Firing of Limestone in JPN Pilot Plant

Jalal Abdullah Azizl, Norashid Aziz2

, Chee Hoong Phang2

ISchool of Mechanical Engineering, 2School of Chemical Engineering,Engineering Campus, Universiti Sains Malaysia, Seri Ampangan, 14300 Nibong Tebal, SPS, Penang.

IntroductionLimestone is a natural occurring

mineral throughout the world. It is widelyused in the cement manufacturing,construction and building, metal refiningprocesses, quicklime industries, agriculturalindustries, flue gases treatments, etc. Thechemical composition of the limestone variesfrom one region to the other region, but itmainly consists of calcium carbonate. In thePeninsular Malaysia, 7 billion tonnes reserveof limestone has been identified by theGeological Survey Department, Malaysia. Itis found that 55% of the limestone inPeninsular Malaysia contains 95% calciumcarbonate (Sulaiman, 1996). It is feasible to beused as raw materials to produce quicklime.

The dissociation of limestone is anenergy intensive process. During dissociationprocess, the carbon dioxide is expelled fromthe limestone to produce quicklime andcarbon dioxide at the theoretical dissociationtemperature of 900°C.

CaC03 (5) + Heat -7 CaO (5) + CO2 (g) -+

Few factors affecting the rate ofdissociation have been outlined in theliterature. These factors are source of thelimestone (Le. structure, composition, anddensity), particle size distribution, shape ofparticles, firing temperature, firing time, rateof heating, partial pressure of gas carbondioxide in the kiln, quality and type of fuelused for firing of limestone (Boynton, 1980;Wingate, 1985; Oates, 1998). Recently,Moropoulou et al. (2000) studied thereactivity of the quicklime produced .fromvarious source of limestone. He reported thatthe quicklime produced from inhomogeneousgrain size limestone with less compactstructure was more reactive than thehomogeneous fine grain size limestone withmore compact and harder structure. He alsofound that quicklime produced at 900°C

showed the higher reactivity and surface areacompared to the quicklime produced at otherhigher temperatures.

However, no studies have beenreported to the temperature profile of thepacked bed in the kiln, which directly affectsthe rate of dissociation and the quality of theproduced quicklime. Therefore, in this work,an attempt was made to design and fabricateJPN Pilot Plant and to study the variousparameters for firing of limestone.

Material and methodsThe high calcium limestone was

selected from Kinta Valley quarry in Perak.The large limestone particles were crushedinto small pieces. Each particle size was 33mm of sphericity 0.64.

The limestone was fired in the pilotplant in Figure 1. This pilot plant consisted offiring kiln fitted with liquefied petroleum gas(LPG) burning system, spiral jacketed bafflecooler, and air intake system. The combustionchamber in Figure 2 was designed speciallyaccording to the flame propagation in order toprovide uniform heat distribution to the wholepacked bed. One thermocouple was fitted inthe combustion chamber and one in the outletof the firing kiln to measure flametemperature and the outlet flue gasestemperature respectively. The firing chamberof 25 em diameter, fitted with 16thermocouples, located across and along thechamber at equal distances of 5 em to recordthe temperature in the chamber.

One to six layers of limestone,weighed 1.49 kg to 8.94 kg, were packed inthe firing chamber. Each layer was made up of58 limestone particles, as shown in Figure 3.

Air/fuel ratio was 32.4/1 in order togenerate 20 kW heat for the packed bed. Thetemperatures were recorded each 5 minutesover various firing time of I, 1.5, 2, 2.5, and 3hours. The conversion of limestone toquicklime and carbon dioxide was calculated.

INCLINED·VERTICALMANOMETER

VALVEORIFICE ue. !l"PLATE

r--,:==:W=====:;I:::;=. 10 FAN MOTOR

Jgj.to·o..,,- .\110000 II

II~~~.~ FIRING~~~~.~-"----+-CHAMBER

I~I COMBUSTION-".--\- CHAMBER

WATERIN

.vALVE

PRESSUREGAUGE

~

COMPUTERPROCESS I!

MONIT.ORING: "

FIRING,KILN S.PlRAL JACKETEDBAFFLE COOLER

FIGURE 1 Schematic diagram of the lPN Pilot Plant for firing of limestone to produce quicklime.

FIGURE 3 Arrangement of the limestone particlesfor each layer.

Results and discussionFigure 4 shows the average

temperatures for the flame, packed bed andoutlet flue gases versus firing time over aperiod of 3 hours. The temperature variationof flame showed to be higher than the packedbed while the outlet flue gases temperatureshowed to be the lowest as expected. Theeffective temperature of the flame to achieve900°C for the packed bed can be deducedfrom the graph to be I 130°C after the elapseof almost I hour of time from the start up ofthe firing. In the following firing time of 2hours, the flame temperature increased to1280°C. A conversion of 87% for thelimestone from the packed bed curves can bealso deduced to be at temperature of 1130°Cafter the elapse of 3 hours of firing time. Thecorresponding temperatures of the outlet fluegases from the kiln deduced from the graphwere varied from 450°C to 690°C over thisperiod of 1-3 hours. This indicatedappreciable sensible heat was leaving the kiln.That could be used as the waste heat for otherpurposes.

Figure 5 shows the averagetemperatures for bottom layer, middle layer,and top layer of the packed bed versus firingtime for firing 5 layers of limestone. Theaverage temperatures of each layer decreasedfrom the bottom layer to the top layer in thepacked bed. The average temperaturedifference between the bottom layer and thetop layer of the packed bed was almost 400°Cover 0.5-3 hours of firing time. At the half anhour of firing time, the average temperatures

ovro

AIR FLOW

LL__--,O-

LPGfLPW

TO SPIRALJACKETED

BAffLE COOLER

INSULATOR(MEDIUM HIGH)NS,lJL~JION

CA"S'TA'BLE CEMENT)

THERMOCO.UPLE +----"-'o'Jl!FIRING

CHAMBER

GRATE:-lt----oI~~~fI

CQMB~STloN -t--"-CHAMBER

BURNER ~q--_~_I

SCALE 1: 1 mm

17 PARTICLES(0.43 kg)

16 PARTICLES(0.42 kg)

FIGURE 2 Schematic diagram of the firing kiln.

3.02.5

-.-.- .. -----...._. --- _....._,

I

1.0 1.5 2.0 2.5 3.0Firing Time (hr)

1.0 1.5 2.0Firing Time (hr)

0.5

. _. _.- ... --- _..__._.- ._.._-- ---I

o .f-----,---,--.----.----,-----1

0.0 0.5

o -.F---.--------.-~-~-~---<

0.0

'"' 1000Ue.....~:=~ 8000)

c.S0)

Eo-<~ 600-r:0);>

<

FIGURE 4 The average temperatures for theflame, packed bed and outlet flue gases versusfiring time for firing 5 layers of limestone.

400-12.5 -7.5 -2.5 2.5 7.5 12.5

Distallee.~~~ ~_~_al ..~ent~ (em) ..JFIGURE 6 The average temperature profiles inthe packed bed versus radial position across thefiring chamber over 3 hours of firing time forfiring 5 layers of limestone.

FIGURE 5 The average temperatures forbottom layer, middle layer, and top layer of thepacked bed versus firing time for firing 5 layersof limestone.

: '"' 1400 .i~: --- 1200 -10)

I~ 1000; I..

i g, 800

I ~" 600- 400-i r:

I ""!~ 200

IIIL . ~ __ . _

of the bottom layer, middle layer, and toplayer of the packed bed increased rapidly fromambient temperature to 930°C, 790°C, and530°C respectively. It showed that only theaverage temperature of the bottom layer of thepacked bed was higher than the theoreticallimestone dissociation temperature of 900°C.Therefore, the significant dissociation oflimestone was only in the first or first twolayers of limestone. The higher layers of thelimestone were not dissociated or thedissociation might have taken place on thesurface of the limestone because of thetemperature below 900°C. After 1.1 hours offiring time, the average temperature of themiddle layer of packed bed had approached to900°C with the conversion of 24%, but it hadincreased rapidly to 84% in the second hour offiring time as shown in Figure 7. When thepacked bed was fired for 3 hours, the averagetemperatures of the bottom layer, middle layerand top layer of the packed bed reached to1250°, 1200°C and 880°C respectively.However, average temperature of the top layerof the packed bed still was under 900°C. Itcaused that the conversion of the top layer tobe only 50%, compared to the other 4 layersof limestone with the conversion of more than90%. It showed that the firing of limestone atthe temperature below 900°C was not feasibleto produce quicklime for commercialpurposes.

Figure 6 shows the averagetemperature profiles in packed bed versusradial position across the firing chamber over3 hours of firing time for firing 5 layers oflimestone. The curves were concavedownwards. This indicated that the conversionof the limestone to the quicklime and carbondioxide was endothermic reaction. Theconcavity of the curves was reduced from halfan hour to 3 hours of firing time. The curveafter 3 hours of firing become almost straightline, indicating the steady state condition forthe endothermic reaction with the uniformaverage temperature. The average temperatureacross the radial of the whole packed bed over3 hours of firing time was 910°C with theconversion of 87%.

r------------.-.-----.----.- ---.------.-._-- .. -. --- ----..-.-.--.. -- -------------- --._--,

o20

60

40

c.~

'"..ell..CQ

U

I1 1.5 2 2.5 3 I

Firing Time (hr) I- ---.--.--.-------.- .-._---.-- ...-.- .'-- -.---..._._.- - --.- -.- -- - _.- ..-. ----...------.. _. ---'---l~ First Layer 13_Second Laye!_ ._ .fi.lII.hi~sL~y~! ~ F~\l!".!b_~yer_. 1S:I FLft_b_ ~Y.~_

. -- - ._----- - ._--_...__.._--. ---- - _._-----_._-----_ ..._--FIGURE 7 Conversion of limestone for each layer when 5 layers of limestone were fired at various firingtime.

Moropoulou, A., Bakolas, A., andAggelakopoulou, E. (2001). The effects oflimestone characteristics and calcinationtemperature to the reactivity of the quicklime.Cement and Concrete Research 31: 633-639.

ConclusionsMaximum conversion of 95% was obtainedby 20 kW of heat over 3 hours in JPN PilotPlant for firing of limestone particles of 4layers, weighed 5.96 kg, average particle size33 mm, and sphericity 0.64.

AcknowledgementsThe financial supports received fromUniversiti Sains Malaysia through IRPAshort-term grant (No.6035025) and Ministryof Science, Technology and Innovation(MOSTI) through National ScienceFellowship awarded to Phang Chee Hoong,are gratefully acknowledged.

Figure 8 shows that the conversion oflimestone versus number of layers oflimestone particles after 3 hours of firing time.The maximum conversion of 95% wasdeduced to be at 4 layers of limestoneparticles. This could be explained by that 20kW of dissipated heat by the burner to thevarious numbers of layers with different heattransfer controlling mode. Meaning that theone layer of limestone was heated by theradiation heat transfer mode more thanconvection and conduction. This was mainlydue to the high voidage of 61%, whichallowed the hot gases to pass through in a veryshort time, causing insufficient time forconvection. As the number of layer increased,the corresponding number and weight of theparticles increased, causing the packed bed tobe denser and denser to a voidage of 44% at 4layers of limestone particles. Therefore, theconvection and conduction heat transfermodes were playing a great role for heating upthe packed bed faster to above 900°C, whilethe temperature rise of the one layer had goneup very slowly to 900°C, confirming theabove discussion.

ReferencesBoynton, R.S. (1980).Technology of Lime andYork: John Wiley & Sons.

Chemistry andLimestone, New

\-\00---------·---------- ----.. -- ---­

I? 95

Ii 90'ell 85

15 80

I 75. 01234567

L~umbe~_~!er~_~f~i_mest~~e.FIGURE 8 The conversion of limestone versusnumber of layers of limestone particles after 3hours firing time.

Oates, J.A.H. (1998). Lime and limestone:chemistry and technology, production anduses, Weinheim: Wiley-VCH.

Sulaiman, Z.B. and Ngah, D.S. (1996).Limestone resources of Peninsular Malaysia.International Symposium on Limestone, 13-15November 1996, Subang Jaya, Malaysia: 77­86.

Wingate, M. (1985), Small-scale Lime­burning: A Practical Introduction, London:Intermediate Technology Publications.