combustion characteristics a front-wall-fired pulverized ......pulverized coal util]ty boiler...

17
C ombL$t. S c i. dild T e ch., 199'1. \ o1 129, pp 2'7'l 293 Reprints available directly from the publisher Photocopying permitted by license only O 1997 OPA (Overseas Publishers Association) Amsterdam B.V. Published under iicense under the Gordon and Breach Science Publishers imPrint Printed in India. Combustion Characteristics of a Front-Wall-Fired Pulverized-Coal 300 MW" UtilitY Boiler M. COSTA*, J. L. T. AZEVEDO and M' G' CARVALHO Mechanical Engineering Department, lnstituto lynerigr T1cnicol Technical aiiviisiiv or tisbon, ni. Rovisco pais, 1096 Lisboa codex, Portugal (Received 21 March 1997; tn final form 26 June 1997) This paper describes the results of an experimental study undertaken in an 300 MW", front-wall- n..J,'pril*"r"a_coal, utility boiler. The data reported include local mean gas species concentra- iio"iJiOr, CO, COr, NO,, anO gas temperatures measured at several ports in the boiler including it or" in tf," torn"l. *giorf una i"""ideni wall heat fluxes taken around the boiler periphery at 39 p.iir. iir" *.iaent will 6.ul nu"". are reported for_two boiler operating conditions During the il;;il;t"l;it report"dhere, a consiierable effort was made to assure minimum variations ;;i;;l;;;p;t"ting conditions and coal chemical and particle size qharacteristics so that the data pi.r"rt"a u." .rp"-"iully o."f.t io, 3-D mathematicai model evaluation and development. The i"rutt, ,.u"^t that: (i) the toonOury air injected below the first row of burners leads to oxidizing "onOiiion, .tor" to ihe back walt; 1i1.1 locai gas temperatures and CO concentrations in the boiler, n.ui tt " urr.rr"rs, ..ached *a*i-u* nalo.Jof about 1470'c and 1.670, respectively; (iii) above the boiler nose the measured No, "on""nt.utions are reasonably uniform with-averaged values of "U"ri eiO ppq and (iv) wall radiant heat fluxes present maximum values at the side wall close to the intermediate row of burners. Keywords:Utilityboiler;pulverized-coal;combustionmeasurements;nearburnerregion INTRODUCTION Inapreviouswolk(Costaetal.,!996),wehavereportedcombustiondatafor an industrial, glass-melting furnace, when we learned, in practice' that consis- tent full-scale data are notoriously difficult to obtain' Apart from the difficul- ties of measuring in industrial furnaces, it is extremely difficult to establish and *Corresponding author' Fax: 3 5 I -1 -841 5545 21'1

Upload: others

Post on 18-Feb-2021

6 views

Category:

Documents


0 download

TRANSCRIPT

  • C ombL$t. S c i. dild T e ch., 199'1. \ o1 129, pp 2'7'l 293Reprints available directly from the publisherPhotocopying permitted by license only

    O 1997 OPA (Overseas Publishers Association)Amsterdam B.V. Published under iicense

    under the Gordon and Breach SciencePublishers imPrint

    Printed in India.

    Combustion Characteristicsof a Front-Wall-Fired Pulverized-Coal300 MW" UtilitY BoilerM. COSTA*, J. L. T. AZEVEDO and M' G' CARVALHO

    Mechanical Engineering Department, lnstituto lynerigr T1cnicol Technicalaiiviisiiv or tisbon, ni. Rovisco pais, 1096 Lisboa codex, Portugal

    (Received 21 March 1997; tn final form 26 June 1997)

    This paper describes the results of an experimental study undertaken in an 300 MW", front-wall-n..J,'pril*"r"a_coal, utility boiler. The data reported include local mean gas species concentra-iio"iJiOr, CO, COr, NO,, anO gas temperatures measured at several ports in the boiler includingit or" in tf," torn"l. *giorf una i"""ideni wall heat fluxes taken around the boiler periphery at 39p.iir. iir" *.iaent will 6.ul nu"". are reported for_two boiler operating conditions During theil;;il;t"l;it report"dhere, a consiierable effort was made to assure minimum variations;;i;;l;;;p;t"ting conditions and coal chemical and particle size qharacteristics so that the datapi.r"rt"a u." .rp"-"iully o."f.t io, 3-D mathematicai model evaluation and development. Thei"rutt, ,.u"^t that: (i) the toonOury air injected below the first row of burners leads to oxidizing

    "onOiiion, .tor" to ihe back walt; 1i1.1 locai gas temperatures and CO concentrations in the boiler,

    n.ui tt "

    urr.rr"rs, ..ached *a*i-u* nalo.Jof about 1470'c and 1.670, respectively; (iii) above theboiler nose the measured No, "on""nt.utions are reasonably

    uniform with-averaged values of

    "U"ri eiO ppq and (iv) wall radiant heat fluxes present maximum values at the side wall close to

    the intermediate row of burners.

    Keywords:Utilityboiler;pulverized-coal;combustionmeasurements;nearburnerregion

    INTRODUCTION

    Inapreviouswolk(Costaetal.,!996),wehavereportedcombustiondataforan industrial, glass-melting furnace, when we learned, in practice' that consis-

    tent full-scale data are notoriously difficult to obtain' Apart from the difficul-

    ties of measuring in industrial furnaces, it is extremely difficult to establish and

    *Corresponding author' Fax: 3 5 I -1 -841 5545

    21'1

  • 278 M. COSTA er al.maintain constant operating conditions over large periods of time. To mini-mize this problem a strong and effective cooperation between the experimentalgroup and the boiler operators is essential. The present research group hasmaintained, lor several years now, a solid collaboration with the people of apower plant located in Sines, about 140 km south of Lisbon on the Atlanticocean coast, property of the portuguese Electricity company (EDp). Theexisting good relations with the operators of Sines power plant have encour-aged the present authors to conduct the study reported here, whose objectivewas to extend the limited data base available for full-scale utility boilers firingpulverized-coal. Such combustion data aremuch needed lor the developmentand evaluation ol 3-D mathematical models.

    The amount of pulverized-coal data available for full-scale utility boilersconcern primarily gross flame properties such as radiative information andflue-gas analysis which does not give the detail required to interpret the morefundamental aspects of the combustion process. Examination of the literaturereveals, however, the existence of a series of remarkable studies from BrighamYoung University (Byu) in the USA (see Butler and webb, 1991; Bonin andQueiroz, 1991,1996; Butler et al, 1992; eueiroz et al., 1993 and Black andMcQuay, 1996, among others) carried out on pulverized-coar, tangentially-fired boilers. Butler and webb (1991) reported measurements of local meantemperatures and incident wall radiant heat fluxes for an g0 MW" pulverized_coal, tangentially-fired boiler. These authors have paid special attention to theburner region. Simultaneously, Bonin and eueiroz (1991) reported the experi-mental results of a parametric study on the same boiler on local mean particlevelocity, size and number density at three ports above the burner level. In thisstudy, variable test parameters included furnace load, excess air and burnertilt. Recently, the BYU research team conducted a notable series of measure-ments on an 160 MW pulverized-coal, tangentially-fired boiler to investigatethe effect of coal type, overfire air and burner tilt (see Black and Mceuay,1996). The data taken during the complete series of tests consisted of particlesize, velocity and concentration, gas temperature and verocity, wall heat fluxes,solids composition, and species concentrations. Experimental data fbr full-scale utility boilers available from other sources have, in general, little detailandlor are limited to locations above the burners, see, for example, Boyd andKent (1986), Fiveland and Latham (1993), Kakaras et al. (7995), Epple et al.(1995), Schnell et al. (1995), Dal secco et al. (1995) and Magel et al. (1996),among others.

    It is clear that little attention has been devoted to the burner region offull-scale utility boilers. This paper reports new combustion data obtained inan 300 MW", flront-wall-fired, pulverized-coal, utility boiler using swirl

  • PULVERIZED COAL UTIL]TY BOILER 219

    burners. Measurements have been made of local mean gas species concentra-

    tions (Or, CO, CO2, NO"), and gas temperatures measured at several ports inthe boiler including those in the burner region, and incident wall heat fluxestaken around the boiler periphery at 39 ports. The incident wall heat fluxes are

    reported for two boiler operating conditions'

    THE UTILITY BOILER

    Sines power plant has four generating units, each rated at 300 MW" and with a

    maximum continuous steam output of 950 ton/h at 535'c and 167 bar- Afterthe high pressure turbine stage, steam is reheated at 43 bar to 535"C. For thepurposes of the present work unit II was considered the more suitable onemainly because (i) EDP has recently conducted extensive global performance

    tests in the unit (ii) the present group (Azevedo et aI', 1996) has conducteddetailed studies for the heat transfer in the convection section ofthe boiler and

    (iii) new burners are now being installed, which will allow us to conduct a new

    experimental comparative study in the future. The boiler of unit II wasconstructed by Mague/Foster Wheeler and is in operation since 1986'

    Figurelpresentsaschematicviewofthefront-wall-fired,pulverized-coalboiler,showingtheburnersandtheinspectionportswheremeasurementshavebeenperformedduringthisstudy.Withtheexceptionofports8,9andl0'each port is represented here by a two-digit figure: the first identifies the level

    from the bottom to the top and the second the position in each level. Theposition in each level is numbered clockrvise from 1 to 8 as indicated in thehg,rr.. The precise elevation of each level is given in Figure 1 with reference to

    ttre bottom of the boiler. The boiler is equipped with a large number ofinspectionportseachofaboutT5mmindiameter.Theinspectionportsavailable at the front wall in level 6 are of difficult access. At the back wa1l,ports are only available at levels 1, 3 and 5 with limited access due to theconvective path of the boiler. Above the boiler nose three ports are available at

    three different levels, numbered in Figure 1 from 8 to 10'The boiler dimensions are approximately 43'5 m height' 15 m wide and

    l|.4mdeep.Itisequippedwith20swirlburnersarrangedinfiverowsoffourburners installed on levels 1 to 5 (see Fig. 1). The distance between the burner

    centerlines is 2.6 m both in the vertical and horizontal directions' The distance

    from the side walls to the centerline of the burners cl0sest to them is 3'6 m' Atlevel 0, just below the first row of burners (level 1), the boiler has two holes for

    boundaryairinjection(seeFig.1)eachwithaninnerdiameterof0'95m.Thedistance from the centerline of these holes to the side walls is 1.15 m' The

  • 280 M. COSTA et a1

    26 v fil

    " v 10:\05 - Pods

    V 25,7

    7.6 v 23,0

    66 e 2O.8

    io i;:tl

    *; *'*f",; t t

    Boundary alr intels all dimensions in mBoundary air

    inlets

    FIGURE 1 Schematic view of the bo'er showing the burners and the inspection ports.

    inspection ports used for measurements are located at a distance of 0.43 m fromthe boiler corner, with the exception of ports 7.2 and7.5 whose distance is 0.63 m.The coal is transported to five Foster wheeler MBF purverizers withgravimetric feeders to supply each burner row. The coarlprimary air stream isintroduced tangentially in an annulus between an inner sleeve (o.d. 0.273 m) andthe secondary air inlet (i.d. 0.50g m). Inside the inner sreeve there are a propaneigniter and an oil burner for start-up purposes. During normal operation coreair is fed through the inner sleeve to prevent coal deposition. The secondary air isfed through a windbox after which it is forced through a set of adjustable vanes.The secondary air enters the boiler through two concentric annular inrets withexternal diameter of 0.74 and 0.99 m, respectively.

    DATA ACQUISITION TECHNIQUES

    Gas Species Concentration and TemperatureThe sampling of gases for the measurement of local mean 02, CO, CO, andNO, concentrations was achieved using a 5 m long, water-cooled, stainless steel

  • PI]I,VERIZED COAL UTILITY BOILER

    probe. It comprised a centrally located 3 mm i.d. tube through which quenchedsamples were evacuated, surrounded by two concentric tubes for probecooling. The gas sample was drawn through the probe and part of thesampling system by an oil-free diaphragm pump. A condenser removed themain particulate burden and condensate. A filter and a drier removed anyresidual moisture and particles so that a constant supply of clean drycombustion gases was delivered to each instrument through a manifold to givespecies concentrations on a dry basis. The probe was cleaned frequently byblowing back with high pressure air to maintain a constant suction flow rate.The analytical instrumentation included a magnetic pressure analyzer(HoribaModel CFA-321A) for O, measurements, and non dispersive infrared gasanalyzers (Horiba Model cFA-311A) for CO, COr, and No- measurements.Zero andspan calibrations with standard mixtures were performed before andafter each measurement session. The maximum drift in the calibration waswithin 1- 2o/o of the full scale. The major sources of uncertainties in theconcentration measurements were associated with the quenching of chemicalreactions and aerodynamic disturbarice of the flow. Quenching of the chemicalreactions was rapidly achieved upon the samples being drawn into the centraltube of the probe due to the high water cooling rate in its surroundingannulus our best estimates indicated quenching rates of about 106 K/sec. Noattempt was made to quantify the probe flow disturbances'

    The literature, referred to in the introduction, reveals that for pulverized-coal fired utility boilers suction pyrometers had been preferred to fine wirethermocouples. This preference cannot be attributed to their superior per-formance but is probably due to the tendency of a thermocouple wire to breakeasily when exposed to hostile environments such as those encountered infurnaces. In addition, suction pyrometers cause greater aerodynamic andthermal disturbances to the flow field and present a tendency for probeblockages when sampling from the burner region as observed by, for example,Butler and Webb (1991). The fine wire thermocouple technique was preferredin the present study in view of the above considerations so that local mean gastemperature measurements were obtained using uncoated 300 pm diameterplatinum/platinum: 13% rhodium thermocouples. The 300 pm diameter wireswere located in a twin-bore alumina sheath with an external diameter of 4 mmand placed inside a 5 m water-cooled stainless steel probe. As radiation lossesrepresent the major source of uncertainty in the mean temperature measure-ments an attempt has been made to quantify them in industrial environmentsin a recent study (Costa et a1.,1996).The calculation have indicated that in theregions of highest temperature the "true" temperature do not exceed themeasured one by more lhansoh.An additional source of uncertainty relates to

    28r

  • 282 M. COSTA et al

    the deposition of ash on the thermocouple wires. This results in an increase inradiation losses owing to the increase in surface area. The error due to ashdeposition on the thermocouple wires was minimized during measurementsby always moving the probe from the boiler walls towards the inner positionswhere the amount of char particles was higher. The condition of the thermo-couple was frequently examined during measurements in these inner positionsand, where necessary, the deposits were carefully removed or the thermo-couple replaced.

    Both probes were mounted on a traverse mechanism which allowed formovements along a line normal to the boiler walls through each inspectionport up to 3 m from the wall. Great care was taken to avoid leakage of air intothe boiler through the inspection ports, especially when probing from posi-tions close to the side or back walls. No thermal distortion of the probes wasobserved and the positioning of them in the boiler was accurate to within* 10 mm. The analog outputs of the analyzers and of the thermocouple weretransmitted via AID boards to a computer where the signals were processedand the mean values computed.

    lncident Wall Heat Flux

    The incident radiation fluxes at the boiler wals were measured using anasymptotic thin film, water-cooled, gas-purged radiometer (HY_CAL EN_GINEERING). The radiometer has a 140'field of view and can measure totalradiant fluxes up to 400 kw/m2 in the wavelength range of 0.5 to 15 pm with aresponse time on the order of 180 milliseconds. The entire radiometer systemwas portable and easily handled by two workers. The radiometer was calib-rated in a black-body furnace according to the procedure set out in chedailleand Braud (r972). In this study, calibrations carried out before and after theexperimental campaign showed differences of less than -f 3"h which gives anindication of the uncertainty of the measurements.

    EXPERIMENTAL CONDITIONS AND DATA RELIABILITY

    During the present study a mixture of two coals has been used: an Americancoal (Anker) and an Australian coal (oak Bridge). The coal mixture propertiesand its size distribution for each row of burners are given in Tables I and II,respectively. Table III summarizes the boiler operating conditions for whichthe results reported herein were obtained. The values given in Table III areaveraged over all the duration of the experimental campaign which has lasted

  • PULVERIZED COAL UTILITY BOILER

    TABLE I Coal characteristics (courtesy of the Sines powerplant)

    Quantity

    283

    Proximate Analysls (u/, as receit:ed):

    MoistureAshVolatilesFixed Carbon

    Higlh Heating Value (MJlkg)

    [Jltimate Analysis ('l as receiued):CarbonHydrogenNitrogenSulfurOxygen

    6.1713.3628.8 551.02

    27.65

    67.684.41.540.945.31

    TABLE II Coal size distribution in the different burner rows+ (cour-tesy of the Sines power plant)

    Grind Rorr 3 Row 4 Row 5

    7o under 16 pm7o under 30.8 pm96 under 79.3 pm7o under 106 pm7u under 204 pmSauter meandiameter (pm)

    32.449.979.686.196.6

    1 3.89

    40.258.682.087.297.6

    1t.73

    34.552.178.885.296.9

    13.36

    36.6 20.155.0 37.281.3 69.486.7 19.596.9 96.512.64 20.89

    *Row I is the bottom row of the burners.

    lor about two weeks for runs I and2.It will be noted in the following sectionthat the data reported for run 2 includes incident wall heat fluxes only.

    The operating conditions of run 1 are frequently used at Sines power plantin order to control NO, formation/emissions from the present burner/boilerconfiguration. In a previous study the use of the upper row of burners (level 5)with overfire air (run 1) was investigated by EDP as a primary measure forNO., reduction. The results were compared with model pfedictions (Coimbraet al.,Igg4),showing that NO" emissions could be reduced with this modifica-tion when compared with the case of all burners in operation (run 2).

    It should be emphasized that during the experimental campaign, a veryconsiderable effort was made by the people involved in the operation of theboiler in order to assure minimum variations on boiler operating conditionsand coal chemical and particle size characteristics. The use of soot blowers inthe lurnace was avoided during the measurements. In support of our claim of

  • 284 M. COSTA er al.

    TABLE III Boiler operating conditions (courtesy of the Sinespower plant)

    Quantity Run 1 Run2

    312180

    1032'7E220

    28870*780

    629 616142* 131+113 9690 t270.6 0.6160 16064 64360 360360 3602.5 2.415.5 15.515 8600 8502150 203033 453.2 2.5

    *Per row; **Swirl number as defined in Be6r and Chigier (1972); swirldirection in each burner is shown in Figure 1.

    good repeatability, example profiles of o, and No" concentrations and tem-perature for run 1 are reported in Figure 2 for several ports along the boilertaken in different days. Excellent repeatability is demonstrated without whichthe usefulness of the data for evaluation and deveiopment of 3-D mathe-matical models would be questionable.

    RESULTS AND DISCUSSION

    Before examining the data in detail, some considerations concerning thenature ofthe flow in the present boiler configuration are required in order to

    Net electrical generation (MW)Steam pressure in the boiler drum (bar)Coal feed rate, as received (ton/h)

    rowlto3(ton/h)row 4 (ton/h)row 5 (ton/h)

    Primary air flow rate (ton/h)rowlto3(ton/h)row 4 (ton/h)row 5 (ton/h)

    Secondary air flow rate (ton/h)rowlto3(ton/h)row 4 (ron/h)row 5 (ton/h)

    Secondary air swirl number**, evaluatedBoundary air flow rate, evaluated (ton/h)Primary airfuel inlet temperature ('C)Secondary air inlet temperature ('C)Boundary air inlet temperature ("C)O, in flue gases (dry volume %)CO, in flue gases (dry volume 7o)CO in flue gases (dry volume ppm)NO, in flue gases (dry volume ppm)SO, in flue gases (mg/Nm3)Particle concentration (mg/Nm3)Carbon in ash (%)

    307

    177

    10522*1722

    33465r7069

  • PULVERIZED COAL UTILITY BOILER 285

    ooE:tGoaE(tF

    Eo.CIq)t6.tlEIG

    o(tcoooz

    1500

    1 200

    900

    600

    300

    0'1500

    1200

    900

    600

    300

    01 500't200

    900

    600

    300

    0

    25.20.15.10,5,0"25.20

    15

    '10

    5

    025

    *(,EE-ttc.9G

    o.Jcooo

    Porl4.6

    Port 3.6

    Port 2.5

    1500

    1 200

    900

    600

    300

    0

    10

    25

    20

    15

    10

    5

    0

    Porl 1.5

    o 40 80 120 160 200 240 2AODistance to the wall (cm)

    .)Oz INO, aT- (*)2dayslaterooz tr No,^ or^

    FIGURE 2 Profiles of local mean gas species concentration and gas temperature for run 1through several ports taken on different days.

    better understand the discussion of the results presented below. In whatfollows the principal qualitative features of the flow are based in a numericalstudy of the present research group for the same boiler and operatingconditions, see Coimbra et aI. (1994). In these calculations, however, theboundary air flow rate was neglected. The flow is mainly characterized by thedeflection of the jets from the burners into two main parts close the back wallbetween levels 2 and 3. The flow deflected upwards presents high velocitiesnear the boiler nose. At the ash pit region of the boiler alatge recirculationzone is formed promoting an upward flow close to the front wall which issuperimposed to the swirling motion created by the burners'

  • 286 M. COSTA et a1

    Combustion in-Furnace DataFigure 3 shows profiles of local mean gas temperature for run 1 throughsymmetric ports where a good degree of symmetry is observed. This will havean obvious influence on the symmetry of the radiative flux incident on thecorresponding walls, as will be shown below. Measurements of local mean gasspecies concentrations through symmetric ports were not performed duringthe present campaign.

    Figures 4 and 5 show profiles of local mean gas species concentration andgas temperature for run 1 through several ports located at the side wall and atthe back wall (ports 1.4 and 3.4), respectively. It will be noted in Figure 4 thatmeasurernents of local mean gas species concentrations through ports 5.5, 5.6,7.5 and 7.6 were not performed during the present campaign. viewed as awhole Figures 4 and 5 show the strong three-dimensional nature of the flow.

    At port 0.6 (Fig. 4), the region between about 100 and 1 60 cm from the boilerside wall presents low temperatures and high o, concentrations which is dueto the mixing effect of the boundary air injected at level 0 (see Fig. 1). Beyond

    '| 500

    1200

    aaa"'C'l'o Port 1.4. Port 1.3

    EcCCECErr.oolo

    o Port 2.6. Port 2.1

    o88'89tStoa a a a C oporro.o

    . port 0,10 40 80 120 160 200 240 280 o

    Distance to the wall (cm)

    FIGURE 3 Profiles of local mean gas temperature for

    .tCCoo"88

    o Porl3.4. Port 3.3

    60Oooo0600ooooo Port 2.5. Potl2.2

    ...r93c8oggggso pod 0.5. port 0,2

    40 80 120 160 200 240 280Distance to the wall (cm)

    run 1 through symmetry ports.

    oooGoILEoF

    01 500

    1200

    300

    0

    300

    0

    600

    300

    01500

    1200

    900

    600

    300

    0'1500

    1200

  • PULVERIZED COAL UTII,ITY BOILER 28',1

    Port 7.6

    Port 5.6

    Port 4.6

    +otr_ffiffi

    Port 3,6

    Port 0.6

    40 80 120 160 200 240 280Distance to the wall (cm)

    Pod 7.5

    Pon 5.5

    Pon 4-5

    Pod 3.5

    Port 2.5

    Port 1.5

    1 500

    't200

    900

    600

    300

    01500

    1200

    900

    600

    300

    0

    1500

    1200

    900

    600

    300

    01500

    1200

    900

    600

    300

    01 500

    1200

    oEsoiE.9EooooooE

    ooo

    600 ^o300 o^;UGrsoo bcE1200 bF900

    Ê600 &o300 El

    o61500 l.tzoo 9

    ceoo €

    Gooo !o3oo eosO. rsoo j2I rzoo

    25 I20.,

    15.10

    0.

    20

    10

    5

    025

    20

    15

    10

    5

    025

    20

    '15

    10

    5

    0

    20

    15

    10

    5

    0

    't500 2512oo 20900 15600 103oo 5o0'1500 251200 20900 15600 10300 5O01500 25lzoo 20900 15600 10300 5o0tsoo 25taoo 2o900 15600 10300 50015oo 2512oo 20900 15600 10300 500

    Pon 0.5

    tr-t_-€_flffit

    1500

    't200

    600

    300

    0o 40 80 120 160 200 240 za'

    Dislance to the wall (cm)

    oo2 rco oco, trNo, aT

    FIGURE4 Profiles of local mean gas speCies concentration and gas temperatufe for run 1through several ports at the side wa11.

    1.6

  • 288 M. COSTA er al.

    Port 3.4

    port 1,4

    Fq-ffi

    ooo(!{)CLEoFco-eoE:otEc.9l!coocoooz

    soE:oEog.96coocooNooltc

    a!oo&o

    25

    20

    tc'10

    5

    0

    25

    20

    15

    10

    5

    o

    1500

    1200

    900

    600

    300

    0

    1500

    1200

    900

    600

    300

    00 40 80 120 160 2oo 240 280Distance to the wall (cm)

    aO, rCO OCO, trNO, oT

    FIGURE 5 Profiles of local mean gas species concentration an

  • :. ! PULVERIZED COAL UTILITY BOILER 289

    concentrations have a similar behavior with increasing values along the boilerheight from ports 1.6 to 4.6. The highest temperatures are observed at ports 3.6and4.6(around 1470"C),above which the gases are rapidly cooled and dilutedby the introduction of air at level 5. At ports 3.6 and 4.6,the CO concentrationare relatively high (maximum around 1.6%) which is the result of the low otconcentrations observed at these locations.

    Ports 0.5 to 4.5 (Fig. 4), close to the back wall, reveal a gradual temperatureincrease and a reduction of the O, concentrations along the boiler height. Thepersistent relatively high O, concentrations observed at all of these portssuggests that the boundary airjets reach the back wall. It is also clear that theinfluence of the boundary air jets is gradually limited to smaller regions alongthe boiler height. In particular, at ports 3.5 and 4.5 the O, concentrationssignificantly decrease towards the inner positions probably due to the upwarddeflection of the flow close to the back wall which occurs between the secondand third row of burners, as mentioned earlier.

    The gas temperatures close to the back wall observed in Figure 4 are higherin the upper burner levels due to the occurrence of combustion close to theback wall. It is interesting to note that gas temperatures do not decreasesignificantly from levels 4 to 7 close to the back wall (ports 4.5 to 7.5) but closeto the front wall (ports 4.6 to 7 .6), the introduction of air in row 5 significantlydecreases the gas temperatufes from about 1450'C at port 4.6 down to about1200"C at port 5.6.

    The measured NO" concentrations showed in Figure 4 follow, in general,the measured co2 concentrations throughout the boiler. close to the backwall, from port 2.5 to 4.5, the NO, concentrations increase due to charnitrogen conversion. Volatile combustion occurring closer to the burnerscould not be characterizedtnthe present work. Nevertheless NO" productionat the flame edges can be observed at ports I.6 and 2'6'

    The profiles showed in Figure 5 for ports 1.4 and 3.4 transverse the boiler inthe direction of the burners. At port 1.4 all profiles are relatively uniformthroughout the region whefe measurements were made. The relatively low COtNO" concentrations and temperature and high O, concentrations result fromboth combustion being in its initial stages and the presence of the boundary airinjected at level 0. At port 3.4, the relatively high o, concentrations and lowtemperatures close to the wall reveal, once again, the presence of boundary air.The decrease in o, concentration and increase in co, and NO" concentrationsand temperature in the direction of the burners are due to the combined effectsof combustion and mixing with the upward flow of hot reacted gases.

    Figure 6 show the proflles of local mean gas species concentrations for run1 for ports 8 to 10. The profiles reveal that the O, concentration increases

  • 290 M. COSTA et al.

    6OoOPortgtrtrtrE!

    a

    gOOOPortg

    gtrotr

    40 80 120 160 200 240 2AODistance to the wall (cm)

    O oz I co O co2 trNo,FIGURE 6 Profiles oflocal mean gas species concentration for run 1 through ports g to 10.

    along the boiler height. At port 8, close to the boiler nose, the o, concentrationsare very low with still significant co concentrations. At ports 9 and 10 themeasured gas species composition results from mixing with gases from the frontwall region of the boiler, where overfire air is injected at level 5. At these ports themeasured No, concentrations are reasonably uniform with averaged values closeto the ones registered onJine by EDP at the economizer outlet (see Tab. III).

    lncident Wall Heat Fluxes

    The measured local incident wall heat fluxes for runs 1 and 2 are listed inFigure 7. In both cases the strong three-dimensional nature of the thermalradiation is clearly observed. In addition, the radiation heat transfer data forboth runs reflect good symmetry with respect to opposite inspection ports atthe same level. For run 1 the trends are somewhat similar to those of thetemperature profiles but with the peaks in the burner levels being moreaccentuated due to the fourth power temperature dependency of radiation.

    15

    EctC!oEEbttc.96coocoooz

    10

    5

    020

    15

    't0

    5

    020

    15

    10.

    5.

    oEeotttoc.9E=oocooootttGooo

    ootro

    O Port 10tr

    ooa

  • . (120) (102) .,

    .t t,a(160) (1471 ,123 115 :1b(231) 084,.

    1 208 158 ;,(268) (173) .r

    1 287 176 ;2(317) (175) ^1 359 21e ;2(27e) 078) ,.s20 209 "

    lt',ii., (pt[ ',j,n r '"- '"" 02.

    Right-SideWall

    .. (107) o16) .i'o 8e 100 ;^ (152) (171) ,,Y_ 1t8 113 :'^., (180) (239)" 14a 209,. (160) (260) n" 193 264 -,45 4.t^ (168) (312) ,:_ 215 s57 ^t5 J.- (1741 (270) ^\, 218 313 "2.5 2.8^ (138) (2341 ,-y- 161 268 :'"1," (gs) (117)l-i.';.5 r35 15e a;)

    Front Wall

    ^ (173) (171) ^; 715 Burners t26 6';o()()()OOQQOOOOOOOOoooo

    BackWall

    o (226).^ 205

    o (235) (230) o^^ 244 245 ^,J,J J.C

    _- (162) (160) ^ii tst 184 i"4

    PULVERIZED COAL UTILITY BOILER

    lrDeprh I wiatn I11.4m I 15.0m I

    -

    KEV:,,:{nuri.2J',....: Riiil,

    FIGURE 7 Measured incident wall heat fluxes (kW/m2) for runs 1 and 2'

    ports located at level 4 present lower heat fluxes as compared with those atlevel 3, despite the similar measured temperatures in both levels. This can beattributed to the higher radiative heat losses to the upper furnace region atlevel 4-last row ofburners in service in run 1'

    overall, the incident wall radiation flux data for run 2 show the same trendas run 1 but with a reduced magnitude except in levels 5 to 7. It is interesting tonote that the higher values of the incident wall radiation fluxes were measured

    in leves 3 for both runs. The differences in the incident wall radiation fluxes forthe two run are minimal at level 7. This is because the gas temperature profiles

    level out very rapidly due to mixing and the smoothing effect of thermairadiation resulting in very small differences between the gas temperatures ofthe runs at level 7.

    CONCLUDING REMARKS

    Measurements have been obtained in an 300 MW", front-wall-fired, pulverized-coal, utility boiler. The results include local mean gas species concentrations of

  • M. COSTA er al

    02, co' co2, NO* and gas temperatures, measured at severar ports in the boilerincluding those in the burner region, and incident wall heat fluxes taken aroundthe boiler periphery at 39 ports. The incident wall heat fluxes have been collectedfor two boiler conditions. The main conclusions drawn are the following:1. The boundary air injected below the first row ofburners leads to oxidizing

    conditions close to the back wall.2' Local gas temperatures and co concentrations in the boiler, near the

    burners, reached maxima of about r4i0"c and r.6yo, respectively.3. The measured No, concentrations follow, in general, the measured co2

    concentrations throughout the boiler. Above the boiler nose the measuredNO, concentrations are reasonably uniform with averaged values of about670 ppm.

    4. wall radiant heat fluxes present maximum values at the side wall close to theintermediate row of burners for runs 7 and2.For run 1, the heat fluxes variedbetween 359 kfim'z at level 3 to 89 kW/m2 near the boiler nose. For run 2,the highest heat flux measured was 3r7 kw lm2 at level 3 and the lowest oneswere 102 and 95 kw lmt near the boiler nose and at level 0, respectively.

    Acknowledgements

    Financial support for this work was provided by the JouLE, III programme ofthe European commission under the contract N" JoF3-cr95-0005 and isacknowledged with gratitude. The authors would like to thank the personnel ofthe Sines power plant for their valuable assistance during all the stages of theexperiments, particularly, Mr. Luis Matos. The authors wish also to thankundergraduate student Paulo Silva and technicians Manuel pratas and RuiMartins who assisted them in conducting the work presented here. Thanks arealso due to Jorge coelho for his cooperation during the preparation olthe figures.

    References

    Azevedo, J. L. T., carnall, F. and carvalho, M. G. (1996) Heat Transfor Assessment of a Boilerconvective section. In New Deo elopments in Heat Exchangers,eds. Afgan, N., carvalho, M. G.,Bar-Cohen, A., Butterworth, D. and Roetzel, W. Gordon and B..u.i'poblirh.rr.Be6r, J. M. and chigier, N. A. (1972) combustion Aerotl.ynamit:s. Applied Science publishers,

    London.Black, D. L. and McQuay, M. Q. (1996) particle Size and velocity Measurements in the RadiantSectionofanlndustrial-Scale,Coal-FiredBoiler:theEffectofCoalType.In CombustionanrlFlre, HTD-vol. 328, pp. 19--27,, eds. Mceuay, M. e. er cl., The American Society ofMechanical Engineers, New york.Bonin, M. P. and Queiroz, M. (1991) Local particle velocity, Size and concentration Measure-

    ments in an Industrial Scale Pulverised Coal Fired Boilei. Combustion antt Ftame,gS,I2l -133.

  • PULVERIZED COAL UTILITY BOILER 293

    Bonin,M.P.andQueiroz,M.(1996)AParametricE-valu^ationofParticle-PhaseDynamlcsrnanlndustrial Pulverized-Coal-Fired Boiler' Fuel' 75, 195-206'

    Boyd,R- K. and Kent,:. H. iiSSOf Three-Dimen^sional Furnace Computer Modelling Proceed-"-' i"gt ofthe 21"t International Synposium on Comhustion'265 214'

    Butler,"B. W. and webb, s. w. (tsg t) iocal Temperature. and wall Radiant Heat Flux Measure-,nents in an Industrial Scaie Coal Fired Boiler' Fuel'1\' I45l 1'464'

    s*r"i, B. W., Wilson, T. and webb, B. W. (1gg2) Measurement of Time-Resolved Local Particle"""'Ci"ra i",,'p.rutur" in u iotiS.ut" Utitiiy noiler. Proceedings oJ the 24'h lnternationalSymposium on Comhustion' 1333 1339'

    crr"JJirr., J. and Braud, v. tts;zl Measurements In Flames, eds. Be6r, J. M. and rhring, M. w.Edward Arnold Ltd., London.

    C"rtr f,A, ftl""iao, M., ialtasar, J. and Carvalho, M'.G' (1996) Combustion-1\4^easurements tnanlndustrialGlass-MeltingFo.nu".. JournalofthelnstituteofEnergy'69'80 86'

    C";;;;; C. F. M., Azevedo,'l:i. i. and Carvalho,. M' G' (1994) 3-D Numerical Model for-''^-i,."ji"ting No_ Emissions from an Industrial pulverised coal combustor. FueI, 73,1128-1134.

    Dal Secco, S., M6chitoua, N. and Simonin, o. (1995) Three Dimensional Modelling of CoalCombustion ana NO, l-ormation in u ZiO UW Coal Fired Furn ace Proceedings of the Thirdlnternational Conlrrnn* oi iotmbustion Technologies for a Clean Enuironment' Papet 3l'2'Lisbon, Portugal, JulY.

    gppliS.,S.ttn"id;, il., Schnell, U. and Hein, K R G (1995) Computerized Analysis of Low-"'"'ilO, Coui_el."O Uiitity noiers. Combustion Science antl Tethnology,l08, 383-401.

    Fiveland, w.A. and I_utti--, c.i. (1gg3) use of N^umerical Modeling in the Design of a' ' -i.r-NO_ Burner for Uiiiity noit"tr. iombustion .Science aru| Technoloay,93,53 12.

    Kakaras, E., *touros,2., e"6ii, H.l. and Papageorgio-u, N. (1995) Influence of the.Quality ofGreek Brown coat, on NO_-SO, Emissioni from Thermal Power Stations,In cowhustionTechnologlies 1o, o

    'itnri E'jt:ironment' eds Carvalho' M G ' Loockwood' F C''

    Fiveland, W. A. and Papadopoulos, C Gordon and Breach Publishers'rntug.f , n-b.,'S.hneider, R ;'R;;iq e'' Stl.tn"ll, U' and Hein' K R G' (1996) Numerical Simula-

    tion of Utility goil"r, *ii[ AJuanced combustion Technologies. Presented at rhe EiqhthIntert,Iati()nal Sytnposi,',n on Transport Phenomena in Combustion (ISTP-VIII), San Fran-cisco, u.S.A, Jul y to_zoi,'loos-.li Tiansport phenomenain Combustion,y o1.2, pp. 1118- 1 129,ed. Chan, S. H. TaYlor and Francis'

    queiior, M.,'Bonin, M. p., sftit"ri^r, J. S. and Dawson, R. W' (1993) Experimentailv Deter-mined particle No-uL, o"ntiiy'statistics in an lndustrial-Scale, Pulverized-Coa1-FiredBorler. Energy and Fuel, 1, 842'851'

    Schneli, U., Schneider, ff., Viug.f, H C, Risio, B, Lepper' J and Hein' J' R' G' (1995)' Numeri-cal Simulation of A,O"ur"fft""l-Fired Comiustion Systems with In-Furnace NO' Controlr""rr"oiogi"r. r ,orrra"iig-, uftn, Third International Conlbrenrc on Cctmbustion Technoloqies

    ;for a Clein EnutronmentlPiper 31'1, Lisbon, Portugal' July 3-6'