An evaluation on rice husks and pulverized coal blends using a drop tube furnace

An evaluation on rice husks and pulverized coal blends using a drop tube furnace

(Parte 1 de 2)

An evaluation on rice husks and pulverized coal blends using a drop tube furnace and a thermogravimetric analyzer for application to a blast furnace

Wei-Hsin Chen*, Jheng-Syun Wu Department of Greenergy, National University of Tainan, Tainan 700, Taiwan, ROC article i nf o

Article history: Received 28 November 2008 Received in revised form 5 June 2009 Accepted 10 June 2009 Available online 17 July 2009

Keywords: Biomass and rice husk Pulverized coal injection (PCI) Blast furnace Drop tube furnace (DTF) Thermogravimetric analysis (TGA) Synergistic effect abstract

To evaluate the potential of pulverized coals partially replaced by rice husks used in blast furnaces, thermal behavior of blends of rice husks and an anthracite coal before and after passing through a drop tube furnace (DTF) was investigated by using a thermogravimetry (TG). For the blends of the raw materials in the TG, fuel reaction with increasing temperature could be partitioned into three stages. When the rice husks were contained in the fuel, a double-peak distribution in the first stage was observed, as a consequence of thermal decompositions of hemicellulose, cellulose and lignin. A linear relationship between the char yield and the biomass blending ratio (BBR) developed, reflecting that synergistic effects in the pyrolytic processes were absent. This further reveals that the coal and the rice husks can be blended and consumed in blast furnaces in accordance with the requirement of volatile matter contained in the fuel. After the fuels underwent rapid heating (i.e. the DTF), a linear relationship from the thermogravimetric analyses of the unburned chars was not found. Therefore, the synergistic effects were observed and they could be described by second order polynomials. When the BBR was less than 50%, varying the ratio had a slight effect on the thermal behavior of the unburned chars. In addition, the thermal reactions of the feeding fuels and of the formed unburned chars behaved like a fingerprint. 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Pulverized coal injection (PCI) is an important technique in the performance of blast furnaces and it has been widely applied in ironmaking processes for reducing the cost of hot metal production [1]. This arises from the fact that the price of coal is lower than that of metallurgical coke. For this reason, one of the essential targets for producing hot metal from blast furnaces is to promote the pulverized coal injection rate [2,3]. As a matter of fact, PCI coal is consumed directly without going through the cokemaking plant.

Hence PCI enables us to reduce CO2 emissions and prolong the coke oven life [4]. In other words, PCI is considered to be environmen- tally friendly compared to the ironmaking process only using coke. Though PCI possesses the aforementioned merits, coal pertains to fossil fuels after all and it is not a renewable energy, therefore CO2 emissions are still inevitable.

Currently, on account of a remarkable fluctuation in oil price and an increase in global temperature, people are deeply aware of the importance of the reduction of CO2 emissions and the development of alternative fuels [5–7]. Reviewing recently developed renewable energy technologies, biomass has received a great deal of attention as a sustainable fuel source. It is known that plants absorb CO2 when theygrow. If the biomass is burned for the purpose of getting heat, power and electricity, the same amount of CO2 will be emitted into the atmosphere. Accordingly, biomass is considered as a renewable and carbon-neutral fuel [8,9]. By virtue of the chemical energy contained in the biomass, it is especially applicable to the processes concerning heat treatment such as gasification and combustion used in power plants [10].

In the past, in order to understand the possible applications of biomass blended with coals in industries, a number of studies concerning co-firing [9,1,12], co-gasification [13–15] and co-pyrolysis [16–19] of blended fuels of coal and biomass have been carried out. For example, in the study of Biagnini et al. [1] concerning the phenomena of co-firing, it was reported that the coal didn’t have a significant effect on the primary reactions of thermal decomposition of biomass fuels. Similarly, the coal did not seem to be influ- encedbythereleaseofvolatilematterfrombiomass.RegardingNOx emissions, Liu et al. [12] stated that lower NOx emissions might be found during co-combustion. This arises from that fact that fuel-N that existed in biomass was predominantly liberated as NH3, which could lead to the formation of NOx and but also acted as a reducing

agent in further reactions with NOx to form N2.

Contents lists available at ScienceDirect

Energy journa l homepage: w.else m/lo cate/energ y

When co-pyrolysis of coal and biomass was considered, Pan et al. [16] carried out a number of experiments of TGA to realize the pyrolytic behavior of pine chips, black coal, waste coal and several biomass-coal blends. It was pointed out that the thermal behavior of the waste coal was poor compared with pine chips and even blackcoal.Whenaminimumof 40%ofpinechipswasblendedwith 60% of waste coal and reacted under the same conditions, the rate of weight loss increased sharply and was similar to that of black coal. However, no interaction took place between biomass and coal in a blend during pyrolysis. Kastanaki et al. [17] used a TG to study co-pyrolysis of coal and four biomass materials. As a result, no substantial interactions were also observed in the solid phase when the coal and the biomass materials were blended. As a consequence, kinetic modeling of the blends could be successfully performed using the kinetic parameters obtained for the separate blend’s components.

The aforementioned studies related to the reactions of blending coal and biomass fuels have provided much information to aid in facilitating the understanding, manipulation and performance of boilers using biomass. In contrast, relatively little research has been performed on blast furnaces using biomass as a fuel. It is known that rice is a very important food and a large amount of rice is grown in many countries. For example, in Taiwan, 783,413 tons of rice were produced in 2007. This implies that the disposal or treatment of the derived rice husks is a notable environmental issue. If the rice husks are utilized in blast furnaces as an alternative fuel, both the imported fuels and CO2 emissions can be reduced to a certain extent. For this reason, the present study is intended to explore the thermal decomposition of a pulverized coal mixed with rice husks at various blending ratios. A drop tube furnace will be employed to produce unburned chars after the blending fuels experiencing rapid heating. The obtained results can provide a useful insight into the application of rice husks in blast furnaces.

2. Experimental

Fig. 1 shows the schematic of pulverized coal injection and internal structure of a blast furnace around the raceway. When coal particles are injected into the blast furnace, they will undergo rapid heating, devolatilization, gas-phase combustion as well as char combustion and gasification [20]. The characteristic times of heating, devolatilization and gas-phase combustion are very short so that they are mainly achieved in the raceway. In contrast, the char combustion and gasification pertain to heterogeneous reactions, rendering that their characteristic times are long compared to the reactions mentioned above [21]. Accordingly, the char combustion and gasification mainly develop near the bird’s nest. In the present study, the synergistic effects of the blended fuels of coal and biomass will be investigated through the TGA to understand the potential of rice husks used in blast furnaces. Similarly, the synergistic effects of unburned char stemming from the fuel’s reactions in a DTF will also be examined through the TGA to realize their phenomena occurring in the bird’s nest.

2.1. Properties and preparation of samples

It is known that the volatile matter and moisture of biomass is generally higher than that of coal [2]; hence, the heating value of biomass is relatively low compared with that of coal. In order to provide a comprehensive study on the blends of coal and biomass, an anthracite coal with very low volatile matter was chosen. With regard to the biomass, rice husks were selected as the fuel to be blended with the coal, as described before. The properties of the coal and the rice husks, such as proximate and ultimate analyses as well as higher heating value, are given in Table 1. As shown in the table, the volatile matter and higher heating value of the coal are 6.32% and 27,773 kJ/kg, respectively, whereas they are 78.93% and 15,841 kJ/kg in the rice husks. It is also noted that rice husks have significantly less fixed carbon and more nitrogen compared to that of the coal. To simulate the situation of fuels used in the raceway of the blast furnace, the coal and the rice husks were individually grounded. The particle sizes of the fuels were controlled between 100 and 200 mesh (74-149 m) in that these particle sizes were frequently employed in PCI. The samples were then mixed in accordance with designed biomass blending ratios (BBRs). In the present study, five BBRs of 0%, 25%, 50%, 75% and 100% were taken into account.

2.2. Thermogravimetric analysis (TGA)

When investigating the characteristics of coal reaction, the devices of thermogravimetry (TG) [19,23] and drop tube furnace (DTF) [24,25] are commonly employed. Corresponding to the TG and DTF the heating rates are approximately in the orders of 10 Kmin 1 [4] and 104-105 K sec 1 [23], respectively. Considering the thermal behavior of the coal, rice husks and unburned char, they were analyzed by means of a TG (PerkinElmer Diamond TG/ DTA). The functions of the TG are to measure and record the

Fig. 1. Schematic of pulverized coal injection and internal structure of a blast furnace around raceway.

dynamics of weight loss as well as endothermic and exothermic reactions with increasing temperature or time. For all the experimental runs, around 5 mg of sample was used. The heating rate of the TG and the flow rate of the carrier gas were 20 Kmin 1 and 500 ccmin 1, respectively. The heating temperature ranged from 25 C to 1400 C. From the recorded distribution of the weight loss, one was able to obtain the thermogravimetric analysis (TGA), derivative thermogravimetric (DTG) analysis and differential thermal analysis (DTA).

2.3. Drop tube furnace (DTF)

Ithasbeenillustratedthattheheatingrateforcoalparticlesinthe raceway is close to that in a DTF [4]. In order to figure out the characteristicsofunburnedcharstemmingfromthereactionsofthecoal and rice husk blends in the raceway, a reaction system along with aDTFwasconducted tosimulatethe reactionprocessesfor thefuels suddenly exposed to a high-temperature environment in the blast furnace. The schematic of the reaction system is shown in Fig. 2.I n brief, the system included a feeding unit, a reactor (viz., the DTF), a particle collection unit and a particle analysis unit. In the feeding unit, fuel was sent bya screw feeder and a carrier gas (N2), with the flow rate of 2000 ccmin 1, was blown to help the transport of particlesintothereactor.Atthesametime,apreheatedgas(N2)with the flow rate of 2000 ccmin 1 also flew into the reactor. The temperature of the preheated gas was controlled at 350 C. The reactor was heated by a SiC heating element, which could heat the furnace up to 1650 C. In the present study, all the cases were performedat thereaction temperature of 1400 C, whichisa typical temperature in the raceway. After the particles passed through the reactor,theunburnedcharparticleswerecapturedviaacyclone.The collected particles were then analyzed in the TG and observed by means of a scanning electron microscope (SEM, Hitachi S4800).

2.4. Measurement quality control

To ensure the measurement quality, before the thermogravimetricanalyseswereperformed,purenitrogenwassentintotheTG for at least 10 min to completely purge the chamber. During this periodtheequilibriumofthebalancemountedinthechambercould beestablishedaswell.AlltheexperimentsintheTGwereperformed over twice and it was found that the obtained results were similar, revealing that the measured results from the TG were reliable. In regardtotheDTF,beforetheexperimentswereperformed,nitrogen with a fixed flow rate was blown into the reaction system and then the flow rate of nitrogen was measured at the system exit. This guaranteed that no gas leakage occurred in the system. In addition, after the fuels undergoing the DTF, ash contained in the unburned char was analyzed. As a result, the relative errors were within 3%, indicating the good reproducibility of the measurements.

3. Results and discussion 3.1. Thermal behavior of raw materials

Thermogravimetric analyses (TGA) and derivative thermogravimetric (DTG) analyses of the fuels at three different BBRs (i.e. 0%,

Fig. 2. Schematic of the reaction system and drop tube furnace.

Table 1 Properties of the investigated coal and biomass.

Fuel Coal (Anthracite) Biomass (rice husks)

50% and 100%) are presented in Fig. 3. As awhole, the TGA curves of the three fuels indicate that the weight-loss processes can roughly be divided intothree stages. The temperature ranges of the first, the second and the third stages are T 500 C, 500 C <T 1000 C and T >1000 C, respectively. For the case of 0% of BBR, Fig. 3a illustrates that the weight of the coal drops from 100% to 96% in the first stage. With increasing BBR, the weight loss in the first stage tends to grow. Specifically, for the BBRs of 50% and 100%, the weights of the fuels decay from 100% to 54% (Fig. 3b) and to 19% (Fig. 3c), respectively. This is attributed to the intrinsically higher volatile matter contained in rice husks, as shown in Table 1. In the second stage, it is noted that, whatever the BBR is, the declining tendency in weight with increasing temperature is relatively slight.

This implies that the fuels are insensitive to the variation of temperature in the second stage. Once the heating temperature goes beyond 1000 C, the trend of weight loss becomes obvious compared to that in the second stage, especially for the BBR of 0%.

In examining the DTG curve shown in Fig. 3a, the weight loss in the first stage is a result of releases of moisture and volatiles, and the local maximum reaction rate occurs at around 360 C. In the second stage, the values in the DTG curve are very low. It is thus known that the coal becomes inert in such thermal environments. In the third stage, seeing that the temperature is sufficiently high, some cracks may happen on coal particles, which results in liberating heavier hydrocarbons or other components with higher bond energy from the coal particles interior [21]. As a consequence, the DTG curve goes up when the temperature rises. With attention shifted to the DTG curves shown in Figs. 3b and 3c, the behavior of weight loss in the first stagebecomes morepronounced. The higher the BBR, the more significant the weight drop is. Unlike only the peripheral parts and the mobile phase of the macromolecular structure, consisting mainly of the polycyclic aliphatic substances [26], to be decomposed from the coal, the main components in the biomass comprises polymers of hemicellulose, cellulose and lignin [18,27,28] and these compositions are decomposed in lower temperatures. This is the reason that increasing BBR substantially increases weight loss in the first stage. On the other hand, it is noteworthy that in the first stage a double-peak distribution is exhibited if the biomass is mixed with coal (Figs. 3b and 3c). It has reported [29,30] that hemicelluloses decomposes mainly between 150 and 350 C, cellulose decomposes between 275 and350 C, and lignin gradually decomposes between 250 and 500 C. In contrast to the distributions shown in Figs. 3b and 3c. it is known that the first peak (T¼217 Cw227 C) induced stems from the thermal decomposition of hemicellulose. Alternatively, the second peak (T¼316 Cw318 C) developed is due to the decomposition of cellulose. The tail behind the second peak corresponds to the decomposition of lignin and the lignin decomposes overlapping the cellulose [28].

3.2. Synergistic effects and thermal reactions of blended fuels

Char yield from a TGA play an important role in accounting for synergistic effects from blending fuels [19,26]. The influences of the BBR on the char yield at the end of each of the three stages (i.e. T¼500 C,1000 C and 1400 C) are demonstrated in Fig. 4, where five BBRs (viz. 0%, 25%, 50%, 75% and 100%) are taken into consideration. Upon inspection of the profiles shown in Fig. 4,i ti s observed that increasing the BBR results in decreased char production and strong linear distributions between the char yield and the BBR are established in that the correlations are high (R2S0.985). Physically, this means that no synergistic effect or chemical interaction takes place when the coal and the rice husks are simultaneously heated. It is known that the volatile matter of the anthracite is fairly low. Therefore, from the viewpoint of fuel used in a blast furnace, the lack of synergistic effect implies that the coal can be blended with the biomass at any BBRs in accordance with the requirement of volatile matter contained in the fuel. In other words, the gas-phase combustion in the raceway can be improved if the biomass is blended with the pulverized coal utilized in PCI.

Despite the lack of a synergistic effect, it should be emphasized that when the BBR is 50%, the char yields are always lower than the linear regression curves (see Figs. 4a, 4b and 4c). That is, the blended fuel with 50% of BBR is characterized by a higher degradation rate or reaction activity. The above observations are consistent with the experimental results of Vuthaluru [18] where coal was individually blended with wood waste and wheat straw.

Fig. 3. Thermogravimetric and derivative thermogravimetric analyses of the raw materials at the biomass blending ratios of (a) 0%, (b) 50% and (c) 100%.

When the DTA curves of the blended fuels with the BBRs of 0% (Fig. 5a), 50% (Fig. 5b) and 100% (Fig. 5c) are examined, as a whole, the thermal reactions proceed fromaweaklyendothermic reaction, a weakly exothermic reaction and then to a strongly endothermic reaction with increasing temperature. It is apparent that, when the BBR is 50%, the exothermic reaction is somewhat stronger than that of the other BBRs, whereas the intensities of the endothermic reactions of the former are smaller than that of the other two cases.

It is the stronger exothermic reaction and weaker endothermic reactions that result in the blended fuel with 50% of BBR being featured by a higher degradation rate. Moreover, it can be seen that thepeak of thestrongendothermicreaction shiftstoward the lower temperature as the BBR increases.

3.3. Thermal behavior of unburned char

After the fuels undergo rapid heating in the DTF, the TGA and DTG curves of the unburned char are displayed in Fig. 6. Seeing that

Fig. 4. Char yield distributions with respect to the biomass blending ratio at the ends of (a) the first stage, (b) the second stage and (c) the third stage.

Fig. 5. Differential thermal analyses of the raw materials at the biomass blending ratios of (a) 0%, (b) 50% and (c) 100%.

the weight loss characteristics of the unburned char are inherently different from that of the raw materials, 400 C rather than 500 C is selected as the end of the first stage. A comparison with Fig. 3a, Fig. 6a depicts that the weight loss in the first stage, stemming from volatile release, is not significant. Undoubtedly, this is attributed to that the volatiles contained in the unburned char are less than that of the coal. The three-stage characteristic is still exhibited. With the BBR of 50%, only one peak developing at 322 C is observed in Fig. 6b. The foregoing behavior may result from that the reaction time of the feeding fuel in the DTF is not sufficiently long so that not all the cellulose is depleted. However, unlike the double-peak obtained in Fig. 3b, the peak of the hemicellulose almost disappears in the unburned char reaction. Furthermore, when the unburned char with the BBR of 100% is examined, Fig. 6c depicts that, in addition to the cellulose, little amounts of hemicellulose is retained as well. It is also noted that the weight of the unburned char drops from 100% to 46% in the first stage, implying that a large portion of the polymers of hemicellulose, cellulose and lignin is still kept in the unburnedchar. This results frommorevolatiles being contained in the feeding fuel where the volatiles are not liberated completely.

Fig. 6. Thermogravimetric and derivative thermogravimetric analyses of the unburned chars at the biomass blending ratios of (a) 0%, (b) 50% and (c) 100%.

Fig. 7. Char yield distributions with respect to the biomass blending ratio at the ends of (a) the first stage, (b) the second stage and (c) the third stage.

3.4. Synergistic effect and thermal reaction of unburned char

Based on the TGA curves obtained in Fig. 6, synergistic effects of the unburned char from various BBRs are examined in Fig. 7. Figs. 7a, 7b and 7c represent the char yields at 400 C,1000 C and 1400 C, respectively. As can be seen in Fig. 7, the distributions of char yield versus the BBR are not characterized by a linear relationship any more. Rather, the synergisticeffects canbe expressed bysecond order polynomials. As awhole, no matter whichstage isconsidered, the distributions can be partitioned into two regimes. When the BBR is less than 50%, an increase in BBR almost has no effect on the char yield. In consequence, in the first regime (BBR 50) the activity of the unburned char is insensitive tothe BBR. In the second regime (BBR> 50), it is obvious that the char yield decays rapidly with increasing the BBR. It is inferred that the polymers of hemicellulose, cellulose and lignin are relatively rich in the feeding fuels so that a portion of the polymers still remain, even though the fuels pass through the DTF. Consequently, the char yield is sensitive to the variation of the BBR.

(Parte 1 de 2)