research article ...d5373. proximate analysis was carried out using astm d5142-90 to acquire...

Research ArticleA Modified Model for Kinetic Analysis of Petroleum Coke

Iman Eslami Afrooz and Dennis Ling Chuan Ching

Department of Fundamental and Applied Sciences Universiti Teknologi PETRONAS Seri Iskandar 32610 Malaysia

Correspondence should be addressed to Iman Eslami Afrooz imaneslamihotmailcom

Received 26 July 2019 Revised 22 September 2019 Accepted 30 September 2019 Published 3 November 2019

Academic Editor Evangelos Tsotsas

Copyright copy 2019 Iman Eslami Afrooz andDennis Ling ChuanChing+is is an open access article distributed under the CreativeCommons Attribution License which permits unrestricted use distribution and reproduction in any medium provided theoriginal work is properly cited

In this study a nonisothermal kinetics analysis of petcoke was performed at heating rates of 10 15 and 20degCmin using thermalgravimetric analysis (TGA)+e behaviour of petcoke at different gasification stages (dewatering volatilization char burning andburnout) was studied +e effect of heating rate on the activation energy of petcoke gasification was also investigated +eactivation energy of petcoke was estimated using different kinetic models that include volume reaction model (VRM) shrinkingcore model (SCM) random pore model (RPM) Coats and Redfern model (CRM) and normal distribution function (NDF) +eNDF model was modified in this study It was found that the experimental data were best fitted with the modified normaldistribution function (MNDF) and SCM+e results also showed that activation energy decreases as heating rate increases leadingto reduction in gasification completion time

1 Introduction

Petcoke is a heavy crude oil refining coproduct It isidentified as a black-colored carbon-rich solid Despite thefew human health or environmental risks posed by theexploitation of petcoke it has a high economic value andmany industrial applications It is mostly used as a boilingand combusting fuel in industrial power generation andcement plants Moreover it can be used as a raw material inmanufacturing processes In addition petcoke is a prom-ising substitute for steam coal in power plants because of itshigher heating value carbon content and low ash comparedto bituminous coals [1 2] However petcoke gasification is adifficult process because of its high content of fixed carbon[3 4] and low volatile matter [5] Moreover the masstransfer of petcoke is influenced by the porosity pore sizeand volume as well as diffusivity and tortuosity of the carbonsubstrate [6 7] To tackle this issue different solutions havebeen studied such as modifying particle size inclusion ofcatalysts and cogasification [2 8] In cogasification as anexample petcoke is combined with additional fuels such ascoal or biomass to improve its low reactivity [9 10] Ap-propriate catalysts can also be incorporated into the gasi-fication of petcoke to improve their low reactivity [4 11 12]

+e other alternative approach to achieve the high carbonconversion is to increase the residence time of the fuelparticles [7]

+e above discussion demonstrates the importance ofpetcoke gasification development +erefore understandingthe gasification kinetics of fuels is key to finding solutions topotential problems of gasification as well as improvingoperating conditions to develop efficient gasification pro-cesses In addition a detailed understanding of reactionkinetics is vital for the feasibility design and scaling ofgasification applications It will also provide valuable in-formation for proper design and operation of gasifiers+erefore the combustion characteristics chemical ele-ments and technical analysis of the fuels fed into the gas-ification reactor must be understood For this purposethermal gravimetric analysis (TGA) and derivative ther-mogravimetry (DTG) curves provide information on fuelreactivity properties such as ignition peak and burnouttemperatures [13 14] +e change in the mass of the samplecaused by devolatilization during thermal decomposition ismonitored by TGA as a function of temperature or time+emaximum reaction rate of the sample can then be obtainedthrough the first derivative of the TGA curve (dxdt) +iscurve is known as DTG

HindawiInternational Journal of Chemical EngineeringVolume 2019 Article ID 2034983 8 pageshttpsdoiorg10115520192034983

In 1975 Tyler and Smith [15] studied the reactivity ofpetcoke with CO2 over a temperature range of 1018ndash1178KA rate order of approximately 06 and activation energies inthe range of 203ndash237 kJmol were obtained for the petcokewith respect to carbon dioxide concentration A constantrate order and activation energy of about 45 with a burn-off range of 21ndash45 were reported Revankar et al [16]investigated the effect of particle size porosity and thicknesson the steam gasification of petcoke where they reportedincrease in the rate constant with decrease in both grain sizeand pellet size with or without a catalyst In contrast ac-tivation energy was found to be independent of particle sizefor the noncatalytic gasification but the frequency factordecreased with particle size However this trend is in reversefor catalytic gasification

Zou et al [17] investigated the kinetic characteristics ofpetcoke gasification with CO2 at 1248ndash1323K and 01MPa+ey proposed a normal distribution function model (NDF)to fit the kinetics data +e gasification rate was observed toincrease with the conversion rate (X) up to the peak ofX 03 +e activation energy and reaction order range of198 kJmol and 054ndash088 were reported respectively for theCO2 gasification of petcoke Yoon et al [18] conducted aTGA study of coal and petroleum coke for cogasification at15degCmin heating rate +e petcoke activation energies of5373 and 4603 kJmol were reported using kinetic modelsof the shrinking core model (SCM) and the integrated modelor the modified volume reaction model (MVRM) re-spectively In a similar study Nemanova et al [19] exploredthe cogasification of petcoke and biomass at a rate of 10degCmin using TGA +ey analyzed TGA data using the volumereaction model (VRM) and determined an activation energyvalue of 1215 kJmol for the petcoke steam gasificationRecently Jayaraman and Gokalp [8] studied the effect ofparticle size on steam gasification of petcoke using TGA andmass spectrometry analysis Relatively high reaction ratesand 90 conversion efficiency were observed for smaller-sized particles (30 μm) +e study proposed an operationtemperature of ge950degC for petcoke gasification under steamor blended steam ambiences with efficient fuel conversion

A summary of the studies discussed above is given inTable 1 Without doubt the review shows there is no discreteactivated state as confirmed by the wide variations of cal-culated activation energies [20] However the existingmodels have some limitations For example the assumptionsof the SCM model may not accurately match the realconditions even though the model is considered the mostappropriate simple representation for the majority ofreacting gas-solid systems [21] According to SCM thereaction occurs first at the ash layer and the outer skin of theparticle and then moves into the nonreacted core Alter-natively the reaction may occur along a diffuse front whichis a kind of intermediate behaviour between the SCM andCRMmodels In addition considering the fast reactions theheat release rate is sufficiently high to cause significanttemperature gradients within the core of the particleHowever there is a lack of kinetic models that accuratelyanalyze the reactions of more complex heat distributions ingas-solid systems +erefore this study investigates the

gasification characteristics and kinetic analysis of petcokeusing TGA and proposed modified normal distributionfunction (MNDF) at differential heating rates Based on thekinetic data collected by TGA some of the common kineticmodels were utilized to estimate the activation energy ofpetcoke +e results were then comparatively analyzed anddiscussed

2 Materials and Methods

21 Sample Preparation For this study petcoke obtainedfrom PETRONAS Melaka Refinery with a HardgroveGrindability Index (HGI) of 111 and a particle size range of5ndash20mm was used +e petcoke was crushed and milled to asize of 07mm It was then dried for 24 hours at 110degC usingan industrial oven Afterward the combustion character-istics of petcoke were investigated using PerkinElmer STA6000 +e initial weights of the petcoke samples were variedsince the TGA results were unaffected by the sample weightTo investigate the effect of heating rate on the gasificationphases TGA experiments were carried out at heating rates of10 15 and 20degCmin In addition an untreated petcokesample (as received) was used to investigate the influence ofsample preparation conditions on TGA curves +e de-scriptions of samples used in this study are outlined inTable 2

22 Ultimate and Proximate Analysis Ultimate analysis wascarried out to determine the percentage of carbon hydrogennitrogen and sulphur chloride of the petcoke using aCHNOS elemental analyzer under consideration of ASTMD5373 Proximate analysis was carried out using ASTMD5142-90 to acquire preliminary results on the moisturecontent volatile matter fixed carbon and ash content ofpetcoke samples In addition a bomb calorimeter withstandard reference of the ASTM D5865 test method wasused to determine the calorific value of the samples Driedpetcoke sample was used for both ultimate and proximateanalysis However the calorific value test was carried out onboth dried and untreated petcoke samples

Table 3 shows the results for ultimate proximate andcalorific value analyses Different heating rates of 10 15 and20degCmin were used while the TGA experiment temperaturewas kept at 800degC +e nitrogen and oxygen flow rates werekept at 100mlmin for all experiments+e change in sampleweight was recorded every second A continuous supply ofnitrogen was maintained for 40 minutes to eliminatemoisture until the reaction temperature specified for thegeneration of char was attained Afterward the effect ofheating rate on the ignition temperature and ignition time ofpetcoke was analyzed by burning each petcoke sample underdifferent heating rates in the presence of oxygen Finally thecombustion characteristics of the char were measured foreach sample

23 Kinetic Analysis Similar to other organic or chemicalmatters like biomass or coal the kinetic of petcoke de-composition can also be predicted by the following equation

2 International Journal of Chemical Engineering

dX

dt Kf(X) (1)

where dXdt K f(X) and X are the isothermal reaction ratethe reaction rate constant the reaction model equation andthe extent of reaction (or the conversion rate) respectively

+e reaction rate K is presented by the Arrhenius ex-pression as follows

K A exp minusEa

RT1113874 1113875 (2)

By substituting equation (2) into equation (1) the ratelaw can be expressed as

dX

dt Kf(X) A exp minus

Ea

RT1113874 1113875f(X) (3)

+e extent of reaction X can be obtained using thefollowing expression

X m0 minus mt

m0 minus mf

(4)

where m0 mt and mf are the initial sample mass samplemass at time t and final sample mass respectively

Some of the important rate models f(X) used to describethe kinetic behaviour of solid-state reactions were listed inref [22] Among them the VRM [23] SCM [21] randompore model (RPM) [24] MVRM [25 26] NDF [17] andCoats and Redfern (CR) model [27 28] have been used byresearchers to predict the kinetic parameters of petcokeBased on the VRM (or homogeneous model) assumptionthe particle density changes uniformly due to the distri-bution of gas inside the solid particles According to theVRM the reaction kinetic expression can be described as

dX

dt Kv(1 minus X) (5)

where (1 minus X) is the remaining fraction of volatile material inthe sample

+e SCM was first established by Yagi and Kunii [29]SCM assumes that the reaction first occurs at the outer skinof the spherical particle and it then moves into the non-reacted core which shrinks in size gradually during reactionprocess According to SCM the reaction kinetic model is asfollows

dX

dt Ks(1 minus X)

23 (6)

+e RPM takes into account the pore structure and itsevolution in the course of reaction Based on the RPM thegasification rate can be written as

dX

dt Kr(1 minus X)(1 minus ψ ln(1 minus X))

12 (7)

where ψ is the particle structure parameter +e RP modelwith the value of ψ equal to zero gives the VR model

Table 2 Samples descriptions

Samples Initial weight (mg) Preparation conditions Heating rate (degCmin) TGA temperature (degC)1 7684 Dried for 24 h 10 8002 8964 Dried for 24 h 15 8003 9963 Dried for 24 h 20 8004 7443 As received 15 800

Table 3 Ultimate proximate and calorific value analysis ofpetcoke

Test Parameter Result

Ultimate analysis (wt)

Carbon 8329Hydrogen 3576Nitrogen 168Sulphur 5528

Proximate analysis (wt)

Moisture 60Volatile matter 148

Ash 014Fixed carbon 7906

Calorific value (Jg) Dried sample 35875As received 36211

Table 1 Summary of kinetic models

Reference Particlesize (mm)

Carbon and ashcontent ()

Gasificationmediumpressure (kPa)

Heating rate(degCmin)

TGA maximumtemperature (degC)

Kineticmodel

Activation energy(kJmol)

[15] 29 09022

964 968 969075 040 043 CO226 and 118 10 745ndash905 mdash 203ndash237

[16] 0037ndash059 9903 Steamatm 20ndash400

25ndash880 695ndash880 SCM 26ndash30634lowastndash52lowast

[17] 9963 9025056 CO2100 25 975ndash1050 NDF 198

[18] 002ndash005 8716025 Airatm 15 1100 1200

1300 1400SCM

MVRM53734603

[19] 1ndash15 92314 Steamatm 10 1250 VRM 1215

lowastWith catalyst

International Journal of Chemical Engineering 3

+e volume reaction model was improved to MVRM (orintegrated model) by Kasaoka et al [25] and Yang et al [26]It was extended for all reaction orders by adding a newparameter (n) as follows

dX

dt Km(1 minus X)

n (8)

Zou et al [17] proposed the normal distribution function(NDF) as a model to predict kinetic parameters +e con-ventional distribution function is approximated by aGaussian distribution that yields a mean value and standarddeviation Based on NDF the reaction kinetic model can bewritten as

dX

dT Kd exp minus

X minus Xm( 11138572

2ω21113888 1113889 (9)

where Xm and ω are the maximal gasification rate and thewidth of the curve respectively

On the contrary the reaction rate can be presented as afunction of temperature for nonisothermal solid-statereactions

dX

dtdT

dt

dX

dT β

dX

dT (10)

where β is the heating rate and dXdT is the nonisothermalreaction rate

Considering equations (3) and (10) the reaction rate canbe represented as

dX

dt β

dX

dT A exp minus

Ea

RT1113874 1113875f(X) (11)

By taking natural logarithms of each side

lndX

dt1113888 1113889 ln β

dX

dT1113888 1113889 ln[Af(X)] minus

Ea

RT (12)

Integration of equation (11) yields

g(X) 1113946X

0

dX

f(X)

A

β1113946

T

0exp minus

Ea

RT1113874 1113875dT (13)

Equation (13) is the integral expression of the rate law Ifwe consider (EaRT) as u equation (13) becomes

g(X) AEa

βR1113946infin

X

expminus u

u2 AEa

βRp(u) (14)

+e p(u) in equation (14) known as the temperatureintegral cannot be integrated by one of the methods ofcalculus However this is not a serious limitation because itcan be approximated via empirical interpolation formulaslike an integral method based on the Coats and Redfern (CR)equation [27 28] +erefore p(u) in equation (14) can beestimated using a Taylor series expansion to yield the fol-lowing expression [22]

g lnminus ln(1 minus X)

T21113888 1113889 lnAR

βEa

1 minus2RT

Ea

1113888 11138891113890 1113891 minusEa

RT (15)

To begin we studied the weight reduction of petcoke fordifferent heating rates Afterward some of the abovementioned

models were used to fit the experimental data obtained byTGA A modified form of NDF was then proposed Finallyusing an Arrhenius plot the accuracy of the kinetic models inpetcoke activation energy approximation was comparativelyevaluated

3 Results and Discussion

31 TGA and DTG Analysis Petcoke was evaluated usingTGA analysis to understand its ignition temperature andcombustion characteristics +e weight reduction for dif-ferent heating rates of 10 15 and 20degCmin is demonstratedin Figure 1 Since there was no evidence of weight loss withinthe first 30 minutes of the TGA Figures 1(a) and 1(b) onlydepict the values that exceed 240degC and 30min respectivelyIt is discernible from the TGA that the dewatering stagestarts around 243degC (at 3213min) 463degC (at 3940min)419degC (at 3000min) and 465degC (at 3951min) and continuestill the temperature reached approximately 518degC (at5771min) 523degC (at 4208min) 525degC (at 3425min) and519degC (at 4198min) for samples 1 to 4 respectively Takinginto cognizance the dewatering stage the weight of allsamples reduced by nearly 6 However the differentialweight values can be attributed to the different appliedheating rates

It is also important to note that the time required forcomplete petcoke dewatering for treated and untreatedsamples (samples 2 and 4) is almost the same +ereforesince the effect of sample drying is negligible more energycan be saved by using untreated samples Volatilization isinitiated immediately after the dewatering stage for allsamples and is completed at about 542degC (at 5838min)548degC (at 4265min) 554degC (at 3485min) and 542degC (at4245min) for samples 1 to 4 respectively Subsequent to thevolatilization stage is the char combustion phase +ecombustion process was finally completed at differenttemperatures of 572degC (at 6116min) 595degC (at 4578min)620degC (at 3821min) and 583degC (at 4483min) for samples 1to 4 respectively +e gray area in Figure 1(b) shows ourimplementations for petcoke combustion which can be usedas a tool to predict the petcoke combustion behaviour undercertain heating rates +is specifically implies the rate ofpetcoke weight loss during the combustion reaction atheating rates between 10 and 20degCmin can be predictedfrom the gray area

To study the effect of reaction rate on different reactionstages the start and end points of time and temperature foreach phase were calculated and the results are presented inTable 4 Irrespective of the heating rates employed the timerequired for total volatilization was the same A similar trendcan be observed for the combustion stage However there isa slight disparity between the temperature values of thevolatilization and combustion stages which can be attrib-uted to the different heating rates applied+is indicates thatthe effect of heating rate on the volatilization and com-bustion stages is negligible Significant deviations are onlyobserved for the time required for the dewatering stage to becompleted +erefore the most pertinent deduction is thatthe time required for moisture removal from petcoke can be


significantly reduced by increasing the heating rate More-over comparison of the heating rates shows 15degCmin is themost effective for all stages of petcoke gasification

+e TGA and DTG curves for the thermal de-composition of petcoke at a heating rate of 15degCmin areshown in Figure 2(a) As earlier explained the TGA con-trasts the weight loss of the petcoke with temperaturechange whereas the DTG represents the time derivative ofthe weight loss for the same temperature change as TGA Itcan be inferred from Figure 2(a) that the rate of weight loss ismaximum at 558degC (peak temperature) Cursory evaluationof the TGA curve indicates the peak temperature occurred inthe char combustion phase+us the rapid weight reductioncorresponds to the combustion phase

It is also important to mention that the change in thevalues of derivative weight as observed from the DTGcurve indicates that the progressive thermal breakdown ofthe organic matter present in the petcoke occurs within therange of 480ndash620degC In addition the burnout temperatureof 620degC represents the temperature where sample oxi-dation is complete +e derivative weight curves for thethermal decomposition of petcoke at different heatingrates are shown in Figure 2(b) It is observed that themaximum weight losses for heating rates between 10 and15degCmin occur at almost the same temperature of 558degCMoreover the maximum weight loss for the treated sample(sample 1) and the untreated sample (sample 2) occurredat 558 and 556degC respectively However at the higherheating rate of 20degCmin the maximum weight loss wasrecorded at 568degC

Conversely the DTG profiles clearly show that the rate ofweight loss slightly decreases with the increase in heatingrate +e maximum conversion rates of minus 3215 minus 2987 andminus 2852minminus 1 were observed for the heating rates of 10 15and 20degCmin respectively Remarkably the DTG profile ofthe untreated sample (sample 4) shows a higher rate ofweight loss (minus 3747minminus 1) than the others It is evidentthat the reaction sensitivity of petcoke to heating rate andsample treatment is very low

32 Arrhenius Plot and Activation Energy +e reactionrates K were calculated using selected kinetic modelsthat include VRM SCM RPM NDF and CRM +eArrhenius plots for the gasification reaction of petcoke(Figure 3) were derived using the logarithmic reactionrate ln (K) minminus 1 versus the reciprocal gasification tem-perature 1000TKminus 1 It is discernible from the Arrheniusplots that the logarithmic gasification reaction rate ex-hibits a linear relationship with temperature for almostall kinetic models and reaction rates with the exceptionof 10degCmin As observed in Figure 3 the slope of ln (K)versus 1000T becomes steeper as the heating rate de-clines Given the proportionality of the activation energy(Ea) with the slope of the Arrhenius plot it can beinferred that the activation energy increases as heatingrate decreases

+e activation energies Ea for different kinetic modelswere determined using equation (16) and results are pre-sented in Table 5 for comparison

0102030405060708090

100110

240 290 340 390 440 490 540 590 640

Wei

ght (

)

Temperature (degC)

Sample 1Sample 2

Sample 3Sample 4

(a)

Sample 1Sample 2

Sample 3Sample 4

0102030405060708090

100110

30 35 40 45 50 55 60 65

Wei

ght (

)

Time (min)

(b)

Figure 1 Rate of petcoke weight loss () vs (a) temperature (degC) and (b) time (min) for different samples +e gray area shows the range inwhich our implementations for petcoke combustion were applied

Table 4 Reaction stages characterization

Sample Heating rate (degCmin)Reaction stage

Dewatering Volatilization CombustionTime (min) Temp (degC) Time (min) Temp (degC) Time (min) Temp (degC)

1 10 2558 275 067 24 278 302 15 268 60 057 25 313 473 20 425 106 06 29 336 66


slope minusEa

R (16)

where slope is the slope of ln (K) versus 1T from theArrhenius plot

Based on the data outlined in Table 5 there is a sig-nificant difference between the activation energy valuesobtained by CRM and NDF and those acquired by the otherkinetic models +is difference can be attributed to sys-tematic errors associated with the calculation of Ea using theintegral isoconversional kinetic models Considering eachkinetic model it can be noted that activation energies ob-tained from different heating rates were not the same andthey all increase with decreased heating rates +is is becausemore energy is required to initiate release of the refractory

molecules enclosed in petcoke in the form of volatiles whenusing lower heating rates +e comparative analysis of Eavalues obtained in this study with those reported in earlierstudies (Table 1) shows that more reliable Ea values can beestimated using SCM RPM and VRM while those obtainedusing NDF and CRM are overestimated +e results in-dicated that increasing the heating rate leads to reduction inactivation energy As a result less energy is needed to ac-tivate atoms or molecules to a state where they can undergo achemical reaction For example using SCM the activationenergy of petcoke is reduced by 7153 with the increase inheating rate from 10 to 20degCmin

33 Modified Normal Distribution Function (MNDF) Asstated above the NDF overestimated the values of activationenergy Nonetheless NDF has some features which make it avaluable approximation function for complex distributionsof variables +e NDF model was then modified by calcu-lating the right values of maximal gasification rate (Xm) andthe curve width (ω) of the normal distributed functionConsidering the normal distribution function (equation (9))its exponential function increases to zero while the value ofω

ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)

10degCmin15degCmin20degCmin

VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)

Figure 3 Comparison of Arrhenius plots for different heating rates of 10 15 and 20degCmin vs (a) VR SC and RP models (b) CR model

Table 5 Activation energy of petcoke

Heating rate (degCmin)Activation energy (kJmol)

VRM SCM RPM CRM NDF MNDF10 186 124 1248 3799 6357 12415 817 545 507 2581 4113 80220 53 353 313 2345 3202 625

0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash40ndash35ndash30ndash25ndash20ndash15ndash10

ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)

Figure 2 Plots of (a) TGA and DTG curves and (b) DTG curves for different heating rates


decreases Meanwhile the value of conversion (X) changesfrom 0 to 1 Consequently the graph of NDF for Xm equal tomean is symmetric with respect to the line X equal to mean(Xmean) Considering these features of NDF choosingthe right values of ω and Xm is necessary to best fit the graphof experimental data with the straight line To this end thevalues of 1 and 0736 are proposed in this study forXm andωrespectively Using these values the followingmodified NDF(MNDF) is proposed

dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)

Plots of minus ln (k) versus 1000T derived from differentheating rates and MNDF and NDF kinetic models are il-lustrated in Figure 4

It is worthwhile to mention that using NDF the loga-rithmic reaction rate results in a straight line with positiveslope+erefore the negative logarithmic function (minus ln) wasused in this study to make the slope of the Arrhenius graphnegative +is change in sign does not affect the results ofactivation energy However this finding was not reported inthe literature [17] It can be clearly seen in Figure 4 that theMNDF yields much tighter curves that establish a better fitwith experimental data than the conventional NDF modelFurthermore compared with the NDF model the MNDFprovided a better minimization of the discrepancy in cal-culated activation energies +e activation energies areconsistent with those extracted from the literature

4 Conclusions

In this study TGA was performed on petcoke to study itscombustion characteristics under different heating rates+edifferent stages of combustion namely dewatering vola-tilization char burning and burnout for petcoke were ex-amined Some common kinetic models (VRM SCM RPMCRM and NDF) were utilized to estimate the activationenergy of the combustion of petcoke +e NDF model wasmodified to best fit the experimental data +e kineticanalysis showed that the combustion results of petcokeunder the heating rate of 10degCmin show a longer transitionstage between dewatering and volatilization It was also

found that the activation energy of petcoke is significantlyaffected by increase in heating rate+e TGA results revealedthat the total time needed for gasification of petcoke wasshortened with the use of 20degCmin heating rate whichindicated that increasing the heating rate leads to reductionin activation energy+us shorter gasification finishing timeis expected while using higher heating rate +e proposedmodified kinetic model (MNDF) was found to accuratelypredict the gasification kinetics the most and achieved 50reduction in activation energy In addition the results showthat the duration for complete petcoke combustion can varyfrom 38 to 61 minutes +e proposed kinetic model providesvaluable information for proper design and operation of thegasifier reactors It also improves the understanding anddevelopment of the gasification process

Data Availability

+e TGA data used to support the findings of this study areavailable from the corresponding author upon request

Conflicts of Interest

+e authors declare that there are no conflicts of interestregarding the publication of this paper

Acknowledgments

+is research was funded by University Technology PET-RONAS (UTP) YUTP grant number 015LC0-078

References

[1] E Furimsky ldquoGasification of oil sand coke reviewrdquo FuelProcessing Technology vol 56 no 3 pp 263ndash290 1998

[2] B N Murthy A N Sawarkar N A Deshmukh T Mathewand J B Joshi ldquoPetroleum coke gasification a reviewrdquo 5eCanadian Journal of Chemical Engineering vol 92 no 3pp 441ndash468 2014

[3] X Zhan J Jia Z Zhou and F Wang ldquoInfluence of blendingmethods on the co-gasification reactivity of petroleum cokeand ligniterdquo Energy Conversion and Management vol 52no 4 pp 1810ndash1814 2011

[4] X Zhan Z Zhou and F Wang ldquoCatalytic effect of blackliquor on the gasification reactivity of petroleum cokerdquo Ap-plied Energy vol 87 no 5 pp 1710ndash1715 2010

[5] J Chen and X Lu ldquoProgress of petroleum coke combusting incirculating fluidized bed boilersmdasha review and future per-spectivesrdquo Resources Conservation and Recycling vol 49no 3 pp 203ndash216 2007

[6] E Furimsky ldquoGasification reactivities of cokes derived fromathabasca bitumenrdquo Fuel Processing Technology vol 11 no 2pp 167ndash182 1985

[7] C Higman and M van der Burgt Gasification Gulf Pro-fessional Publications Amsterdam Netherlands 2nd edition2008

[8] K Jayaraman and I Gokalp ldquoGasification characteristics ofpetcoke and coal blended petcoke using thermogravimetryand mass spectrometry analysisrdquo Applied 5ermal Engi-neering vol 80 pp 10ndash19 2015

10degCmin15degCmin20degCminNDF

MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)

Figure 4 Comparison of Arrhenius plots for MNDF and NDF


[9] J Fermoso B Arias M G Plaza et al ldquoHigh-pressure co-gasification of coal with biomass and petroleum cokerdquo FuelProcessing Technology vol 90 no 7-8 pp 926ndash932 2009

[10] J Fermoso B Arias B Moghtaderi et al ldquoEffect of co-gas-ification of biomass and petroleum coke with coal on theproduction of gasesrdquo Greenhouse Gases Science and Tech-nology vol 2 no 4 pp 304ndash313 2012

[11] Z-J Zhou Q-J Hu X Liu G-S Yu and F-C Wang ldquoEffectof iron species and calcium hydroxide on high-sulfur pe-troleum coke CO2 gasificationrdquo Energy amp Fuels vol 26 no 3pp 1489ndash1495 2012

[12] S H Lee S J Yoon H W Ra Y I Son J C Hong andJ G Lee ldquoGasification characteristics of coke and mixturewith coal in an entrained-flow gasifierrdquo Energy vol 35 no 8pp 3239ndash3244 2010

[13] C Wang F Wang Q Yang and R Liang ldquo+ermogravi-metric studies of the behavior of wheat straw with added coalduring combustionrdquo Biomass and Bioenergy vol 33 no 1pp 50ndash56 2009

[14] F Trejo M S Rana and J Ancheyta ldquo+ermogravimetricdetermination of coke from asphaltenes resins and sedimentsand coking kinetics of heavy crude asphaltenesrdquo CatalysisToday vol 150 no 3-4 pp 272ndash278 2010

[15] R J Tyler and I W Smith ldquoReactivity of petroleum coke tocarbon dioxide between 1030 and 1180 krdquo Fuel vol 54 no 2pp 99ndash104 1975

[16] V V S Revankar A N Gokarn and L K DoraiswamyldquoStudies in catalytic steam gasification of petroleum coke withspecial reference to the effect of particle sizerdquo Industrial ampEngineering Chemistry Research vol 26 no 5 pp 1018ndash10251987

[17] J H Zou Z J Zhou F C Wang et al ldquoModeling reactionkinetics of petroleum coke gasification with CO2rdquo ChemicalEngineering and Processing Process Intensification vol 46no 7 pp 630ndash636 2007

[18] S J Yoon Y-C Choi S-H Lee and J-G Lee ldquo+ermog-ravimetric study of coal and petroleum coke for co-gasifi-cationrdquo Korean Journal of Chemical Engineering vol 24no 3 pp 512ndash517 2007

[19] V Nemanova A Abedini T Liliedahl and K Engvall ldquoCo-gasification of petroleum coke and biomassrdquo Fuel vol 117pp 870ndash875 2014

[20] P D Garn ldquoKinetic parametersrdquo Journal of 5ermal Analysisand Calorimetry vol 13 no 3 pp 581ndash593 1978

[21] C N Satterfield ldquoChemical reaction engineering OctaveLevenspiel Wiley New York (1972) 578 pages $1695rdquoAIChE Journal vol 19 no 1 pp 206-207 1973

[22] J E White W J Catallo and B L Legendre ldquoBiomasspyrolysis kinetics a comparative critical review with relevantagricultural residue case studiesrdquo Journal of Analytical andApplied Pyrolysis vol 91 no 1 pp 1ndash33 2011

[23] M Ishida and C Y Wen ldquoComparison of zone-reactionmodel and unreacted-core shrinking model in solidmdashgasreactionsmdashI isothermal analysisrdquo Chemical Engineering Sci-ence vol 26 no 7 pp 1031ndash1041 1971

[24] S K Bhatia and D D Perlmutter ldquoA random pore model forfluid-solid reactions I Isothermal kinetic controlrdquo AIChEJournal vol 26 no 3 pp 379ndash386 1980

[25] S Kasaoka Y Sakata and C Tong ldquoKinetic evaluation ofreactivity in carbon dioxide-gasification of various coal charsand comparison with steam-gasificationrdquo Journal of the FuelSociety of Japan vol 62 no 5 pp 335ndash348 1983

[26] X F Yang J Zhou X Gong and Z Yu ldquoKinetic and char-acteristic study of char-H2O gasification by isothermal

thermogravimetryrdquo Coal Conversion vol 2003 no 4 pp 46ndash50 2003

[27] A W Coats and J P Redfern ldquoKinetic parameters fromthermogravimetric datardquoNature vol 201 no 4914 pp 68-691964

[28] A W Coats and J P Redfern ldquoKinetic parameters fromthermogravimetric data IIrdquo Journal of Polymer Science Part BPolymer Letters vol 3 no 11 pp 917ndash920 1965

[29] S Yagi and D Kunii ldquoStudies on combustion of carbonparticles in flames and fluidized bedsrdquo Symposium (In-ternational) on Combustion vol 5 no 1 pp 231ndash244 1955


International Journal of

AerospaceEngineeringHindawiwwwhindawicom Volume 2018

RoboticsJournal of

Hindawiwwwhindawicom Volume 2018


Active and Passive Electronic Components

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwwwhindawicom

Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of



Journal ofEngineeringVolume 2018

SensorsJournal of



RotatingMachinery


Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia

Submit your manuscripts atwwwhindawicom

In 1975 Tyler and Smith [15] studied the reactivity ofpetcoke with CO2 over a temperature range of 1018ndash1178KA rate order of approximately 06 and activation energies inthe range of 203ndash237 kJmol were obtained for the petcokewith respect to carbon dioxide concentration A constantrate order and activation energy of about 45 with a burn-off range of 21ndash45 were reported Revankar et al [16]investigated the effect of particle size porosity and thicknesson the steam gasification of petcoke where they reportedincrease in the rate constant with decrease in both grain sizeand pellet size with or without a catalyst In contrast ac-tivation energy was found to be independent of particle sizefor the noncatalytic gasification but the frequency factordecreased with particle size However this trend is in reversefor catalytic gasification

Zou et al [17] investigated the kinetic characteristics ofpetcoke gasification with CO2 at 1248ndash1323K and 01MPa+ey proposed a normal distribution function model (NDF)to fit the kinetics data +e gasification rate was observed toincrease with the conversion rate (X) up to the peak ofX 03 +e activation energy and reaction order range of198 kJmol and 054ndash088 were reported respectively for theCO2 gasification of petcoke Yoon et al [18] conducted aTGA study of coal and petroleum coke for cogasification at15degCmin heating rate +e petcoke activation energies of5373 and 4603 kJmol were reported using kinetic modelsof the shrinking core model (SCM) and the integrated modelor the modified volume reaction model (MVRM) re-spectively In a similar study Nemanova et al [19] exploredthe cogasification of petcoke and biomass at a rate of 10degCmin using TGA +ey analyzed TGA data using the volumereaction model (VRM) and determined an activation energyvalue of 1215 kJmol for the petcoke steam gasificationRecently Jayaraman and Gokalp [8] studied the effect ofparticle size on steam gasification of petcoke using TGA andmass spectrometry analysis Relatively high reaction ratesand 90 conversion efficiency were observed for smaller-sized particles (30 μm) +e study proposed an operationtemperature of ge950degC for petcoke gasification under steamor blended steam ambiences with efficient fuel conversion

A summary of the studies discussed above is given inTable 1 Without doubt the review shows there is no discreteactivated state as confirmed by the wide variations of cal-culated activation energies [20] However the existingmodels have some limitations For example the assumptionsof the SCM model may not accurately match the realconditions even though the model is considered the mostappropriate simple representation for the majority ofreacting gas-solid systems [21] According to SCM thereaction occurs first at the ash layer and the outer skin of theparticle and then moves into the nonreacted core Alter-natively the reaction may occur along a diffuse front whichis a kind of intermediate behaviour between the SCM andCRMmodels In addition considering the fast reactions theheat release rate is sufficiently high to cause significanttemperature gradients within the core of the particleHowever there is a lack of kinetic models that accuratelyanalyze the reactions of more complex heat distributions ingas-solid systems +erefore this study investigates the

gasification characteristics and kinetic analysis of petcokeusing TGA and proposed modified normal distributionfunction (MNDF) at differential heating rates Based on thekinetic data collected by TGA some of the common kineticmodels were utilized to estimate the activation energy ofpetcoke +e results were then comparatively analyzed anddiscussed

2 Materials and Methods

21 Sample Preparation For this study petcoke obtainedfrom PETRONAS Melaka Refinery with a HardgroveGrindability Index (HGI) of 111 and a particle size range of5ndash20mm was used +e petcoke was crushed and milled to asize of 07mm It was then dried for 24 hours at 110degC usingan industrial oven Afterward the combustion character-istics of petcoke were investigated using PerkinElmer STA6000 +e initial weights of the petcoke samples were variedsince the TGA results were unaffected by the sample weightTo investigate the effect of heating rate on the gasificationphases TGA experiments were carried out at heating rates of10 15 and 20degCmin In addition an untreated petcokesample (as received) was used to investigate the influence ofsample preparation conditions on TGA curves +e de-scriptions of samples used in this study are outlined inTable 2

22 Ultimate and Proximate Analysis Ultimate analysis wascarried out to determine the percentage of carbon hydrogennitrogen and sulphur chloride of the petcoke using aCHNOS elemental analyzer under consideration of ASTMD5373 Proximate analysis was carried out using ASTMD5142-90 to acquire preliminary results on the moisturecontent volatile matter fixed carbon and ash content ofpetcoke samples In addition a bomb calorimeter withstandard reference of the ASTM D5865 test method wasused to determine the calorific value of the samples Driedpetcoke sample was used for both ultimate and proximateanalysis However the calorific value test was carried out onboth dried and untreated petcoke samples

Table 3 shows the results for ultimate proximate andcalorific value analyses Different heating rates of 10 15 and20degCmin were used while the TGA experiment temperaturewas kept at 800degC +e nitrogen and oxygen flow rates werekept at 100mlmin for all experiments+e change in sampleweight was recorded every second A continuous supply ofnitrogen was maintained for 40 minutes to eliminatemoisture until the reaction temperature specified for thegeneration of char was attained Afterward the effect ofheating rate on the ignition temperature and ignition time ofpetcoke was analyzed by burning each petcoke sample underdifferent heating rates in the presence of oxygen Finally thecombustion characteristics of the char were measured foreach sample

23 Kinetic Analysis Similar to other organic or chemicalmatters like biomass or coal the kinetic of petcoke de-composition can also be predicted by the following equation


dX

dt Kf(X) (1)



K A exp minusEa

RT1113874 1113875 (2)


dX


Ea

RT1113874 1113875f(X) (3)


X m0 minus mt

m0 minus mf

(4)



dX




dX

dt Ks(1 minus X)

23 (6)


dX


12 (7)


















Kineticmodel


[15] 29 09022




[17] 9963 9025056 CO2100 25 975ndash1050 NDF 198

[18] 002ndash005 8716025 Airatm 15 1100 1200

1300 1400SCM

MVRM53734603


lowastWith catalyst



dX

dt Km(1 minus X)

n (8)


dX

dT Kd exp minus


2ω21113888 1113889 (9)



dX

dtdT

dt

dX

dT β

dX

dT (10)



dX

dt β

dX

dT A exp minus

Ea

RT1113874 1113875f(X) (11)


lndX

dt1113888 1113889 ln β

dX

dT1113888 1113889 ln[Af(X)] minus

Ea

RT (12)


g(X) 1113946X

0

dX

f(X)

A

β1113946

T

0exp minus

Ea

RT1113874 1113875dT (13)


g(X) AEa

βR1113946infin

X

expminus u

u2 AEa

βRp(u) (14)



T21113888 1113889 lnAR

βEa

1 minus2RT

Ea

1113888 11138891113890 1113891 minusEa

RT (15)














0102030405060708090

100110

240 290 340 390 440 490 540 590 640

Wei

ght (

)

Temperature (degC)

Sample 1Sample 2

Sample 3Sample 4

(a)

Sample 1Sample 2

Sample 3Sample 4

0102030405060708090

100110

30 35 40 45 50 55 60 65

Wei

ght (

)

Time (min)

(b)





1 10 2558 275 067 24 278 302 15 268 60 057 25 313 473 20 425 106 06 29 336 66


slope minusEa

R (16)





ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)


VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)





0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)


ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)




dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


dX

dt Kf(X) (1)



K A exp minusEa

RT1113874 1113875 (2)


dX


Ea

RT1113874 1113875f(X) (3)


X m0 minus mt

m0 minus mf

(4)



dX




dX

dt Ks(1 minus X)

23 (6)


dX


12 (7)


















Kineticmodel


[15] 29 09022




[17] 9963 9025056 CO2100 25 975ndash1050 NDF 198

[18] 002ndash005 8716025 Airatm 15 1100 1200

1300 1400SCM

MVRM53734603


lowastWith catalyst



dX

dt Km(1 minus X)

n (8)


dX

dT Kd exp minus


2ω21113888 1113889 (9)



dX

dtdT

dt

dX

dT β

dX

dT (10)



dX

dt β

dX

dT A exp minus

Ea

RT1113874 1113875f(X) (11)


lndX

dt1113888 1113889 ln β

dX

dT1113888 1113889 ln[Af(X)] minus

Ea

RT (12)


g(X) 1113946X

0

dX

f(X)

A

β1113946

T

0exp minus

Ea

RT1113874 1113875dT (13)


g(X) AEa

βR1113946infin

X

expminus u

u2 AEa

βRp(u) (14)



T21113888 1113889 lnAR

βEa

1 minus2RT

Ea

1113888 11138891113890 1113891 minusEa

RT (15)














0102030405060708090

100110

240 290 340 390 440 490 540 590 640

Wei

ght (

)

Temperature (degC)

Sample 1Sample 2

Sample 3Sample 4

(a)

Sample 1Sample 2

Sample 3Sample 4

0102030405060708090

100110

30 35 40 45 50 55 60 65

Wei

ght (

)

Time (min)

(b)





1 10 2558 275 067 24 278 302 15 268 60 057 25 313 473 20 425 106 06 29 336 66


slope minusEa

R (16)





ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)


VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)





0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)


ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)




dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



dX

dt Km(1 minus X)

n (8)


dX

dT Kd exp minus


2ω21113888 1113889 (9)



dX

dtdT

dt

dX

dT β

dX

dT (10)



dX

dt β

dX

dT A exp minus

Ea

RT1113874 1113875f(X) (11)


lndX

dt1113888 1113889 ln β

dX

dT1113888 1113889 ln[Af(X)] minus

Ea

RT (12)


g(X) 1113946X

0

dX

f(X)

A

β1113946

T

0exp minus

Ea

RT1113874 1113875dT (13)


g(X) AEa

βR1113946infin

X

expminus u

u2 AEa

βRp(u) (14)



T21113888 1113889 lnAR

βEa

1 minus2RT

Ea

1113888 11138891113890 1113891 minusEa

RT (15)














0102030405060708090

100110

240 290 340 390 440 490 540 590 640

Wei

ght (

)

Temperature (degC)

Sample 1Sample 2

Sample 3Sample 4

(a)

Sample 1Sample 2

Sample 3Sample 4

0102030405060708090

100110

30 35 40 45 50 55 60 65

Wei

ght (

)

Time (min)

(b)





1 10 2558 275 067 24 278 302 15 268 60 057 25 313 473 20 425 106 06 29 336 66


slope minusEa

R (16)





ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)


VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)





0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)


ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)




dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia








0102030405060708090

100110

240 290 340 390 440 490 540 590 640

Wei

ght (

)

Temperature (degC)

Sample 1Sample 2

Sample 3Sample 4

(a)

Sample 1Sample 2

Sample 3Sample 4

0102030405060708090

100110

30 35 40 45 50 55 60 65

Wei

ght (

)

Time (min)

(b)





1 10 2558 275 067 24 278 302 15 268 60 057 25 313 473 20 425 106 06 29 336 66


slope minusEa

R (16)





ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)


VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)





0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)


ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)




dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


slope minusEa

R (16)





ndash55

ndash53

ndash51

ndash49

ndash47

ndash45

ndash43

ndash41

1190 1195 1200 1205 1210 1215 1220 1225 1230 1235

ln k

(min

ndash1)

1000T (Kndash1)


VRMSCMRPM

(a)

ndash160

ndash155

ndash150

ndash145

ndash140

ndash135

ndash130

1190 1200 1210 1220 1230 1240

ln [ndash

ln (1

ndash X

)T2 ] (

min

ndash1)

1000T (Kndash1)


(b)





0

20

40

60

80

100

120W

eigh

t (

)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)

ndash3068

ndash2568

ndash2068

ndash1568

ndash1068

ndash568

ndash068

432

Der

ivat

ive w

eigh

t (

min

)

TGADTA

(a)

50 100

150

200

250

300

350

400

450

500

550

600

650

700

750

800

8500

Temperature (degC)


ndash505

Der

ivat

ive w

eigh

t (

min

)

Sample 1Sample 2

Sample 3Sample 4

(b)




dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



dX

dt Kd exp minus

(X minus 1)2

108341113890 1113891 (17)



4 Conclusions



Data Availability




Acknowledgments


References










MNDF

1195 1200 1205 1210 1215 1220 1225 1230 12351190

1000T (Kndash1)

10152025303540455055

ndashln

(K) (

min

ndash1)




























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia



























RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


research article ...d5373. proximate analysis was carried out using astm d5142-90 to acquire...

Documents