Pure oxygen fixed-bed gasification of wood under high temperature (>1000 °C) freeboard conditions

Applied Energy 191 (2017) 153–162

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Pure oxygen fixed-bed gasification of wood under high temperature (>1000 °C) freeboard conditions Henrik Wiinikka ⇑, Jonas Wennebro, Marcus Gullberg, Esbjörn Pettersson, Fredrik Weiland SP Energy Technology Center AB, Box 726, S-941 28 Piteå, Sweden

h i g h l i g h t s
Oxygen blown fixed bed and entrained flow gasification of wood were compared.
Both gasifiers efficiently produced a high quality syngas with high CO and H2 yields.
Generation of high quality syngas is not restricted to entrained flow gasification.
Instead, it is technology independent and coupled to high process temperature.

a r t i c l e

i n f o

Article history: Received 7 September 2016 Received in revised form 4 January 2017 Accepted 23 January 2017 Available online 1 February 2017 Keywords: Gasification Oxygen blown Biomass Fixed bed Entrained flow

a b s t r a c t In this paper, the performance (syngas composition, syngas production and gasification efficiency) of an 18 kW atmospheric fixed bed oxygen blown gasifier (FOXBG) with a high temperature (>1000 °C) freeboard section was compared to that of a pressurized (2–7 bar) oxygen blown entrained flow biomass gasifier (PEBG). Stem wood in the form of pellets (FOXBG) or powder (PEBG) was used as fuel. The experimentally obtained syngas compositions, syngas production rates and gasification efficiencies for both gasification technologies were similar. Efficient generation of high quality syngas (in terms of high concentration and yield of CO and H2 and low concentration and yield of CH4, heavier hydrocarbons and soot) is therefore not specific to the PEBG. Instead, efficient gasification seems to be linked to high reactor process temperatures that can also be obtained in a FOXBG. The high quality of the syngas produced in the FOXBG from fuel pellets is promising, as it suggests that in the future, much of the cost associated with milling the fuel to a fine powder will be avoidable. Furthermore, it is also implied that feedstocks that are nearly impossible to pulverize can be used as un-pretreated fuels in the FOXBG. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction High temperature (>1000 °C) oxygen blown entrained flow gasification is a well proven gasification technology for the production of high quality synthetic gas (syngas). The produced syngas typically exhibits high concentrations of CO and H2 and low concentrations of undesirable products, such as CH4 and heavier hydrocarbons, tar and soot. Due to its favorable composition, the syngas produced in oxygen blown entrained flow gasification can be used in the synthesis of motor fuels and other chemicals [1–3]. Other potential areas of application for the syngas produced in high temperature gasification systems include the steel industry, where the syngas can be used to generate fuel for heat treatment

⇑ Corresponding author. E-mail address: [email protected] (H. Wiinikka). http://dx.doi.org/10.1016/j.apenergy.2017.01.054 0306-2619/Ó 2017 Elsevier Ltd. All rights reserved.

furnaces [4]. However, in order to achieve high fuel conversion, throughput and gas quality in the entrained flow gasification process, the fuel needs to be introduced in fine powder form or as droplets into the gasifier. For most solid fuels, this requires the extensive pretreatment of the feedstock, including drying, milling or liquefaction. Depending on the feedstock, especially in the case of biomass and waste raw materials, the pretreatment procedure may be very complicated and expensive, which lowers the overall efficiency of the gasification process [5,6]. Therefore, there should be an incentive to develop a high temperature gasification technology that does not require the extensive pretreatment of the feedstock. Air, oxygen or steam blown fixed bed gasifiers can be operated with a wide range of different feedstocks (coal, biomass and waste) and are reasonably insensitive to fuel pretreatment [7–15]. Fixed bed gasifiers can be either updraft (fuel enters at the top, while the gasifying agent is introduced at the bottom) or downdraft (both

154

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

fuel and gasification agent enter from the bottom, with the fuel coming from a lock-hopper system [2]). Regardless of fuel type, a combustion zone is formed in the fuel bed where the temperature reaches 1000–1400 °C, depending on the gasification agent. The drawback with traditional fixed bed gasifiers is that the hot gas produced in the combustion zone cools significantly as it percolates through the reduction, pyrolysis and drying zones (updraft gasifiers) or reduction zone (downdraft gasifiers). The gas produced during traditional fixed bed gasification therefore contains significant amounts of tar and has a relatively high CH4 concentration. Without extensive gas cleaning, these properties make the gas unfavorable for upgrading to motor fuels. The hypothesis explored in this paper is that the syngas produced in oxygen blown fixed bed gasification has comparable qualities to the syngas produced in entrained flow gasification, if the fixed bed gasifier is constructed with a high temperature (>1000 °C) freeboard section. To the best of the authors’ knowledge, this has not been investigated in the open literature before and therefore also defines the novelty of the work. The work was carried out by (i) designing and constructing an oxygen blown fixed bed gasifier with a high temperature freeboard, (ii) determining the characteristics of the gasifier (gas composition, gas production, and gasification efficiency) and (iii) comparing the results, when possible, with earlier published results obtained during oxygen blown entrained flow gasification [16] using the same feedstock, stem wood from pine and spruce as fuel. 2. Fixed bed gasification with high temperature freeboard The general behavior of an oxygen blown, entrained flow gasifier in thermodynamic equilibrium has been described by Weiland et al. [16]. The results can be generalized as design criteria for oxygen blown fixed bed gasifiers and are therefore shortly reviewed here. Theoretical and experimental investigations have shown that the most important operating parameter is the O2 stoichiometric ratio, k, since it affects the temperature and composition of the produced gas and also the efficiency of the gasification process [16,17]. Therefore, in this work, the performances of both the entrained and the fixed bed gasifier were experimentally evaluated as a function of k. The stoichiometric ratio k is defined as the ratio _ O2 (kg/h), and the stoichiometric O2 demand of the supplied O2, m _ O2 ;stoich (kg/h) as the following: for complete combustion, m

k¼

_ O2 m : _ O2 ;stoich m

ð1Þ

The efficiency of a gasification process is often described by the cold gas efficiency, CGE which is the ratio of the amount of chemical energy stored in the produced cooled syngas and the amount of energy introduced into the gasifier with the fuel. The CGE is calculated as the following:

CGE ¼

_ cg LHV cg m ; _ fuel LHV fuel m

ð2Þ

_ cg (kg/s) and m _ fuel (kg/s) are the mass flows of cold gas from where m the gasifier and the mass flow of fuel input, and LHVcg (MJ/kg) and LHVfuel (MJ/kg) are the lower heating values (LHV) of the cold gas and the fuel, respectively. Two different definitions of CGE are used in this work. The CGEpower is calculated using all energy containing species in the syngas, whereas CGEfuel only uses CO and H2 concentrations. The CGEpower represents a case where the produced gas is used for heat and power generation in a downstream combustion process (e.g., gas engine or gas turbine), while CGEfuel is more practical in cases where the syngas is used for motor fuel production in a downstream synthesis process, with only CO and H2 as active species. For the wood used in the present work, at low k (<0.25), the

amount of oxygen and the resulting process temperature are too low for complete gasification, which affects the CGE negatively, due to the presence of carbon residues. Increasing k beyond 0.25 increases the process temperature and gas production, and the maximum CGE occurs at k values between 0.27 and 0.3. Further increasing k results in a further increase of the process temperature due to oxidation reactions until the temperature is high enough for dissociation of CO2 and H2O. As a consequence of the oxidation the CGE is also reduced. The theoretical operating window in terms of k for oxygen blown gasifiers is thereby between 0.27 and 0.6. In the following, the theory specific to the design of oxygen blown, fixed bed gasification processes of wood with a high temperature freeboard is described. A simplified sketch of the process is presented in Fig. 1 and important reactions are listed in Table 1. Fuel is introduced into the gasifier from the top and the gasification agent (pure O2) is introduced from the bottom into three different regions: (i) primary, (ii) secondary and, (iii) tertiary oxygen supply zones. As explained above, the gasifier can therefore be classified as an updraft gasifier. The char left after pyrolysis of the fuel is converted to gas in two steps. Close to the bottom of the gasifier where primary O2 is introduced, the carbon in the bottom layer reacts with O2 producing CO2 and heat (char combustion region, R1). Due to the exothermic reactions, the temperature rises to above 2000 °C. After the O2 has been consumed, the remaining char in the top layer of the char bed reacts with the CO2 percolating through the bed from the char combustion region forming CO (char gasification region, R2). The reactions specific to this process are endothermic, thus the temperature decreases to approximately 750 °C by the end of the char gasification region. The amount of O2 supplied through the primary oxygen inlet corresponds to the amount of O2 needed to convert all of the carbon in the char to CO. Above the char gasification zone, the wood pellets dry (R3) and pyrolyze (R4) and the produced gas (CO, CO2, H2, H2O and CH4) from the pyrolysis (R4) and tar destruction (R5) reacts with the O2 supplied through the secondary oxygen inlet producing CO, CO2, H2, and H2O (R6–R9). The amount of O2 supplied here corresponds to the amount that is needed to raise the temperature to 1000 °C (k  0.3). In this work, the O2 supply was not intended to be reduced below 0.3, since when the temperature decreases below 1000 °C, tar destruction becomes incomplete, leading to an undesirable gas composition. The important high temperature freeboard zone is found above the bed zone. In the high temperature freeboard zone, the intent is to destroy all light and heavy hydrocarbons and to reduce the amount of formed soot. The extra amount of O2 which is needed to increase the temperature above 1000 °C is supplied through the tertiary oxygen inlet. Depending on k, different process temperatures are reached in the freeboard zone, as indicated in Fig. 1. With a temperature above 1400 °C, significant CH4 destruction is also possible in an O2-free H2O and CO2 environment [18], probably due to the steam and CO2 reforming reactions (R10 and R11). Furthermore, there is a coupling between the tar and soot in biomass gasification since soot is assumed to form from the thermal decomposition of tars (R5). Several investigations in drop tube furnaces simulating entrained flow gasification have shown that as the temperature increases above 1000 °C, the tar yield in the gas decreases, while the soot yield increases significantly [19,20]. Maximum soot production occurs at a temperature of approximately 1200 °C [20]. Again, a temperature above 1400 °C is needed to reduce the amount of soot to trace levels in the gas [21]. Given the above, one can therefore assume that roughly the same process conditions (temperature above 1400 °C, k > 0.36) are needed in the freeboard zone during fixed bed gasification in order to significantly reduce the production of CH4 and soot. Finally, the resulting gas composition in the freeboard zone is believed to be controlled by the water gas shift reaction (R12).

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

155

Fig. 1. Schematic of oxygen blown fixed bed gasification with the high temperature freeboard zone.

Table 1 Simplified chemical reaction scheme of the reactions involved during oxygen blown fixed bed gasification. Reaction

Description/Comment

Ref

C(s) + O2 ? CO2 C(s) + CO2 ? 2CO H2O(l) + heat ? H2O(g) Biomass + heat ? CO + H2 + CO2 + H2O + CH4 + Tars Tars + heat ? CO + H2 + CO2 + H2O + CH4 + C(s) H2 + ½O2 ? H2O CO + ½O2 ? CO2 CH4 + ½O2 ? CO + 2H2 CH4 + 2O2 ? CO2 + 2H2O CH4 + H2O ? CO + 3H2 CH4 + CO2 ? 2CO + 2H2 CO + H2O M CO2 + H2

Char combustion, 394 kJ/mol Char gasification (Boudouard reaction), +172 kJ/mol Drying, +44 kJ/mol Pyrolysis Thermal destruction of tars forming gas and soot Hydrogen combustion, 242 kJ/mol Carbon monoxide combustion, 283 kJ/mol Methane partial combustion, 36 kJ/mol Methane combustion, 803 kJ/mol Methane steam reforming, +206 kJ/mol Methane carbon dioxide reforming, 248 kJ/mol Water gas shift, 41 kJ/mol

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12

3. Experimental 3.1. The gasifiers The fixed bed oxygen blown gasifier (FOXBG) was built at SP Energy Technology Center in Piteå during the first half of 2016. The gasifier operated at atmospheric pressure with a high temperature (>1000 °C) freeboard zone according to the principle described in Section 2. The FOXBG was designed for a fuel feed rate between 2 and 4 kg/h of wood pellets corresponding to a thermal power of 10–20 kWth. Fig. 2 shows a schematic and a photo of the FOXBG which mainly consists of four important parts, the fuel feeding system, the ceramic lined reactor, the gas cooler and the flare. The reactor was a vertical cylinder with an outer diameter of 0.7 m and a height of 1.75 m. The reactor consists of two main parts: a short, conically shaped bed section at the bottom of the gasifier and a long, cylinder shaped freeboard section (see Fig. 3). The inside of the gasifier is lined with a ceramic material and has an inner diameter of 0.19 m in the freeboard zone. The layer of the ceramic material that is exposed to the syngas consist of Vibron

175B (85% Al2O3, 12% SiO2, <1% Fe2O3). Oxygen was introduced to the reactor through three different inlets (primary, secondary and tertiary). The inlets were small alumina (Al2O3) tubes that were inserted through the ceramic lining. The vertical primary inlet contained five holes (3 mm in diameter) and was located underneath the bed zone. The secondary inlet contained four holes (0.8 mm in diameter) that were located just below the freeboard zone with an injection angle of 30° (horizontally) downwards, towards the bed (see Fig. 3). The secondary oxygen inlets were also directed tangentially, at an angle of 14°, in order to create a rotating flow field above the fuel bed (see Fig. 3). Tertiary oxygen was injected horizontally and towards the centerline through 6 holes evenly distributed around the perimeter of the reactor, each hole with a diameter of 0.8 mm. The tertiary inlet was located between the bed zone and the freeboard zone. The process temperatures were measured with four S type thermocouples, one in the bed zone and three in the freeboard zone of the gasifier. The thermocouples were inserted approximately 10 mm into the gas flow and were protected by sintered alumina tubes (8 mm diameter). The gas outlet was located at the top of the freeboard zone where the process gas exited horizontally

156

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

Fig. 2. Schematic and photo of the fixed bed oxygen blown gasifier. The thermocouple used to measure the process temperature is marked with (X).

Fig. 3. Detailed schematic of the reactor part of the fixed bed oxygen blown gasifier showing the primary oxygen inlets (red arrows), secondary oxygen inlets (blue arrows) and tertiary oxygen inlets (green arrows). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

through a ceramic lined part with an inner diameter of 100 mm before entering a water cooled gas cooler which reduced the gas temperature to 30 °C. The gas cooler was designed so that the process gas flowed through 12 pipes, each with an inner diameter

of 19 mm, that were surrounded by water (i.e., there was no direct contact between the process gas and the cooling water) with a temperature of 15 °C. The pipes in the gas cooler were 600 mm long. The condensed water from the process gas was collected in

157

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

a container underneath the gas cooler. After the gas cooler, the temperature of the gas was increased to 100 °C by an air heated tubular heat exchanger in order to prevent condensation of residual water in the syngas pipe. Finally, the produced syngas was flared inside another ceramic lined combustion chamber. An oil burner was used as ignition source. The fuel (wood pellets) was introduced from the top of the gasifier above the freeboard zone through a mechanical feeding system. A small nitrogen purge flow (1–2 l/min) was introduced to the pipe just below the fuel feeding system to serve as protection from the hot uprising gas from the reactor. The pressure inside the gasifier was controlled by an air driven ejector located just before the combustion chamber. The Pressurized Entrained flow Biomass Gasifier (PEBG) pilot plant has been described in detail elsewhere [22], therefore only a brief description is given here. The PEBG gasifier consists of a refractory lined reactor (inner diameter 0.52 m, length 1.52 m) followed by a bubbling 2-level water sprayed quench for syngas cooling and for filtering smelt and particles. Through the top-fed burner, a jet of fuel particles and oxygen is injected. A small amount of nitrogen is also injected with the wood powder in order to limit the risk of backfiring. Five ceramic shielded S-type thermocouples measured the process temperature. After the hot reactor, water sprays followed by a bubbling quench cool and clean the gas from ash, char and soot particles. Finally, the produced syngas is flared on the top of the building. Table 2 summarises the most important design and operation parameters of the two gasifiers. It is important to note here that both gasifiers were autothermal (i.e., the heat in the gasifier is produced by the chemical reactions itself). Due to the higher operating pressure, the reactor energy density (defined as the energy released per unit volume of the reactor, assuming complete combustion of the fuel) was higher in the case of the PEBG, than in the case of the FOXBG. The plug flow residence time was also significantly higher in the PEBG, especially at higher operational pressures. One important parameter for the performance of a gasifier is the heat losses from the walls of the reactor to the surroundings [16,23]. According to simple calculations, taking into account radiation and natural convection, the relative heat losses were estimated to be 5% for the PEBG and 9% for the FOXBG. 3.2. Fuels Stem wood from pine and spruce in the form of pellets (FOXBG) or powder (PEBG) were used as fuel. The physical and chemical compositions of the fuels are summarized in Table 3. The wood pellet was delivered by Bioenergi i Luleå AB. The fuel used in the PEBG was also delivered as wood pellets either by Glommers MiljöEnergi AB or Stenvalls Trä AB. Wood powder was thereafter produced by milling the wood pellets in a hammer mill. The fuel was

Table 2 Comparison between the PEBG and the FOXBG.

Autothermal/Allothermal Pressure (bar) Thermal power (kW)a Reactor volume (m3) Energy density (kW/m3)a Rector residence time (s)b Heat losses (kW) Relative heat losses (%)c a b c

PEBG

FOXBG

Autothermal 2–7 211–613 0.365 577–1677 3.3–20.3 25–35 4–9

Autothermal 1 18 0.043 417 3.0–6.0 1.7 9

Assuming complete combustion of the fuel. Plug-flow assumption. Relative heat losses = Heat losses/Thermal power.

Table 3 Physical and chemical characterization of the fuels. Physical analysis (mm)

PEBG

FOXBG

d50 d90 Diameter Length

0.13–0.14 0.24 – –

– – 6 11.2 ± 3.2

Proximate analysis (wt-% as received) Moisture Volatile matter Fixed carbon Ash

6.2 ± 0.7 77.4 ± 0.7 15.0 ± 1.4 0.3 ± 0.0

5.9 ± 0.5 – – 0.3

Ultimate analysis (wt-% dry) Carbon Hydrogen Oxygen Nitrogen Chlorine Sulphur

51.0 ± 0.3 6.3 ± 0.1 42.1 ± 0.5 <0.1 <0.2 0.02

51.3 6.3 42.0 <0.1 – 0.013

Calorimetric analysis (MJ/kg dry) Lower heating value (LHV)

19.55 ± 0.07

19.18

milled until no significant amount of particles larger than 0.75 mm remained. 3.3. Gaseous, particle and PAH measurements The syngas composition from the FOXBG was determined both by using an on-line micro Gas Chromatograph (lGC) and by taking bag samples. Similar measurements with the same instruments were carried out in the PEBG as well [16]. The sampled gas was cooled and dried by using a water condenser before entering the lGC (see Fig. 2). The lGC (490 lGC, Varian) was configured with two columns using argon as carrier gas and using Thermal Conductivity Detectors (TCD). The molecular sieve separates He, H2, O2, N2, CH4 and CO while the column (PoraPlot U) separates CO2, C2H4, C2H6, C2H2, H2S and COS. The gas to the lGC was further cooled by a Peltier cooler (not shown in Fig. 2) to about 4 °C and was cleaned with a glass wool filter. The cycle time of the lGC was less than four minutes. The bags (10 l aluminum foil Tedlar bags) were filled by a pump in about two minutes and the samples were analyzed with an ordinary Gas Chromatograph (GC). The GC (CP 3800, Varian) was equipped with columns for separating the same compounds as in the case of the lGC except He. The GC was also equipped with additional channels for hydrocarbons using a Flame Ionization Detector (FID) and sulphur compounds using Pulsed Flame Photometric Detector (PFPD). The GCs were calibrated by using a calibration gas with a known concentration of H2, CH4, CO, CO2, C2H6/C2H2 and H2S/COS. Hot gas from the outlet of the FOXBG was withdrawn isokinetically by using a suction probe. The total mass of particulate matter (PM) in the gas was collected on a 47-mm quartz filter that was weighed before and after the experiment. The length of the pipe from the sampling position to the filter was 0.95 m. To avoid condensation of water vapor on the collected particles, the temperature of the filter holder was always kept above 120 °C. Particle sampling in the PEBG was non-isokinetic. The samples were collected the center line of the gasifier by using a nitrogen dilution probe. A low pressure impactor was used to determine PM mass concentration and size distribution [24]. Since the particle concentration measurements are sensitive for the sampling method, the results regarding PM from the FOXBG and the PEBG may not be directly comparable. The concentrations of 16 PAH species in the condensed water from the gas cooler (FOXBG) or the quench (PEBG) were analyzed

158

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

by an accredited laboratory (ALS Scandinavia AB, Sweden) using a method based on US EPA 8270 and CSN EN ISO 6468. 3.4. Operating conditions, experimental and analytical procedure This chapter describes the process operating conditions and experimental procedure for the FOXBG. The experimental procedure for the PEBG has been described elsewhere [16]. An electrical heater (spiraled silicon carbide element, 8 kW) was installed vertically in the center of the freeboard zone in order to heat up the ceramic lining to approximately 1300 °C prior to the experimental campaigns. At the start of the experiment the electrical heater was removed and the fuel feeding system was installed. The process was started by feeding the desired amount of oxygen through the oxygen inlets, followed by starting the fuel feeding system. The system was then operated in combustion mode for a short time (30 s) until the desired k was reached. Between each operating case, the fuel feeding was stopped while the oxygen was still flowing to burn out residual char in the fuel bed. After changing operating conditions, steady state (i.e., in terms of process temperature in the freeboard and gas composition) was reached after 45– 60 min of operation. Process parameters, mainly temperatures but also reactor pressure and water flow to the gas cooler was logged by an Agilent 34970A data logger. The oxygen mass flow was controlled by mass flow controllers (Bronkhorst F201AV), one for each oxygen register. The fuel feed rate was determined by weighing the amount of fuel entering the reactor over 20 min intervals. The fuel feeder was calibrated before starting the experiments. The feed rate over the 20 min periods remained well controlled during the campaigns; however, it is important to point out that maintaining an even feed rate when using wood pellets is difficult, thus fuel mass flow rates fluctuated over shorter time scales. Two, approximately identical experimental campaigns were carried out in this study in order to estimate process variation. The process operating conditions and the resulting process temperatures and mass balance closures are shown in Table 4. The parameter that was varied during the experiments was the O2 feed rate (kg/h), while the rest of the parameters were kept constant. The aim was to operate the gasifier at a fuel feed rate of 3.5 kg/h (corresponding to about 18 kW); however, the actual feed rate, measured every 20th minute, varied between 3.25 and 3.81 kg/h. As a consequence, the actual k varied between 0.282 and 0.474 instead of the intended 0.3 and 0.5. The O2 mass flow rates through the primary and secondary inlets were constant at all operating conditions (corresponding to a k value of approximately 0.3). Varying k was achieved by increasing the amount of O2 to the tertiary inlets. The fuel feeding system was shrouded from the syngas by injecting a small amount of N2

(1–2 l/min) through the top of the freeboard zone. The system pressure was held at ±20 Pa relative to atmospheric pressure throughout the whole campaign. After the gas cooler, He was injected to the gas stream (1 l/min) in order to be able to measure the amount of produced gas. By measuring the amount of He in the produced gas with the lGC it was possible to calculate the actual syngas flow rate. The yield of the different gas components, the elemental mass balance and the CGEs were calculated from the known gas composition, syngas flow and the fuel feed rate. The CGEs in this work were defined based on the lower heating values (LHV) of the fuel and the syngas, respectively. For each operating condition, the condensate (containing water and small amounts of PAHs) from the gas cooler was collected in a bin for 10 min. After collection, the condensate was poured into a glass bottle for weighing PAH analysis. Prior to the analysis, the condensate was stored in a refrigerator at 5 °C. The mass balance closure of C, O and H (see Table 4) was in general very good and realistic (<1.0), which implied that the results presented herein for the FOXBG are reliable. At the lowest k, the mass balance closure for C (i.e., the carbon conversion) was looser. This was expected, due to significant formation of unconverted carbon products (char and soot). 4. Results and discussion 4.1. Major gas components The dry and nitrogen free gas compositions (vol-%) and the gas production rates (mole/kg fuel) for the major gas components (CO, H2, CO2, and CH4) as a function of k for both the PEBG and the FOXBG are presented in Fig. 4. As consequence of different nitrogen injections with the fuel and different gas productions, the N2 concentration in the syngas varied between 2.5 and 9.6% for the FOXBG and between 5.0 and 13.5% for the PEBG. There were no significant differences in either the major gas composition or gas production between the two gasification technologies over the whole range of investigated k values. The concentrations of CH4 and CO2 were significantly affected by increasing k, but only small changes were observed for CO and H2. For both gasifiers, the highest production rates of CO and H2 (i.e., the important species for motor fuels) occurred at k values between 0.35 and 0.4. For lower k, the C and H atoms are stored in hydrocarbons or in incomplete gasification products (tar, soot or char) which lower the production of CO and H2. For k > 0.4, the CO and H2 are instead oxidized to increase the production of CO2 and H2O at the expense of CO and H2. The increase of CO2 production is clearly seen in Fig. 4, while the postulated increase in H2O concentration is not shown since the composition data are presented on a dry basis.

Table 4 Process operating conditions, resulting k values and temperatures, and mass balance closure for the FOXBG. Fuel feeding (kg/h)

a

O2 flow Primary (kg/h)

O2 flow Secondary (kg/h)

O2 flow Tertiary (kg/h)

Lambda (–)

Process temp (°C)

Mass balance closure (–) Ca

O

H

Day 1

3.69 3.59 3.44 3.53 3.25

0.667 0.667 0.667 0.667 0.667

0.762 0.762 0.762 0.762 0.762

0.953 0.715 0.477 0.238 0

0.474 0.44 0.407 0.347 0.323

1169 1138 1093 1046 1003

0.91 0.92 0.95 0.94 0.91

0.93 0.96 0.97 0.99 0.97

0.88 0.89 0.91 0.94 0.94

Day 2

3.81 3.63 3.56 3.52 3.74

0.667 0.667 0.667 0.667 0.667

0.762 0.762 0.762 0.762 0.762

0.953 0.715 0.477 0.238 0

0.46 0.435 0.393 0.348 0.282

1200 1134 1103 1026 971

1.00 0.99 0.95 0.97 0.88

0.98 1.00 0.97 0.97 0.97

0.98 0.99 0.95 1.01 0.97

Can also be called ‘‘carbon conversion”.

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

Fig. 4. Dry and nitrogen free concentrations (vol-%) and the gas production (mol/kg fuel) of the major gas components (CO, H2, CO2 and CH4) in the syngas as a function of k for both the FOXBG and PEBG [16].

Although the two gasifiers behaved similarly, a small difference in gas composition and production for CO, H2 and CO2 may be observed. Compared to the PEBG, the FOXBG produced a slightly higher amount of H2 and CO2 and a slightly lower amount of CO. This small difference can most likely be explained by the water gas shift reaction (R12). For similar k, the process temperature measured by the first thermocouple (see Fig. 2) in the high temperature freeboard region of the FOXBG was slightly lower compared to the process temperature in the PEBG. When the process temperature is reduced, the exothermic nature of the water gas shift reaction shifts the gas composition towards more H2 and CO2 at the expense of CO and H2O. The high concentration of CO and H2 (70–80%) in the syngas from both the FOXBG and the PEBG makes the gas an excellent material for producing synthetic motor fuels and chemicals. For example, the gas from the 3 MWth LTU Green Fuels black liquor gasifier has a CO and H2 concentration of only 61% before it is upgraded to dimethyl ether (DME) in the downstream synthesis plant [25]. The LTU Green Fuels gasifier has been operated for more than 25,000 h [25], during which the usability of syngas for motor fuel production was clearly demonstrated. It is therefore reasonable to argue that upgrading the higher quality gas produced in the FOXBG and PEBG to DME or other synthetic products is possible. 4.2. Minor gas components and particles The concentrations (vol-%, dry basis) and gas production rates (mol/kg fuel) of the minor gas components (C2H4, C2H2, and

159

Fig. 5. Dry and nitrogen free concentration (vol-%) or gas production (mol/kg fuel) of C2H2, C2H4, C2H6, C6H6 and particles (g/Nm3 or g/kg fuel) as a function of k for both the FOXBG and PEBG. The gaseous concentrations and production from PEBG are adopted from [16] and the particle concentration and production from [24].

C6H6) are presented in Fig. 5, along with the particle concentrations (g/Nm3) and particle production rates (g/kg fuel). Unfortunately, gas concentration and production data for C2H4 and C2H2 from the PEBG have not been published elsewhere and are therefore not available for comparison with FOXBG results. The concentration and production of both the minor gas species and particles reduced significantly when k was increased in the FOXBG. For k > 0.45, only trace levels (100 ppm) of minor gas species were detected in the produced gas. At k > 0.4, the gas concentration and production of C6H6 for both gasifiers was in the same order of magnitude, but at low k (below 0.35) the concentration and production of C6H6 was about twice as high in the FOXBG compared to that in the PEBG. One possible explanation for the observed trend may be a non-ideal flow pattern being present in the freeboard zone of the FOXBG when the amount of tertiary oxygen is reduced, leading to a slip of C6H6. The particle (i.e., soot) concentration and production in the FOXBG was higher compared to that in the PEBG. However, this observation may be due to the fact that different sampling systems were used in the two gasifiers, as mentioned previously. The particle production from the FOXBG (1–8 g/kg fuel) was also in the same order magnitude, or slightly lower than that (10–40 g/kg fuel) obtained during high-temperature (1000–1350 °C) entrained flow gasification of wood, performed in an electrical heated laboratory (0.55–0.57 kg/h fuel supply) reactor [19]. This indicates that the performance of the FOXBG in terms of particle production is competitive compared to the entrained flow gasification method.

160

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

4.3. Polyaromatic hydrocarbons Since the process temperatures in both the FOXBG and PEBG were significantly above 900 °C, the tar produced during gasification consists of PAHs [26], and therefore analysis of the concentration of PAHs also yields insight on tar concentration. The PAH species identified in the condensates from the FOXBG were naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, and pyrene at high k (0.436 and 0.46). In addition to these PAH species, benz[b]fluoranthene and benz [k]fluoranthene were also identified at low k (0.282 and 0.348). The total concentration of PAHs (mg/l) in the condensate from the FOXBG and the PEBG are presented in Table 5. To make the PAH concentrations directly comparable, the PEBG data presented in [27] have been corrected for the dilution of water caused by the water sprays in the PEBG quench. For both gasifiers, the trend is that the concentration of PAHs in the condensate was reduced as k increased. However, due to the complicated sampling procedure and PAH analysis methodology, the data fluctuated. At low k, the PAH concentrations in the condensate were within the same order of magnitude for both gasifiers. However, at higher k, the PAH concentration in the condensate from the PEBG was an order of magnitude lower compared to that in the FOXBG condensate. The PAH concentration in the hot gas (mg/Nm3) prior to the gas cooler in the FOXBG (see Table 5) was calculated from the known PAH concentration in the condensate and from the condensate production rate, assuming that all of the PAHs produced during gasification were absorbed in the condensate. Naturally, one may question this assumption; however, when the gas cooler and the pipe after the gas cooler were cleaned after a day of operation, only deposits of dry particles (i.e., soot) were observed – this indicates that the production of tars was very low. Compared to other types of gasifiers, such as ordinary fixed bed [2,9,14], indirect fluidized bed [28] or two stage gasifiers [29], the tar concentration and production (i.e., the PAH concentration) of the FOXBG were several orders of magnitude lower than those of the PEBG. According to Sikarwar et al. [2], the allowable tar levels in the gas depend on the downstream application: values around 50 mg/Nm3 for gas engines, 5 mg/Nm3 for gas turbines, 1 mg/Nm3 for fuels cells and between 0.1 and 1 mg/Nm3 for Fisher-Tropsch or methanol synthesis are typical. This indicates that the gas produced in the FOXBG can be used directly as a fuel in gas engines and gas turbines without extensive treatment for separating the tar, but for synthetic fuel production the tar levels needs to be reduced by an order of magnitude or more. Since the gas after the gas cooler is at room temperature [30], reducing tar levels can most likely be carried out by using different cold gas cleanup technologies (e.g., wet scrubbing with liquid absorbent, wet electrostatic precipitator). 4.4. Gasification efficiency The CGEs for both gasifiers as a function of the investigated k are presented in Fig. 6. As a consequence of similar gas production (see Fig. 5) in the PEBG and the FOXBG, the CGEs values were also similar. For both gasifiers, the optimal CGEs seem to occur at k values

Fig. 6. CGEs for the FOXBG and the PEBG [16] as a function of k.

around 0.35. For the FOXBG, the energy content of unconverted particles (assuming heating value for graphite) corresponds to 0.3–1.3% of the energy input by the fuel. This means that the CGEs could be increased by up to 1.3% if all particles were converted to gas instead of leaving the reactor as carbon residue. Following the work of Weiland et al. [5], it must be remembered that the PEBG technology requires additional fuel preparation (milling), while the FOXBG does not which lowers the plant efficiency. The plant efficiency (gplant), assuming that the energy used to mill the fuel is generated from power generation by burning a part of the syngas in a gas turbine or gas engine is defined as [5]:

gplant ¼

_ cg LHV cg  Wgmilling m _ fuel m el ; _ fuel LHV fuel m

ð3Þ

where Wmilling (MJ/kgfuel) is the prime electrical power consumption for milling and gel is the efficiency of the electrical power generation process (gel = 0.4 in this work). Weiland et al. [5] estimated that the gplant can be reduced by up to 9%-units for woody biomass by taking into account the milling step. This means that there is great potential in the FOXBG concept, since fuel milling is not a requirement in that process. 4.5. Heating value of the gas In addition to syngas production, there is also a possibility to use the gas as fuel for heat and power production. For power production, where the gas is burned in gas engines or gas turbines, a high heating value of the produced gas is a crucial performance parameter since it facilitates ignition of the and increases the thermodynamic efficiency of the process. Furthermore, high flame temperatures are important in the metallurgical industry. Fig. 7 presents the lower heating value (LHV) of the gas (dry) from the FOXBG. The LHV is inversely proportional to k and varies between 8.7 and 11.3 MJ/Nm3 – this value is significantly higher compared

Table 5 Concentration and production of PAHs. FOXBG

PEBG

k

0.282

0.348

0.435

0.46

0.35

0.425

0.50

PAH in gas condensate (mg/l) PAH in gas (mg/Nm3) PAH production (mg/kg fuel)

38 4.8 5.7

63 6.4 8.5

14 1.7 2.5

27 3.8 5.1

56.5 7.1 9.5

0.4 0.1 0.1

1.4 0.4 0.5

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

161

Acknowledgments This work has been founded by Bio4Energy and RISE – the institutions are highly acknowledged by the authors of this work. Calle Yllipää and Matthias Lundgren are also acknowledged for their invaluable assistance during the experiments. Furthermore, Dr. Roger Molinder and Dr. Pal Toth are acknowledged for language editing.

References

Fig. 7. LHV (MJ/Nm3) of the produced gas (dry, not nitrogen free) from the FOXBG.

to that of the fuel gas generated during air blown gasification, which often is around 4–7 MJ/Nm3 [2]. The heating value of the gas from the FOXBG is also competitive compared to the gas produced from the PEBG which has a heating value of 7.7–11.2 MJ/ Nm3 [31]. One can therefore conclude that the gases generated in both the FOXBG and the PEBG are suitable for heat and power applications. 4.6. Optimal operation window for the FOXBG To a small extent, the optimal operation window for the FOXBG depends on the application. For heat and power production, the optimal k value is around 0.3 since both the gasification efficiency (i.e., CGEpower) and the LHV of the gas reaches its maximum value close to that operation point (see Figs. 6 and 7). For motor fuel production, a slightly higher k is desired (0.4–0.45), since the CGEfuel is relatively constant at those conditions (see Fig. 6), while at the same time, the concentration of CH4 and heavier hydrocarbons is reduced at k values above 0.4 (see Fig. 5). 5. Conclusions In this work, a laboratory scale (18 kW), atmospheric, oxygen blown, fixed bed gasifier with a high temperature freeboard zone was characterized with respect to syngas composition, syngas production, and gasification efficiency. The results were compared to those obtain during the operation of a pilot scale (200–600 kW), pressurized (2–7 bar), oxygen blown, entrained flow gasifier using the same feedstock (wood pellets and wood powder) as fuel. The results presented in the current work clearly show that it is possible to produce a syngas of similarly high quality (high concentration and production of CO and H2, low concentration and production of CH4, heavier hydrocarbons and soot) and gasification efficiency during fixed bed gasification as during entrained flow gasification. The generation of a high quality syngas is therefore not specific to the entrained flow gasification technology; instead, it is coupled to a high overall process temperature (above 1000 °C) and aerodynamics (good mixing) in the gasification reactor, which, as shown here, can also be obtained in an oxygen blown fixed bed gasifier with high temperature freeboard conditions. As fixed bed gasification does not require extensive fuel pretreatment (milling), this observation opens the possibility for the more efficient production of high quality syngas in the future, since the pretreatment cost can be significantly reduced. Furthermore, feedstocks that are difficult to pulverize can also be used as fuel in a fixed bed gasifier.

[1] Higman C, van der Burgt M. Gasification. Burlington, Massachusetts: Gulf Professional Publishing; 2008. [2] Sikarwar VS, Zhao M, Clough P, Yao J, Zhong X, Memon MZ, et al. An overview of advances in biomass gasification. Energy Environ Sci 2016;9:2939–77. [3] Andersson J, Lundgren J. Techno-economic analysis of ammonia production via integrated biomass gasification. Appl Energy 2014;130:484–90. [4] Gunarathne DS, Mellin P, Yang W, Pettersson M, Ljunggren R. Performance of an effectively integrated biomass multi-stage gasification system and a steel industry heat treatment furnace. Appl Energy 2016;170:353–61. [5] Weiland F, Nordwaeger M, Olofsson I, Wiinikka H, Nordin A. Entrained flow gasification of torrefied wood residues. Fuel Process Technol 2014;125:51–8. [6] Leijonhorst EJ, Assink D, van de Beld L, Weiland F, Wiinikka H, Carlsson P, et al. Entrained flow gasification of straw- and wood-derived pyrolysis oil in a pressurized oxygen blown gasifier. Biomass Bioenergy 2015;79:166–76. [7] Na JI, Park SJ, Kim YK, Lee JG, Kim JH. Characteristics of oxygen-blown gasification for combustible waste in a fixed-bed gasifier. Appl Energy 2003;75:275–85. [8] Skhonde MP, Matjie RH, Bunt JR, Strydom AC, Schobert H. Sulfur behavior in the Sasol-Lurgi fixed-bed dry-bottom gasification process. Energy Fuels 2009;23:229–35. [9] Umeki K, Yamamoto K, Namioka T, Yoshikawa K. High temperature steam-only gasification of woody biomass. Appl Energy 2010;87:791–8. [10] Thanapal SS, Annamalai K, Sweeten JM, Gordillo G. Fixed bed gasification of dairy biomass with enriched air mixture. Appl Energy 2012;97:525–31. [11] Yoon SJ, Son YI, Kim YK, Lee JG. Gasification and power generation characteristics of rice husk and rice husk pellet using a downdraft fixed-bed gasifier. Renewable Energy 2012;42:163–7. [12] Ouadi M, Brammer JG, Kay M, Hornung A. Fixed bed downdraft gasification of paper industry wastes. Appl Energy 2013;103:692–9. [13] Pu G, Zhou HP, Hao GT. Study on pine biomass air and oxygen/steam gasification in the fixed bed gasifier. Int J Hydrogen Energy 2013;38:15757–63. [14] Cerone N, Zimbardi F, Villone A, Strjiugas N, Kiyikci EG. Gasification of wood and torrefied wood with air, oxygen, and steam in a fixed bed pilot plant. Energy Fuels 2016;30:4034–43. [15] Lenis YA, Pérez JF, Melgar A. Fixed bed gasification of Jacaranda Copaia wood: effect of packing factor and oxygen enriched air. Ind Crops Prod 2016;84:166–75. [16] Weiland F, Wiinikka H, Hedman H, Wennebro J, Pettersson E, Gebart R. Influence of process parameters on the performance of on oxygen blown entrained flow biomass gasifier. Fuel 2015;153:510–9. [17] Adeyemi I, Janajreh I, Arink T, Ghenai C. Gasification behavior of coal and woody biomass: validation and parametrical study. Appl Energy 2017;185:1007–18. [18] Dufor A, Valin S, Castelli P, Thiery S, Boissonnet G, Zoulalian A, et al. Mechanisms and kinetics of methane thermal conversion in a syngas. Ind Eng Chem Res 2009;48:6564–72. [19] Qin K, Lin W, Jensen PA, Jensen AD. High-temperature entrained flow gasification of biomass. Fuel 2012;93:589–600. [20] Septien S, Valin S, Dupont C, Peyrot M, Salvador S. Effect of particle size and temperature on woody biomass fast pyrolysis at high temperature (1000– 1400°C). Fuel 2012;97:202–10. [21] Billuad J, Valin S, Peyrot M, Salvador S. Influence of H2O, CO2 and O2 addition on biomass gasification in entrained flow reactor conditions: experiments and modelling. Fuel 2016;166:166–78. [22] Weiland F, Hedman H, Marklund M, Wiinikka H, Öhrman O, Gebart R. Pressurized oxygen blown entrained-flow gasification of wood powder. Energy Fuels 2013;27:932–41. [23] Wiinikka H, Johansson AC, Wennebro J, Carlsson P, Öhrman OGW. Evaluation of black liquor gasification intended for synthetic fuel or powder production. Fuel Process Technol 2015;139:216–25. [24] Weiland F, Sweeney DJ, Wiinikka H. Extractive sampling of gas and particulates from the reactor core of an entrained flow biomass gasifier. Energy Fuels 2016;30:6405–12. [25] Jafri Y, Furusjö E, Kirtania K, Gebart R. Performance of a pilot-scale entrained flow black liquor gasifier. Energy Fuels 2016;30:3175–85. [26] Milne TA, Evans RJ, Abatzoglou N. Biomass gasifiers ‘‘tars”: their nature, formation, and conversion. NREL, Golden, Colorado; November 1998. [27] Molinder R, Öhrman OGW. Characterization and cleanup of wastewater from pressurized entrained flow biomass gasification. ACS Sustain Chem Eng 2014;2:2063–9.

162

H. Wiinikka et al. / Applied Energy 191 (2017) 153–162

[28] Larsson A, Israelsson M, Lind F, Seemann M, Thunman H. Using ilmenite to reduce the tar yield in a dual fluidized bed gasification system. Energy Fuels 2014;28:2632–44. [29] Zeng X, Wang F, Li H, Wang Y, Dong L, Xu G. Pilot verification of a low-tar twostage coal gasification process with a fluidized bed pyrolyzer and fixed bed gasifier. Appl Energy 2014;115:9–16.

[30] Abdoulmoumine N, Adhikari S, Kulkarni A, Chattanathan S. A review on biomass gasification syngas cleanup. Appl Energy 2015;155:294–307. [31] Weiland F. Pressurized entrained flow gasification of pulverized biomass: Experimental characterization of the process performance. PhD Thesis, Department of Energy Science and Mathematics, Luleå University of Technology, Luleå; 2015.