- Email: [email protected]
Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter
Journal of the Energy Institute xxx (2015) 1e10
Contents lists available at ScienceDirect
Journal of the Energy Institute journal homepage: http://www.journals.elsevier.com/journal-of-the-energyinstitute
Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter Adisak Pattiya*, Suntorn Suttibak Bio-Energy and Renewable Resources Research Unit, Faculty of Engineering, Mahasarakham University, Kamriang, Kantharawichai, Maha Sarakham 44150, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 July 2015 Received in revised form 29 September 2015 Accepted 6 October 2015 Available online xxx
Sugarcane residues e leaves (SL) and tops (ST) from Thailand were pyrolysed in a fluidised bed reactor incorporated with a hot filter. The aim was to investigate the effects of reaction temperatures and hot filtration on pyrolysis products. The bio-oils were characterised by water, solids and ash contents, density, heating value, pH, viscosity and stability. The optimum pyrolysis temperatures for SL and ST were 429 C and 403 C, which gave maximum bio-oil yields of 52.5 wt% and 59.0 wt%, respectively. Using the hot filter reduced bio-oil yield by 7e8 wt%. However, the filtered bio-oils had better viscosity, solids and ash contents and stability. © 2015 Energy Institute. Published by Elsevier Ltd. All rights reserved.
Keywords: Sugarcane leaves Sugarcane tops Bio-oil Fast pyrolysis Hot vapour filtration
1. Introduction It is generally accepted that the fuel shortage and the environmental problems are one of the major crises that need urgent attention. A solution to these problems is the exploitation of renewable energy technology such as fast pyrolysis. Fast pyrolysis is mostly applied to biomass to convert it to a liquid form, termed “bio-oil”. Many researchers studied the production of bio-oil from various types of biomass by many fast pyrolysis reactor configurations. The bio-oil yield could be as high as 80 wt% on dry biomass feed [2] depending on the type of biomass and the reactor unit. The lower heating value (LHV) of bio-oil is typically about 14e18 MJ/kg [2]. Bio-oil can be used as an alternative fuel in furnaces and engines to produce heat and power. In addition, bio-oil can be a raw material for chemical production. Recently, bio-oil has been proposed as a feed for biorefineries to optimise the production of fuels and chemicals to get the highest value out of the bio-oil in social, environmental and economic aspects [1]. Sugarcane is a perennial plant of the genus Saccharum. The main sources of sugarcane are Brazil, India, China, Thailand, Pakistan and Mexico. Typically, sugarcane is grown for sugar production. Lately, sugarcane has become an ethanol production feedstock. The demand for sugar and ethanol is increasing. Therefore, it is expected that the residues will increase accordingly. Residues from sugarcane plantations include bagasse, leaves and tops. The bagasses are traditionally used as a fuel in the sugar factories to generate electricity for their own use. The leaves and tops are mostly burnt in the fields and are not efficiently used for energy. In Thailand, some farmers burn the leaves before harvesting the crops, leading to air pollution. Only a small portion of the leaves and tops are used as a compost and animal feed. Globally, sugarcane production is ~2 109 Tonnes per year, whereas in Thailand the production is ~1 108 Tonnes per year, or 5% of the global production [6]. Assuming a crop to residue ratio of 0.302 for sugarcane residues (leaves and tops) [15], the residues would be ~6 108 Tonnes per year or 96 MTOE (Million Tonnes of Oil Equivalent). This would cost ~$US75,000 million based on a $107/barrel crude oil price. Therefore, fast pyrolysis technology applied to sugarcane leaves and tops for bio-oil production would have two advantages e the fuel value and the environmental impact. Recently Xu et al. [21], investigated bio-oil production by flash pyrolysis of sugarcane residues using a bubbling fluidised bed pyrolyser with a focus on the post treatments of the aqueous phase. They studied the effect of pyrolysis temperature and vapour residence time and
* Corresponding author. Tel.: þ66 43 754321x3036; fax: þ66 43 754316. E-mail address: [email protected] (A. Pattiya). http://dx.doi.org/10.1016/j.joei.2015.10.001 1743-9671/© 2015 Energy Institute. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
2
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
found that maximum liquid yields were for sugarcane bagasse e 58.1%, external e 52.6% and whole plant e 55.5% which were obtained at the pyrolysis temperature of 400 C. Although Xu et al. [21] pyrolysed sugarcane residues in a fluidised bed reactor, they did not fully examine the properties of the liquid product. Only the heating value and water content were used to calculate the energy contained in the bio-oil. Subsequently, in 2012, our group reported the fast pyrolysis of sugarcane and cassava residues in a free-fall reactor [16]. The sugarcane residues bio-oils were measured for water content, solids content, pH, specific gravity, elemental composition, heating value and chemical composition by GC/MS. Nevertheless, one of the most important properties of bio-oil e its stability e was not mentioned due to the small quantity of bio-oil produced. Therefore, the current paper continues the investigation of fast pyrolysis of sugarcane agricultural residues by assessing the impacts of reactor types and in-situ upgrading by hot vapour filtration. Hot filtration can be achieved by candle, granular or fixed bed filters. Previous studies applied candle filters to fluidised bed reactor for production of bio-oil from rice straw [7,9], radiata pine [8,14], Bamboo sawdust [7] and Oriental white [13]. A granular filter is another type that has been tested for bio-oil upgrading [4,12,20]. A hot filter based on a fixed bed of glass wool was also tested with a fluidised-bed fast pyrolyser using rice straw and husk [17] and cassava rhizome [18] and found that the bio-oil solids content could be significantly reduced, although the filter design is relatively simple. In this work, a fluidised bed reactor was used and the products yields and properties were compared with previously reported results using a free fall reactor [16]. This work also assessed the influence of a fixed-bed hot filter on bio-oil properties, especially the solids content, viscosity and storage stability. 2. Experimental 2.1. Biomass feedstock Biomass samples were sugarcane leaves (SL) and tops (ST) from plantations in north-east Thailand. The samples were sundried, ground and sieved to a 250e425 mm range particle size. Prior to experiments, they were dried in an oven at 105 C for 24 h to reduce the moisture content to below 10 wt%. Drying the samples also eased the feeding and could reduce the bio-oil water content. SL and ST samples were tested for their basic properties including proximate and ultimate analyses as well as heating value. The proximate analysis determined the moisture, volatile matter, fixed carbon and ash contents according to ASTM (E1756-01, E872-82 and E1755-01). The ultimate analysis determined carbon (C), hydrogen (H), nitrogen (N), sulphur (S) and oxygen (O) contents using a ‘Leco CHN/ S Determinator’ analyser according to ISO/IEC Guide 22 and EN 45014 standards at the Laboratory Equipment Center, Mahasarakham University, Thailand. The heating values were calculated based on the ultimate analysis results and equations (1) and (2). The higher heating value (HHV) of biomass was calculated from a correlation developed by Sheng and Azevedo [19]:
HHV
MJ kg
¼ 1:3675 þ 0:3137C þ 0:7009H þ 0:0318O
(1)
where C, H are percentages on dry basis of carbon, hydrogen, respectively and O* is 100-C-H-Ash. The lower heating value (LHV) was calculated from HHV and the hydrogen content:
LHV
MJ kg
H ¼ HHV 21:82171 100
(2)
The biomass analysis results are summarised in Table 1. 2.2. Fast pyrolysis apparatus Fast pyrolysis of biomass was carried out in a fluidised bed reactor unit. The unit consisted of a pre-heater, a biomass hopper, a twostaged feeder, a fluidised bed reactor, two cyclone separators, a hot filter and a bio-oil product collection unit (see Fig. 1). The reactor Table 1 Characteristics of agricultural residues from sugarcane plantations in Thailand. Analysis Proximatea (wt%) Moisture Volatile matter Fixed carbonb Ash Ultimate (wt%, dry, ash-free basis) Carbon (C) Hydrogen (H) Nitrogen (N) Sulphur (S) Oxygenb (O) H/C molar ratio O/C molar ratio Molecular formula Heating value (MJ/kg, dry basis) HHV LHV a b
Sugarcane leaves
Sugarcane tops
3.5 76.5 16.1 7.4
2.7 78.5 14.6 6.9
50.8 5.0 0.4 0.2 43.6 1.2 0.6 CH1.19O0.64
48.6 4.7 0.6 0.2 45.8 1.2 0.7 CH1.16O0.71
17.8 16.8
17.2 16.2
Moisture content is on wet basis, whereas volatile matter, fixed carbon and ash contents are on dry basis. Calculated by difference.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
3
was made from a 304 stainless-steel pipe with its internal diameter 50 mm and height 450 mm. Silica sand with a 250e425 mm particle size was used as the fluidising and heat transfer medium. The pyrolysis vapour produced was cleaned up in the cyclone separators and a hot filtration unit prior to condensation. The hot filtration unit was made from a 304 stainless-steel pipe with an internal diameter 50 mm and height 450 mm. Approximately 10 g of glass wool was loosely packed inside the filter and a K-type thermocouple was installed at the middle of the glass wool bed. The nitrogen pre-heater, pyrolysis reactor and hot filter were heated by heating wires and the temperatures were adjusted using PID controllers and monitored using K-type thermocouples. Since the heated zones were insulated with ceramic fibre blankets to minimise heat loss, the temperatures along the reactor and filter length were ±5 C. Pyrolysis vapour was condensed and captured into liquid product by a water-cooled heat exchanger, an electrostatic precipitator (ESP), a series of two dry ice/acetone condensers and a cotton wool filter. 2.3. Fast pyrolysis experiments There were three independent variables in the fast pyrolysis experiments, namely biomass types, pyrolysis temperatures and the use of hot vapour filtration unit. The biomass types investigated were sugarcane leaves and tops. Experiments were run at temperatures of 375 C, 400 C, 425 C, 450 C and 500 C. The fast pyrolysis system had 2 configurations e with and without the hot vapour filtration unit. For the without-hot-filter setup, the pyrolysis vapour exiting the second cyclone went directly to the condenser, leading to shorter vapour residence time. The dependent variables were product yields and physicochemical properties of bio-oil e water content, solids content, ash content, density, heating value, pH, viscosity and stability. The controlled variables were a nitrogen gas flow rate of 7 L/min, a biomass feed rate of ~300 g/h, a temperature at the pre-heater, cyclones, hot filter and transfer line of 420 C, a cooling water temperature of 30 C and a dry-ice/ acetone condenser temperature of 78 C and a run duration of 1 h. It is important to note that the temperature of the pre-heater, cyclones, hot filter and transfer line are controlled at 420 C because at higher than this temperature, it is likely that vapour can undergo thermal cracking leading to lower bio-oil yield. If the temperature is lower than this, the pyrolysis vapour could re-condense or re-polymerise to solid, which could also reduce the liquid yield. Typically, it took 1.5 h to preheat the system from room temperature to steady state. The run duration was timed only when biomass was fed to the system. 2.4. Mass balance calculation The main products from fast pyrolysis process are liquid bio-oil, solid char and gases. The yields of each product were calculated by weighing all parts of the fast pyrolysis system e biomass hopper, silica sand, fluidised bed reactor, cyclone separators, hot filter and product collection unit e before and after each experiment. The bio-oil yields were the combined weight of the liquid from the product collection unit. The char yields were the combined weight of the solid from the reactor, the cyclones, the hot filter, transfer line and the solids in bio-oil. The gas yields were calculated by difference. 2.5. Analysis of pyrolysis products 2.5.1. Bio-oil analysis The bio-oil was characterised by measuring water, solids and ash contents, density, pH, elemental composition, heating value, viscosity and stability. Each analysis was performed in triplicate. 2.5.1.1. Water content. Water content of bio-oil was determined by the Renewable Laboratory, National Metal and Materials Technology Center (MTEC), Pathum Thani, Thailand, using a Karl-Fischer (KF) titration technique.
Fig. 1. Schematic diagram of the fast pyrolysis unit.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
4
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
2.5.1.2. Solids content. Solids content of bio-oil was determined by vacuum filtration. The solids in bio-oil were defined as ethanol insolubles. About 2e3 g of bio-oil was dissolved in ethanol and filtered through a pre-dried and pre-weighed 6 mm mean pore size qualitative filter paper (Whatman No. 3). The liquid was then washed with excess of ethanol until the filtrate was clear to ensure that there was no organic liquid left on the paper. The filter paper with the solids was air-dried for approximately 15 min and further dried in an oven at 105 C for 30 min. Then the paper was cooled in a desiccator and weighed. This method was suggested by Oasmaa and Peacocke [11]. 2.5.1.3. Ash content. Ash content of bio-oil was determined as the amount of residue when heating bio-oil to 775 C in an oxygen atmosphere. Direct heating of bio-oil would result in foaming and splashing due to the high water contents. Thus, the first controlled evaporation of water at 105 C is needed before rapid heating to 775 C [11]. 2.5.1.4. Density. The density of bio-oil was measured using a density bottle at room temperature (about 30 C). 2.5.1.5. pH value. The pH value of bio-oil was measured with a pH metre (COMBI pH/mV/Temp Bench Metre) at room temperature. Before the measurement, the instrument was calibrated with liquid calibration standards of pH 4 and 7. 2.5.1.6. Elemental composition. The elemental composition of bio-oil was determined by using the same technique as described in section 2.1. 2.5.1.7. Heating value. Higher heating value on dry basis (HHVdry) was calculated based on the elemental analysis results using a correlation developed by Channiwala and Parikh [3] e see equation (3). The lower heating value on dry basis (LHVdry) was calculated from equation (2). For heating values on wet basis (HHVwet and LHVwet), equations (4) and (5) were used by taking into account the water content in bio-oil (H2O).
HHV dry
MJ kg
¼ 0:3491C þ 1:1783H þ 0:1005S 0:10340O 0:0211A
(3)
C, H, S, O and A in equation (3) are percentages of carbon, hydrogen, sulphur, oxygen and ash in bio-oil on dry basis.
HHV wet LHV wet
MJ kg
MJ kg
H O ¼ HHV dry 1 2 100
(4)
H O H2 O ¼ LHV dry 1 2 2:442 100 100
(5)
2.5.1.8. Viscosity. Bio-oil kinematic viscosity was measured at 40 C using a Cannon-Fenske Routine Viscometer with a SDM viscosity bath (Art. 370) according to ASTM D445 and D446. 2.5.1.9. Stability. Bio-oil stability is an important quality indicator. The stability index is defined as the change in bio-oil viscosity after accelerated ageing ðnstored Þ compared to its initial viscosity ðnfresh Þ [11]:
StabilityIndex ¼
nstored nfresh nfresh
(6)
nstored is the viscosity of bio-oil stored in an oven at 80 C for 24 h as being equivalent to bio-oil stored at room temperature for 1 year [10], whereas the nfresh is the viscosity of bio-oil within 24 h of production. 2.5.2. Char analysis The char samples obtained from cyclone collectors from the experimental runs that gave highest bio-oil yields were tested for elemental composition, heating value, ash content and particle size distribution (PSD). The methods used for determining the elemental composition were the same as those used for biomass and bio-oil analysis. The char heating values were calculated according to the equations (2) and (3). The procedure for ash content analysis was the same as that for biomass ash content, see section 2.1. The particle size distribution of char was obtained by sieving approximately 30 g of char into size ranges: 0e75, 75e212, 212e250, 250e300, 300e425, 425e500 and >500 mm. Each test sieve was weighed before and after sieving and the results were plotted to obtain the PSD.
3. Results and discussion 3.1. Effect of pyrolysis temperature on product distribution The effect of pyrolysis temperature on yields is shown by Fig. 2 e for sugarcane leaves (SL) e and Fig. 3 for sugarcane tops (ST). The setting temperatures are 375, 400, 425, 450 and 500 C, but the actual average temperatures recorded and calculated throughout the experimental runs are plotted in the graphs. The actual temperatures are within ±5 C of the setting values. It can be seen that increasing the temperature from 375 C to 500 C, decreased the char yield and increased the gas yield for both feedstocks. This trend was also reported by Pattiya et al. [16] when SL and ST were pyrolysed in a free fall reactor and by Xu et al. [21] when sugarcane residues (internal bagasse) was pyrolysed in a bubbling fluidised bed reactor. The maximum bio-oil yields when pyrolysing SL and ST using the fluidised bed reactor were 52.5 wt% at 429 C and 59.0 wt% at 403 C, respectively. When the same biomass species were pyrolysed in a free fall reactor, a similar maximum yield of Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
5
Fig. 2. Effect of temperature on product yields derived from fast pyrolysis of sugarcane leaves in a fluidised bed reactor unit without the hot vapour filter.
Fig. 3. Effect of temperature on product yields derived from fast pyrolysis of sugarcane tops in a fluidised bed reactor unit without the hot vapour filter.
53.3% was obtained at 400 C for SL [16]. However, the maximum bio-oil yield for ST pyrolysed with the free fall reactor was only 44.1 wt% at 350 C. This shows that different yields of bio-oil may be obtained if a different reactor configuration is applied, even though the same biomass type is used. This is supported by the results of Xu et al. [21] when sugarcane residues (external and whole plant) were pyrolysed in a fluidised bed reactor and 52.6e55.5 wt% of maximum bio-oil yield could be obtained at 400 C. Based on these findings on fast pyrolysis of sugarcane residues, the fluidised bed configuration is more effective than the free fall type for obtaining high bio-oil yield. This is likely related to the heat transfer to biomass particles. For a fluidised bed reactor, biomass particles were rapidly heated by direct contact with the heat transfer medium or by thermal conduction as well as by hot fluidising gas or thermal convection. The heat transfer medium usually has high surface area because of the small particle size. In comparison to the fluidised bed type, the free fall reactor heats the biomass particles mostly through the reactor internal wall. The hot wall heats the carrier gas by convection. When a biomass particle dropped into the reactor, it was heated by the direct contact with the reactor wall (conduction) and with the hot carrier gas (convection). 3.2. Effect of hot vapour filtration on the product distribution Fig. 4 shows that for both biomass types, the bio-oil yields reduced due to the use of the hot filtration unit. Bio-oil yields reduced for SL e from 53 wt% to 46 wt% e and ST e from 59 wt% to 51 wt%. This reduction occurred in conjunction with the increase in the gas yield, whereas the char yield was nearly constant. This reflected that part of the bio-oil was transformed into permanent gas when the pyrolysis vapour went through the fixed-bed hot filter. Similar findings were reported for fast pyrolysis of rice straw, rice husk, cassava rhizome and cassava stalk in fluidised bed reactors using either a fixed-bed glass wool hot filter [17,18] or a ceramic candle filter [5]. For fixed-bed glass wool hot filtration, when using paddy residues, the bio-oil yield decreased 4e7 wt% on dry biomass basis, whereas when cassava residues were Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
6
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
Fig. 4. Effect of hot vapour filtration on product yields derived from fast pyrolysis of sugarcane leaves (SL) and sugarcane tops (ST).
pyrolysed, yield decreased 6e7 wt%. Chen et al. [5] reported only 2.2 wt% decrease of bio-oil yield when applying a ceramic candle filter. Our results are consistent with those from previous studies: the use of hot vapour filtration could lead to the decrease in 2e8 wt% bio-oil yield on dry biomass basis. This is equivalent to 5e14% of bio-oil yield reduction. Since the reduction of bio-oil yield when applying the hot filter corresponds to the increase of the gas yield, it is expected that the pyrolysis vapour may undergo secondary cracking. The cracking could be thermal, catalytic or both. The thermal cracking occurred when pyrolysis vapour was exposed to high temperatures for longer time than normal conditions. When pyrolysis vapour passed through the filter, it was in direct contact with the filter medium, which may be at elevated temperature. Also, when a hot filter is installed downstream, this allows the vapour to spend more time in the hot zone prior to condensation. This time is typically called hot vapour residence time. The exposure of the vapour for long time increases the probability of thermal cracking. If the vapour is in contact with some solid particles that are catalytically active, a second type of cracking reactions (catalytic cracking) may take place. The in-situ catalysts in fast pyrolysis are, for example, alkali and alkaline earth metals such as sodium, potassium, magnesium and calcium. These catalysts exist in the form of char fines. If the char fines accumulate on the filter medium surface becoming char cake, when the vapour passes through, secondary catalytic cracking could occur, thus lowering the bio-oil yield and increasing the gas yield. 3.3. Characterisation of pyrolysis products 3.3.1. Bio-oil Table 2 summarises the properties of SL and ST bio-oils produced at their optimum pyrolysis temperatures with and without the use of the hot filter: water, solids and ash contents, density, pH value, elemental composition, heating value, viscosity and stability. 3.3.1.1. Water content. The standard specification for pyrolysis liquid bio-oil (ASTM D7544-12) states that the bio-oil water content should not exceed 30%. Table 2 shows that the majority of the bio-oil produced from sugarcane residues met this requirement except for the filtered ST bio-oil which had slightly higher water content than the standard requirement. It can also be noticed that the use of the hot filter resulted in the increase in the water contents of 11.9% for SL and 13.4% for ST. Pattiya et al. [16] reported that the water contents of SL and ST bio-oil produced by a free fall reactor with a small hot filter were 18% and 24%. These values are in the range of the filtered and non-filtered bio-oils results from the current study. For both reactor types applied, the SL bio-oils appeared to have lower water content than the ST ones. 3.3.1.2. Solids content. Following ASTM D7544-12, the pyrolysis liquid biofuels should contain no more than 2.5 wt% for grade G and 0.25 wt % for grade D. Grade G is intended for use in industrial burners, whereas grade D is for use in commercial or industrial burners requiring low solids and ash content. Table 2 shows that without the hot filter, the SL and ST bio-oil solids content was 3.1e3.5 wt%, failing the ASTM requirement. With the hot filter, the solids content reduced to 0.8 wt%, equivalent to 74e77% solids reduction, thus meeting the Grade G requirement. This reduction is equivalent to 74e77%. The solids content of SL and ST bio-oils produced by the free fall reactor was 2.9e3.7 wt % [16]. In addition, the solids content of cassava and rice residue bio-oils produced by the fluidised bed reactor was 3.3e3.9 wt% without the hot filter and 0.5e0.8 wt% with the hot filter [17,18]. 3.3.1.3. Ash content. The ash content of SL and ST bio-oils was reduced from 0.3-0.4 wt% to 0.01 wt% by the aid of the filter. The ASTM D754412 standard requirements for the ash content are 0.25 wt% for grade G and 0.15 wt% for grade D biofuels. Therefore, our results show that only the filtered bio-oil meets this requirement. The current bio-oils had much lower ash content than those produced in the free fall reactor since the ash content of ~1 wt% was reported [16]. 3.3.1.4. Density. The density of all bio-oils produced in this work was 1.1 g/ml. This is the same as that produced by the free fall reactor [16]. Therefore, it is apparent that neither the reactor configuration nor the hot filtration affected bio-oil density. The density of pyrolysis liquid biofuels is expected to be in the range of 1.1e1.3 kg/m3 according to the ASTM D7544-12. The current results are just sufficient to meet the lower end of the standard. 3.3.1.5. pH value. The pH value of bio-oils produced in this work was in the range of 3.0e4.2. This is similar to the pH value range (3e4) of the bio-oils produced from the free fall reactor [16] using the sugarcane residues or from the fluidised bed reactor using cassava and rice Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
7
Table 2 Properties of bio-oil obtained from pyrolysis of sugarcane leaves and sugarcane tops. Feedstock
Sugarcane leaves With hot filter
432 Pyrolysis temperature ( C) Elemental composition (wt%, dry, ash-free basis) Carbon 65.1 Hydrogen 7.2 Nitrogen 1.0 a Oxygen 26.7 H/C molar ratio 1.3 O/C molar ratio 0.3 Molecular formula CH1.34O0.31 Water content (wt%) 23.6 Density (g/ml) 1.1 Heating value by calculation method (MJ/kg) HHV (water-free basis) 28.5 HHV (as-produced basis) 21.7 LHV (water-free basis) 26.9 LHV (as-produced basis) 20.0 pH value 3.0 Kinematic viscosity @ 40 C (cSt) Fresh bio-oilb 16.3 Aged bio-oilc 50.6 Solids content (wt%) 0.8 Ash (wt%) 0.01 Stability index 2.1 a b c
Sugarcane tops Without hot filter
With hot filter
Without hot filter
429
404
403
59.9 7.7 0.9 31.1 1.5 0.4 CH1.55O0.39 11.7 1.1
71.3 5.9 1.0 21.7 1.0 0.2 CH1.00O0.23 32.4 1.1
62.9 6.6 0.6 29.6 1.3 0.4 CH1.26O0.35 19.0 1.1
26.7 23.6 25.1 21.9 3.0
29.6 20.0 28.3 18.3 4.2
26.6 21.6 25.2 19.9 3.8
23.2 705.4 3.5 0.41 29.5
2.8 6.5 0.8 0.01 1.3
19.7 455.2 3.1 0.30 22.1
Calculated by difference. Fresh bio-oil is the one produced within 24 h. Aged bio-oil is bio-oil stored in an oven at 80 C for 24 h.
residues [17,18] and slightly higher than typical wood-derived bio-oil pH of 2.5 [1]. Since the ASTM D7544-12 specification for pyrolysis liquid biofuel does not mandate a minimum pH, the current bio-oils are acceptable with respect to pH. 3.3.1.6. Elemental composition. Table 2 shows that ST bio-oils contained higher carbon and lower hydrocarbon and oxygen contents than the SL bio-oils. The same trend for carbon and oxygen contents was reported when processing these biomass samples in the free fall pyrolyser [16]. The carbon contents of ST and SL bio-oils were 65.4 wt% and 60.0 wt%, respectively, whereas the oxygen content was 24.2 wt% for ST bio-oil and 29.5 wt% for SL bio-oil [16]. The difference in the carbon and oxygen contents of the two bio-oils was even higher for hot filtered bio-oils. In other words, the use of the hot filtration unit led to an increase of the carbon content and a reduction of the hydrogen and oxygen content e see Table 2. This implies that deoxygenation reactions were involved in the secondary cracking of the pyrolysis vapour while passing through the fixed bed of glass wool filter. The changes in the elemental composition of bio-oil when applying the hot filter could also be noticed when pyrolysing different biomass species. The current findings are consistent with those using cassava stalk as raw material [18]. However, the opposite trend was observed when cassava rhizome, rice straw and rice husk were pyrolysed as carbon content was reduced and oxygen content was increased when the hot filter was equipped with the fluidised bed reactor. Therefore, the changes of the elemental composition with the hot filtration are dependent on biomass type as well. Based on the elemental composition, the H/C and O/C molar ratios could be calculated and the molecular formulae of the bio-oils were generated. When comparing the ratios and the formulae for sugarcane residues bio-oils produced in this work with those of the bio-oil previously produced from Thai agricultural residues such as cassava stalk, cassava rhizome rice straw and rice husk applying both the fluidised bed reactor with and without the hot filter and the free fall reactor [16e18], H/C ratios of 0.92e1.58 and O/C ratios of 0.14e0.71 were found. On average, the molecular formula for Thai agricultural residues bio-oil is CH1.3O0.3. 3.3.1.7. Heating value. The heating values of SL and ST bio-oils produced with and without the hot filter are summarised in Table 2. Both higher heating value (HHV) and lower heating value (LHV) are shown on water-free and as-produced bases. The water-free basis is important for understanding the heating value of the organic part or non-aqueous part of the bio-oil. According to the ASTM D7544-12, the minimum gross heat of combustion, or HHV (on as-produced basis), for pyrolysis liquid biofuel is 15 MJ/kg. Since the HHV on as-produced basis of the SL and ST bio-oils were in the range of 20.0e23.6 MJ/kg, the bio-oils produced met the ASTM requirement for heating value. On as-produced basis, the heating value of SL bio-oils was higher than those of the ST bio-oils. This result is different from the heating values of bio-oils produced from the free fall reactor, although the same types of biomass were pyrolysed. The HHV of ST and SL bio-oils produced from the free fall reactor were 29.3 MJ/kg and 26.7 MJ/kg, respectively [16], even though the free fall ST bio-oil had approximately 6% higher water content than the SL bio-oil. This means that the organic part of ST bio-oil had a higher heating value than that of the SL one. This is due to the lower oxygen content of the ST bio-oils both produced with and without the use of the hot vapour filtration unit. When considering the heating values of bio-oils produced from the fluidised bed reactor on a water-free basis, it is apparent that the ST pyrolysed in this work gave bio-oils of higher heating values than the SL. This result is different from the as-produced basis data due to the influence of the bio-oils water content. The impact of the hot filtration on the bio-oil heating values can also be seen from Table 2. On as-produced basis, the heating values of unfiltered bio-oils appeared to be higher than those of the filtered ones. This is because the water content of the filtered bio-oils was higher Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
8
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10 Table 3 Properties of chars obtained from pyrolysis of sugarcane leaves and sugarcane tops. Feedstock Elemental composition (wt%, dry basis) Carbon Hydrogen Nitrogen Oxygena Ash H/C molar ratio O/C molar ratio Molecular formula Heating value (MJ/kg, dry basis) HHV LHV a
Sugarcane leaves
Sugarcane tops
57.7 3.1 0.6 13.1 25.5 0.6 0.2 CH0.65O0.17
60.6 2.7 1.4 9.3 26.0 0.5 0.1 CH0.54O0.12
21.9 21.2
22.8 22.2
Calculated by difference.
than the unfiltered ones. This finding is in good agreement with those bio-oils produced from fast pyrolysis of rice straw and husk [17] and cassava stalk and rhizome [18]. However, on a water-free basis, the opposite result was observed. The filtered bio-oils gave higher heating values due to their lower oxygen content on a dry basis. This trend is similar to cassava stalk bio-oils produced by a fluidised bed reactor [18]. However, the heating values of the bio-oils produced from cassava rhizome, rice straw and rice husk were higher when the hot filter was not applied. This is owing to the lower oxygen content and the higher carbon content of the unfiltered bio-oils [17,18]. Part of the carbon is likely to come from the solid char present in the unfiltered bio-oils. 3.3.1.8. Viscosity and stability. Table 2 shows that the ST gave lower viscosity for all bio-oil samples than the SL. This is due to the higher water content of the ST bio-oils. Since ASTM D7544-12 requires a maximum kinematic viscosity at 40 C of 125 mm2/s or cSt, the majority of the fresh and aged bio-oils produced in this work met the requirements with the exception of the aged unfiltered bio-oils. This demonstrates the significance of the hot vapour filtration. It is also obvious that the bio-oils produced with the use of hot filter had lower viscosity for both fresh and aged bio-oils. This is related to the higher water content and lower solids content of the filtered bio-oil. It is also expected that when the vapour passed through the filter, cracking of heavy organic components took place giving off lower molecular weight compounds and water as products. This finding is in good agreement with previous studies on different biomass types (cassava and paddy residues) [17,18]. A stability index was calculated from the fresh and aged bio-oils viscosity. The higher the index, the less stable the bio-oil. It is apparent that the filtered bio-oils were more stable than the unfiltered ones, and the SL bio-oils were more stable than the ST ones. The stability indices of the filtered SL and ST bio-oils were rather similar to those produced from cassava and paddy residues, which were in the range of 0.7e1.8 [17,18]. 3.3.2. Char The SL and ST char samples were analysed for elemental composition, ash content, heating value and particle size distribution. Based on the elemental composition, the H/C and O/C molar ratios and the molecular formulae were determined. It can be noticed from Tables 1 and 3 that the molecular formulae of the SL and ST biomass, CH1.19O0.64 and CH1.16O0.71, respectively, were changed into CH0.65O0.17 and CH0.54O0.12 when becoming SL and ST char, respectively. This shows that fast pyrolysis could reduce the molecular proportion of the H and O elements and the char became rich in carbon. The change of the molecular formulae from biomass to char when applying the fluidised bed reactor is similar when using the free fall reactor [16]. On average, the molecular formulae of the char from Thai agricultural residues from sugarcane [16] and cassava [18] are CH0.6O0.1. This is similar to rice straw char molecular formula, CH0.69O0.11 [7], and is slightly different from radiate pine, CH0.51O0.24 [8], and bamboo sawdust char, CH0.61O0.57 [7]. Table 3 also shows that the higher heating value of the SL and ST char were ~22e23 MJ/kg. This is much higher than lignite and is comparable to subbituminous coal. Therefore, the char produced may be used as a fuel to replace or substitute traditional coal for heat and power generation. In addition, the char may be used as a soil amendment so that carbon would be returned to the ground, thus supporting the carbon negative concept. The char particle size distribution is shown in Fig. 5. The original biomass size range before being pyrolysed was 250e425 mm. The char particle size was obviously reduced in size for both feedstocks. The ST char (Fig. 5 (b)) seemed to be smaller in size than the SL char (Fig. 5 (a)). The majority of the char particles had a 143.5e362.5 mm size range. This size range of char may be suitable for replacing the coal in pulverised coal gasification. Alternatively, the char could be made to pellets or briquettes to increase the energy density.
4. Conclusions Bio-oils from fast pyrolysis of sugarcane leaves (SL) and tops (ST) were produced by a fluidised bed reactor and were characterised for their basic properties. The SL gave maximum bio-oil yield of 52.5 wt% at 429 C, whereas the ST gave 59.0 wt% at 403 C. When the hot filter was applied, these bio-oil yields were decreased to 45.6 wt% for SL and to 50.7 wt% for ST. The optimum temperatures for both biomass types found in this work were higher than those for a free fall reactor. The bio-oil properties showed that only the SL bio-oil produced with the use of the hot filter met the ASTM D7544-12 requirements for water, solids and ash contents, density, heating value and viscosity. The filtered ST bio-oil met the requirements for all of these properties except for water content. Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
9
Fig. 5. Particle size distribution (PSD) of char derived from fast pyrolysis of sugarcane leaves (a) and sugarcane tops (b).
Acknowledgement Financial support from the Faculty of Engineering, Mahasarakham University, Thailand (2014 incoming budget, contract no: ENGEN 57002/Dec2013) and the Energy Policy and Planning Office (EPPO), Ministry of Energy, Royal Thai Government is gratefully acknowledged. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
A.V. Bridgwater, Review of fast pyrolysis of biomass and product upgrading, Biomass Bioenergy 38 (0) (2012) 68e94. A.V. Bridgwater, D. Meier, D. Radlein, An overview of fast pyrolysis of biomass, Org. Geochem. 30 (12) (1999) 1479e1493. S.A. Channiwala, P.P. Parikh, A unified correlation for estimating HHV of solid, liquid and gaseous fuels, Fuel 81 (8) (2002) 1051e1063. J.-Y. Chen, T.-Y. Yang, H.-P. Wan, I.-L. Chung, H.-T. Lee, Removal of char fines from biomass fast pyrolysis vapors b moving granular bed filter, in: Proceeding of the 13th Asia Pacific Confederation of Chemical Engineering Congress, APCChE 2010, Taipei, 2010. T. Chen, C. Wu, R. Liu, W. Fei, S. Liu, Effect of hot vapor filtration on the characterization of bio-oil from rice husks with fast pyrolysis in a fluidized-bed reactor, Bioresour. Technol. 102 (10) (2011) 6178e6185. FAO, Food and Agriculture Organization of the United Nations for a World without Hunger (FAOSTAT), 2013 (Cited 29 September 2015), Available from: http://faostat.fao. org. S.-H. Jung, B.-S. Kang, J.-S. Kim, Production of bio-oil from rice straw and bamboo sawdust under various reaction conditions in a fast pyrolysis plant equipped with a fluidized bed and a char separation system, J. Anal. Appl. Pyrolysis 82 (2) (2008) 240e247. B.-S. Kang, K.H. Lee, H.J. Park, Y.-K. Park, J.-S. Kim, Fast pyrolysis of radiata pine in a bench scale plant with a fluidized bed: influence of a char separation system and reaction conditions on the production of bio-oil, J. Anal. Appl. Pyrolysis 76 (1e2) (2006) 32e37. K.-H. Lee, B.-S. Kang, Y.-K. Park, J.-S. Kim, Influence of reaction temperature, pretreatment, and a char removal system on the production of bio-oil from rice straw by fast pyrolysis, using a fluidized bed, Energy Fuels 19 (5) (2005) 2179e2184. A. Oasmaa, E. Kuoppala, Fast pyrolysis of forestry residue. 3. Storage stability of liquid fuel, Energy Fuels 17 (4) (2003) 1075e1084. A. Oasmaa, C. Peacocke, Properties and Fuel Use of Biomass Derived Fast Pyrolysis Liquids, VTT Publications, 2010. C. Paenpong, S. Inthidech, A. Pattiya, Effect of filter media size, mass flow rate and filtration stage number in a moving-bed granular filter on the yield and properties of bio-oil from fast pyrolysis of biomass, Bioresour. Technol. 139 (2013) 34e42. H.J. Park, Y.-K. Park, J.-I. Dong, J.-S. Kim, J.-K. Jeon, S.-S. Kim, J. Kim, B. Song, J. Park, K.-J. Lee, Pyrolysis characteristics of oriental white oak: kinetic study and fast pyrolysis in a fluidized bed with an improved reaction system, Fuel Process. Technol. 90 (2) (2009) 186e195. H.J. Park, Y.-K. Park, J.S. Kim, Influence of reaction conditions and the char separation system on the production of bio-oil from radiata pine sawdust by fast pyrolysis, Fuel Process. Technol. 89 (8) (2008) 797e802. A. Pattiya, Thermochemical characterization of agricultural wastes from Thai cassava plantations, Energy Source Part A Recovery Util. Environ. Eff. 33 (8) (2011) 691e701.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
10
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
[16] A. Pattiya, S. Sukkasi, V. Goodwin, Fast pyrolysis of sugarcane and cassava residues in a free-fall reactor, Energy 44 (1) (2012) 1067e1077. [17] A. Pattiya, S. Suttibak, Influence of a glass wool hot vapour filter on yields and properties of bio-oil derived from rapid pyrolysis of paddy residues, Bioresour. Technol. 116 (2012a) 107e113. [18] A. Pattiya, S. Suttibak, Production of bio-oil via fast pyrolysis of agricultural residues from cassava plantations in a fluidised-bed reactor with a hot vapour filtration unit, J. Anal. Appl. Pyrolysis 95 (2012b) 227e235. [19] C. Sheng, J.L.T. Azevedo, Estimating the higher heating value of biomass fuels from basic analysis data, Biomass Bioenergy 28 (5) (2005) 499e507. [20] L.E. Whitmer, Removal of particulate matter from condensable vapors using a moving bed granular filter, in: Mechanical Engineering, Vol. Master of Science, Iowa State, 2011, p. 163. [21] R. Xu, L. Ferrante, C. Briens, F. Berruti, Bio-oil production by flash pyrolysis of sugarcane residues and post treatments of the aqueous phase, J. Anal. Appl. Pyrolysis 91 (1) (2011) 263e272.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
Contents lists available at ScienceDirect
Journal of the Energy Institute journal homepage: http://www.journals.elsevier.com/journal-of-the-energyinstitute
Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter Adisak Pattiya*, Suntorn Suttibak Bio-Energy and Renewable Resources Research Unit, Faculty of Engineering, Mahasarakham University, Kamriang, Kantharawichai, Maha Sarakham 44150, Thailand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 16 July 2015 Received in revised form 29 September 2015 Accepted 6 October 2015 Available online xxx
Sugarcane residues e leaves (SL) and tops (ST) from Thailand were pyrolysed in a fluidised bed reactor incorporated with a hot filter. The aim was to investigate the effects of reaction temperatures and hot filtration on pyrolysis products. The bio-oils were characterised by water, solids and ash contents, density, heating value, pH, viscosity and stability. The optimum pyrolysis temperatures for SL and ST were 429 C and 403 C, which gave maximum bio-oil yields of 52.5 wt% and 59.0 wt%, respectively. Using the hot filter reduced bio-oil yield by 7e8 wt%. However, the filtered bio-oils had better viscosity, solids and ash contents and stability. © 2015 Energy Institute. Published by Elsevier Ltd. All rights reserved.
Keywords: Sugarcane leaves Sugarcane tops Bio-oil Fast pyrolysis Hot vapour filtration
1. Introduction It is generally accepted that the fuel shortage and the environmental problems are one of the major crises that need urgent attention. A solution to these problems is the exploitation of renewable energy technology such as fast pyrolysis. Fast pyrolysis is mostly applied to biomass to convert it to a liquid form, termed “bio-oil”. Many researchers studied the production of bio-oil from various types of biomass by many fast pyrolysis reactor configurations. The bio-oil yield could be as high as 80 wt% on dry biomass feed [2] depending on the type of biomass and the reactor unit. The lower heating value (LHV) of bio-oil is typically about 14e18 MJ/kg [2]. Bio-oil can be used as an alternative fuel in furnaces and engines to produce heat and power. In addition, bio-oil can be a raw material for chemical production. Recently, bio-oil has been proposed as a feed for biorefineries to optimise the production of fuels and chemicals to get the highest value out of the bio-oil in social, environmental and economic aspects [1]. Sugarcane is a perennial plant of the genus Saccharum. The main sources of sugarcane are Brazil, India, China, Thailand, Pakistan and Mexico. Typically, sugarcane is grown for sugar production. Lately, sugarcane has become an ethanol production feedstock. The demand for sugar and ethanol is increasing. Therefore, it is expected that the residues will increase accordingly. Residues from sugarcane plantations include bagasse, leaves and tops. The bagasses are traditionally used as a fuel in the sugar factories to generate electricity for their own use. The leaves and tops are mostly burnt in the fields and are not efficiently used for energy. In Thailand, some farmers burn the leaves before harvesting the crops, leading to air pollution. Only a small portion of the leaves and tops are used as a compost and animal feed. Globally, sugarcane production is ~2 109 Tonnes per year, whereas in Thailand the production is ~1 108 Tonnes per year, or 5% of the global production [6]. Assuming a crop to residue ratio of 0.302 for sugarcane residues (leaves and tops) [15], the residues would be ~6 108 Tonnes per year or 96 MTOE (Million Tonnes of Oil Equivalent). This would cost ~$US75,000 million based on a $107/barrel crude oil price. Therefore, fast pyrolysis technology applied to sugarcane leaves and tops for bio-oil production would have two advantages e the fuel value and the environmental impact. Recently Xu et al. [21], investigated bio-oil production by flash pyrolysis of sugarcane residues using a bubbling fluidised bed pyrolyser with a focus on the post treatments of the aqueous phase. They studied the effect of pyrolysis temperature and vapour residence time and
* Corresponding author. Tel.: þ66 43 754321x3036; fax: þ66 43 754316. E-mail address: [email protected] (A. Pattiya). http://dx.doi.org/10.1016/j.joei.2015.10.001 1743-9671/© 2015 Energy Institute. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
2
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
found that maximum liquid yields were for sugarcane bagasse e 58.1%, external e 52.6% and whole plant e 55.5% which were obtained at the pyrolysis temperature of 400 C. Although Xu et al. [21] pyrolysed sugarcane residues in a fluidised bed reactor, they did not fully examine the properties of the liquid product. Only the heating value and water content were used to calculate the energy contained in the bio-oil. Subsequently, in 2012, our group reported the fast pyrolysis of sugarcane and cassava residues in a free-fall reactor [16]. The sugarcane residues bio-oils were measured for water content, solids content, pH, specific gravity, elemental composition, heating value and chemical composition by GC/MS. Nevertheless, one of the most important properties of bio-oil e its stability e was not mentioned due to the small quantity of bio-oil produced. Therefore, the current paper continues the investigation of fast pyrolysis of sugarcane agricultural residues by assessing the impacts of reactor types and in-situ upgrading by hot vapour filtration. Hot filtration can be achieved by candle, granular or fixed bed filters. Previous studies applied candle filters to fluidised bed reactor for production of bio-oil from rice straw [7,9], radiata pine [8,14], Bamboo sawdust [7] and Oriental white [13]. A granular filter is another type that has been tested for bio-oil upgrading [4,12,20]. A hot filter based on a fixed bed of glass wool was also tested with a fluidised-bed fast pyrolyser using rice straw and husk [17] and cassava rhizome [18] and found that the bio-oil solids content could be significantly reduced, although the filter design is relatively simple. In this work, a fluidised bed reactor was used and the products yields and properties were compared with previously reported results using a free fall reactor [16]. This work also assessed the influence of a fixed-bed hot filter on bio-oil properties, especially the solids content, viscosity and storage stability. 2. Experimental 2.1. Biomass feedstock Biomass samples were sugarcane leaves (SL) and tops (ST) from plantations in north-east Thailand. The samples were sundried, ground and sieved to a 250e425 mm range particle size. Prior to experiments, they were dried in an oven at 105 C for 24 h to reduce the moisture content to below 10 wt%. Drying the samples also eased the feeding and could reduce the bio-oil water content. SL and ST samples were tested for their basic properties including proximate and ultimate analyses as well as heating value. The proximate analysis determined the moisture, volatile matter, fixed carbon and ash contents according to ASTM (E1756-01, E872-82 and E1755-01). The ultimate analysis determined carbon (C), hydrogen (H), nitrogen (N), sulphur (S) and oxygen (O) contents using a ‘Leco CHN/ S Determinator’ analyser according to ISO/IEC Guide 22 and EN 45014 standards at the Laboratory Equipment Center, Mahasarakham University, Thailand. The heating values were calculated based on the ultimate analysis results and equations (1) and (2). The higher heating value (HHV) of biomass was calculated from a correlation developed by Sheng and Azevedo [19]:
HHV
MJ kg
¼ 1:3675 þ 0:3137C þ 0:7009H þ 0:0318O
(1)
where C, H are percentages on dry basis of carbon, hydrogen, respectively and O* is 100-C-H-Ash. The lower heating value (LHV) was calculated from HHV and the hydrogen content:
LHV
MJ kg
H ¼ HHV 21:82171 100
(2)
The biomass analysis results are summarised in Table 1. 2.2. Fast pyrolysis apparatus Fast pyrolysis of biomass was carried out in a fluidised bed reactor unit. The unit consisted of a pre-heater, a biomass hopper, a twostaged feeder, a fluidised bed reactor, two cyclone separators, a hot filter and a bio-oil product collection unit (see Fig. 1). The reactor Table 1 Characteristics of agricultural residues from sugarcane plantations in Thailand. Analysis Proximatea (wt%) Moisture Volatile matter Fixed carbonb Ash Ultimate (wt%, dry, ash-free basis) Carbon (C) Hydrogen (H) Nitrogen (N) Sulphur (S) Oxygenb (O) H/C molar ratio O/C molar ratio Molecular formula Heating value (MJ/kg, dry basis) HHV LHV a b
Sugarcane leaves
Sugarcane tops
3.5 76.5 16.1 7.4
2.7 78.5 14.6 6.9
50.8 5.0 0.4 0.2 43.6 1.2 0.6 CH1.19O0.64
48.6 4.7 0.6 0.2 45.8 1.2 0.7 CH1.16O0.71
17.8 16.8
17.2 16.2
Moisture content is on wet basis, whereas volatile matter, fixed carbon and ash contents are on dry basis. Calculated by difference.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
3
was made from a 304 stainless-steel pipe with its internal diameter 50 mm and height 450 mm. Silica sand with a 250e425 mm particle size was used as the fluidising and heat transfer medium. The pyrolysis vapour produced was cleaned up in the cyclone separators and a hot filtration unit prior to condensation. The hot filtration unit was made from a 304 stainless-steel pipe with an internal diameter 50 mm and height 450 mm. Approximately 10 g of glass wool was loosely packed inside the filter and a K-type thermocouple was installed at the middle of the glass wool bed. The nitrogen pre-heater, pyrolysis reactor and hot filter were heated by heating wires and the temperatures were adjusted using PID controllers and monitored using K-type thermocouples. Since the heated zones were insulated with ceramic fibre blankets to minimise heat loss, the temperatures along the reactor and filter length were ±5 C. Pyrolysis vapour was condensed and captured into liquid product by a water-cooled heat exchanger, an electrostatic precipitator (ESP), a series of two dry ice/acetone condensers and a cotton wool filter. 2.3. Fast pyrolysis experiments There were three independent variables in the fast pyrolysis experiments, namely biomass types, pyrolysis temperatures and the use of hot vapour filtration unit. The biomass types investigated were sugarcane leaves and tops. Experiments were run at temperatures of 375 C, 400 C, 425 C, 450 C and 500 C. The fast pyrolysis system had 2 configurations e with and without the hot vapour filtration unit. For the without-hot-filter setup, the pyrolysis vapour exiting the second cyclone went directly to the condenser, leading to shorter vapour residence time. The dependent variables were product yields and physicochemical properties of bio-oil e water content, solids content, ash content, density, heating value, pH, viscosity and stability. The controlled variables were a nitrogen gas flow rate of 7 L/min, a biomass feed rate of ~300 g/h, a temperature at the pre-heater, cyclones, hot filter and transfer line of 420 C, a cooling water temperature of 30 C and a dry-ice/ acetone condenser temperature of 78 C and a run duration of 1 h. It is important to note that the temperature of the pre-heater, cyclones, hot filter and transfer line are controlled at 420 C because at higher than this temperature, it is likely that vapour can undergo thermal cracking leading to lower bio-oil yield. If the temperature is lower than this, the pyrolysis vapour could re-condense or re-polymerise to solid, which could also reduce the liquid yield. Typically, it took 1.5 h to preheat the system from room temperature to steady state. The run duration was timed only when biomass was fed to the system. 2.4. Mass balance calculation The main products from fast pyrolysis process are liquid bio-oil, solid char and gases. The yields of each product were calculated by weighing all parts of the fast pyrolysis system e biomass hopper, silica sand, fluidised bed reactor, cyclone separators, hot filter and product collection unit e before and after each experiment. The bio-oil yields were the combined weight of the liquid from the product collection unit. The char yields were the combined weight of the solid from the reactor, the cyclones, the hot filter, transfer line and the solids in bio-oil. The gas yields were calculated by difference. 2.5. Analysis of pyrolysis products 2.5.1. Bio-oil analysis The bio-oil was characterised by measuring water, solids and ash contents, density, pH, elemental composition, heating value, viscosity and stability. Each analysis was performed in triplicate. 2.5.1.1. Water content. Water content of bio-oil was determined by the Renewable Laboratory, National Metal and Materials Technology Center (MTEC), Pathum Thani, Thailand, using a Karl-Fischer (KF) titration technique.
Fig. 1. Schematic diagram of the fast pyrolysis unit.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
4
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
2.5.1.2. Solids content. Solids content of bio-oil was determined by vacuum filtration. The solids in bio-oil were defined as ethanol insolubles. About 2e3 g of bio-oil was dissolved in ethanol and filtered through a pre-dried and pre-weighed 6 mm mean pore size qualitative filter paper (Whatman No. 3). The liquid was then washed with excess of ethanol until the filtrate was clear to ensure that there was no organic liquid left on the paper. The filter paper with the solids was air-dried for approximately 15 min and further dried in an oven at 105 C for 30 min. Then the paper was cooled in a desiccator and weighed. This method was suggested by Oasmaa and Peacocke [11]. 2.5.1.3. Ash content. Ash content of bio-oil was determined as the amount of residue when heating bio-oil to 775 C in an oxygen atmosphere. Direct heating of bio-oil would result in foaming and splashing due to the high water contents. Thus, the first controlled evaporation of water at 105 C is needed before rapid heating to 775 C [11]. 2.5.1.4. Density. The density of bio-oil was measured using a density bottle at room temperature (about 30 C). 2.5.1.5. pH value. The pH value of bio-oil was measured with a pH metre (COMBI pH/mV/Temp Bench Metre) at room temperature. Before the measurement, the instrument was calibrated with liquid calibration standards of pH 4 and 7. 2.5.1.6. Elemental composition. The elemental composition of bio-oil was determined by using the same technique as described in section 2.1. 2.5.1.7. Heating value. Higher heating value on dry basis (HHVdry) was calculated based on the elemental analysis results using a correlation developed by Channiwala and Parikh [3] e see equation (3). The lower heating value on dry basis (LHVdry) was calculated from equation (2). For heating values on wet basis (HHVwet and LHVwet), equations (4) and (5) were used by taking into account the water content in bio-oil (H2O).
HHV dry
MJ kg
¼ 0:3491C þ 1:1783H þ 0:1005S 0:10340O 0:0211A
(3)
C, H, S, O and A in equation (3) are percentages of carbon, hydrogen, sulphur, oxygen and ash in bio-oil on dry basis.
HHV wet LHV wet
MJ kg
MJ kg
H O ¼ HHV dry 1 2 100
(4)
H O H2 O ¼ LHV dry 1 2 2:442 100 100
(5)
2.5.1.8. Viscosity. Bio-oil kinematic viscosity was measured at 40 C using a Cannon-Fenske Routine Viscometer with a SDM viscosity bath (Art. 370) according to ASTM D445 and D446. 2.5.1.9. Stability. Bio-oil stability is an important quality indicator. The stability index is defined as the change in bio-oil viscosity after accelerated ageing ðnstored Þ compared to its initial viscosity ðnfresh Þ [11]:
StabilityIndex ¼
nstored nfresh nfresh
(6)
nstored is the viscosity of bio-oil stored in an oven at 80 C for 24 h as being equivalent to bio-oil stored at room temperature for 1 year [10], whereas the nfresh is the viscosity of bio-oil within 24 h of production. 2.5.2. Char analysis The char samples obtained from cyclone collectors from the experimental runs that gave highest bio-oil yields were tested for elemental composition, heating value, ash content and particle size distribution (PSD). The methods used for determining the elemental composition were the same as those used for biomass and bio-oil analysis. The char heating values were calculated according to the equations (2) and (3). The procedure for ash content analysis was the same as that for biomass ash content, see section 2.1. The particle size distribution of char was obtained by sieving approximately 30 g of char into size ranges: 0e75, 75e212, 212e250, 250e300, 300e425, 425e500 and >500 mm. Each test sieve was weighed before and after sieving and the results were plotted to obtain the PSD.
3. Results and discussion 3.1. Effect of pyrolysis temperature on product distribution The effect of pyrolysis temperature on yields is shown by Fig. 2 e for sugarcane leaves (SL) e and Fig. 3 for sugarcane tops (ST). The setting temperatures are 375, 400, 425, 450 and 500 C, but the actual average temperatures recorded and calculated throughout the experimental runs are plotted in the graphs. The actual temperatures are within ±5 C of the setting values. It can be seen that increasing the temperature from 375 C to 500 C, decreased the char yield and increased the gas yield for both feedstocks. This trend was also reported by Pattiya et al. [16] when SL and ST were pyrolysed in a free fall reactor and by Xu et al. [21] when sugarcane residues (internal bagasse) was pyrolysed in a bubbling fluidised bed reactor. The maximum bio-oil yields when pyrolysing SL and ST using the fluidised bed reactor were 52.5 wt% at 429 C and 59.0 wt% at 403 C, respectively. When the same biomass species were pyrolysed in a free fall reactor, a similar maximum yield of Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
5
Fig. 2. Effect of temperature on product yields derived from fast pyrolysis of sugarcane leaves in a fluidised bed reactor unit without the hot vapour filter.
Fig. 3. Effect of temperature on product yields derived from fast pyrolysis of sugarcane tops in a fluidised bed reactor unit without the hot vapour filter.
53.3% was obtained at 400 C for SL [16]. However, the maximum bio-oil yield for ST pyrolysed with the free fall reactor was only 44.1 wt% at 350 C. This shows that different yields of bio-oil may be obtained if a different reactor configuration is applied, even though the same biomass type is used. This is supported by the results of Xu et al. [21] when sugarcane residues (external and whole plant) were pyrolysed in a fluidised bed reactor and 52.6e55.5 wt% of maximum bio-oil yield could be obtained at 400 C. Based on these findings on fast pyrolysis of sugarcane residues, the fluidised bed configuration is more effective than the free fall type for obtaining high bio-oil yield. This is likely related to the heat transfer to biomass particles. For a fluidised bed reactor, biomass particles were rapidly heated by direct contact with the heat transfer medium or by thermal conduction as well as by hot fluidising gas or thermal convection. The heat transfer medium usually has high surface area because of the small particle size. In comparison to the fluidised bed type, the free fall reactor heats the biomass particles mostly through the reactor internal wall. The hot wall heats the carrier gas by convection. When a biomass particle dropped into the reactor, it was heated by the direct contact with the reactor wall (conduction) and with the hot carrier gas (convection). 3.2. Effect of hot vapour filtration on the product distribution Fig. 4 shows that for both biomass types, the bio-oil yields reduced due to the use of the hot filtration unit. Bio-oil yields reduced for SL e from 53 wt% to 46 wt% e and ST e from 59 wt% to 51 wt%. This reduction occurred in conjunction with the increase in the gas yield, whereas the char yield was nearly constant. This reflected that part of the bio-oil was transformed into permanent gas when the pyrolysis vapour went through the fixed-bed hot filter. Similar findings were reported for fast pyrolysis of rice straw, rice husk, cassava rhizome and cassava stalk in fluidised bed reactors using either a fixed-bed glass wool hot filter [17,18] or a ceramic candle filter [5]. For fixed-bed glass wool hot filtration, when using paddy residues, the bio-oil yield decreased 4e7 wt% on dry biomass basis, whereas when cassava residues were Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
6
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
Fig. 4. Effect of hot vapour filtration on product yields derived from fast pyrolysis of sugarcane leaves (SL) and sugarcane tops (ST).
pyrolysed, yield decreased 6e7 wt%. Chen et al. [5] reported only 2.2 wt% decrease of bio-oil yield when applying a ceramic candle filter. Our results are consistent with those from previous studies: the use of hot vapour filtration could lead to the decrease in 2e8 wt% bio-oil yield on dry biomass basis. This is equivalent to 5e14% of bio-oil yield reduction. Since the reduction of bio-oil yield when applying the hot filter corresponds to the increase of the gas yield, it is expected that the pyrolysis vapour may undergo secondary cracking. The cracking could be thermal, catalytic or both. The thermal cracking occurred when pyrolysis vapour was exposed to high temperatures for longer time than normal conditions. When pyrolysis vapour passed through the filter, it was in direct contact with the filter medium, which may be at elevated temperature. Also, when a hot filter is installed downstream, this allows the vapour to spend more time in the hot zone prior to condensation. This time is typically called hot vapour residence time. The exposure of the vapour for long time increases the probability of thermal cracking. If the vapour is in contact with some solid particles that are catalytically active, a second type of cracking reactions (catalytic cracking) may take place. The in-situ catalysts in fast pyrolysis are, for example, alkali and alkaline earth metals such as sodium, potassium, magnesium and calcium. These catalysts exist in the form of char fines. If the char fines accumulate on the filter medium surface becoming char cake, when the vapour passes through, secondary catalytic cracking could occur, thus lowering the bio-oil yield and increasing the gas yield. 3.3. Characterisation of pyrolysis products 3.3.1. Bio-oil Table 2 summarises the properties of SL and ST bio-oils produced at their optimum pyrolysis temperatures with and without the use of the hot filter: water, solids and ash contents, density, pH value, elemental composition, heating value, viscosity and stability. 3.3.1.1. Water content. The standard specification for pyrolysis liquid bio-oil (ASTM D7544-12) states that the bio-oil water content should not exceed 30%. Table 2 shows that the majority of the bio-oil produced from sugarcane residues met this requirement except for the filtered ST bio-oil which had slightly higher water content than the standard requirement. It can also be noticed that the use of the hot filter resulted in the increase in the water contents of 11.9% for SL and 13.4% for ST. Pattiya et al. [16] reported that the water contents of SL and ST bio-oil produced by a free fall reactor with a small hot filter were 18% and 24%. These values are in the range of the filtered and non-filtered bio-oils results from the current study. For both reactor types applied, the SL bio-oils appeared to have lower water content than the ST ones. 3.3.1.2. Solids content. Following ASTM D7544-12, the pyrolysis liquid biofuels should contain no more than 2.5 wt% for grade G and 0.25 wt % for grade D. Grade G is intended for use in industrial burners, whereas grade D is for use in commercial or industrial burners requiring low solids and ash content. Table 2 shows that without the hot filter, the SL and ST bio-oil solids content was 3.1e3.5 wt%, failing the ASTM requirement. With the hot filter, the solids content reduced to 0.8 wt%, equivalent to 74e77% solids reduction, thus meeting the Grade G requirement. This reduction is equivalent to 74e77%. The solids content of SL and ST bio-oils produced by the free fall reactor was 2.9e3.7 wt % [16]. In addition, the solids content of cassava and rice residue bio-oils produced by the fluidised bed reactor was 3.3e3.9 wt% without the hot filter and 0.5e0.8 wt% with the hot filter [17,18]. 3.3.1.3. Ash content. The ash content of SL and ST bio-oils was reduced from 0.3-0.4 wt% to 0.01 wt% by the aid of the filter. The ASTM D754412 standard requirements for the ash content are 0.25 wt% for grade G and 0.15 wt% for grade D biofuels. Therefore, our results show that only the filtered bio-oil meets this requirement. The current bio-oils had much lower ash content than those produced in the free fall reactor since the ash content of ~1 wt% was reported [16]. 3.3.1.4. Density. The density of all bio-oils produced in this work was 1.1 g/ml. This is the same as that produced by the free fall reactor [16]. Therefore, it is apparent that neither the reactor configuration nor the hot filtration affected bio-oil density. The density of pyrolysis liquid biofuels is expected to be in the range of 1.1e1.3 kg/m3 according to the ASTM D7544-12. The current results are just sufficient to meet the lower end of the standard. 3.3.1.5. pH value. The pH value of bio-oils produced in this work was in the range of 3.0e4.2. This is similar to the pH value range (3e4) of the bio-oils produced from the free fall reactor [16] using the sugarcane residues or from the fluidised bed reactor using cassava and rice Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
7
Table 2 Properties of bio-oil obtained from pyrolysis of sugarcane leaves and sugarcane tops. Feedstock
Sugarcane leaves With hot filter
432 Pyrolysis temperature ( C) Elemental composition (wt%, dry, ash-free basis) Carbon 65.1 Hydrogen 7.2 Nitrogen 1.0 a Oxygen 26.7 H/C molar ratio 1.3 O/C molar ratio 0.3 Molecular formula CH1.34O0.31 Water content (wt%) 23.6 Density (g/ml) 1.1 Heating value by calculation method (MJ/kg) HHV (water-free basis) 28.5 HHV (as-produced basis) 21.7 LHV (water-free basis) 26.9 LHV (as-produced basis) 20.0 pH value 3.0 Kinematic viscosity @ 40 C (cSt) Fresh bio-oilb 16.3 Aged bio-oilc 50.6 Solids content (wt%) 0.8 Ash (wt%) 0.01 Stability index 2.1 a b c
Sugarcane tops Without hot filter
With hot filter
Without hot filter
429
404
403
59.9 7.7 0.9 31.1 1.5 0.4 CH1.55O0.39 11.7 1.1
71.3 5.9 1.0 21.7 1.0 0.2 CH1.00O0.23 32.4 1.1
62.9 6.6 0.6 29.6 1.3 0.4 CH1.26O0.35 19.0 1.1
26.7 23.6 25.1 21.9 3.0
29.6 20.0 28.3 18.3 4.2
26.6 21.6 25.2 19.9 3.8
23.2 705.4 3.5 0.41 29.5
2.8 6.5 0.8 0.01 1.3
19.7 455.2 3.1 0.30 22.1
Calculated by difference. Fresh bio-oil is the one produced within 24 h. Aged bio-oil is bio-oil stored in an oven at 80 C for 24 h.
residues [17,18] and slightly higher than typical wood-derived bio-oil pH of 2.5 [1]. Since the ASTM D7544-12 specification for pyrolysis liquid biofuel does not mandate a minimum pH, the current bio-oils are acceptable with respect to pH. 3.3.1.6. Elemental composition. Table 2 shows that ST bio-oils contained higher carbon and lower hydrocarbon and oxygen contents than the SL bio-oils. The same trend for carbon and oxygen contents was reported when processing these biomass samples in the free fall pyrolyser [16]. The carbon contents of ST and SL bio-oils were 65.4 wt% and 60.0 wt%, respectively, whereas the oxygen content was 24.2 wt% for ST bio-oil and 29.5 wt% for SL bio-oil [16]. The difference in the carbon and oxygen contents of the two bio-oils was even higher for hot filtered bio-oils. In other words, the use of the hot filtration unit led to an increase of the carbon content and a reduction of the hydrogen and oxygen content e see Table 2. This implies that deoxygenation reactions were involved in the secondary cracking of the pyrolysis vapour while passing through the fixed bed of glass wool filter. The changes in the elemental composition of bio-oil when applying the hot filter could also be noticed when pyrolysing different biomass species. The current findings are consistent with those using cassava stalk as raw material [18]. However, the opposite trend was observed when cassava rhizome, rice straw and rice husk were pyrolysed as carbon content was reduced and oxygen content was increased when the hot filter was equipped with the fluidised bed reactor. Therefore, the changes of the elemental composition with the hot filtration are dependent on biomass type as well. Based on the elemental composition, the H/C and O/C molar ratios could be calculated and the molecular formulae of the bio-oils were generated. When comparing the ratios and the formulae for sugarcane residues bio-oils produced in this work with those of the bio-oil previously produced from Thai agricultural residues such as cassava stalk, cassava rhizome rice straw and rice husk applying both the fluidised bed reactor with and without the hot filter and the free fall reactor [16e18], H/C ratios of 0.92e1.58 and O/C ratios of 0.14e0.71 were found. On average, the molecular formula for Thai agricultural residues bio-oil is CH1.3O0.3. 3.3.1.7. Heating value. The heating values of SL and ST bio-oils produced with and without the hot filter are summarised in Table 2. Both higher heating value (HHV) and lower heating value (LHV) are shown on water-free and as-produced bases. The water-free basis is important for understanding the heating value of the organic part or non-aqueous part of the bio-oil. According to the ASTM D7544-12, the minimum gross heat of combustion, or HHV (on as-produced basis), for pyrolysis liquid biofuel is 15 MJ/kg. Since the HHV on as-produced basis of the SL and ST bio-oils were in the range of 20.0e23.6 MJ/kg, the bio-oils produced met the ASTM requirement for heating value. On as-produced basis, the heating value of SL bio-oils was higher than those of the ST bio-oils. This result is different from the heating values of bio-oils produced from the free fall reactor, although the same types of biomass were pyrolysed. The HHV of ST and SL bio-oils produced from the free fall reactor were 29.3 MJ/kg and 26.7 MJ/kg, respectively [16], even though the free fall ST bio-oil had approximately 6% higher water content than the SL bio-oil. This means that the organic part of ST bio-oil had a higher heating value than that of the SL one. This is due to the lower oxygen content of the ST bio-oils both produced with and without the use of the hot vapour filtration unit. When considering the heating values of bio-oils produced from the fluidised bed reactor on a water-free basis, it is apparent that the ST pyrolysed in this work gave bio-oils of higher heating values than the SL. This result is different from the as-produced basis data due to the influence of the bio-oils water content. The impact of the hot filtration on the bio-oil heating values can also be seen from Table 2. On as-produced basis, the heating values of unfiltered bio-oils appeared to be higher than those of the filtered ones. This is because the water content of the filtered bio-oils was higher Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
8
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10 Table 3 Properties of chars obtained from pyrolysis of sugarcane leaves and sugarcane tops. Feedstock Elemental composition (wt%, dry basis) Carbon Hydrogen Nitrogen Oxygena Ash H/C molar ratio O/C molar ratio Molecular formula Heating value (MJ/kg, dry basis) HHV LHV a
Sugarcane leaves
Sugarcane tops
57.7 3.1 0.6 13.1 25.5 0.6 0.2 CH0.65O0.17
60.6 2.7 1.4 9.3 26.0 0.5 0.1 CH0.54O0.12
21.9 21.2
22.8 22.2
Calculated by difference.
than the unfiltered ones. This finding is in good agreement with those bio-oils produced from fast pyrolysis of rice straw and husk [17] and cassava stalk and rhizome [18]. However, on a water-free basis, the opposite result was observed. The filtered bio-oils gave higher heating values due to their lower oxygen content on a dry basis. This trend is similar to cassava stalk bio-oils produced by a fluidised bed reactor [18]. However, the heating values of the bio-oils produced from cassava rhizome, rice straw and rice husk were higher when the hot filter was not applied. This is owing to the lower oxygen content and the higher carbon content of the unfiltered bio-oils [17,18]. Part of the carbon is likely to come from the solid char present in the unfiltered bio-oils. 3.3.1.8. Viscosity and stability. Table 2 shows that the ST gave lower viscosity for all bio-oil samples than the SL. This is due to the higher water content of the ST bio-oils. Since ASTM D7544-12 requires a maximum kinematic viscosity at 40 C of 125 mm2/s or cSt, the majority of the fresh and aged bio-oils produced in this work met the requirements with the exception of the aged unfiltered bio-oils. This demonstrates the significance of the hot vapour filtration. It is also obvious that the bio-oils produced with the use of hot filter had lower viscosity for both fresh and aged bio-oils. This is related to the higher water content and lower solids content of the filtered bio-oil. It is also expected that when the vapour passed through the filter, cracking of heavy organic components took place giving off lower molecular weight compounds and water as products. This finding is in good agreement with previous studies on different biomass types (cassava and paddy residues) [17,18]. A stability index was calculated from the fresh and aged bio-oils viscosity. The higher the index, the less stable the bio-oil. It is apparent that the filtered bio-oils were more stable than the unfiltered ones, and the SL bio-oils were more stable than the ST ones. The stability indices of the filtered SL and ST bio-oils were rather similar to those produced from cassava and paddy residues, which were in the range of 0.7e1.8 [17,18]. 3.3.2. Char The SL and ST char samples were analysed for elemental composition, ash content, heating value and particle size distribution. Based on the elemental composition, the H/C and O/C molar ratios and the molecular formulae were determined. It can be noticed from Tables 1 and 3 that the molecular formulae of the SL and ST biomass, CH1.19O0.64 and CH1.16O0.71, respectively, were changed into CH0.65O0.17 and CH0.54O0.12 when becoming SL and ST char, respectively. This shows that fast pyrolysis could reduce the molecular proportion of the H and O elements and the char became rich in carbon. The change of the molecular formulae from biomass to char when applying the fluidised bed reactor is similar when using the free fall reactor [16]. On average, the molecular formulae of the char from Thai agricultural residues from sugarcane [16] and cassava [18] are CH0.6O0.1. This is similar to rice straw char molecular formula, CH0.69O0.11 [7], and is slightly different from radiate pine, CH0.51O0.24 [8], and bamboo sawdust char, CH0.61O0.57 [7]. Table 3 also shows that the higher heating value of the SL and ST char were ~22e23 MJ/kg. This is much higher than lignite and is comparable to subbituminous coal. Therefore, the char produced may be used as a fuel to replace or substitute traditional coal for heat and power generation. In addition, the char may be used as a soil amendment so that carbon would be returned to the ground, thus supporting the carbon negative concept. The char particle size distribution is shown in Fig. 5. The original biomass size range before being pyrolysed was 250e425 mm. The char particle size was obviously reduced in size for both feedstocks. The ST char (Fig. 5 (b)) seemed to be smaller in size than the SL char (Fig. 5 (a)). The majority of the char particles had a 143.5e362.5 mm size range. This size range of char may be suitable for replacing the coal in pulverised coal gasification. Alternatively, the char could be made to pellets or briquettes to increase the energy density.
4. Conclusions Bio-oils from fast pyrolysis of sugarcane leaves (SL) and tops (ST) were produced by a fluidised bed reactor and were characterised for their basic properties. The SL gave maximum bio-oil yield of 52.5 wt% at 429 C, whereas the ST gave 59.0 wt% at 403 C. When the hot filter was applied, these bio-oil yields were decreased to 45.6 wt% for SL and to 50.7 wt% for ST. The optimum temperatures for both biomass types found in this work were higher than those for a free fall reactor. The bio-oil properties showed that only the SL bio-oil produced with the use of the hot filter met the ASTM D7544-12 requirements for water, solids and ash contents, density, heating value and viscosity. The filtered ST bio-oil met the requirements for all of these properties except for water content. Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
9
Fig. 5. Particle size distribution (PSD) of char derived from fast pyrolysis of sugarcane leaves (a) and sugarcane tops (b).
Acknowledgement Financial support from the Faculty of Engineering, Mahasarakham University, Thailand (2014 incoming budget, contract no: ENGEN 57002/Dec2013) and the Energy Policy and Planning Office (EPPO), Ministry of Energy, Royal Thai Government is gratefully acknowledged. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
A.V. Bridgwater, Review of fast pyrolysis of biomass and product upgrading, Biomass Bioenergy 38 (0) (2012) 68e94. A.V. Bridgwater, D. Meier, D. Radlein, An overview of fast pyrolysis of biomass, Org. Geochem. 30 (12) (1999) 1479e1493. S.A. Channiwala, P.P. Parikh, A unified correlation for estimating HHV of solid, liquid and gaseous fuels, Fuel 81 (8) (2002) 1051e1063. J.-Y. Chen, T.-Y. Yang, H.-P. Wan, I.-L. Chung, H.-T. Lee, Removal of char fines from biomass fast pyrolysis vapors b moving granular bed filter, in: Proceeding of the 13th Asia Pacific Confederation of Chemical Engineering Congress, APCChE 2010, Taipei, 2010. T. Chen, C. Wu, R. Liu, W. Fei, S. Liu, Effect of hot vapor filtration on the characterization of bio-oil from rice husks with fast pyrolysis in a fluidized-bed reactor, Bioresour. Technol. 102 (10) (2011) 6178e6185. FAO, Food and Agriculture Organization of the United Nations for a World without Hunger (FAOSTAT), 2013 (Cited 29 September 2015), Available from: http://faostat.fao. org. S.-H. Jung, B.-S. Kang, J.-S. Kim, Production of bio-oil from rice straw and bamboo sawdust under various reaction conditions in a fast pyrolysis plant equipped with a fluidized bed and a char separation system, J. Anal. Appl. Pyrolysis 82 (2) (2008) 240e247. B.-S. Kang, K.H. Lee, H.J. Park, Y.-K. Park, J.-S. Kim, Fast pyrolysis of radiata pine in a bench scale plant with a fluidized bed: influence of a char separation system and reaction conditions on the production of bio-oil, J. Anal. Appl. Pyrolysis 76 (1e2) (2006) 32e37. K.-H. Lee, B.-S. Kang, Y.-K. Park, J.-S. Kim, Influence of reaction temperature, pretreatment, and a char removal system on the production of bio-oil from rice straw by fast pyrolysis, using a fluidized bed, Energy Fuels 19 (5) (2005) 2179e2184. A. Oasmaa, E. Kuoppala, Fast pyrolysis of forestry residue. 3. Storage stability of liquid fuel, Energy Fuels 17 (4) (2003) 1075e1084. A. Oasmaa, C. Peacocke, Properties and Fuel Use of Biomass Derived Fast Pyrolysis Liquids, VTT Publications, 2010. C. Paenpong, S. Inthidech, A. Pattiya, Effect of filter media size, mass flow rate and filtration stage number in a moving-bed granular filter on the yield and properties of bio-oil from fast pyrolysis of biomass, Bioresour. Technol. 139 (2013) 34e42. H.J. Park, Y.-K. Park, J.-I. Dong, J.-S. Kim, J.-K. Jeon, S.-S. Kim, J. Kim, B. Song, J. Park, K.-J. Lee, Pyrolysis characteristics of oriental white oak: kinetic study and fast pyrolysis in a fluidized bed with an improved reaction system, Fuel Process. Technol. 90 (2) (2009) 186e195. H.J. Park, Y.-K. Park, J.S. Kim, Influence of reaction conditions and the char separation system on the production of bio-oil from radiata pine sawdust by fast pyrolysis, Fuel Process. Technol. 89 (8) (2008) 797e802. A. Pattiya, Thermochemical characterization of agricultural wastes from Thai cassava plantations, Energy Source Part A Recovery Util. Environ. Eff. 33 (8) (2011) 691e701.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001
10
A. Pattiya, S. Suttibak / Journal of the Energy Institute xxx (2015) 1e10
[16] A. Pattiya, S. Sukkasi, V. Goodwin, Fast pyrolysis of sugarcane and cassava residues in a free-fall reactor, Energy 44 (1) (2012) 1067e1077. [17] A. Pattiya, S. Suttibak, Influence of a glass wool hot vapour filter on yields and properties of bio-oil derived from rapid pyrolysis of paddy residues, Bioresour. Technol. 116 (2012a) 107e113. [18] A. Pattiya, S. Suttibak, Production of bio-oil via fast pyrolysis of agricultural residues from cassava plantations in a fluidised-bed reactor with a hot vapour filtration unit, J. Anal. Appl. Pyrolysis 95 (2012b) 227e235. [19] C. Sheng, J.L.T. Azevedo, Estimating the higher heating value of biomass fuels from basic analysis data, Biomass Bioenergy 28 (5) (2005) 499e507. [20] L.E. Whitmer, Removal of particulate matter from condensable vapors using a moving bed granular filter, in: Mechanical Engineering, Vol. Master of Science, Iowa State, 2011, p. 163. [21] R. Xu, L. Ferrante, C. Briens, F. Berruti, Bio-oil production by flash pyrolysis of sugarcane residues and post treatments of the aqueous phase, J. Anal. Appl. Pyrolysis 91 (1) (2011) 263e272.
Please cite this article in press as: A. Pattiya, S. Suttibak, Fast pyrolysis of sugarcane residues in a fluidised bed reactor with a hot vapour filter, Journal of the Energy Institute (2015), http://dx.doi.org/10.1016/j.joei.2015.10.001