MPS pulverizers. The unitexperienced frequent deratings to below the 600 MW level due to high furnace exit gas temperatures anddeposits plugging the convection pass. 192 wall blowers were replaced with the 4 water cannons and theamount of surface cleaned increased. Data was recovered from plant data loggers for approximately a yearof operation before the cannons were installed. This was compared to data from the first year of operationof the cannons.As can be seen from the following graphs FEGT was reduced by over and average of 100 oF . There was areduction in NOx emissions of 8 to 10%.

Belle River Unit 1 Hourly Average FEGT vs Main Steam Flow Pre vs Post Water Cannon Periods

2,075

2,125

2,175

2,225

2,275

2,325

2,375

4.30 4.35 4.40 4.45 4.50 4.55

Main Steam Flow (M lb/hr)

FE

GT

(Deg

F)

Pre

Post

Poly. (Pre)

Poly. (Post)

Measured Average FEGT at 600 MW (Net) - Unit 1 Days After Boiler Cleaning

2,20

1

2,25

8

2,26

8

2,30

8

2,33

5

2,34

6

2,34

1 2,36

2

2,37

2

2,19

6

2,22

1

2,21

5

2,28

8

2,25

7

2,22

0

2,24

5

2,23

5

2,26

4

2,100

2,150

2,200

2,250

2,300

2,350

2,400

12 daysafter

cleaning

36 daysafter

cleaning

57 daysafter

cleaning

65 daysafter

cleaning

81 daysafter

cleaning

91 daysafter

cleaning

106 daysafter

cleaning

114 daysafter

cleaning

124 daysafter

cleaning

FEG

T (D

eg F

)

Before Cannon

After Cannon

Cannon Problems

Delta =5 Deg F

Delta = 37 Deg F

Delta =53 Deg F

Delta =20 Deg F

Delta =78 Deg F

Delta =126 Deg F

Delta =96 Deg F

Delta =127 Deg F

Delta =108 Deg F

Unit 1 Measured Hourly Average NOx vs Main Steam FlowPre and Post Water Cannon

0.1000

0.1400

0.1800

0.2200

0.2600

0.3000

0.3400

0.3800

0.4200

0.4600

0.5000

0.5400

4.00 4.05 4.10 4.15 4.20 4.25 4.30 4.35 4.40 4.45 4.50 4.55 4.60

Main Steam Flow (Mlb/hr)

NO

x (lb

/MB

tu)

99NOx

01NOx

Expon. (99NOx )

Expon. (01NOx)

SmartGauge Testing

An installation of the patented SmartGauge deposit monitoring system was installed at a southern powerplant. The unit is a 880 MW plant with a CE designed boiler. The boiler has a group of reheat pendantssuspended just in front of the nose of the furnace. This location promotes slag and deposit formation. Aset of strain gages were installed on the rods that suspend the pendants. The gages are specially selected tobe able to measure the weight of the pendant and the relatively small change in the load due to slag and ashdeposits. Also two gages were installed on the rods that suspend the economizer. There was a knownaccumulation of ash on the economizer. The gages were installed during a major outage, so the boiler wasrelatively clean when it returned to service. It is not necessary to have an outage to install the gages as theyare installed in the area above the penthouse. The location of the gages is shown on the two charts below.

Reheater Gage Location Economizer Gage Location

37

Reheater4

8

2 Blower No.11

125

Economizer

10

Individual readings from each gage on the reheater pendants are shown on the chart below. Also shown isoperation of retractable sootblowers. The operation of the blowers coincides with reduction of weight onthe gages.

SmartGauge Individual Support Rod Data

-500

-400

-300

-200

-100

0

100

200

1/8/03 0:00 1/10/03 0:00 1/12/03 0:00 1/14/03 0:00 1/16/03 0:00 1/18/03 0:00 1/20/03 0:00

Str

ain

(mic

ro in

/in)

Guage 6 Guage 7 Guage 8 Guage 9

Guage 10 Guage 11

4 8 7&8 8 7 8 8 4 7 8 7&2 7Retract No.

The output from the gages was converted from strain to stress using the stress-strain relationship for steel.Then with the diameter of the rod know the stress was converted to load in pounds. The data from eachgage was set to zero at its low point. Then the data was combined to give a total weight for the six rods thatwere measured. The following graph is the result.

Economizer rodsCable 5 Cable 4Channel 2 Channel 1

Hot Reheater RodsCable 6 Cable 11 Wireless Cable 7 Wireless Cable 10 Cable 9 Cable 8Channel 3 Channel 8 Channel 4 With 2 gagesChannel 7 Channel 6 Channel 5

Front Wall of Boiler

Birds eye view looking down on boiler roof

Data acquisition system

The gages installed on the economizer were able to detect the increase in the weight of the ash deposit.This weight gain was not affected by blowing of the retractable soot blowers or by the sonic horns that areinstalled in this area. There are known obstacles to flow in this area that allow the ash to accumulate.

SmartGauge on Economizer

-100

-80

-60

-40

-20

0

20

40

60

80

100

120

1/7/03 0:00 1/9/03 0:00 1/11/03 0:00 1/13/03 0:00 1/15/03 0:00 1/17/03 0:00 1/19/03 0:00 1/21/03 0:00 1/23/03 0:00

Str

ain

(m

icro

in/in

)

Guage 4 Guage 5 Linear (Guage 5 ) Linear (Guage 4 )

12R&L 12 R&L 12 R&LRetract No.

Reheat Pendant Ash Load from SmartGauge

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

1/13/03 0:00 1/14/03 0:00 1/15/03 0:00 1/16/03 0:00 1/17/03 0:00 1/18/03 0:00 1/19/03 0:00 1/20/03 0:00 1/21/03 0:00

Del

ta L

oad

on g

ages

(lb

s.)

7&84 8R 8 7 8 8 4 7 8 7&2 8 Retract No.

Conclusion

The various tests demonstrate the benefits of intelligent sootblowing. Water cannon cleaning of the furnacecan reduce furnace exit gas temperatures. The tests also demonstrate improved control of both main steamand reheat steam temperatures. Reduction in reheat attemperation spray flow improves cycle efficiencyand heat rate. Reduction in NOx production is demonstrated over long periods of time. The use of sensorsin the convection pass is also demonstrated. This technology has future uses in control of retractablesootblowers and optimization of the cleaning devices in the convection pass.

References

Furnace Cleaning using Water Cannons at Detroit Edison, Belle River Station, Peter Kohlert, Presented atthe 2002 Water Cannon Users Conference

Furnace Cleaning Using Water Cannons, Charlie Breeding, Presented at Electric Power 2002

Service Report No. 2K2-003, Alabama Power Miller Steam Plant Unit One, High Velocity ThermocoupleTesting, Marcus Mathews, Innovative Combustion Technologies, Inc., February 2002

Demonstration of Clyde Bergemann Water Cannons at Miller Unit 1, Charles Bookhaker, Brian Mead,Mike Carlisle, John Sorge; Southern Company, EPRI Heat Rate Conference, Birmingham, AL, January2003

Results of Performance Tests of Water Cannon Furnace Cleaning at Springerville, Chris Patterson, TucsonElectric Power, Charlie Breeding, Sr. Engineer, Clyde Bergemann; Presented at the EPRI Heat RateConference, Birmingham, AL., Jan. 2003

United States Patent Number 6,323,442 issued November 27, 2001