Comparisons of Fly Ash and Deposition Between Air and Oxy-Fuel Combustion in Bench-Scale Fluidized Bed with Limestone Addition

，，，，

（School of Energy Science and Engineering，Harbin Institute of Technology，Harbin 150001，China）

Comparisons of Fly Ash and Deposition Between Air and Oxy-Fuel Combustion in Bench-Scale Fluidized Bed with Limestone Addition

Zhimin Zheng，Hui Wang∗，Yongjun Guo，Li Yang，Shuai Guo and Shaohua Wu

（School of Energy Science and Engineering，Harbin Institute of Technology，Harbin 150001，China）

In Oxy-fuel circulating fluidized bed，the residual CaO particles may react with high concentration of CO2in flue gas to form bonded deposit on heat transfer surfaces in backpass when limestone is used as a sorbent to capture SO2.In this paper，experiments were designed on ash deposition in a bench-scale fluidized bed under oxy-fuel and air atmosphere.A novel ash deposit sampling probe was used to simulate the tubes of tail surfaces. The chemical composition of fly ash and ash deposit from both air-firing and oxy-fuel firing cases were analyzed by Inductively Coupled Plasma-Atomic Emission Spectrometry（ICP-AES）and Scanning Electron Microscopy（SEM），respectively.The degrees of carbonation reaction of ash deposits were measured by Thermo Gravimetric Analysis.The results showed that there are distinct differences in fly ash deposition rate between oxy-fuel and air firing cases，and oxy-fuel combustion with limestone addition can affect chemical composition of fly ash and ash deposit，especially for elements of Ca，Na，K，and S.However，the carbonation reaction degree of ash deposits is found weak，which is due to the relatively low CaO content in ash deposit or not long enough of the sampling time.

carbonation；ash deposit；fly ash；oxy-fuel；CFB

1 Introduction

Circulating Fluidized Bed（CFB）boiler technology is feasible for oxy-fuel combustion due to its operation flexibility and efficient SO2capture while in-situ injection of calcium-based sorbents.However，there is a potential problem due to ash deposition when limestone is in use［1-2］.In oxy-fuel combustion，when the bed temperature reaches 870℃［3］，limestone captures SO2via indirect sulfation instead of direct sulfation mechanism.If excess calcium oxide exists，it will be carbonated to form calcium carbonate at specific conditions.The phenomenon of carbonation of fly ash can occur at some locations（e. g.，cyclone，dip-leg，seal-pot，heat exchanger surface）. The deposition of carbonated fly ash may cause severe fouling.Carbonation of fly ash was observed in the pilot scale tests under CFB oxy-fuel combustion［4］.The experimental results from Beisheim et al.［5］showed that after an exposure time of 30 h in the simulated oxy-fuel combustion atmosphere，the deposition of particles produced the bonded deposits due to carbonation reaction. The reaction is given as Eq.（1）.This reaction is affected by many factors，such as temperature，water vapor concentration，and CO2concentration［6］.

In our previous study，the effects of oxy-fuel combustion on ash deposition and fly ash properties without limestone addition were investigated.Results showed that oxy-fuel combustion has obvious higher ash deposition propensity and fine ash concentrations［7］.In addition，initial stage of ash deposition behavior were investigated［8］.Due to the complexity of ash deposition problem，deeper research needs to be carried out.

The part of experimental results on ash deposition during the oxy-fuel combustion has been published in the previous paper［8］.The characteristics of fly ash and ash depositsin both air and oxy-fuel combustion were investigated at the different Ca／S molar ratios to thoroughly compare the deposition behavior in these two combustion cases.Ash deposition experiments were carried out in the bench-scale reactor.A novel ash deposit sampling probe was designed to simulate the heat exchanger surfaces and this probe can easily control its surface temperature.The temperature range of the surface is 200-700℃.Understanding the difference of ash deposition between air and oxy-fuel combustion conditions is important for building or retrofitting industrial CFB boilers.

2 Experimental

2.1 FBC Experimental Setup and Operating Conditions

The details of the experimental setup have been discussed in detail in two previous published papers［7-8］and thus only a brief description will be given here.The experimental setup is presented in Fig.1（a）.

Fig.1 Schematic of the bench-scale fluidized bed combustor and ash deposit sampling system

This setup mainly contained a heater，a furnace，a cyclone，a convective section，and a filter bag.The furnaces were equipped with the several electric heaters.A screw feeder is controlled between 7.5 and 13.6 g／min for the mixture of coal and limestone.The flow rates of air，O2and CO2were controlled by three MFCs.The fluidization velocity was kept constant for comparison reasons.Two excess air ratios（1.4 and 1.28）were chosen to ensure similar O2concentrations at the outlet.This paper aims to understand dynamic ash deposition rate when the heat flux through the ash deposit significantly varied with deposition time［9］. Time range needs to be long enough for stable state to be reached and enough ash samples to be collected for further analysis.Therefore，one hour is chosen as the deposition time for all cases.The experimental conditions are listed in Table 1.

2.2 Material

An anthracite coal（AC）produced at Jincheng and a limestone produced at Shou County in China were used.As shown in Fig.2，the ranges of particle sizes for coal and limestone were 0-2.36 mm，and 0-1.18 mm，respectively.The average diameter of the limestone was 564 μm.The proximate and ultimate analyses of AC，the composition of ash，and composition of limestone are listed in Tables 2-4，respectively.In Table 4，the determination of calcium oxide and magnesium oxide contentin limestone by methods of chemical analysis（GB／T3286.1-2012）. The content of CaCO3and MgCO3is calculated by the content of CaO and MgO，respectively.The particle sizes of the quartz sand used as bed material ranged from 0.18 to 0.55 mm.

Fig.2 PSDs of the coal and limestone

2.3 Ash Sampling and Analysis

In order to collect the typical ash deposit samples，a novel deposit sampling probe was designed.The dimension of this probe has been given in the previous paper.A hole was drilled parallel to the axial and close to the outer surface.A K-type thermocouple of 1 mm O.D.was used to measure the surface temperature of the probe，which was fixed into the hole near theoutside surface of the probe.The probe surface temperature can be controlled in the range from 200℃to 700℃with a standard deviation of approximate 2℃by a temperature-controlled system.As shown in Fig.1（b），the ash deposit sampling system consists of a probe，a temperature-controlled system and a telescopic sleeve and so on.The telescopic sleeve was used as a protective cover to protect the probe head from exposure to the flue gas before the combustion conditions met the demand.The sleeve with 38 mm O.D.and 100 mm length can be easily moved by controlling the motion of the ring.When it was pulled off，the probe was exposed to flue gas and its surface temperature reached the desired temperature rapidly. Fig.1（c）shows the set temperatures.Each experimental case was carried out more than three times.The chemical compositions of collected ash samples were analyzed by an Inductively Coupled Plasma-Atomic Emission Spectrometry（ICP-AES）（IRIS INTREPID 2 XDL，Thermo Elemental）.The morphology of ash deposits was tested by an Environment Scanning Electron Microscopy（ESEM）（Quanta 200FEG，FEI Corporation）.The carbonation reaction degree was tested by Thermo gravimetric Analysis（TGA）（Mettle-Toledo，TGA／SDTA851e）. Ash deposits of 10 mg or so were gradually heated to 900℃with a heating rate of 20℃／min and then kept at 900℃for 40 min.Since the resolve temperature of limestone used in this study was approximately 600℃，the weight loss from 600 to 900℃divided by the total mass of ash sample was defined as degree of carbonation.The curve of weight loss is similar to the curves of the published paper［8］but not listed here.

Table 1 Experimental conditions

Table 2 Proximate and ultimate analyses of AC

Table 3 Chemical compositions of the resulting ash of AC（wt%）

Table 4 Chemical compositions of the limestone as sorbent（wt.%）

3 Results and Discussion

3.1 Ash Deposit

3.1.1 Deposition propensity

The deposition rate（DR）and deposition propensity（DP）are normally used to quantify fly ash deposition，as illustrated in formulas（2）and（3），respectively.In combustion experiments，when different Ca／S ratios were used，different amounts of ash were generated.To eliminate the influence caused by ash amount，formula（3）is selected to evaluate the ash deposition.

Fig.3 shows the comparison on the deposition propensity between air and oxy-fuel combustion at different Ca／S ratios.It can be observed thatDPhas an increasing trend with increasing Ca／S ratios for air combustion cases，but the trend is just contrary to that of oxy-fuel combustion.MeanwhileDPin air is higher than that in oxy-fuel combustion，and the gap ofDPbetween two combustion cases increases with the increase of Ca／S ratios.For Ca／S＝0，it should be noted that there is a slight difference for deposition propensity between oxy-fuel combustion and air combustion.The result differs from previous results whereDPwas obvious higher in oxy-fuel combustion［7］.It is possibly due to the deposition time with one hour in our experiment，rather than four hours in the previous experiments.Actually，ash deposition may vary as deposition time increases because the ash deposition is a dynamic process［10-12］.One hour is long enough to form a thin deposit layer on the surface of this probe，which composes the so-called initial deposit layer.It is close to wall surface of the tube and distinct from the outer layer of the thick ash deposit，because it mainly consists of fine particles less than 10 μm in diameter［9］.Generally，the results also show that limestone addition has an obvious effect on ash deposition in the initial stage of ash deposit formation for both air combustion and oxy-fuel combustion but the effects for two combustion case are different.According to ash deposition mechanism，three possible reasons for that are illustrated here.Firstly，ash formation can be affected including particle emission and chemical compositions after limestone addition［13-14］.Secondly，fine particle emission can be increased in oxy-fuel combustion with high O2concentration［7，15］.Special emphasis was given to that because fine particle emission has a significant influence on initial deposit layer［9］.Thirdly，the fragmentation of limestone can be affected by oxy-fuel combustion［16］，which can affects the fly ash compositions，especially for high Ca／S ratio.

3.1.2 Deposit morphology

The morphologies of ash deposit were observed by the photographic way and SEM，respectively，as illustrated in Fig.4.As shown in the photographs，it is noted that the deposits on surface are loose and evenly distributed for both air and oxy-fuel combustion.Most of these deposits are concentrated on the windward side of the probe，and little ash appears on the two sides. There is no clear vision of two sides because the photographs were taken just above the ash deposits. Large particles with great kinetic energy deposit easily on the windward in inertial deposition mechanism，while fine particles deposit on sides mainly due to thermophoresis［17］.Seeing from SEM images，most of particles are less than 10μm.In the previous research［8］，deposition rate was found to be affected obviously by the probe surface temperature.In this experiment，it was concluded that thermophoresis plays an important role in ash deposition.

Fig.3 Deposition propensities under oxy-fuel and air combustion conditions at different Ca／S ratios

Fig.4 Deposit morphology of deposits resulting from oxy-fuel and air combustion conditions at Ca／S＝2.5

3.1.3 Deposit chemical composition

It should be noted that for both air and oxy-fuel combustion，most of element compounds show the same changing trends，as shown in Fig.5，except elements of K，Na，and S.It can be noticed that the changes of deposition chemical compositions are within the variation of error bars.Under this situation，we only talk about the average values to analyze its tendency. For oxy-fuel combustion，the contents of both K2O and Na2O decrease with the increase of Ca／S ratios，but for air combustion，they decrease at first and then slightly increase.For the same Ca／S ratio，differences can be found for air and oxy-fuel combustion，and it is especially obvious for the content of K2O，Na2O，and SO3.As for the K2O content，it is found that oxy-fuel combustion has slightly higher content than that of air combustion at the same Ca／S ratio from 0 to 3.5.As for the contents of Na2O and SO3，the comparisons are uncertain.It should be also noted that there are only minor differences but with relatively big errors（Standard Deviation）for the contents of some elements such as SO3.It is analyzed that main error resources come from the deposit samples and measurement method.For each case，at least three samples are collected with small differences for deposition mass and then they are analyzed independently by the ICP-AES. It can reduce the uncertainties from the ash samples to some content.The relative standard deviations（RSD）of the elements measured by ICP-AES range from 0.13%to 2.19%.For example，the RSD for SO3is 0.54%.Carbonation degrees of ash deposits from oxyfuel and air combustion conditions measured by TGA at Ca／S＝2.5 are 1.38 and 1.54%，respectively. Carbonation degree of ash deposits areweak，which is related to low content of CaO in fly ash and short deposition time.

3.2 Fly Ash

3.2.1 ICP-AES analysis

Ash deposition contents strongly depend on chemical compositions of fly ash，although the detained relationship between them has not been determined.But it is certain that some elements（eg.Na，K，S and Ca，etc.）have significant effect on ash deposition.It is because some compounds composed of those elements usually have relative low melting point，which will result in easy sticking on the tube and facilitating ash deposits growth.Some compounds also react with flue gas，which can lead to sintering of ash deposits due to chemical reaction［5］.As shown in Fig.6，for the major elements of Si and Al，they decrease slightly as Ca／S ratio increases from 0 to 3.5 both in oxy-fuel combustion and air combustion conditions，in contrast that the content of Ca element increase obviously.But the contents of Fe and P elements seem not to be affected by limestone addition for oxy-fuel combustion and air combustion.For the content of MgO，there seems to have a reverse changing trend as Ca／S ratio increases for both conditions.But now it is difficult to explain the phenomenon.The elements of K2O，Na2O，and SO3were paid more attention than other elements because they are the main reasons of fouling and corrosion in the boilers.For K2O and Na2O elements under oxy-fuel combustion condition，they have an obvious decreasing trend at Ca／S＝2.5 but it does not be enhanced further by higher Ca／S ratio when it is 3.5.The contents of K2O and Na2O elements under air combustion condition have a slightly decreasing trend with increasing of Ca／S ratio.And for SO3content，it increases first and then decrease for both conditions.However，it is not excluded that the trend can be affected by measurement errors.Generally，the effect of limestone addition on the elements contents in fly ash really exists in both oxy-fuel combustion and air combustion conditions.The changing trends for most of the elements seem to be similar.

Fig.5 Chemical compositions of deposits resulting from oxy-fuel and air combustion conditions at different Ca／S ratios

Fig.6 Chemical compositions of fly ash resulting from oxy-fuel and air combustion conditions at different Ca／S ratios

3.2.2 Particle size distribution

The PSD of fly ash can strongly affect the deposition behavior［18-19］.According to ash deposition mechanism［17］，inertial impaction is dominant mechanism for big particles whose diameter more than 10μm，but thermophoresis becomes the main deposition mechanism for small particles whose diameter less than 5 μm.For inertial impaction，particle size is an important factor which determines whether the particles can deposit on the surface of the tube.For thermophoresis deposition mechanism，the temperature gradient near the probe is a dominant factor affecting fine particles deposition.Meanwhile，chemical compositions can show obvious differences for ash particles with different sizes，especially for fine particles due to different formation mechanism. Therefore，the PSDs of fly ash are taken for measurement here.The result of the PSD of fly ash at Ca／S＝0 has been discussed in the previous published paper［7］.In Fig.7，it should be noted that the average particle is 7.9 μm from oxy-fuel combustion compared with 5.1μm of air combustion at Ca／S＝0，and at Ca／S＝2.5，the average particle for oxy-fuel combustion and air combustion is 7.1 μ m and 6.0 μm，respectively.The results seem to support the conclusion that oxy-fuel combustion has a slightly greater PSD compared with air combustion with or without limestone.In addition，it is found that limestone addition has a slight influence on the PSD of fly ash under both conditions.However，it is difficult to build the direct relation between the PSD of fly ash and deposit propensity due to lacking of enough data. Actually ash deposition rate is also affected by other factors，such as chemical compositions of fly ash and property of flow field.In addition，an important point needs to be noted that oxy-fuel combustion with limestone may change the distribution of bottom ash，coarse ash from the cyclone，and fly ash，which can also affect the ash deposition.Fryda et al.［20］found that oxy-fuel combustion could change ash distribution between ash deposit and filter ash in a lab scale pulverized coal combustor.But no researches on ash distribution under CFB oxy-fuel combustion were investigated.Therefore，further study needs to be done to investigate the effect of CFB oxy-fuel combustion on fly ash to understand ash deposition behavior.

Fig.7 PSDs of fly ash resulting from oxy-fuel and air combustion conditions at Ca／S＝0 and 2.5

4 Conclusions

Limestone addition has an effect on deposit propensity，chemical compositions，PSD of fly ash for both oxy-fuel combustion and air combustion.The deposit propensity has a decreasing trend for oxy-fuel combustion with the increase of Ca／S ratios from 0 to 3.5，but it is not obvious for air combustion.The difference gap of deposit propensity for both conditions increases with the increase of Ca／S ratio.Oxy-fuel combustion with limestone addition can affect chemical composition of fly ash and ash deposit，especially for Ca，Na，K，and S elements.But for some elements，the variation tendencies with the increase Ca／S ratio are the same for both conditions.Carbonation degree of ash deposits is weak，which is related to low content of CaO in fly ash and short deposition time.The average particle size of fly ash particles under oxy-fuel combustion case is slightly higher than that under air combustion case.But due to the complexity of ash deposition process，further study will be required to explain the ash deposit behavior.

［1］Wall T，Liu Yinhui，Bhattacharya S.A Scoping Study on Oxy-CFB Technology as an Alternative Carbon Capture Option for Australian Black and Brown Coals.Melbourne：Monash University，2012.

［2］Anthony E J.Oxyfuel CFBC：status and anticipated development.Greenhouse Gases-Science and Technology，2013，3（2）：116-123.

［3］De Diego L F，Rufas A，García-Labiano F，et al.Optimum temperature for sulphur retention in fluidised beds working under oxy-fuel combustion conditions.Fuel，2013，114：106 -113.

［4］Nsakala N Y，Liljedahl G N，Turek D G.Commercialization development of oxygen fired CFB for greenhouse gas control. Bethlehem，PA：ALSTOM Power Inc，2007.

［5］Beisheim T，Zieba M，Scheffknecht G.Experimental investigations on deposit formation by recarbonation on external heat exchanger surfaces in oxyfuel CFB boilers. Energy Procedia，2013，37：1283-1290.

［6］Wang Chunbo，Jia Lufei，Tan Yewen，et al.Carbonation of fly ash in oxy-fuel CFB combustion.Fuel，2008，87（7）：1108-1114.

［7］Zheng Zhimin，Wang Hui，GuoShuai，et al.Fly ash deposition during oxy-fuel combustion in a bench-scale fluidized-bed combustor.Energy＆Fuels，2013，27（8）：4609-4616.

［8］Wang Hui，Zheng Zhimin，GuoShuai，et al.Investigation of the initial stage of ash deposition during oxy-fuel combustion in a bench-scale fluidized bed combustor with limestone addition.Energy＆Fuels，2013，28（6）：3623-3631.

［9］Shimogori M，Mine T，Ooyatsu N，et al.Effects of fine ash particles and alkali metals on ash deposition characteristics at the initial stage of ash deposition determined in 1.5 MWth pilot plant tests.Fuel，2012，97：233-240.

［10］Luan Chao，You Changfu，Zhang Dongke.An experimental investigation into the characteristics and deposition mechanism of high-viscosity coal ash.Fuel，2013，119：14-20.

［11］Li Gengda，Li Shuiqing，Xu，Xiaoguang，et al.Dynamic behavior of biomass ash deposition in a 25 kW onedimensional down-fired combustor.Energy＆Fuels，2014，1（28）：219-227.

［12］Li Gengda，Li Shuiqing，Dong Ming，et al.Comparison of particulate formation and ash deposition under oxy-fuel and conventional pulverized coal combustions.Fuel，2013，106：544-551.

［13］Bäfvera L S，Rönnbäcka M，Lecknerb B，et al.Particle emission from combustion of oat grain and its potential reduction by addition of limestone or kaolin.Fuel Processing Technology，2009，90（3）：353-359.

［14］Wu Yinghai，Wang Chunbo，TanYewen，et al.Characterization of ashes from a 100 kWthpilot-scale circulating fluidized bed with oxy-fuel combustion.Applied Energy，2011，88（9）：2940-2948.

［15］Qu Chengrui，Zhao Changsui，Zhou Wu，et al.Emission properties of PM2.5 derived from CFB under O2／CO2atmosphere.Proceedings of the Combustion Institute，2011，33：2829-2835.

［16］CarlosL，Fabrizio S，Piero S，et al.Primary fragmentation of limestone under oxy-firing conditions in a bubbling fluidized bed.Fuel Processing Technology，2011，92：1449-1456.

［17］Baxter L L.Ash deposition during biomass and coal combustion：A mechanistic approach.Biomass and Bioenergy，1993，4（2）：85-102.

［18］Yu Dunxi，Morris W J，Erickson R，et al.Ash and deposit formation from oxy-coal combustion in a 100 kW test furnace.International Journal of Greenhouse Gas Control，2011，5：S159-S167.

［19］Sathyanarayanarao Subbarao K K，Rindt C C M，van Steenhoven A A.Growth rates of dry particulate fouling under variable process conditions.Proceedings of the International Conference on Heat Exchanger Fouling and Cleaning.Crete Island，2011.210-218.

［20］Fryda L，Sobrino C，Cieplik M，et al.Study on ash deposition under oxyfuel combustion of coal／biomass blends.Fuel，2010，89：1889-1902.

TK224.9

：

：1005-9113（2015）05-0078-07

10.11916／j.issn.1005-9113.2015.05.012

2015-03-18.

Sponsored by the Foundation for Innovative Research Groups of the National Natural Science Foundation of China（Grant No.51421063），and the Key Technologies Research and Development Program of China（Grant No.2012BAA02B01-04），and the Collaborative Innovation Center of Clean Coal Power Plant with Poly-generation.

∗Corresponding author.E-mail：wanghui_hb＠hit.edu.cn.