7/28/2019 TwinHearthFurnace

1/3October 2012 Steel Times International

Refractories

Improving roof life of the twin hearthfurnaces at Bhilai Steel Plant

BHILAI Steel Plant at Bhilai, ChhattisgarhState, India is a unit of the Steel Authority ofIndia Ltd. It is an integrated steel plant ofcapacity 5.4Mt/y of crude steel. Production ofsteel is through two routes an ingot route ofcapacity 2.5Mt of crude steel using four twinopen hearth furnaces (THF) and an oxygenconverter (BOS) continuous caster route of

annual capacity 3Mt.Steel Melting Shop-1 (SMS-1) houses thefour twin hearth furnaces (THFs) each of 2 x250t capacity. At any one time, only three of theTHF are operating as one is kept in coldreserve. SMS-1 is equipped with six chargingmachines each of 10t capacity and five over-head hot metal cranes each of capacity 125/30t.

The Twin hearth furnace consist of twohearths separated by a bridge wall with a com-mon roof. The Twin hearth furnace is workedby synchronizing operations in the two hearthssuch that each are at different stages of opera-tion as they proceed. While one is in the solidperiod the other hearth will be in liquid period.Oxygen lances through ports in the roof are

used to accelerate processing the charge. In therecent past there has been a falling trend in thelife of the roof refractories as indicated in Figs 1& 2.

The decrease in roof life led to a decrease inproductivity of the steel shop and an increase inrefractory consumption. To investigate the rootcause of this various investigations were carriedout to find out how the furnace operatingparameters affect roof life. The quality of brickused and procedure of manufacture were alsostudied and necessary modifications carried outto improve the brick quality. The present paperfocuses on the different technological interven-tion carried out to stabilize and improve the

furnace campaign.Few Twin hearth furnace operate today asonly 1.2% (18.16Mt) of global steel was pro-duced in open hearth furnaces in 2011, mainlyin CIS countries. In 2011, India produced justunder 1Mt of steel by this method. From a lit-

There has been a gradual decrease in thelining life of the roofs of the Twin Hearthfurnaces at Bhilai Steel Plant due to acombination of operating practice andquality of refractory materials. This paperfocuses on the interventions carried outto improve the roof life which haveincreased the campaign life from an aver-age of 368 heats in 2010-11 to an aver-age of 436 heats with a record life of545 heats in 2011-12.By Somnath Kumar, S K Garai, K KKeshari, A K Bhattacharya, N KGhosh, I Roy & S Roy

erature survey of past and present operations,roof life is reported to vary from plant to plantand depends upon furnace design (principallythe height of the roof), operating conditions,and furnace maintenance. Deviation from theoptimum parameters may lead to a reduction inroof life.

One of the main causes for shortened rooflife is the corrosive action from dust evolvingfrom the bath during oxygen blowing. Blowingof metal with oxygen is always accompanied byan abundant release of dust from the bath. Thisdust consisting mainly of iron oxides (mag-netite, hematite, wustite), and settles on the

working surface of the roof in large quantities.The dust, thereby enriches the basic phases inthe refractory brock (periclase and chromite)with iron oxide which lowers the melting pointby 300 to 400C, and leads to rapid fusion andwear of the roof(1).

The roof of twin hearth furnace is lined witha magnesite chromite brick. This refractory isvery sensitive to temperature fluctuations whichcause sudden changes in thermal stress duringprocessing the charge. Structural stresses alsoact on the roof. Thus forces are constantlychanging on the magnesite chromite roofdepending on the temperature of the roof andduration of melting. Short period changes inthe temperature of the roof which arise from

the regular reversal of the flue gases betweenchambers also cause changes in these thrustforces. The roof brick spalls when there is asudden and prolonged change in the value of

the thrust force (such as during repairs, charg-ing of burden, and hot patching of the fur-nace)(2).

The life of the roof also depends to a consid-erable extent on the workmanship in buildingthe structure. Too dense and non-uniform lay-ing in the bricks in the walls of the roof archescan lead to a maximum concentration of thrustforces in a section(3).

Summarizing the various factors which canaffect the life of the roof are:

Changes in roof configuration resulting fromthermal expansion as the temperature rises

causing unevenness in the roof wear so lead-ing to loss of static stability.

Thermal stability and open porosity of theroof refractories. Also a substantial lack ofphysical and chemical uniformity in therefractories through the thickness of the roofmay also lead to variable thrust forcesbecause of differential expansion of bricks.The photographs in Figs 3 & 4 illustrate thedeterioration of the roof brickwork.

Overcooling of the roof during hot repairs,hearth repairs, emergency cut off fuel andduring charging (especially when workingwith solid charges), and also when the burn-

er flame is reversed.

Duration of charging, heating and pouring ofmolten pig iron. Also, the thermal load dur-

www.steeltimesint.com

The authors are at the R & D Centre for Iron and Steel, Steel Authority of India Ltd., Ranchi - 834002

500

448

425

457

426

368

450

450

400

400

350

350

300

300

340

365

337329

406

318

408

338

358

250

Apr 11 May June July Aug Sept Oct Nov Avg06-07 07-08 08-09

Year

09-10 10-11

Rooflife

(heats)

Fig 1 Average annual life of the roof from2006 to 2011

Fig 2 Average life of the roof from Apr 2011to Nov 2011

Fig 3 Twin hearth furnace roof duringdismantling

Fig 4 Brick failure: Slag Infiltration & ThermalSpalling

7/28/2019 TwinHearthFurnace

2/3Steel Times International October

heating is completed, liquid iron (around 200-220t) from the blast furnace is poured into thecharged bath and oxygen blowing is com-menced to refine the metal. The direction ofthe gases in the furnace are reversed by gatevalves at this stage, and the second bath istapped. The cycle is then repeated.

Measures for operationalimprovement

Reversal of gas flowA twin hearth steelmaking furnace is the first

industrial structure representing a proper steel-making unit without regenerators, designed foroperation with intensive oxygen blowing in thebath. The fundamental principle of its opera-tion is the utilization of the physical and chem-ical heat of gases formed during blowing theadjacent hearth to heat the solid charge materi-als in the hearths twin. In normal practice thelatent and chemical heat of the flue gas is usedto preheat the cold charge in the adjacenthearth and then sucked out through the oppo-site down-take.

Improper reversal practice will lead to cool-ing of the hearth roof above where the coldcharge is located. During the trial it was

ensured that the gases generated during blow-ing should be reversed to the adjacent hearthfor proper heating of the scrap and fluxes.

Overlapping of oxygen blowOverlapping in the twin hearth furnace meansthat both the hearths of the furnace containmetal in liquid form. While one hearth contains

metal in almost a refined condition, the otherhearth contains hot metal where the refiningprocess has just started. During overlapping ofhearths the overall gas load on the furnace roofincreases and corrosive action of iron oxidedust in the gas affects the roof bricks. Duringthe trial the period of overlap of the oxygenblow was restricted to a maximum of 30 min-

utes (Fig 6).

Furnace draftThe primary requirement in improving twinhearth furnace operation is to substantially cutdown leakage of air into the furnace chamber.This will greatly increase efficiency in post com-bustion of carbon monoxide to further heat thecharge. This will also help in extending the rooflife and increasing the yield of the furnace(4).The furnace damper is used for reversing theflue gas and the draft is controlled by an ID fan.Under regular plant operation, the furnacedraft was slightly negative in the range of (-)32to 34mmWC (millimetre of water column) andin some cases as high as (-)38mmWC. To avoidingress of cold air, which adversely affects theroof life by cooling down the roof, the furnacedraft was reduced to (-)28-29mmWC. A morenegative furnace draft will also have implicationon the movement of the furnace gases. Theincreased wear of the roof brickwork can beattributed to an increase in the rate of move-ment of these gases, which contain a largeamount of melting dust and slag.

Solid period/charging periodOvercooling of the roof occurs during thecharging period, especially during charging ofsolid charge ie scrap, limestone and iron ore.This causes sharp fluctuations in the roof tem-

perature over a prolonged period leading to theredistribution of thrust forces in the deeper lay-ers of the roof brick. As a result, cracks andspalling of the roof bricks occurs from areas atcontact between the softened and the hard lay-ers of brick. An increase in the time taken tocharge will lead to overcooling of the roofbricks, and thus affect the performance of theroof. To avoid overcooling of the roof duringcharging the solids, emphasis was given onkeeping the time taken to less than 2.5 hours(Fig 7).

Measures to improve brick quality

Input raw material & brick making

Around 420t of Mag-Chrome bricks arerequired for one set of roof lining of the twinhearth furnace at Bhilai Steel Plant. The bricksare made at the SAIL Refractory Unit (SRU) atBhilai, India from Dead Burnt Magnesia(DBM) and Chromite ore in various propor-tions and bonded with molasses and dextrin asadditives. The other materials used are salvagedMag-Chrome and Chrome Mag brick-bats(commonly known as Blast) generated from thesteel plant in varying quantities.

Samples of all input raw materials ie deadburnt magnesia, chromite, and blast materialwere tested with respect to chemical analysis.The analysis of blast material indicated SiO2,

11.57% to 11.73% and Fe2O3, 6.96% (Table 1).Higher SiO2 and Fe2O3 in the matrix con-tribute to formation of low melting compoundsresulting in faster erosion of the bricks.

Phase analysis of normal bricksThe phase analysis study of Mag-Chrome nor-mal roof bricks was carried out using XRD and

ing the period of charging, heating and pour-ing of pig iron.

Melting and refining time and the intensity ofthe oxygen blow during the refining period.

THF operationThe Twin-bath furnace is a modified version ofthe traditional open hearth (Siemens Martin)furnace. The Twin hearth furnace is a doublebath furnace without regenerators. It isdesigned to operate with intensive oxygenblowing into the bath through multiple roof

lances (Fig 5). The fundamental principle of itsoperation is the utilization of the physical andchemical heat from the gases evolved from thebath during blowing to heat the solid charge inthe adjoining hearth. To achieve this there is agap between the two hearths and the commonroof for the transfer of combustion productsfrom one hearth to the other.

The first operation after tapping is fettlingthe refractories which is carried out using sin-tered dolomite or magnesite. When fettling iscompleted the hearth is charged first with partof the scrap load, then with limestone and thenagain with the remaining scrap. Thus, limestoneis sandwiched between the scrap. Around 40-

50t of scrap in total and 10-12t of limestone arecharged. After charging this solid material, thishearth is preheated by directing the combustiongases from the adjacent hearth which is in itsoxygen blowing refining stage, across the solidcharge. These gases impart both their heat con-tent and chemical heat, by ensuring post com-bustion of the CO present, to the solids. When

Refractories

www.steeltimesint.com

Lances

Slag Slag

srs srs

Flue

Flue burners

To gcp

Waste gas & Fce draftmeasuring instruments

Chimney damper

Chimney burner

D

o

wn

D

o

w

n

Lances

Burners

Roof

Fig 5 Twin HearthFurnace at BhilaiSteel Plant

250 400

350

300

250

200

150

100

50

0

237355

30 20 31

113

42 44

200

150

100

50

0

Overlapping duration in minutes Solid period duration in hours

No

ofheats

No

ofheats

7/28/2019 TwinHearthFurnace

3/3October 2012 Steel Times International

the results showed that the presence of thedesirable phases ie Spinel (MgCr2O4) was verylow at 1.0 to 3.1% and Forsterite (2MgO.SiO2)was low at around 2.0-3.7%, whereas the unde-sirable Donahite (Fe.Mg)(CrFe2O4) phase waspresent at 2.3 to 10.2% as indicated in Table 2.The silica in the bricks if not combined to forma high melting point phase (eg Forsterite 2MgO.SiO2) will instead form low meltingcompounds with MgO such as Donahite.Similarly, the iron oxide will also form an unde-sirable phase.

Formulation of modified brickcompositionThe study of the mag-chrome brick used in thetwin hearth showed that the bricks are madeusing Grade-III and Grade IV Dead BurntMagnesia (DBM) which has higher amount of

SiO2 (6.5-7.5%) and thus the resulting brickcannot meet specification. Further, the higherSiO2 in the brick matrix will contribute to theformation of a low melting point compoundcausing further erosion of the bricks(5). Mag-Chrome blast material, which is a mixture ofcrushed, used Mag-Chrome bricks from THF,BSP and of rejected bricks from the SAILRefractory Unit (SRU), Bhilai, has SiO2 in therange of 11-11.5%. This was used in manufac-turing of 20-30% of the roof bricks.

Based on the study of brick making practicesat SRU, and the analysis of both fresh and usedbricks and input raw materials, it was proposedthat the following procedure be adopted for themanufacture of Mag-Chrome roof bricks:

Use DBM of grade III with a SiO2 content of6.07-6.55% to avoid formation of low melt-ing point phases which may arise with gradeIV and grade V type DBM which have SiO2contents of 7.5 to 8.5%.

Continue using grade I type chromite fines(Cr2O3-52-54%, SiO2-5%).

Avoid using blast material in the brick composition, which may contribute to higher

SiO2 (11.57-11.73%) and Fe2O3 (6.96%)and results in the formation of unwantedDonahite phase (6.6-10.2%).

Replacement of 10% DBM micro fines withpurer magnesia (SWM); and

Bricks to be fired with extended soakingtimes of 5 hours 20 minutes minimum at1600C.

ExperimentalTrials were conducted after modification ofoperational factors (as discussed above) usingthe modified brick composition A few sets ofbricks were made to the following composition.

Result and discussionThe new formulation helped in decreasing the

porosity as well as improved the spalling resist-ance of the bricks (Table 4):Improvement in operational practices in

combination with improvement in brick qualityhelped in increasing the roof life form an aver-age of 368 heats to 436 heats with a maximumlining life of 545 heats during the FY year 2010-11. This resulted in an increased furnace avail-ability to 73.0%. Refractory consumption wasalso decreased by 15% ie 4.3kg/t from the pre-vious level of 5.09kg/t. All these factors eventu-ally led to increase in productivity of the fur-naces from 101.22 t/h to 107.54 t/h.

The improved practice led to large monetarybenefits because of the increase in productivityand decrease in refractory consumption.

ConclusionsA decrease in the roof life of THF may be

attributed to deviations from standard oper-ating practices as well to a deterioration inbrick quality.

Frequent and prolonged deviation from theoptimum level of overlapping in which solidcharge remains in both furnaces, willadversely affect the roof life.

Refractories

www.steeltimesint.com

Raw material Composition (%)

Table 3 Composition of Bricks For Trial

DBM (Grade-III) 40-45

Chinese DBM 16-20

Purer Magnesia (SWM) 8-12 (as micro fines)

DBM (Grade-III) 8-12 (as micro fines)

Chromite 18-22

Fired at 1610C and soaked for 5 hours

Properties Earlier Modified

Table 4 Properties of Modified bricks v/s Normal Bricks

Bulk Density (gm/cc) 2.82-2.86 2.82-2.95

Apparent Porosity (%) 20.7-22.6 18.0-22.2

Ref. Under Load (oC) 1590 1600-1640

CCS (kg/cm2) 450 500 455 830

PLC (%) 1600 X 3 hrs -0.03 to +0.14 +0.06 to +0.26

Spalling (1000oC), Air - + 35 cycles, hairline crack

Samples Periclase Spinel Forsterite Chronium Donahite(MgO) (MgO.Cr2O3) (2MgO,SiO2) Oxide (Cr2O3) (Fe.Mg) (CrFe2O4)

Fresh Bricks

(Hot Face) 91.9 2.3 2.0 NIL 3.8

Fresh Bricks

(Cold Face) 83.7 3.1 3.7 0.2 9.3

Blast Material 81.1 - 7.0 - 11.8

Table 2. Phase Analysis of Mag-Chrome Bricks and Input Materials

Parameters % MgO CaO SiO2 Fe2O3 Cr2O3Table 1 Chemicalanalysis of inputmaterials and bricksamples (wt%)

DBM 91.15 1.51 6.55 0.57 -

Chromite 10.46 - 5.54 15.69 51.47

Salvaged Brick Bats 66.46 1.32 11.57 6.96 9.90

Brick Sample-1 66.52 1.68 8.20 6.86 11.65

Brick Sample-2 68.54 1.4 6.8 6.62 11.14

Furnace draft is an important parameterwhich also affects the roof performance.Regular analysis of waste gas will help inmaintaining an optimum draft.

Brick quality especially porosity and thermalstability are important refractory characteris-tic for obtaining good furnace roof life.

Both the silica (SiO2) and iron oxide (Fe2O3)content in the refractory raw materials mustbe controlled to promote the formation ofthe desirable phases, spinel and forsterite inthe fired bricks.

Use of salvaged magchrome brick is detri-mental to the quality of the brick due to con-tamination with free iron, SiO2 etc, but;

To promote the formation of desirable phas-es, the use of around 10% of salvaged mag-chrome bricks is preferable. In such casespurity level of other raw material viz deadburnt magnesia (DBM) and chromite mustbe increased.

References1 M B Zilbershtein, M G Kostyanoi et al Protectionof the Main Roof of Open- Hearth Furnaces byMeans of Air (Under Conditions of Intensification ofMelting by Compressed Air), Translated fromMetallurg, No 1, pp 20-22, January,1969 2 G T TuK,A A Mukhlynin et al, Increasing the Life of Roofs ofOpenHearth Furnaces, Translated from Metallurg,No 3, pp 16-21, March,19693 I P Gerasimenko, peration of Open HearthFurnace Roofs, Translated from Metallurg, No 3, pp20-23, March, 19744 K M Trubetskov, L M Efimov, and M M Privalov,

The Present Situation and Future Prospects forTwin-Bath-Furnace Steel Production Translated fromMetallurg, No 11, pp 32-34, November, 1974 5 0 JMosser, G Brchebner, K Dosinger, RHI Refractories