laboratory-study-of-spontaneous-combustion-of-coal
- 格式:pdf
- 大小:243.17 KB
- 文档页数:8
A laboratory study of spontaneous combustion of coal:the influence ofinorganic matter and reactor sizeWiwik Sujanti,Dong-ke Zhang*CRC for New Technologies for Power Generation from Low-rank Coal,Department of Chemical Engineering,The University of Adelaide,Adelaide,SA5005,AustraliaReceived8May1998;received in revised form30September1998;accepted9October1998AbstractA laboratory investigation into the effect of inorganic matter on spontaneous combustion behaviour of a Victorian brown coal.Fourteen samples were prepared,namely,the raw coal,water-washed coal,acid-washed coal,and acid-washed coal doped with11additives.Each of the samples was then tested in an isothermal reactor to obtain its critical ambient temperature,above which spontaneous combustion occurs. The relative effectiveness of the additives was determined by comparing their critical ambient temperatures with those of the raw coal and the acid-washed coal.Potassium chloride,Montan powder,and sodium chloride were found to be the most effective inhibitors,followed by magnesium acetate,and calcium chloride.The presence of sodium nitrate and ammonium chloride in the coal samples did not show any significant influence on the spontaneous combustion.However,calcium carbonate,sodium acetate,potassium acetate,and pyrite promoted the spontaneous combustion.The effect of additive loading was also investigated for an inhibition agent(KCl)and a promotion agent (NaAc).It was revealed that the effectiveness of these promotion and inhibition agents was enhanced with an increase in the additive loading. Low-temperature oxidation kinetics were also estimated by an energy balance approach and compared with the self-heating potential of these samples.The effects of reactor size and reactor specific surface area on the critical ambient temperatures are also discussed.᭧1999Elsevier Science Ltd.All rights reserved.Keywords:Additives;Coal combustion;Reactivity;Inorganic matter;Self-heating;Spontaneous combustion1.IntroductionSelf-heating and spontaneous combustion of coal have long been known to pose a serious problem for coal produ-cers and users[1].This is because spontaneous combustion of coal can lead to loss of desirable coal properties and product,and raise concerns about safety,especially in coal stockpiles,transportation over long distances,and in underground mining[1–3].A number of studies were carried out to determine the best method of preventing the spontaneous combustion of coal[4–6].Compaction of coal piles and inerting the atmo-sphere were applied[4].Additives,such as carbamide, diammonium phosphate,gel solution from water glass[5], aqueous solutions of inorganic salts,and salt–clay mixtures were also examined as potential inhibition agents to spon-taneous combustion[6].Smith,et al.[4]evaluated the effec-tiveness of10additives to suppress the self-heating process and revealed that sodium nitrate,sodium chloride,and calcium carbonate inhibited the spontaneous combustion, whereas sodium formate and sodium phosphate stimulated the self-heating process.The main aim of the work described in this paper was to examine the effect of inherent inorganic matter and different additives on spontaneous combustion so as to obtain a better understanding of the role of the inorganic matter in the low-temperature oxidation processes.An isothermal reactor was used to determine the critical ambient temperature,above which spontaneous ignition(thermal runaway)occurs.The critical ambient temperatures were then used to compare the tendency of various coal samples toward spontaneous combustion.An increase in the critical ambient temperature due to the use of an additive may indicate the effectiveness of the additive to suppress the self-heating potential of the coal.The effect of additive loading on the critical ambient temperature for spontaneous combustion was investigated for typical promotion and inhibition additives.Low-temperature oxidation kinetics of these coal samples are also estimated from the temperature measurements.The effects of reactor size on the critical ambient temperaturesFuel78(1999)549–556 0016-2361/99/$-see front matter᭧1999Elsevier Science Ltd.All rights reserved. PII:S0016-2361(98)00188-4*Corresponding author.Tel.:ϩ61-8-8303-5085;fax.:ϩ61-8-8303-6081.and the estimated low-temperature oxidation kinetics are also examined.2.Experimental 2.1.Coal samplesA Victorian brown coal was selected for this study,Table 1features the proximate,ultimate and inorganic analyses of the coal.The freshly picked raw coal was crushed and sieved and the 0.25–1.00mm size fraction was retained for preparation of all other samples.Double-deionised water-washed and acid-washed samples [7,8]were prepared to investigate the effect of inherent inorganic matter on the spontaneous combustion.Double-deionised waterwashing removed the water-soluble inorganic matter and acid-washing (37%HCl acid solution)removed most of the inorganic matter from the raw coal To investigate the effect of individual inorganic compo-nents in coal on spontaneous combustion,11additives were selected and,respectively,added into the acid-washed coal.They were pyrite (FeS 2),potassium acetate (KAc),sodium acetate (NaAc),calcium carbonate (CaCO 3),sodium nitrate (NaNO 3),ammonium chloride (NH 4Cl),calcium chloride (CaCl 2),magnesium acetate [Mg(Ac)2],sodium chloride (NaCl),potassium chloride (KCl),and Montan powder [CaCl 2powder ϩ3%sodium dodecyl sulphate (SDS)].Water soluble additives [KAc,NaAc,NaNO 3,NH 4Cl,CaCl 2,Mg(Ac)2,NaCl,and KG]were dissolved in double-deionised water to form solutions with a concentration of4.76wt%.2.5l of each solution and 600g of acid-washed coal were mixed to form slurry.The mass ratio of the addi-tive in a solution to the acid-washed coal was 25%.The slurry was continuously stirred for 24h.The liquid in the slurry was then allowed to drain through a filter,and the resulting paste was dried in nitrogen at 105ЊC for 8–10h.Chemical analysis confirmed that the samples prepared in such a way contain about 5wt.%additive.Water insoluble additives (FeS 2,CaCO 3,and Montan powder)were added to the coal as solid fine powder and the mixtures were well mixed.The concentration of water insoluble additives in the coal samples was 5wt.%.All the samples were stored in tightly sealed plastic bags to avoid pre-oxidation before experimentation.2.2.Experimental proceduresA sample to be tested was loosely placed in an isothermal reactor,a stainless-steel cylindrical reactor of 8cm height and 4cm diameter.Four thermocouples were inserted inside the reactor to monitor temperatures along the reactor axis.The thermocouples had a precision of ^0.1ЊC.A gas (N 2or air)was preheated in an 18m long copper tube to the oven temperature,and then passed through the coal sample from the bottom of the reactor.The oven was first preheated to a preselected temperature.During this stage,nitrogen gas passed through the reactor at 100ml min Ϫ1(N.T.P.)to minimise any pre-oxidation.Once the desired temperature was reached,air then replaced the nitrogen at the same flow rate and was maintained through-out the experiment.A data acquisition system was activated to record the temperatures.A schematic diagram of experi-mental apparatus used in this work is shown in Fig.1.In preliminary experiments,attempts were also made to follow the gas products in the exiting gas from the reactor using a gas chromatograph.However,no detectable CO or CO 2was noticed,nor was there any depletion of oxygen at temperatures of interest.This was apparently due to the low reaction temperatures (thus slow reaction rates)and the small quantity of coal samples used.Therefore,temperature measurement was chosen as the sole technique to monitor development of spontaneous combustion in these experi-ments.It was also confirmed that the four thermocouples gave exactly the same temperature reading.The experiment was terminated either when a thermal runaway occurred or when the temperature of the sample continued to level off towards the oven temperature.Typical temperature profiles of the Mg(Ac)2-added-coal are shown in Fig.2.It can be seen that at temperatures below 155ЊC,thermal runaway did not occur,so that the experiments were terminated when the average temperature along the reactor showed a trend to level off.However,at 155ЊC,thermal runaway occurred.Experiments were repeated for different oven temperatures,at a 5ЊC interval,until the critical ambi-ent temperature,T crit ,of a sample was determined.Repeat-ability of the determined T crit values was confirmed byW.Sujanti,D.Zhang /Fuel 78(1999)549–556550Table 1Proximate and ultimate of coal used [7]Proximate analysis (wt%)Moist VM (db) F.C.(db)Ash (db)61.348.348.7 3.0Ultimate analysis (wt%db)C H N O Organic sulfur S Total 67.4 4.70.726.90.260.26Fig.1.Schematic diagram of the experimental apparatus.repeating the experiments and essentially the same T crit value was obtained.Thus,for each sample,the critical ambient temperature was taken as the average of the lowest temperature at which the thermal runaway occurred and the highest temperature at which the thermal runaway did not occur,with an accuracy of^2.5ЊC.3.Results and discussion3.1.The effect of inherent inorganic matterThe presence of inorganic matter in coal may catalyse the oxidation of coal[7],and removing the inorganic matter may change the coal reactivity.Table2shows the inorganic analysis of raw,water-washed,and acid-washed coal.It is shown that water-washing only removes small amount of Na,Fe and K,while acid-washing removes most of this inorganic matter.Therefore,in this instance,the influence of inherent inorganic matter on the spontaneous combustion can be substantially eliminated by treating the raw coal with an acid.Table3shows that the critical ambient temperatures of the water-washed sample(145ЊC)and the acid-washed sample(147.5ЊC)are higher than that of the raw coal (142.5ЊC).The differences are small in relation to the accu-racy of determination but the increase in critical ambient temperature indicates that removing the inherent inorganic matter from coal helps reduce its self-heating potential.3.2.The effect of additivesTable3also shows that among the11additives used, FeS2,KAc,NaAc,and CaCO3promote the spontaneous combustion of the coal used as the relevant critical ambient temperatures are considerably lower than that of the acid-washed coal.NH4Cl and NaNO3are effectively neutral,but CaCl2and Mg(Ac)2increase the critical ambient tempera-ture for spontaneous combustion by5ЊC–152.5ЊC,and Montan powder,KCl,and NaCl push the critical ambient temperature to157.5ЊC,indicating that these additives inhi-bit the spontaneous combustion.For additives which show the same critical ambient temperatures of the samples,such as Montan powder,KCl,and NaCl,or CaCl2and Mg(Ac)2, their abilities to suppress the spontaneous combustion are compared,based on the rates of temperature rise at their critical ambient temperatures.It has been acknowledged[1,2,7,9]that FeS2in coal cata-lyses the oxidation reaction.In addition,in a moist air,FeS2 itself also experiences an exothermic oxidation reaction which helps accelerate the coal self-heating[1,2,7,9]. Sodium and potassium also catalyse the spontaneous combustion especially when they are present as organic salts[7,9].The results for FeS2,KAc and NaAc added samples used in the present study are in agreement with previously published work.The addition of CaCO3to the acid-washed coal decreases the critical ambient temperature of the acid-washed coal by 10ЊC to137.5ЊC,which could be due to the catalytic effectW.Sujanti,D.Zhang/Fuel78(1999)549–556551Fig.2.Temperature histories of Mg(Ac)2-added-coal sample at several oven temperatures.Table2Inorganic analyses of raw coal,water-washed coal,and acid-washed coalSamples Inorganic analysis(wt%db)Na Mg Fe Si Al K Ca Raw coal0.090.280.29Ͻ0.05Ͻ0.050.0070.62 Water-washed coal0.060.280.22Ͻ0.05Ͻ0.050.0060.62 Acid-washed coalϽ0.0010.0010.03Ͻ0.05Ͻ0.05Ͻ0.0020.004of the calcium on the spontaneous combustion.NaNO3does not show any effect,suggesting that sodium in an inorganic salt would have little catalytic effect when compared with its organic salts.However,these results do not agree with the previousfindings[4]that NaNO3and CaCO3inhibited spon-taneous combustion.NH4Cl was also found to have no significant effect on the spontaneous combustion behaviour of the coal,but this may be a consequence of the decom-position of the salt at100ЊC,which would eliminate any catalytic activity of the NHϩ4.In general,alkali and alkaline-earth metals in organic salts tend to promote the spontaneous combustion while inorganic salts would inhibit spontaneous combustion. However,Mg(Ac)2seems to be an exception,as it inhibits the spontaneous combustion.It was suggested[10,11]that Mg can enhance oxidation of carbon,probably at high temperatures.However,Mg2ϩmay have no catalytic effect or be a negative catalyst for the low-temperature(Ͻ200ЊC) oxidation reaction of coal.3.3.The effect of additive loadingTo investigate the effect of additive loading on promotion or inhibition of spontaneous combustion,samples with1,5, and10wt%of NaAc and KCl were prepared and tested. Table4presents the critical ambient temperatures of these samples.In general,the critical ambient temperature decreases with increasing NaAc loading,but increases with KAc loading,indicating that promotion and inhibition effects of these additives on spontaneous combustion depend on their loading levels.Although1%and5wt% NaAc loadings in the samples offered a similar critical ambient temperature,within the accuracy(^2.5ЊC)of the experiments,a closer examination of the temperature profiles during the self-heating revealed that for the5wt% NaAc loading a higher rate of temperature rise was experi-enced.Similarly,in the presence of1wt%KCl the sample showed the same critical ambient temperature as the acid-washed coal,but its rate of temperature rise was lower due to the inhibition effect of KCl.3.4.Estimation of low-temperature oxidation kineticsTo compare reactivities of the samples and correlate the reactivities with their self-heating behaviour observed in the experiments,low-temperature oxidation kinetics wereW.Sujanti,D.Zhang/Fuel78(1999)549–556552Table3Summary of critical ambient temperatures,heat of oxidation reaction,and apparent kinetic constants estimated for various samples Sample T crit(ЊC)Q(MJ kgϪ1)E(kJ molϪ1)A(kg m-3sϪ1PaϪ1)FeS2ϩcoal110.0^2.518.879.6 4.91×103KAcϩcoal125.0^2.521.8111.9 3.53×107NaAcϩcoal137.5^2.522.2109.5 1.79×107CaCO3ϩcoal137.5^2.521.2189.6 1.83×1017Raw coal142.5^2.522.9144.59.84×1010Water-washed coal145.0^2.523.7239.7 1.55×1022Acid-washed coal147.5^2.524.2164.4 2.43×1013NH4Clϩcoal147.5^2.522.6160.2 4.21×1012NaNO3ϩcoal147.5^2.522.3203.5 1.42×1018CaCl2ϩcoal152.5^2.520.3145.6 6.53×1010Mg(Ac)2ϩcoal152.5^2.522.6103.2 5.07×105Montan powderϩcoal157.5^2.520.4102.9 3.52×105KClϩcoal157.5^2.521.5109.1 2.00×106NaClϩcoal157.5^2.515.7219.57.35×1019Table4Summary of the critical ambient temperatures and kinetic constants estimated for samples with various additive loadings Sample T crit(ЊC)Q(MJ kgϪ1)E(kJ molϪ1)A(kg m-3sϪ1PaϪ1)Acid-washed coal137.5^2.524.2103.6 1.62×1061%NaAc-added-coal132.5^2.523.1180.18.92×10155%NaAc-added-coal132.5^2.522.2208.0 4.34×101910%NaAc-added-coal122.5^2.521.0256.5 5.78×10261%KCl-added-coal137.5^2.522.4164.09.03×10135%KCl-added-coal142.5^2.521.5179.5 5.64×101510%KCl-added-coal147.5^2.520.4179.1 2.45×1015estimated by an energy balance approach.The heat balance for a sample in an isothermal reactor at a point when the coal temperature is equal to the oven temperature(T o),under which condition there is no heat loss from the sample to the oven environment,can be written as:c sd Td t!oQA eϪE R T o P n O21where c is the specific heat capacity of coal[7](J kgϪ1KϪ1), 1550J kgϪ1KϪ1;s is the bulk density of coal(kg mϪ3), 1350kg mϪ3;T is the coal temperature(K)and t is time (s);Q is the heat of oxidation reaction(J kgϪ1)which was determined experimentally for each of the samples.A is the apparent pre-exponential,factor(kg mϪ3sϪ1PaϪ1)and E is the apparent activation energy(J molϪ1);R is the universalgas constant.P O2is the partial pressure of oxygen gas,which is21kPa in the present case.The apparent reaction order,n,is assumed to be1.According to Eq.(1),from a linear plot of ln[d T/d t]o versus1/T o the apparent kinetic constants,A and E,can be determined.Fig.3shows typical plots of ln[d T/d t]o versus1/ T o for the raw and acid-washed coal samples.Since it is evident that a linear relation between ln[d T/d t]o versus1/ T o does exist,and that the temperature inside the reactor is noticed to be uniform,Eq.(1)can be used as a technique for estimation of apparent kinetics of coal in such spontaneous combustion tests.Fig.4compares the Arrhenius plots of several samples in the presence of promotion agents,raw coal,water-washed, and acid-washed coal.The presence of FeS2in the coal results in the highest reactivity which corresponds to the lowest critical ambient temperature observed.The reactivity of the samples with the presence of KAc and NaAc are lower than that of the sample with FeS2,but higher than those of raw coal,water-washed,and acid-washed coal samples.However,the reactivity of the water-washed coal is higher than that of the raw coal,which is not in agreementW.Sujanti,D.Zhang/Fuel78(1999)549–556553Fig.3.Typical linear plots of ln[d T/d t]versus1/T o for the raw and acid-washed coal samples for estimation of apparent oxidation kinetics ofcoal.parison of reaction rates for various samples in temperature range of110ЊC–150ЊC,assuming an oxygen partial pressure of21kPa.with their critical ambient temperature data.To further clar-ify this,the reaction rates estimated using the calculated kinetics are plotted against the critical ambient temperatures for all samples studied in Fig.5.In general,Fig.5shows an overall trend that the lower the critical ambient temperature of a sample,the higher its oxidation reactivity.However,the data are more scattered for those samples containing the inhibition agents.The reactivities of samples with the inhi-bition agents do not show a clear trend with their critical ambient temperatures.It is possible that the assumed sample densities and specific heat capacities of the samples used in Eq.(1)may have contributed to the inconsistency in the reactivity–critical ambient temperature correlation.The varying packing density of the samples in the reactors could also have contributed to this inconsistency.However,reproducible critical ambient temperature measurements reject this argument,although the effect of packing density is of interest on its own,which will be further investigated in future studies.Further,the coal treatments,i.e.washing and additive loading,may have changed the coal structure,and thus the coal reactivity.Fig.6compares the Arrhenius plots of several samples with the presence of NaAc and KCl in various loadings.Fig.6shows that 10wt%of NaAc in the coal sample leads to the highest reactivity of the NaAc loaded samples,which corre-sponds to the critical ambient temperatures observed.Fig.6also shows that although,within the experimental accuracy,both 1and 5wt%NaAc-added-coal samples have the same critical ambient temperature,the 5wt%NaAc-added-coal sample is more reactive.The reactivities of the samples with KCl in them show the opposite trend;the higher KCl load-ing,the lower its reactivity.W.Sujanti,D.Zhang /Fuel 78(1999)549–556554Fig.5.Low-temperature oxidation rates versus the critical temperature for various samples in air at 140ЊC.Fig.6.The effect of additive loading on the estimated oxidation rates for NaAc and KCl added coal samples.4.The effect of reactor sizesTo examine the effect of reactor size on the self-heating process,three larger isothermal reactors were constructed,with 27,15and 12cm height and 9.8,8and 6cm diameter,respectively.The four isothermal reactors,R-1,R-2,R-3,and R-4,respectively,were filled with 1000,350,140and 30g of coal,and the critical ambient temperature of each reactor was determined.In general,it was observed that the critical ambient temperatures decreased with increasing coal mass.However,the correlation between the critical ambient temperatures and the coal mass is not rectilinear.Whether a spontaneous combustion occurs or not is deter-mined by two competing processes:the rate of heat genera-tion due to oxidation,and the rate of heat loss to the surroundings which depends on the external surface area.Thus,Fig.7shows that the critical ambient temperature increases linearly with specific external surface area.This suggests that for a given coal,the spontaneous combustion behaviour depends on the external heat transfer environ-ment.The apparent kinetic parameters,E and A ,from these four reactors (R-1,R-2,R-3and R-4)were also determined.It was found that the values of E obtained from these reactors are similar,i.e.106.16,105.49,107.37and 107.10kJ mol Ϫ1,respectively.However,the values of A are significantly different,i.e. 2.99×108,1.97×108,1.09×108and 1.60×107kg m Ϫ3s Ϫ1Pa Ϫ1,respectively.Fig.8shows that the smaller the reactor,thus the larger the specific exter-nal surface area,the lower the reactivities.As summarised by Chen [12]the value of A varies with temperature,and extent of reaction.It was observed during the experiments that the smaller the reactor,the higher the critical ambient temperature required for the thermal runaway to occur,and thus the more rapid the oxidation of the coal.The pre-oxidation before the thermal runaway is detected,could have caused a loss of the reactivity for smaller reactors.W.Sujanti,D.Zhang /Fuel 78(1999)549–556555Fig.7.A plot of critical temperature versus reactor specific external surfacearea.Fig.8.The effect of reactor size on the estimated reactivity of coal oxidation in air between 70ЊC and 140ЊC.5.ConclusionsWater-washing only removes a small amount of inherent inorganic matter,while acid-washing removes most of the inorganic matter in this coal.The critical ambient tempera-ture of acid-washed coal is higher than those of raw and water-washed coals,indicating the catalytic effect of inher-ent inorganic matter in the coal.Of the eleven additives applied to the acid-washed coal,FeS2,KAc,NaAc and CaCO3were found to promote the spontaneous combustion, while Montan powder,KCl,NaCl,CaCl2,and Mg(Ac)2 inhibit the spontaneous combustion.NaNO3and NH4Cl, have very little effect.The low-temperature oxidation kinetics of the samples were also estimated,and whilst the reactivities show a general trend of lower critical ambient temperature,the higher the reactivity of the sample,further investigation is required in order to improve the accuracy of the kinetic estimation.The promotion and inhibition effects of NaAc and KCl,respectively,on spontaneous combustion are enhanced with an increase in their loadings.The critical ambient temperatures were also determined for various reactor sizes.The critical ambient temperature showed a linear correlation with the reactor specific external surface area,indicating the important role of external heat transfer in the spontaneous combustion process. AcknowledgementsThe authors gratefully acknowledge thefinancial and other support received for this research from the Coopera-tive Research Centre(CRC)for New Technologies for Power Generation from Low-rank Coal,which is estab-lished and supported under the Australian Government’s Cooperative Research Centres program.Sujanti would also like to thank the CRC for a postgraduate scholarship. The authors also wish to thank Prof.Chuguang Zheng of Huazhong University of Science and Technology,China, for measurements of the heating values of the coal samples. References[1]Carras JN,Young BC.Prog Energy Combust Sci1994;20:1.[2]Glasser D,Bradshaw S.Advances in reactor design and combustionscience.In:Handbook of heat and mass transfer,vol.4.1990.p.1071.[3]Kim AG,Chaiken RF.Fire in abandoned coal mines and waste banks.Information Circular,USA Bureau of Mines,1993.[4]Smith AC,Miron Y,Lazzara CP.Inhibition of spontaneous combus-tion of coal.Report of Investigation—USA Bureau of Mines,RI 9196,1988.[5]Evseev V.Twenty-first International Conference of Safety in MinesResearch Institute,Sydney,1985.p.481.[6]Chamberlain EAC.Colliery Guardian1974;March:79.[7]Durie RA.The science of Victorian brown coal:structure,propertiesand consequences for utilisation.Oxford:Butterworth-Heinemann, 1991.p.267.[8]Ye DP.Gasification of South Australian Lignite.PhD thesis,TheUniversity of Adelaide,Adelaide,1994.p.82.[9]Banerjee SC.Spontaneous combustion of coal and minefires.Rotter-dam:A.A.Balkema,1985.p.7.[10]Jackson PJ.In:Lahaye J,Prado G,editors.Fundamentals of thephysical-chemistry of pulverised coal combustion.Dordrecht:Marti-nus Nijhoff,1987.p.273.[11]Emeleus HL,Sharpe AG.Advances in inorganic chemistry and radio-chemistry,vol.1.New York:Academic Press,1959.p.180. [12]Chen XD.Twenty-fourth Australian and New Zealand ChemicalEngineering Conference and Exhibition,vol.6.Sydney,1996.p.109.W.Sujanti,D.Zhang/Fuel78(1999)549–556 556。