RST-66

UPDATE OF NOX CONTROLTECHNOLOGIES AT RILEY STOKER

byR.A. LISAUSKAS

andC.E . McHALE

RILEY STOKER CORPORATIONWORCESTER, MASSACHUSETTS

Presented at the1987 JOINT SYMPOSIUM ON

STATIONARY COMBUSTION NOX CONTROLNew Orleans, Louisiana

MARCH 23-26, 1987

A'STE;R M 1 4;1615-0014Q,-

cstcyrNow Part of Babcock Power Inc.www.babcockpower.com

UPDATE OF NO)(., CONTROLTECHNOLOGIES AT KILEY STOKER

C .E . McHaleR .A . Lisauskas

Riley Stoker Corporation

ABSTRACT

Recent design and operating experience with Riley Stoker low-NOXcombustion systems in pilot scale and commercial furnaces isreviewed . The performance of several commercial low-NO X burnerinstallations in wall- and Turbo-fired furnaces is described . Bothemissions reductions from uncontrolled levels and the impact of thecombustion process modifications on furnace temperatures arediscussed . Pilot scale results focus on in-furnace NO, and S0 2control processes such as reburning and sorbent injection .

Recent activities include CCV burner design refinements, anddeveloping staged combustion systems for utility boilers, industrialstoker-fired boilers and circulating fluidized bed combustors . Sincejet aerodynamics influence combustion and NOx reduction efficiencies,two-phase jets in furnace enclosures are discussed .

Paper - 3B

INTRODUCTION I

In the past several years, interest in NO I control has broadened toinclude a wider range of technologies and Ifuels .

Because of therelatively stable boiler population, retrdfit NOx controls have becomeincreasingly important for both large industrial and utilityboilers . Combustion controls such as low NOx burners and overfire airsystems appear to be significantly more cost effective than postcombustion control techniques . The regent promulgation of new federalNO x standards for boilers below 250x10 Btu/hr in size will requirethe implementation of controls on smaller industrial boilers . RileyStoker Corporation has recently field tested new NO x controlcombustion systems on industrial coal-fired Turbo, as well as straightwall boiler designs . As a result of major changes in fuel prices,there is also renewed interest in NOx controls for gas and oil-firedboilers . This interest applies to both the industrial and utilitysectors .

LOW NOx BURNER FIELD TESTS

In addition to traditional boiler systems,i,new technologies are beingdeveloped to utilize other types of fuels .recovery boilers are being constructed andwaste fuels such as municipal solid wastefuels (RDF) . The number of such resourceis expected to increase dramatically overyears . Commercial fluidized bed boiler sy~both industrial and utility boiler applications . Fluidized bedcombustion systems have the unique ability to burn a wide variety offuels with reduced emissions . There is a rowing awareness that alongwith the destruction of harmful toxic materials, air emissions such asNO x must be controlled during waste incineration .

New stoker-fired resourcebrought on-line to burnMSW) and refuse derivedecovery systems in the U .S .he next five to tentems are also emerging in

Recent commercial applications of combustion modifications for NOxreduction have raised system operation and'design concerns . Low NOoperation has the potential to adversely affect both performance an~unit life . Pilot and full-scale efforts are underway to identify andassess these impacts, and to develop practical solutions .The following is a summary of recent Riley experience and efforts todevelop NOx control technology for a varie y of boiler combustionsystems :

Riley Stoker manufacturers two low-NOX burner systems : the ControlledCombustion Venturi (CCV) burner (U .S . Patent No . 4,479,442) for wall-fired boilers, and the Tertiary Staged Venturi (TSV) system (U .S .Patend No . 4,517,904) for Turbo furnaces . Both systems have beenoperating at utility and industrial installations for several years .Typical NOx emissions achieved with these systems are summarized inFigure l . Details of recent field experience with these burners

fol low .

The Controlled Combustion Venturi (CCV) burner was originallydeveloped for retrofit on pulverized coal wall-fired boilers . NO xcontrol is achieved through a patented venturi coal nozzle design,which utilizes both a venturi section and , a four bladed coal spreaderto control fuel/air mixing . This nozzle design provides swirl to theprimary air/coal stream, while separating Ithe stream into fuel-richand fuel-lean layers before mixing with the secondary air . Results ofa recent utility test program on the CCV burner are presented inFigure 2 . Since this unit also has an overfire air system, weobtained data at primary zone stoichiometriies between 1 .0 and 1 .2 .The data demonstrate that the burners are opergting in the field aspredicted by recent Riley, EPA and EPRI 10x10 Btu/hr test programs(1,2,3) .

Mechanical improvements to the CCV burnerThis burner was originally designed as a rspreader for our pre-NSPS Flare burner desthe Flare burner secondary air system are

re also being developed .placement coal nozzle andign . No modifications inade in the retrofit . We

ping a new secondary air register and control system fthe CCV burner . Our objective is to design a new generation burnerthat is easier to operate and reliable over a wide range of registersettings . This is important since pilot aithat achieving low-NO emissions and effecupon consistently controlling secondary ainew generation CCV burner system will be tcombustion test facility this year . A sininstalled in a 300 MW boiler and operatedreliability in a commercial application .

Experience with the CCV burner and staged ¢ombustion systems led tothe development of a low-NO burner for Turbo fired systems, theTertiary Staged Venturi (TSf) burner . The TSV burner has a coalnozzle and spreader assembly resembling th CCV burner . The TSVburner is designed for two stage combustio~

It operates with a fuelrich recirculation zone near the burner throat (40 to 70% oftheoretical air) . Four tertiary air portsvanes are used to raise the total burner f70 and 100% of theoretical air . The balanprovided through the overfire and underfirFigure 3 . NOx emissions of 300 ppm have cwith this system in large industrial boile

IMPROVED OVERFIRE AIR SYSTEMS

d field testing have shownive carbon burnout dependsdesign conditions . Thested in our 100x106Btu/hrle burner will then beor a year to evaluate its

equipped with directionalont stoichiometry betweene of the combustion air isair systems shown innsistently been achievedoperations .

d ,

Overfire air is an important application of staged combustion toreduce NOx emissions in utility and industial boilers . It is alsoused to control furnace temperatures and burnout in stoker-fired

industrial systems . In all of these appli''Ications, a thoroughunderstanding of the mixing process between the overfire air andcombustion zone products is critical to achieving desired systemperformance . Riley has been employing both computational fluiddynamics and laboratory flow modeling to design overfire air systemsfor a variety of firing systems .

Riley will install a new air staging system on a 425 MWe Turbo firedboiler in April . The design objectives were : 1) increase boiler loadto 105° Magximum Continuous Rating (MCR) ; 2) reduce NOx emissions to0 .65 lb/10 Btu while overfiring ; and 3) achieve efficient carbonburnout with the low NO combustion system. An additional objectivewas utilizing as much of the existing equi went as possible tominimize retrofit costs .

Our design to achieve these objectives requires modification of theexisting 24 Directional Flame burners, and diversion and control ofadditional combustion air to the air staging system . The existingoverfire air system is being upgraded to include separate overfire airductwork with a new air control system.

Onderfire air ports arebeing installed to provide additional stag ng and improve lowerfurnace combustion efficiency . Extensive lomputer simulations wereperformed to design the staging system and1predict performance .

The location of the overfire air ports in he furnace venturi,combined with the Turbo furnace burner flow pattern, readily allowsseparating the final OFA combustion zone f om the primary burnerzone . However, the proximity of the under ire air (UFA) ports to theburners is critical . If the underfire air ports are too close to theburners, UFA will be entrained rapidly int the main combustionproduct flow . If this occurs, minimal staging will be achieved andfinal NO x reductions could be disappointing . Burnout considerationsrequire well-controlled mixing of all three flow streams . To furtherimprove burnout efficiency, additional win overfire air ports havebeen installed near the furnace side walls . Flow model studies (4)suggest that the increased mixing achievedwith these wing portspromotes carbon burnout under staged firing conditions .

A computational fluid dynamics code that predicts temperatures, flowstreamlines, turbulence parameters, and pot-combustion oxygenconcentrations was used to analyze the furnace performance impacts ofthese combustion system modifications . Anlisometric view of thefurnace section studied in the computer model is shown in Figure 4 .Figure 5 presents some results of this analysis . The temperatureprofiles shown predict no major changes in .furnace temperatures willoccur when this retrofit is completed . Similar profiles of staged airjet penetration and jet velocities were generated during the designstudy .

RESOURCE RECOVERY

Riley has been active in the design and manufacture of resourcerecovery boiler systems for many years .

~n 1968, Riley supplied amunicipal solid waste (MSW) fired stoker and recovery boiler to theTown of Braintree, Masachusetts . The design of stoker-fired systemsfor waste fuels is subject to several constraints . Waste fuels suchas MSW have widely variable heating value and moisture contents .Refuse systems, therefore, require a flexible combustion air system toinsure complete combustion and to minimize the formation of pollutantssuch as CO and NOx . Overfire air is ofte used in these stoker-firedsystems to enhance incineration, combustio flexibility, andcontrol . The same computer code used to analyze OFA systems forsuspension fired systems is also being used to evaluate theeffectiveness of OFA stoker fired refuse s stems . Figure 6 representsoxygen profiles computed for a traveling gate stoker firing a refusederived fuel (RDF) . Comparing these oxyge~ concentration profiles isan invaluable tool in evaluating furnace m xing and in designingefficient combustion systems for the destruction of toxic constituentsin these waste fuels .

I

In 1984, Riley obtained the exclusive Amerstep grate stoker refuse combustion systemJapan . Since 1963, Takuma has installed sMSW systems throughout Japan . These facilcapacity in excess of 50,000 tons per day .,boiler system is currently undergoing star-up and testing inOlmstead, Minnesota .

i

The Riley/Takuma combustion system includes both a step grate stokerdesign and an automatic combustion controldiagrams of the combustion system, and thecontrol system are shown in Figures 7a andindividual undergrate air control systems,and ignition, combustion, and burnout . Approximately 20% of thecombustion air is provided by the overfire air system. Multiplelevels of overfire air jets are used to control mixing above the grateand furnace temperatures . The combustion Ontrol system combined withgrate speed and furnace control systems provide the flexibility toburn a variety of MSW fuels regardless of ource, season, andgeographic origin .

To insure complete combustion, MSW systems operate with excess airlevels of 80% or more . The automatic oxygen control system servesminimize furnace 02 levels for improved combustion efficiency andlower NO x emissions . NOx emissions of 1001ppm (at 12% 02) and lowerhave been acheived under this mode of operation . A number ofcombustion modification techniques can also be used on commercial MSWsystems to further control NO

20V and

injection above the grate hasachieved NOx reductions of 20V and greater

The same level of NOx

ican license for an advancedfrom Takuma Co . Ltd . ofme 250 resource recoveryties represent a combinedThe first Riley/Takuma

system (5) . Schematiccombustion air and oxygen7b . Three grates, withare used to promote drying

to

control has also been attained with 20 to'30q flue gas recirculation(FGR) . FGR has the added benefit reducin excess air requirements .Takuma is also evaluating and testing sev ral in-furnace NOx reductiontechniques including ammonia and UREA injection . In addition, Rileyand Takuma are currently evaluating an integrated NOx control conceptbased on natural gas co-firing .

FLUIDIZED BED BOILERS

Atmospheric fluidized bed combustion offers another route forachieving both low NO and SO2 emissions without post combustion fluegas treatment .

Rileyxcurrently designs and manufacturers fluidizedbed boiler systems based on an advanced ci ;culating fluidized bedcombustion technology known as Multi-Solid~ Fluidized Bed Combustion(MSFBC) . MSFBC boiler systems are supplied in the U .S . by Riley undera license agreement with Battelle Development Corporation .

The MSFBC system, shown in Figure 8, is a wo stage combustion process(6) . It consists of an entrained bed of fine ash and limestoneparticles superimposed on a fluidized dens bed of large particles .The lower dense bed of large particles

ser es to promote mixing and

increase the residence time of entrained a h, fuel and limestone . Thecombustor operates in a fast fluidized bed mode with superficial gasvelocities of 30 to 35 ft/sec . Heat trans er and combustion aredecoupled in this system . Heat recovery o curs in an external heatexchange and a convective boiler . The exte nal heat exchanger actuallyconsists of heat exchange tubes immersed i a conventional lowvelocity fluidized bed . Combustion temper ture is controlled between1500 and 1700°F by recycling entrained ash through the external heatexchanger .

Low combustion temperatures and a two stage design offer the potentialfor achieving NO x emissions significantly 1requirements . MSFBC systems are operatedthe combustor under fuel rich conditions .i n the upper oxidation zone . As in otherthe three most important variables affectirlsystem include : (1) primary zone stoichiometry ; (2) primary zoneresidence time, and (3) primary zone temperature .

ower than current NSPS,ith the lower dense bed ofCarbon burnout is completedtaged combustion systems,g NO x control in a MSFBC

n MSFBC NOx emissions isdata and data from awide range of operating

residence time under properpproaching 0 .1 lb/106Btu/hr

The importance of each of these variablesshown in Figure 9 . Coal-fired pilot scale50x106 Btu/hr field unit are presented forconditions . Given sufficient primary zonefuel rich conditions, NO emission levelscan be achieved . In MSFBC systems, primary zone temperature iscontrolled by adjusting the solids recyclecombustor . In addition to these primary cdsecondary NOx dependence has been observed'

rate to the lowerntrol variables, awith fuel particle size,

fuel nitrogen content, and excess air level .

Riley continues to design and develop MSFBC systems for new andincreasingly larger boiler applications . (MSFBC systems are capable ofburning a wide range of waste materials, as well as coal .

It's twostage design makes MSFBC well suited for the combustion of difficultto burn high nitrogen species . Currently, there are ten MSFBC boilerinstallations, ranging in size from 50,000 to 625,000 lb/hr of steameither operating, undergoing start up, or under design andconstruction throughout the world .

PILOT SCALE STUDIES

Riley is continuing to evaluate combustionthat can be retrofit to existing industriasystems . In addition to the 100x106 Btu/hr(CBTF), the Riley Research Center recently3x106Btu/hr Pilot Scale Combustion Facili

modification techniquesand utility boilerCoal Burner Test Facilitybegan operating ay tNS~r~ . ine NSL;r is

designed to simulate thermal conditions throughout the radiant andconvective passages of a boiler . This new,facility is equipped withair cooled tube banks to represent various convective heat transfersurfaces from superheater to the economize . The PSCF has been usedto investigate both in-furnace SO

and NO X control techniques .

S02removal characteristics for both furnace injected limestone anddolomite as measured in the PSCF are shown in Figure 10 . Thetemperature profile and sorbent injection location maintained duringcombustion tests on a high sulfur (3 .9%) Kentucky bituminous coal isshown in Figure 11 . Quench rates in the s lfactor reaction window(2300 to 1600°F) were on the order of 300*F/sec . This condition istypical of many full scale furnaces .

The PSCF has also been used to evaluate advanced NOx control conceptssuch as reburning . Figure 12 summarizes t'e results of NO emissionswhen natural gas is injected as the reburn fuel above a pulverizedcoal flame . NOx emissions are shown as a unction of reburning zonestoichiometry .

Uncontrolled NOx emissions for these tests wereapproximately 550 ppm . Reburning zone residence times ranged from 400to 700 mills . Figure 10 also compares PSCr results with resultsobtained in Riley Research's 100x106 Btu/h~ combustion burner testfacility using coal as the reburn fuel (2) . Final NO x emissions inboth cases are similar . However, in coal-fired applications whereupper furnaces residence times are limited, the use of gas or oil asthe reburning fuel may be preferable since faster combustion rates canbe obtained .

SUMMARY

Riley Stoker is developing NOx control systems for a wide range ofboiler designs and fuel types .

Combustion controls such as low NO ~,burners and overfire air systems are being; installed on utility anaindustrial wall-fired and Turbo-fired units . In addition to thesetraditional boiler systems, new technologies which utilize other typesof fuels, are being developed . These include a circulating fluidizedbed and stoker-fired resource recovery boilers to burn waste fuels .

Combustion process modification techniques such as air staging arecommon to all of these boiler systems . Evaluations of stagedcombustion operation are focusing on two areas : determining theoptimum amount of staging air to limit pollutant formation, andoptimizing the staging system design to ma ntain burnout and furnacetemperature requirements . Three dimensional computer simulations ofthe turbulent mixing processes in overfire'air applications are beingused to establish commercial system designs . In additon, sorbentinjection systems for in-furnace sulfur emissions control andreburning with natural gas as the staging fuel have been evaluated inour pilot scale combustion facility . Future plans include evaluatingstoker firing systems in this test furnaces This combination ofexperimental and analytical evaluations is enabling Riley to designcommercial NOx control systems for a broad range of combustion systemconfigurations and fuel types .

REFERENCES

1 . Penterson, R .A . Development of an Economical Low NOx FiringSystem for Coal Fired Steam Generators . Presented at the ASMEJoint Power Generation Conference, Denver, October, 1982 .

2 . Lisauskas, R ., Snodgrass, R ., Johnson, S ., and Eskinazi, D .Experimental Investigation of Retrofit Low NOx CombustionSystems . In : Proceedings of the 1985 Joint Symposium onStationary Combustion NOx Control .

3 . Lisauskas, R ., Itse, D ., and Masser, C .Performance from Single Burner Tests tProceedings of the 1985 Joint SymposiumNOx Control .

4 . Lisauskas, R ., McHale, C ., Afonso, R .,Development of Overfire Air Design GuiFired Boilers . Presented at the 1987Stationary Combustion NOx Control, New

5 . Palacios, F .A . and Wakamura, Y . AutomMass Fired Refuse Incineration Plant .

Extrapolation of BurnerField Operation . In :on Stationary Combustion

and Eskinazi, D .elines for Single Walloint Symposium onOrleans, March, 1987 .

tic Combustion Control ofPresented at the

International Public Works Congress and Equipment Show, LosAngeles, September, 1985 .

6 . Nack, H ., Litt, R .D . and Kim, B .C . Multisolid Fluidized BedCombustion . Chemical Engineering Progress, January 1984 .

NOX (ppm)

1000 r-

800

600

400

200 0

MEMO

FIELD

RILEY CBTF

MITSUI PILOT FURNACE

dFlare

CCV CCV-MS Directional

TSV

wFlame

WALL-FIREDTURBO

Figure 1 . Burner Performance Comparison

NOx , ppm

600

400

200

0 .6

0 .8

1 .0

1 .2

Primary Zone Stoichiometry

Figure 2 . CCV Burner Performance at a 360 MWe Unit

Figure 3a . Advanced Turbo FurnaceStaging System

TSV Burner

0 .8

1 .0

1 .2

Burner

Zone

Stolchio m etry

Figure 3b . NOx vs . Burner Zone Stoichiometryfor the TSV System (Coal Firing at100% and 60% Load)

Figure 4 . Isometric View of Furnace Modeling Domain

Figure 5 . Temperatures for Original and Staged Operating Conditions

OFA -+~(out of plane)

.06

`--~OFA

q$-0FA

Figure 6 . Predicted Oxygen Mass Concentration Profiles for anRDF-Fired Boiler with Multiple Levels of Overfire Air Injection

uli1i11 i IqIIIIII I',LIIIIIillllla"j

11,1)

NI'~1 I~l, I I lj

_

lli

REFUSE CHARGINGHOPPER AND CHUTE

16

Figure 7a . Riley/Takuma Municipal Solid Waste

Combustion System

AIR EXCESS RATIO SETTINGFUELINPUT SETTINGCALORIFIC VALUE SETTING

Fiqure 7b . Schematic Diagram of Riley/Takuma Combustion Air and LowOxygen Control System

Fluidizin g Air

Solids RecycleA Coalp Sand and Limestone

Ash

Figure 8 . Multi-Solid Fluidized Bed Combustion Process

Figure 9 . NOx Emissions Data from MSFBC Systems

0 .2

0.4

0.6

0.8

1 .0

1.2

1 . . 0 .0

0.2

0.4

0.6

0.8

1 .0

1 .2 1400

1500

1600

1700

1600

PRIMARY AIR / STOICHIOMETRIC AIR

REDUCING ZONE RESIDENCE TIME (s)

TEMPERATURE (°F)

0 .6 0.6 - 0.6f

fSymbol Scale

0 .5 (106 Scale0 .5 Symbol Scale 0 .5 -'w Symbol Scale(10 6 Btu/hr) (106 Btu/hr)O 50 .0m 0 50 .0 0 .4 a O 50 .00 .4 1 .0 " 0 .40 11 1 .0 4 .0 w I 1 1 .0

0 .4O

4 .0 D 0 .4 w 1 .0 v 0 .4 o. v 4 .00

1 .0 F F

a WEO v

?IO w) F vw

F 0 .3 'x 0 .3 " o n 0 .3 w . vOIOF WI

OO

J

y'J OI

0 .2F

0.2 0.2

0 °1 0Z od .10 .1

PRDA NOx LIMIT - 0 .1 PRDA NOx LIMIT I PRDA NOx LIMIT" -O-O 0 .1 -s- O-

O' OI

00 I

0 .0 I 0 .0 0 .0 1 I

1870 ° F

1970 ° F --!

2110°F

2390 ° F

3 .5 FT-310

4.5 FT

ASHPIT f

Figure 10 . 3 x 106 Btu/hr . Pilot-Scale Combustion Facility

20

SORBENTINJECTION

BURNER

1450°F

zO

100

U 600w

NO

40

20

1

2

3

4Ca/S, MOLAR RATIO

(a) SULFUR CAPTURE

Figure 11 . Comparison of Sorbents Tested in the RileyResearch Pilot-Scale Furnace . SorbentInjection at 21000F, 2 .8 x 10 6 Btu/hr,Kentucky No . 9 Coal

NOx, ppm

400

300

200

100

O CBTF Tests coal/coal

A PSCF Tests coal/gas

-I

i -A

t

1

1

I

t

'

i

0

0 .2 0 .4 0 .6 0 .8 1 .0

Stoichiometric Ratio, SR2

Figure 12 . Reburning Test Results - 100 x 106 Btu/hr CBTF and 3 x 106 Btu/hr PSCF