- Email: [email protected]
Gasification of a solid recovered fuel in a pilot scale fluidized bed reactor
Fuel 117 (2014) 528–536
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Gasification of a solid recovered fuel in a pilot scale fluidized bed reactor Umberto Arena a,b,⇑, Fabrizio Di Gregorio b a b
Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, Via Vivaldi 43, 81100 Caserta, Italy AMRA s.c.a r.l., Analysis and Monitoring of Environmental Risk, Via Nuova Agnano 11, 80125 Napoli, Italy
h i g h l i g h t s Pilot-scale investigation of fluidized bed gasification of a solid recovered fuel. Tests under conditions of thermal/chemical steady state at various equivalence ratios. Complete composition of the syngas, including tar, particulate and acid/basic gases. Tar characterization according to the ECN classification. Partitioning of main inorganic elements in the entrained fines and bed particles.
a r t i c l e
i n f o
Article history: Received 7 August 2013 Received in revised form 14 September 2013 Accepted 17 September 2013 Available online 27 September 2013 Keywords: Gasification Solid recovered fuel Fluidized bed reactor Waste-to-energy
a b s t r a c t The paper investigates the technical feasibility of an air gasification process of a Solid Recovered Fuel (SRF) obtained from municipal solid waste. A pilot scale bubbling fluidized bed gasifier, having a feedstock capacity of about 70 kg/h and a maximum thermal output of about 400 kW, provided the experimental data: the complete composition of the syngas (including the tar, particulate and acid/basic gas contents), the chemical and physical characterization of the bed material and that of entrained fines collected at the cyclone. The experimental runs were carried out by reaching a condition of thermal and chemical steady state under values of equivalence ratio ranging from 0.25 to 0.33. The results indicate that the selected SRF can be conveniently gasified, yielding a syngas of valuable quality for energy applications. The rather high content of tar in the syngas indicates that the more appropriate plant configuration should be that of a ‘‘thermal gasifier’’, with the direct combustion of the syngas in a burner ad hoc designed, coupled with an adequate energy-conversion device. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction and framework The amount of the municipal solid waste (MSW) increases all over the world as a consequence of growing urbanization and the standard of living. A recent study indicates that urban areas consume about 75% of natural resources and produce 50% of the total amount of waste in the world [1]. A remarkable amount of this waste contains materials (such as paper, glass, metals, wood and plastics), mainly deriving from packaging, which can efficiently be recycled for resource recovery. But not all the MSW can be recycled and, moreover, the waste recycling processes always generate significant amounts of residues, in some cases having very high heating values [2,3]. It is today recognized that in a fully sustainable waste management system no one process is suitable for all waste streams [4–8]. In other words, no single waste management
⇑ Corresponding author at: Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, Via Vivaldi 43, 81100 Caserta, Italy. Tel.: +39 0823 274414; fax: +39 0823 274592. E-mail address: [email protected] (U. Arena). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2013.09.044
practice (i.e. landfill, recycling, biochemical or thermochemical conversion) can handle the full array of waste types and, at the same time, satisfy the recognized criteria of an integrated and sustainable management system [3,9]: (i) to minimize use of landfills and ensure that no landfilled waste is biologically active or contains mobile hazardous substances; (ii) to minimize operations that entail excessive consumption of raw materials and energy without yielding an overall environmental advantage; (iii) to maximize recovery of materials, albeit in respect of the previous point; and (iv) to maximize energy recovery for materials that cannot be efficiently recycled, in order to save both landfill volumes and fossil-fuel resources. With reference to the latter criterion, two remarkable waste streams have a great potential of being used in sustainable energy recovery processes: the unsorted residual waste (i.e. the dry organic fraction of MSW that cannot conveniently be recycled and it is not suitable for biological treatments) and the high heating value residues produced by some recycling chains (that in some countries cannot be landfilled [10]). These non-hazardous waste fractions can be turned into a Solid Recovered Fuel (SRF), i.e. in a sufficiently homogeneous waste-derived
529
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
fuel, obtained as the result of a mechanical process to comply with a CEN standard [11]. The European standardization defines different grades of SRF on the basis of specific composition ranges, mainly low heating value and chlorine and mercury contents. An efficient way for thermochemical exploitation of SRFs is fluidized bed gasification. Gasification is the conversion of solid fuel to a synthesis-gas through gas-forming reactions occurring in presence of an amount of oxidant lower than that required for the stoichiometric combustion. The resulting fuel gas (named ‘‘producer gas’’ or ‘‘syngas’’) can be utilized in a separate energy conversion device: it contains large amounts of not completely oxidized products (mainly CO, H2 and lower amounts of CH4), together with different organic (tar) and inorganic (H2S, HCl, NH3, HCN, alkali metals) impurities and particulates [12]. Fluidization is one of the most interesting gasification technologies, mainly for the high quality of gas–solid contact and the very efficient mass and heat transfers, but also for its good process flexibility, which accommodates variation in fuel quality and allows to utilize different fluidizing agents, reactor temperatures and gas residence times, to add reagents along the reactor height and to operate with or without a specific catalyst. A large-scale fluidized bed gasifier has been recently put in operation in Lahti, southern Finland: it is able to treat 250 kt/y of SRF derived from household waste (origin sorted) together with industrial waste, demolition wood and waste wood from industry and to produce 50 MW of power and 90 MW of district heat [13]. But there is also a great interest for SRF gasification for energy production in small and medium scale fluidized bed plants [14–16]. The paper describes the results of a research program aimed at assessing the technical feasibility of a fluidized bed gasifier able to treat 5000 t/y of a SRF, obtained as one of the output solid streams of a sorting and mechanical treatment of MSW collected in a urban area of the Middle of Italy. To this aim, a number of tests with this SRF were carried out in a pilot scale bubbling fluidized bed gasifier (BFBG). The obtained experimental results were processed by mass and energy balances, in order to obtain data and information useful to define both a suitable plant configuration and the related design solutions for a fully sustainable energy generation.
2. The fluidized bed gasifier, the experimental procedures and the material tested 2.1. The fluidized bed gasifier The design and operating parameters of the pilot scale bubbling fluidized bed gasifier are reported in Table 1. The plant has a feedstock capacity between 30 and 100 kg/h, depending on the type of fuel, and a maximum thermal output of about 400 kW. It is composed of three main sections, as it is sketched in Fig. 1: the feeding system, the fluidized bed gasifier and the syngas cleaning unit. The first can be divided in the blast and fuel feeding. The blast feeding is heated up to about 150 °C by a first electric heater, then sent to a mixing point with an optional stream of steam at about 150 °C, and finally heated by a second electric heater up to the desired inlet temperature at the fluidized bed bottom. In all the experiments reported here, only air was used as blast agent and always injected at the bed bottom while the fuel was always fed by means of an overbed system. The gasification section is composed of a cylindrical reactor, 0.381 m ID and 5.90 m high, which is heated up to the reaction temperature by the sensible heat of pre-heated blast gases and by a set of three external electrical furnaces. The syngas cleaning section is composed of a high efficiency cyclone (for abatement of fine particles), a simple wet scrubber (for removal of tar, residual dust, acid gases and ammonia) and a flare.
Table 1 Main design and operating parameters of the pilot scale BFB gasifier. Geometrical parameters
Feedstock capacity Thermal output Typical bed amount Oxidizing agent Feeding system Range of bed temperatures Range of fluidizing velocities Produced gas treatments Safety equipments
Internal diameter: 0.381 m Total height: 5.90 m Reactive zone height: 4.64 m Wall thickness: 12.7 mm Up to 100 kg/h (on the basis of the fuel heating value) Up to about 400 kW 145 kg Air (but also oxygen, steam, and their mixtures) Over-bed water-cooled screw feeder 700–950 °C 0.3–1 m/s Cyclone, scrubber, flare Water seal, safety valves, rupture disks, alarms, nitrogen line for safety inerting
2.2. The experimental procedures Each run has a start-up of about 3 h, during which three electric heaters located along the reactor lead the temperature up to about 700 °C, while the bed is fluidized at a fixed velocity. At this point, the flow rates of SRF and air are adjusted in order to obtain the desired value of the equivalence ratio ER, defined as the ratio between the oxygen content of air supply and that required for the stoichiometric complete combustion of the fuel effectively fed to the reactor. Under the selected operating conditions of ER and air preheating temperature, and without any more thermal assistance of external heaters, the reactor gradually reaches a thermal and chemical steady state. The gas and solids sampling procedures are then activated and measurements of pressure, temperature, blast flow rates and syngas composition upstream of and downstream of the wet scrubber are taken and averaged over the whole period of steady state, which usually is kept for about 2 h. Fig. 2 shows the successive steps of the described procedure, reporting the time profile of gas composition and bed temperature for a typical run. In order to increase the reliability of measurements, in all the runs the composition of the syngas downstream of the cleaning section was on-line monitored by means of two systems: a series of IR analysers (Horiba VA-3115 for CO, CO2 and O2, Horiba VA3001 for CH4 and Teledyne Anal. Instr.-2000 for H2) and an Agilent 3000 gas-chromatograph equipped with 4 different columns (Mol Sieve, PoraPlot, OV, Alumina) for the detection of a wide spectrum of syngas compounds. Gas was also sampled in Tedlar bags at two other points (2.3 m from the reactor bottom and the gasifier exit) and sent to off-line measurements. Two methods were used to evaluate the amount and composition of tar, due to the relevance of this measurements for the technical and economic feasibility of the process. The first assumes that tar is composed of all organic compounds with a molecular weight larger than benzene, excluding soot and char: then it conservatively imputes to the tar amount the whole carbon loading which, as a result of a mass balance on atomic species, cannot be attributed either to the producer gas or to the solids collected at the cyclone or present inside the bed. The second method samples the condensable species by means of a system composed of four in-series cooling coils, a suction pump and a flow meter operated with a syngas flow rate of about 300 dm3N /h for 30 min: the condensed hydrocarbons are then off-line analyzed with a specific pre-treatment in a Perkin–Elmer Clarus 500 gas chromatograph coupled with a mass spectrometer (GC–MS). The first conservative method was utilized for quantitative determination of tar concentration in the obtained producer gas. The second procedure was instead used to detect tar
530
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
°
Fig. 1. Schematic of the pilot-scale fluidized bed gasifier, with the indication of some operating parameters. Fi, with i = 1–18, indicates different gas and solid streams.
Fig. 2. Time profile of gas composition and bed temperature in a typical run with the SRF.
compounds belonging to the classes between 2 and 5 of the classification proposed by the Energy research Center of The Netherlands (ECN) [17]. Hydrogen chloride, hydrogen sulphide and ammonia were measured by bubbling the producer gas through two pairs of gas stripping bubblers, connected in series and containing basic and acid solutions respectively, and subsequently analyzed by means of a Dionex DX-120 ion chromatograph. Data obtained from on-line and off-line gas measurements and from chemical analyses of solid samples were processed to develop mass balance on atomic species and the related energy balance for each test. The flow rate of produced syngas was determined by the ‘‘tie-component method’’ [18] applied to the value of nitrogen content in the dry syngas, as obtained by (on-line and off-line) GC measurements, and adequately corrected to take into account the nitrogen fed into the gasifier with the waste and that leaving it as ammonia. 2.3. The fuel and the bed material The BFBG was fed with an SRF produced by a commercial sorting platform located in the Middle of Italy. The received SRF was mechanically processed to obtain mono-size cylindrical pellets in order to avoid any effect related to fuel size and to increase the density of material to be fed into the gasifier. The size of pellets, 5 mm of diameter and 30 mm of length, is compatible with
BFBGs: for these reactors the maximum size should be smaller than 10–20 cm, in order to avoid floating effects, and it is preferable to avoid fuels with a too high percentage of very small particles, in order to avoid lifting out of too many fines out of the bed. Table 2 reports the ultimate analysis and low heating value of the pelletized SRF, together with the composition of its inorganic fraction, as obtained by analysing the different stocks utilized in the experimental runs. The reported range of variation is rather limited, even though the original unsorted residual waste was collected at different time and in different catchment areas. Data reported in Table 2 indicate the SRF pellets as a good quality fuel [19] with an ash fusion temperature in the range 1400– 1700 °C, which is sufficiently higher than that typical of FBG operating conditions [20]. The bed material used during the experimental runs is olivine, due to its good performance as tar removal agent, reported in several studies on gasification of biomass [21–23] and polyolefin plastic wastes [14,24]. The utilized olivine has a particle size range of 200–400 lm and a particle density of 2900 kg/m3. It is a neo-silicate of Fe and Mg coming from Austria and composed mainly of MgO, SiO2, and Fe2O3 but even Al2O3, Cr2O3, Mg3O4 and CaO. It was pre-calcined at the production site and further calcined just before each test for about 8 h in oxidizing atmosphere at 800 °C, in order to improve their potential catalytic activity as tar removal agent [21,23].
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Cn Hm þ n=2O2 ! nCO þ m=2H2
Table 2 Main chemical properties of the tested SRF. Ultimate analysis (%wt, C H N S Cl O (by diff.) Moisture Ash
ar)
41.2–45.4 6.0–6.5 0.66–0.70 0.1–0.3 0.1–0.2 22.9–24.2 3.7–9.1 18.5–20.4
Heating value (MJ/kgfuel,ar) Low heating value (LHV) a
18,600–21,300
Ash composition (mg/kgdb) Aluminum as Al2O3 Antimony as Sb Arsenic as As Cadmium as Cd Calcium as CaO Chrome as Cr Cobalt as Co Copper as Cu Iron as Fe2O3 Lead as Pb Magnesium as MgO Manganese as MnO Mercury as Hg Nickel as Ni Phosphorus as P2O5 Potassium as K2O Silicon as SiO2 Sodium as Na2O Vanadium as V Zinc as Zn Chlorine as Cl
10,500–17,900 <50 <50 <50 62,500–82,600 <50–130 <50 106–319 8480–11,600 87–162 3130–6720 185–282 <50 <50–67 2460–3200 5410–9070 31,100–35,200 <50 <50 314–386 1200–1950
Ash melting temperature Half fluid temperature (HFT)b (°C)
1420–1730
ar = as received; db = on dry basis. a Channiwala and Parikh [19]. b Kim et al. [20].
3. Results and discussion 3.1. Composition of the producer gas The pilot scale BFBG was operated by injecting the SRF into a bed of olivine particles fluidized at a fixed superficial gas velocity of about 0.7 m/s and temperatures of about 850 °C, under different values of equivalence ratio, in order to quantitatively assess its behavior in the fluidized bed gasification process. Table 3 lists the operating conditions of all tests, together with the main experimental results. In particular, some variables, such as the bed temperature at steady state, the syngas temperature at the reactor exit and the overall fluidizing velocity (calculated by taking into account the flow rate of the produced syngas and assuming that it was completely generated inside the bed) have been more properly considered as state variables. They are the answers of the FBG system to the set of operating variables, mainly the equivalence ratio, fuel heating value and air preheating temperature. Data in Table 3 also shows that the flow rate of entrained fines is in the order of 0.1 kg/kgfuel, while the flow rate of ash injected with the fuel is about 0.2 kg/kgfuel: thereby, about a half of ashes escapes the reactor as entrained fines while the other part remains into the bed and is periodically extracted from the bottom of the fluidized bed. Fig. 3 reports the composition of the obtained producer gas in all the air gasification tests as a function of equivalence ratio ER. An increase of ER implies a larger amount of oxygen available for reaction with volatiles in the pyrolysis zone [24,25], and then a greater extension of partial oxidation reactions:
531
ð1Þ
Accordingly, the diagrams show a limited reduction of the contents of methane, ethylene and CnHm hydrocarbons (with n = 2–4) and a corresponding increase of that of carbon monoxide. The latter could be also related to the (much) slower reactions of steam and dry reforming:
Cn Hm þ nH2 O ! nCO þ ðn þ m=2ÞH2
ð2Þ
Cn Hm þ nCO2 ! 2nCO þ m=2H2
ð3Þ
The concentration of hydrogen shows a very limited variations for the opposite effects of H2 generated by the recalled reactions of hydrocarbons and that consumed by its own oxidation. Similarly, the CO2 concentration as a function of ER appears to have limited variations, as a consequence of the opposite effects of Boudouard reaction:
C þ CO2 ! 2CO
ð4Þ
and CO partial oxidation. The greater extension of exothermic partial oxidation reactions also generates an overall effect of reactor temperature increase, which is described by the average values of bed temperatures reported in Fig. 4. They range from the lowest value of 849 °C at ER = 0.256 to the highest value of 932 °C at ER = 0.332. As mentioned above, since the tests were carried out under thermal and chemical steady state conditions, the bed temperature is a state variable of the system, i.e. its answer to the set of operating conditions. The air preheating temperature was gradually reduced to compensate this increase, even though its role appears limited, as it is suggested by the data related to the two tests at ER = 0.256. It is noteworthy that the minimum value of ER to efficiently operate the pilot scale gasifier without additional heating was 0.25: lower values would reduce too much the carbon conversion efficiency. A single test was carried out by using a mixture of air and steam as fluidizing stream, with a steam-to-fuel ratio of 0.69 kgsteam/ kgwaste, i.e. a steam-to-carbon ratio of 1.46 kgsteam/kgC-waste. The process performance for this test greatly got worse: this is likely related to lowering of reactor temperature (in our test down to 802 °C) determined by the heat requirement of endothermic steam reforming reactions. This consequently induces a reduction of carbon conversion or the necessity of more oxygen to maintain the thermal level (which in turn implies a reduction of the heating value of the producer gas) [25]. A specific and deeper investigation is necessary to better understand this aspect. 3.2. Process performance parameters The four diagrams of Figs. 5 and 6 report the values of the main process performance parameters, as a function of equivalence ratio: – The LHV of the producer gas (excluding the contribution of tar compounds). – Its specific energy, i.e. the chemical energy of the producer gas per kg of SRF. – The Carbon Conversion Efficiency, CCE, defined as the carbon flow rate converted to gaseous products with respect to that fed to the reactor with the SRF. – The Cold Gas Efficiency, CGE, defined as the fraction of the chemical energy of the waste which is transferred to the producer gas. As expected on the basis of previous studies on biomass and waste-derived fuel [14–16,21,22,24,25], the LHVs reduce as ER
532
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Table 3 Operating conditions and experimental results of all the tests. Operating conditions Test# ER () Air/fuel (A/F) (kgair/kgfuel) Bed amount (kg) SRF flow rate (kgfuel/h) Fluidizing velocity (m/s) Air preheating temperature (°C) Process results and parameters Overall fluidizing velocitya (m/s) Bed temperature at steady state (°C) Syngas temperature at reactor exit (°C) Syngas production (volumetric basis) (m3N /kgfuel) Syngas production (mass basis) (kgsyngas/kgfuel) Syngas LHV (kJ/m3N ) Specific energy (kW h/kgfuel) CCE, CGE, -
4 0.256 1.47 146 61.4 0.70 375
6 0.255 1.54 146 58.6 0.70 520
3 0.272 1.56 146 57.65 0.71 355
5 0.302 1.73 146 51.9 0.72 329
1 0.318 1.82 146 49.4 0.73 542
2 0.332 1.90 146 47.40 0.75 273
0.77 849 807 1.62
0.79 852 780 1.73
0.78 869 857 1.72
0.79 879 801 1.91
0.82 898 822 2.04
0.79 932 815 2.00
1.99 5390
2.18 5652
2.10 5550
2.34 5160
2.56 4910
2.46 4690
2.42 0.70 0.53
2.71 0.80 0.56
2.66 0.75 0.58
2.78 0.81 0.61
2.79 0.92 0.61
2.61 0.79 0.57
Syngas composition (%) N2 CO2 CO H2 CH4 C2H4 C2H6 C2H2 C3H6 C6H6 Elutriated fines (g/kgfuel) Carbon elutriated fines (gC/kgC-fuel) Tar (g/m3N )b Naphthalene (% of identified tar species) Pyrene (% of identified tar species) HCl (mg/m3N )c
61.77 13.36 9.74 8.48 4.18 2.19 0.04 0.20 0.02 0.20 87.0 23.7 73
60.64 15.69 8.98 7.11 4.20 2.72 0.08 0.12 0.10 0.32 67.3 19.4 47
61.67 12.28 10.47 8.44 4.57 2.14 0.04 0.21 0.03 0.15 64.9 32.5 58
61.86 12.83 10.40 8.24 4.35 2.05 0.04 0.16 0.02 0.10 102.8 20.2 39
60.66 14.04 12.73 7.08 3.33 1.78 0.07 0.18 0.02 0.08 62.7 24.2 5
64.61 11.66 1058 7.68 3.78 1.27 0.02 0.18 0.00 0.21 60.7 22.7 42
29 51 10.5
38 3.0 n.a.
65 3.2 3.1
12 60 117.5
76 3.5 2.1
80 1.3 21.7
H2S (mg/m3N )c
7.4
n.a.
0.2
39.1
0.5
0.6
NH3 (mg/m3N )c
74.2
n.a.
134
39.6
2.7
18.7
a
This value was calculated taking into account the produced syngas (then assuming that it was completely generated inside the bed). This value was conservatively determined on the basis of mass balance on atomic species. The samples of gas for measurements of NH3, HCl and H2S were all taken at the reactor exit and then before the cleaning gas system, with the exclusion of test#1 for which they were taken downstream of the wet scrubber. b
c
Fig. 3. Composition of the producer gas in all the gasification tests. Fig. 4. Bed temperature and air preheating temperature in all the gasification tests.
increases. This is the result of opposite effects: the strong reduction of content of hydrocarbons having very high heating values (mainly methane, ethylene and acetylene), and the corresponding increase of carbon monoxide contents. The values reported in Fig. 5 range between a minimum of 4.7 MJ/m3N and a maximum of 5.7 MJ/m3N , and grow up to more than 7 MJ/m3N if the energy content of the tar compounds is taken into account (assuming all the tar as naphthalene) [15]. In the range of experimental conditions tested, the specific energy of the producer gas slightly increases,
from 2.4 kW h/kgwaste to 2.8 kW h/kgwaste, as a consequence of ER increasing. This is related to the corresponding slight increase of both CCE and CGE as a function of ER in the tested range, which is shown in the diagrams of Fig. 6. In particular, higher values of CCE are mainly due to the greater extension of oxidation reactions of carbon fines, being char conversion not much affected by oxygen [25]. The increase of CGE is instead related to the reduced production of condensable heavy hydrocarbons (tar) that is shown in
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
533
Fig. 7. Tar content in the producer gas at the reactor exit in all the gasification tests.
Fig. 5. Low heating value and specific energy of the obtained producer gas in all the gasification tests.
Fig. 6. Carbon conversion and cold gas efficiencies in all the gasification tests.
Fig. 7. The values of all these process parameters, together with the other results of the experimental activity, seem to suggest that the best process performance should be obtained for value of equivalence ratio in the range 0.30–0.33. 3.3. Tar concentration and characterization The amount and composition of tar contained in the producer gas is another crucial parameter of the gasification process. The
decreasing trend with ER and the values of tar concentration reported in Fig. 7 are in accordance with the results recently reported for a different SRF fed into a laboratory scale FBG reactor [15]. It is noteworthy that Fig. 7 reports the tar content in the producer gas at the reactor exit, i.e. before the syngas cleaning unit, since the aim of the investigation was providing data for designing of all the components of a commercial FBG plant, and those of the cleaning section in particular. For this reason, the tar composition has been investigated by means of the described GC–MS procedure. The results reported in Fig. 8 indicate that four main identifiable compounds form the largest fractions of these condensable hydrocarbons: naphthalene, C10H8 (that is generally between 30% and 80% of the identified species of tar, as reported in Table 3), fluorene, C13H10, phenanthrene, C14H10, and pyrene, C16H10. The first three are light PAHs (with 2 or 3 aromatic rings) that belong to the class IV of the recalled ECN taxonomy, i.e. that of tar condensing at intermediate temperatures (218 °C, 295 °C and 336 °C, respectively) even at relatively low concentration; pyrene is instead a heavy PAH (with 5 rings) that belongs to class V, i.e. that of tar condensing at relatively high-temperatures (404 °C for pyrene) at low concentrations. In cold gas applications, such as gas turbines and internal combustion engines, the producer gas requires compression before its injection, becoming saturated when the tar vapor pressure overcomes the tar saturation pressure. This means that tar classes 4 and 5 are a major cause of condensation, and then of tar plugging in pipelines and fouling in energy conversion devices [26,27]: the selectivity of treatment to remove or convert tar is then of crucial relevance. The state-of-the-art of mechanical, catalytic and thermal methods for tar reduction indicates the strong difficulties in selecting and designing a proper fuel gas cleaning system, in terms of plant complexity and availability, overall costs and efficiency [26–28]. This suggests that the final configuration of the FBG for the tested SRF must carefully take into account this aspect and it should evaluate the possibility of a ‘‘thermal gasifier’’ configuration, where the hot producer gas is directly sent to an adequate gas burner, with the additional advantage of a potentially complete exploitation of the tar heating content [12]. This option appears even more relevant taking into account that some specific energy conversion devices, such as mild combustors [29] or externally fired gas turbines [30], have recently reached a good degree of development and are now available, or close to be available, on the market. The first are characterized by an elevated temperature of reactants and low process temperature increase, in order to reduce the maximum temperature of the whole process below the value of main pollutants’ formation and to increase the stability of the combustion process [31]. The EFGTs utilize ambient air as working fluid and a gas–gas high temperature exchanger to provide heat addition: the separation of the working fluid from the combustion fumes preserves the rotating parts from fouling and plugging, while the use of the exhaust clean
534
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Fig. 8. Tar composition at the reactor exit in all the gasification tests.
hot air from the turbine outlet as the oxidizing gas in the syngas combustion, assures high thermodynamic efficiencies [32]. These devices could be coupled to a low-medium scale FBG to obtain an efficient conversion of all the combustible components of the producer gas, including tar compounds. This option, if applied to the tested SRF, could greatly could grow up the maximum value of syngas specific energy, up to a value of 3.6 kW h/kgfuel.
of particles sampled from the bed, as obtained by means of X-ray Fluorescence (XRF) analysis. The table also reports, for each chemical element, two dimensionless parameters: the ratio Qi,fines/Qi,fuel between the mass flow rate of the element that escapes the reactor as fines and that of the same element that enters into the reactor inside the fuel, and the enrichment factor EF in the ash, defined as [33]:
EF ¼ ½ðelement concentration in ashÞ= 3.4. Bed material and entrained fines characterization
ðelement concentration in fuelÞ ½% ash in fuel=100:
Table 4 reports, for a couple of tests carried out in the same day and with the same stock of SRF, the complete characterization of inorganic fractions of the fines collected at the cyclone and that
Data reported in the table could be utilized for analysis of element partitioning during the gasification process of SRF [34], highlighting in particular that QZn,fines/QZn,fuel and QCu,fines/QCu,fuel are
Table 4 Composition of the inorganic fraction of fines collected at the cyclone and material sampled from the bed, during air-gasification tests 2 and 3, as obtained by means of X-ray Fluorescence (XRF) analysis. Element
Al Sb As Cd Ca Co Cr Fe P Mg Mn Hg Ni Pb K Cu Si Na V Zn Chloride
Units
g/100 g mg/kg mg/kg mg/kg g/100 g mg/kg mg/kg g/100 g g/100 g g/100 g g/100 g mg/kg mg/kg mg/kg g/100 g mg/kg g/100 g g/100 g mg/kg mg/kg mg/kg
SRF stock A (units)
Test#2 – cyclone Collected fines (units)
EF ()
Qi,fines/ Qi,fuel ()
Test#3 – cyclone Collected fines (units)
EF ()
Qi,fines/ Qi,fuel ()
Test#2 – bed Bed particles (units)
EF ()
Test#3 – bed Bed particles (units)
EF ()
as
1.79 50 50 50 6.25 50 129 1.16 0.246 0.672 0.0185 50 67.1 162 0.541 319 3.52 0.005 50 314 1200
11.5 118 50 50 32.9 50 370 5.23 1.39 1.77 0.0736 50 567 1700 3.13 3670 18.1 0.005 50 5610 43800
1.19 0.44 0.19 0.19 0.97 0.19 0.53 0.83 1.05 0.49 0.74 0.19 1.56 1.94 1.07 2.13 0.95 0.19 0.19 3.31
0.39 0.14 0.06 0.06 0.32 0.06 0.17 0.27 0.34 0.16 0.24 0.06 0.51 0.64 0.35 0.70 0.31 0.06 0.06 1.07
11.9 172 50 50 36.1 50 252 4.47 1.39 0.0122 0.0665 50 241 1280 3.36 4850 17.6 0.005 50 5390 41200
1.23 0.64 0.19 0.19 1.07 0.19 0.36 0.71 1.05 0.00 0.67 0.19 0.66 1.46 1.15 2.81 0.93 0.19 0.19 3.18
0.43 0.22 0.06 0.06 0.37 0.06 0.13 0.25 0.37 0.001 0.23 0.06 0.23 0.51 0.40 0.99 0.32 0.06 0.06 1.11
1.61 50 50 50 3.95 195 2090 10.7 0.264 44.2 0.123 50 2580 50 0.265 738 36.7 0.005 50 1064 89.6
0.17 0.18 0.18 0.18 0.12 0.72 2.99 1.71 0.20 12.2 1.23 0.18 7.11 0.06 0.09 0.43 1.93 0.18 0.18 0.63
2.17 50 50 50 6.18 205 2690 10.9 0.416 41.1 0.126 50 2620 50 0.238 1770 34 0.005 50 1490 136
0.22 0.18 0.18 0.18 0.18 0.76 3.85 1.74 0.31 11.3 1.26 0.18 7.22 0.06 0.08 1.03 1.79 0.18 0.18 0.88
Al2O3 Sb As Cd CaO Co Cr Fe2O3 P2O5 MgO MnO Hg Ni Pb K2O Cu SiO2 Na2O V Zn Cl
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
equal to about 1. Taking into account the confidence interval of data, mainly related to the fluctuations in feeding rates of trace elements, this means that most of zinc and copper are entrained together with elutriated fines, as well as about half of lead. This appears to be the consequence of low-boiling point of the element (as in the case of zinc) or that of the chemical compounds formed in the reducing atmosphere of the reactor (as hydroxides or sulfates in case of copper and chlorides in case of lead). The corresponding values of EF in the collected fines are larger than 1, indicating the ash enrichment for these elements. The high or very high values of EF evaluated in the bed material for Fe, Si, Cr and Mg (about 1.7, 1.9, 3 and 11.5, respectively) are instead influenced by the composition of olivine particles. Data related to iron in the elutriated fines collected at the cyclone can suggest some observations about the catalytic activity of olivine. When olivine provides a great catalytic support for tar cracking and carbonization, fines collected at the cyclone contain greatly larger quantities of iron than those entering the reactor with the fuel. Several tests [14] highlighted that part of the elemental iron of olivine, which as a consequence of catalytic action is linked with the coke layer by coordination complexes, can be detached from the particle by mechanical attrition and entrained out of the reactor in the syngas. Consequently, the values of QFe,fines/ QFe,fuel were estimated to be greatly larger than 1, typically 100 or more [14]. Data in Table 4 for the runs carried out with SRF at ER = 0.332 and ER = 0.272 indicate instead that QFe,fines/QFe,fuel is equal to about 0.25. This suggests that the catalytic action of olivine is not, or is only partially, present during air gasification of SRF, as indirectly confirmed by the absence of coke over the external surface of olivine particles in all the experimental tests. It is likely that the catalytic support to the cracking and isomerization determined by magnesium (that is 43% on mass basis of olivine) is always active, so that the heavier fragments are broken and a number of unsaturated hydrocarbons with two or three carbon atoms are formed. On the contrary, the catalytic enhancement of the dehydrogenation and carbonization determined by active sites of elemental iron is absent, and tar formation is not inhibited. There are two main possible reasons: (i) considering the very high content of Fe and Al in the SRF, the ferrous and non-ferrous metals in the ash can act as competitors of the iron oxides on the external surface of olivine particles, so avoiding their reduction to elemental Fe; (ii) the high content of oxygen in the SRF could prevent the reduction of the same iron oxides or may generate CO which, in turn, binds the iron to different kinds of carbonyls (such as Fe2(CO)9, Fe3(CO)12, and their hydrides): in both cases, iron catalytic activity is inhibited.
4. Concluding remarks The technical feasibility of air gasification of a solid recovered fuel in a bubbling fluidized bed gasifier has been investigated in a pilot scale reactor, having a maximum thermal output of 400 kW. The unit was operated under conditions of thermal and chemical steady state, for values of equivalence ratio ranging from 0.25 to 0.33. The temperature of gasification reaction zone varied in the range 850–930 °C, then it was always safely lower than the ash fusion temperature and sufficiently high to allow an efficient carbon conversion. The control of the optimal temperature can be achieved mainly by modulating the equivalence ratio and, just for a small extent, the air preheating temperature. The best process performances have been obtained by fluidizing the reactor with air and for values of equivalence ratio in the range 0.30–0.33. The obtained syngas has a LHV of about 5 MJ/m3N , that increases up to more than 7 MJ/m3N if the energy content of the tar is also taken into account. The cold gas efficiency, i.e. the
535
fraction of chemical energy of the SRF that is transferred to the syngas, can be as high as 60%. The carbon conversion efficiency, i.e. the percentage of inlet carbon that is converted to CO, CO2 and gaseous hydrocarbons, is between 0.8 and 0.9. The remaining part of carbon flow rate is present in the producer gas as condensable heavy hydrocarbons (tar) and carbonaceous fine particles escaped from the bed as entrained fines. The tests on the pilot scale reactor measured the flow rates of fines and tar at the reactor exit. These values indicate that, even under the optimal set of operating conditions, the concentrations of dusts and tars in the obtained syngas make very difficult to select and design an adequate fuel gas cleaning system and then could not allow the utilization of an internal combustion engine or a gas turbine, even because the catalytic action of olivine is only partially present during the SRF gasification process. It is likely that a proper configuration for small- or medium-size plants should directly sent the hot producer gas to a specifically designed gas burner. In this way, there will be no more necessity of a (probably complex and expensive) gas cleaning unit, even though some conditioning on the producer gas can be always operated before the producer gas combustion, and the plant will also value the heating content of tar compounds.
Acknowledgments The study was carried out with the financial support of Smaltimenti Sud s.r.l. Data reported in the paper are original calculations developed by the authors and cannot be considered as official information of Smaltimenti Sud s.r.l. The authors are greatly indebted with Mr. Onofrio Annoscia and Mr. Gianfranco De Troia of Eco-Engineering Impianti s.r.l. for their technical assistance during all the gasification tests and with Dr. Donato Santoro, which performed large part of the off-line laboratory analyses.
References [1] World Bank. What a waste_a global review of solid waste management. Urban development urban series knowledge papers 68135, 2012,. [2] Garg A, Smith R, Hill D, Longhurst PJ, Pollard SJT, Simms NJ. An integrated appraisal of energy recovery options in the United Kingdom using solid recovered fuel derived from municipal solid waste. Waste Manage 2009;29:2289–97. [3] Arena U, Di Gregorio F. A waste management planning based on substance flow analysis. Res., Conserv. Recycl., . [4] McDougall F, White P, Franke M, Hindle P. Integrated solid waste management: a life cycle inventory. 2nd ed. Blackwell Science Ltd.; 2001. [5] EPA-United States Environmental Protection Agency_Office of Solid Waste. Beyond RCRA_Waste and Materials Management in the Year 2020. Report EPA530-R-02-009, 2003, . [6] ISWA. Waste and climate change. 2009, . [7] Brunner PH. Clean cycles and safe final sinks. Waste Manage Res 2009;28:575–6. [8] Bosmans A, Vanderreydt I, Geysenc D, Helsen L. The crucial role of waste-toenergy technologies in enhanced landfill mining: a technology review. J Clean Prod 2013;55:10–23. [9] ECD-European Commission Directive 2008/98/EC of the European Parliament and of the Council of 19 November 2008 on waste and repealing certain directives (waste framework directive). Off J 22/11/2008; L312: 3. [10] ECD-European Council Directive 1999/31/EC of 26 April 1999 on the landfill of waste. Off J 16/07/1999; L182: 1–19. [11] European Committee for Standardisation. CEN/TC343/WG 2, Solid recovered fuels – specifications and classes. Draft European Standard, 2005. [12] Arena U. Process and technological aspects of municipal solid waste gasification. A review. Waste Manage 2012;32:625–39. [13] Hankalin V, Honkola T. High efficiency power production by gasification. In: Arena U, Chirone R, Miccio M, Salatino P editors. Proc. of the 21st international conference on fluidized bed combustion. Naples: ISBN: 978-88-89677-83-4; June 4–6 2012. [14] Arena U, Zaccariello L, Mastellone ML. Fluidized bed gasification of wastederived fuels. Waste Manage 2010;30:1212–9.
536
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
[15] Dunnu G, Panopoulos KD, Karellas S, Maier J, Touliou S, Koufodimos G, et al. The solid recovered fuel Stabilat: characteristics and fluidised bed gasification tests. Fuel 2012;93:273–83. [16] Di Gregorio F, Zaccariello L. Fluidized bed gasification of a packaging derived fuel: energetic, environmental and economic performances for waste-toenergy plants. Energy 2012;42:331–41. [17] Van Paasen SVB, Kiel JHA. Tar formation in a fluidized bed gasifier: impact of fuel properties and operating conditions. Report ECN-C-04-013, 2004, Petten, the Netherlands. [18] Felder RM, Rousseau RW. Elementary principles of chemical processes. 3rd ed. J. Wiley & Sons; 2000. [19] Channiwala SA, Parikh PP. A unified correlation for estimating HHV of solid, liquid and gaseous fuels. Fuel 2002;81:1051–63. [20] Kim M-R, Jang J-G, Lee S-K, Hwang BY, Lee JK. Correlation between the ash composition and melting temperature of waste incineration residue. Korean J Chem Eng 2010;27(3):1028–34. [21] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar elimination in biomass gasification processes. Biomass Bioenergy 2003;24:125–40. [22] Arena U, Di Gregorio F, Santonastasi M. A techno-economic comparison between two design configuration for a small scale, biomass-to-energy gasification based system. Chem Eng J 2010;162:580–90. [23] Kirnbauer F, Wilk V, Hofbauer H. The positive effects of bed material coating on tar reduction in a dual fluidized bed gasifier. Fuel 2012;95:553–62. [24] Mastellone ML, Arena U. Olivine as a tar removal catalyst during fluidized bed gasification of plastic waste. AIChE J 2008;54(6):1656–67.
[25] Gómez-Barea A, Ollero P, Leckner B. Optimization of char and tar conversion in fluidized bed biomass gasifiers. Fuel 2013;103:42–52. [26] Li C, Suzuki K. Tar property, analysis, reforming mechanism and model for biomass gasification – an overview. Renew Sustain Energy Rev 2009;13:594–604. [27] Anis S, Zainal ZA. Tar reduction in biomass producer gas via mechanical, catalytic and thermal methods: a review. Renew Sustain Energy Rev 2011;15:2355–77. [28] Brown D, Gassner M, Fuchino T, Maréchal F. Thermo-economic analysis for the optimal conceptual design of biomass gasification energy conversion systems. Appl Therm Eng 2009;29:2137–52. [29] Cavaliere A, de Joannon M. Mild combustion. Prog Energy Combust Sci 2004;30:329–66. [30] Cocco D, Deiana P, Cau G. Performance evaluation of small size externally fired gas turbine (EFGT) power plants integrated with direct biomass dryers. Energy 2006;31(10-11):1459–71. [31] Milani A, Saponaro A. Diluted combustion technologies. IFRF Combust J 2001;1:1–32. [32] Arena U, Di Gregorio F, Amorese C, Mastellone ML. A techno-economic comparison of fluidized bed gasification of two mixed plastic wastes. Waste Manage 2011;31:1494–504. [33] Zevenhoven R, Kilpinen P. Control of Pollutants in flue gases and fuel gases. Turku, Finland: Åbo Akademi University; 2001.. [34] Arena U, Di Gregorio F. Element partitioning in combustion- and gasificationbased waste-to-energy units. Waste Manage 2013;33:1142–50.
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
Gasification of a solid recovered fuel in a pilot scale fluidized bed reactor Umberto Arena a,b,⇑, Fabrizio Di Gregorio b a b
Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, Via Vivaldi 43, 81100 Caserta, Italy AMRA s.c.a r.l., Analysis and Monitoring of Environmental Risk, Via Nuova Agnano 11, 80125 Napoli, Italy
h i g h l i g h t s Pilot-scale investigation of fluidized bed gasification of a solid recovered fuel. Tests under conditions of thermal/chemical steady state at various equivalence ratios. Complete composition of the syngas, including tar, particulate and acid/basic gases. Tar characterization according to the ECN classification. Partitioning of main inorganic elements in the entrained fines and bed particles.
a r t i c l e
i n f o
Article history: Received 7 August 2013 Received in revised form 14 September 2013 Accepted 17 September 2013 Available online 27 September 2013 Keywords: Gasification Solid recovered fuel Fluidized bed reactor Waste-to-energy
a b s t r a c t The paper investigates the technical feasibility of an air gasification process of a Solid Recovered Fuel (SRF) obtained from municipal solid waste. A pilot scale bubbling fluidized bed gasifier, having a feedstock capacity of about 70 kg/h and a maximum thermal output of about 400 kW, provided the experimental data: the complete composition of the syngas (including the tar, particulate and acid/basic gas contents), the chemical and physical characterization of the bed material and that of entrained fines collected at the cyclone. The experimental runs were carried out by reaching a condition of thermal and chemical steady state under values of equivalence ratio ranging from 0.25 to 0.33. The results indicate that the selected SRF can be conveniently gasified, yielding a syngas of valuable quality for energy applications. The rather high content of tar in the syngas indicates that the more appropriate plant configuration should be that of a ‘‘thermal gasifier’’, with the direct combustion of the syngas in a burner ad hoc designed, coupled with an adequate energy-conversion device. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction and framework The amount of the municipal solid waste (MSW) increases all over the world as a consequence of growing urbanization and the standard of living. A recent study indicates that urban areas consume about 75% of natural resources and produce 50% of the total amount of waste in the world [1]. A remarkable amount of this waste contains materials (such as paper, glass, metals, wood and plastics), mainly deriving from packaging, which can efficiently be recycled for resource recovery. But not all the MSW can be recycled and, moreover, the waste recycling processes always generate significant amounts of residues, in some cases having very high heating values [2,3]. It is today recognized that in a fully sustainable waste management system no one process is suitable for all waste streams [4–8]. In other words, no single waste management
⇑ Corresponding author at: Department of Environmental, Biological and Pharmaceutical Sciences and Technologies, Second University of Naples, Via Vivaldi 43, 81100 Caserta, Italy. Tel.: +39 0823 274414; fax: +39 0823 274592. E-mail address: [email protected] (U. Arena). 0016-2361/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.fuel.2013.09.044
practice (i.e. landfill, recycling, biochemical or thermochemical conversion) can handle the full array of waste types and, at the same time, satisfy the recognized criteria of an integrated and sustainable management system [3,9]: (i) to minimize use of landfills and ensure that no landfilled waste is biologically active or contains mobile hazardous substances; (ii) to minimize operations that entail excessive consumption of raw materials and energy without yielding an overall environmental advantage; (iii) to maximize recovery of materials, albeit in respect of the previous point; and (iv) to maximize energy recovery for materials that cannot be efficiently recycled, in order to save both landfill volumes and fossil-fuel resources. With reference to the latter criterion, two remarkable waste streams have a great potential of being used in sustainable energy recovery processes: the unsorted residual waste (i.e. the dry organic fraction of MSW that cannot conveniently be recycled and it is not suitable for biological treatments) and the high heating value residues produced by some recycling chains (that in some countries cannot be landfilled [10]). These non-hazardous waste fractions can be turned into a Solid Recovered Fuel (SRF), i.e. in a sufficiently homogeneous waste-derived
529
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
fuel, obtained as the result of a mechanical process to comply with a CEN standard [11]. The European standardization defines different grades of SRF on the basis of specific composition ranges, mainly low heating value and chlorine and mercury contents. An efficient way for thermochemical exploitation of SRFs is fluidized bed gasification. Gasification is the conversion of solid fuel to a synthesis-gas through gas-forming reactions occurring in presence of an amount of oxidant lower than that required for the stoichiometric combustion. The resulting fuel gas (named ‘‘producer gas’’ or ‘‘syngas’’) can be utilized in a separate energy conversion device: it contains large amounts of not completely oxidized products (mainly CO, H2 and lower amounts of CH4), together with different organic (tar) and inorganic (H2S, HCl, NH3, HCN, alkali metals) impurities and particulates [12]. Fluidization is one of the most interesting gasification technologies, mainly for the high quality of gas–solid contact and the very efficient mass and heat transfers, but also for its good process flexibility, which accommodates variation in fuel quality and allows to utilize different fluidizing agents, reactor temperatures and gas residence times, to add reagents along the reactor height and to operate with or without a specific catalyst. A large-scale fluidized bed gasifier has been recently put in operation in Lahti, southern Finland: it is able to treat 250 kt/y of SRF derived from household waste (origin sorted) together with industrial waste, demolition wood and waste wood from industry and to produce 50 MW of power and 90 MW of district heat [13]. But there is also a great interest for SRF gasification for energy production in small and medium scale fluidized bed plants [14–16]. The paper describes the results of a research program aimed at assessing the technical feasibility of a fluidized bed gasifier able to treat 5000 t/y of a SRF, obtained as one of the output solid streams of a sorting and mechanical treatment of MSW collected in a urban area of the Middle of Italy. To this aim, a number of tests with this SRF were carried out in a pilot scale bubbling fluidized bed gasifier (BFBG). The obtained experimental results were processed by mass and energy balances, in order to obtain data and information useful to define both a suitable plant configuration and the related design solutions for a fully sustainable energy generation.
2. The fluidized bed gasifier, the experimental procedures and the material tested 2.1. The fluidized bed gasifier The design and operating parameters of the pilot scale bubbling fluidized bed gasifier are reported in Table 1. The plant has a feedstock capacity between 30 and 100 kg/h, depending on the type of fuel, and a maximum thermal output of about 400 kW. It is composed of three main sections, as it is sketched in Fig. 1: the feeding system, the fluidized bed gasifier and the syngas cleaning unit. The first can be divided in the blast and fuel feeding. The blast feeding is heated up to about 150 °C by a first electric heater, then sent to a mixing point with an optional stream of steam at about 150 °C, and finally heated by a second electric heater up to the desired inlet temperature at the fluidized bed bottom. In all the experiments reported here, only air was used as blast agent and always injected at the bed bottom while the fuel was always fed by means of an overbed system. The gasification section is composed of a cylindrical reactor, 0.381 m ID and 5.90 m high, which is heated up to the reaction temperature by the sensible heat of pre-heated blast gases and by a set of three external electrical furnaces. The syngas cleaning section is composed of a high efficiency cyclone (for abatement of fine particles), a simple wet scrubber (for removal of tar, residual dust, acid gases and ammonia) and a flare.
Table 1 Main design and operating parameters of the pilot scale BFB gasifier. Geometrical parameters
Feedstock capacity Thermal output Typical bed amount Oxidizing agent Feeding system Range of bed temperatures Range of fluidizing velocities Produced gas treatments Safety equipments
Internal diameter: 0.381 m Total height: 5.90 m Reactive zone height: 4.64 m Wall thickness: 12.7 mm Up to 100 kg/h (on the basis of the fuel heating value) Up to about 400 kW 145 kg Air (but also oxygen, steam, and their mixtures) Over-bed water-cooled screw feeder 700–950 °C 0.3–1 m/s Cyclone, scrubber, flare Water seal, safety valves, rupture disks, alarms, nitrogen line for safety inerting
2.2. The experimental procedures Each run has a start-up of about 3 h, during which three electric heaters located along the reactor lead the temperature up to about 700 °C, while the bed is fluidized at a fixed velocity. At this point, the flow rates of SRF and air are adjusted in order to obtain the desired value of the equivalence ratio ER, defined as the ratio between the oxygen content of air supply and that required for the stoichiometric complete combustion of the fuel effectively fed to the reactor. Under the selected operating conditions of ER and air preheating temperature, and without any more thermal assistance of external heaters, the reactor gradually reaches a thermal and chemical steady state. The gas and solids sampling procedures are then activated and measurements of pressure, temperature, blast flow rates and syngas composition upstream of and downstream of the wet scrubber are taken and averaged over the whole period of steady state, which usually is kept for about 2 h. Fig. 2 shows the successive steps of the described procedure, reporting the time profile of gas composition and bed temperature for a typical run. In order to increase the reliability of measurements, in all the runs the composition of the syngas downstream of the cleaning section was on-line monitored by means of two systems: a series of IR analysers (Horiba VA-3115 for CO, CO2 and O2, Horiba VA3001 for CH4 and Teledyne Anal. Instr.-2000 for H2) and an Agilent 3000 gas-chromatograph equipped with 4 different columns (Mol Sieve, PoraPlot, OV, Alumina) for the detection of a wide spectrum of syngas compounds. Gas was also sampled in Tedlar bags at two other points (2.3 m from the reactor bottom and the gasifier exit) and sent to off-line measurements. Two methods were used to evaluate the amount and composition of tar, due to the relevance of this measurements for the technical and economic feasibility of the process. The first assumes that tar is composed of all organic compounds with a molecular weight larger than benzene, excluding soot and char: then it conservatively imputes to the tar amount the whole carbon loading which, as a result of a mass balance on atomic species, cannot be attributed either to the producer gas or to the solids collected at the cyclone or present inside the bed. The second method samples the condensable species by means of a system composed of four in-series cooling coils, a suction pump and a flow meter operated with a syngas flow rate of about 300 dm3N /h for 30 min: the condensed hydrocarbons are then off-line analyzed with a specific pre-treatment in a Perkin–Elmer Clarus 500 gas chromatograph coupled with a mass spectrometer (GC–MS). The first conservative method was utilized for quantitative determination of tar concentration in the obtained producer gas. The second procedure was instead used to detect tar
530
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
°
Fig. 1. Schematic of the pilot-scale fluidized bed gasifier, with the indication of some operating parameters. Fi, with i = 1–18, indicates different gas and solid streams.
Fig. 2. Time profile of gas composition and bed temperature in a typical run with the SRF.
compounds belonging to the classes between 2 and 5 of the classification proposed by the Energy research Center of The Netherlands (ECN) [17]. Hydrogen chloride, hydrogen sulphide and ammonia were measured by bubbling the producer gas through two pairs of gas stripping bubblers, connected in series and containing basic and acid solutions respectively, and subsequently analyzed by means of a Dionex DX-120 ion chromatograph. Data obtained from on-line and off-line gas measurements and from chemical analyses of solid samples were processed to develop mass balance on atomic species and the related energy balance for each test. The flow rate of produced syngas was determined by the ‘‘tie-component method’’ [18] applied to the value of nitrogen content in the dry syngas, as obtained by (on-line and off-line) GC measurements, and adequately corrected to take into account the nitrogen fed into the gasifier with the waste and that leaving it as ammonia. 2.3. The fuel and the bed material The BFBG was fed with an SRF produced by a commercial sorting platform located in the Middle of Italy. The received SRF was mechanically processed to obtain mono-size cylindrical pellets in order to avoid any effect related to fuel size and to increase the density of material to be fed into the gasifier. The size of pellets, 5 mm of diameter and 30 mm of length, is compatible with
BFBGs: for these reactors the maximum size should be smaller than 10–20 cm, in order to avoid floating effects, and it is preferable to avoid fuels with a too high percentage of very small particles, in order to avoid lifting out of too many fines out of the bed. Table 2 reports the ultimate analysis and low heating value of the pelletized SRF, together with the composition of its inorganic fraction, as obtained by analysing the different stocks utilized in the experimental runs. The reported range of variation is rather limited, even though the original unsorted residual waste was collected at different time and in different catchment areas. Data reported in Table 2 indicate the SRF pellets as a good quality fuel [19] with an ash fusion temperature in the range 1400– 1700 °C, which is sufficiently higher than that typical of FBG operating conditions [20]. The bed material used during the experimental runs is olivine, due to its good performance as tar removal agent, reported in several studies on gasification of biomass [21–23] and polyolefin plastic wastes [14,24]. The utilized olivine has a particle size range of 200–400 lm and a particle density of 2900 kg/m3. It is a neo-silicate of Fe and Mg coming from Austria and composed mainly of MgO, SiO2, and Fe2O3 but even Al2O3, Cr2O3, Mg3O4 and CaO. It was pre-calcined at the production site and further calcined just before each test for about 8 h in oxidizing atmosphere at 800 °C, in order to improve their potential catalytic activity as tar removal agent [21,23].
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Cn Hm þ n=2O2 ! nCO þ m=2H2
Table 2 Main chemical properties of the tested SRF. Ultimate analysis (%wt, C H N S Cl O (by diff.) Moisture Ash
ar)
41.2–45.4 6.0–6.5 0.66–0.70 0.1–0.3 0.1–0.2 22.9–24.2 3.7–9.1 18.5–20.4
Heating value (MJ/kgfuel,ar) Low heating value (LHV) a
18,600–21,300
Ash composition (mg/kgdb) Aluminum as Al2O3 Antimony as Sb Arsenic as As Cadmium as Cd Calcium as CaO Chrome as Cr Cobalt as Co Copper as Cu Iron as Fe2O3 Lead as Pb Magnesium as MgO Manganese as MnO Mercury as Hg Nickel as Ni Phosphorus as P2O5 Potassium as K2O Silicon as SiO2 Sodium as Na2O Vanadium as V Zinc as Zn Chlorine as Cl
10,500–17,900 <50 <50 <50 62,500–82,600 <50–130 <50 106–319 8480–11,600 87–162 3130–6720 185–282 <50 <50–67 2460–3200 5410–9070 31,100–35,200 <50 <50 314–386 1200–1950
Ash melting temperature Half fluid temperature (HFT)b (°C)
1420–1730
ar = as received; db = on dry basis. a Channiwala and Parikh [19]. b Kim et al. [20].
3. Results and discussion 3.1. Composition of the producer gas The pilot scale BFBG was operated by injecting the SRF into a bed of olivine particles fluidized at a fixed superficial gas velocity of about 0.7 m/s and temperatures of about 850 °C, under different values of equivalence ratio, in order to quantitatively assess its behavior in the fluidized bed gasification process. Table 3 lists the operating conditions of all tests, together with the main experimental results. In particular, some variables, such as the bed temperature at steady state, the syngas temperature at the reactor exit and the overall fluidizing velocity (calculated by taking into account the flow rate of the produced syngas and assuming that it was completely generated inside the bed) have been more properly considered as state variables. They are the answers of the FBG system to the set of operating variables, mainly the equivalence ratio, fuel heating value and air preheating temperature. Data in Table 3 also shows that the flow rate of entrained fines is in the order of 0.1 kg/kgfuel, while the flow rate of ash injected with the fuel is about 0.2 kg/kgfuel: thereby, about a half of ashes escapes the reactor as entrained fines while the other part remains into the bed and is periodically extracted from the bottom of the fluidized bed. Fig. 3 reports the composition of the obtained producer gas in all the air gasification tests as a function of equivalence ratio ER. An increase of ER implies a larger amount of oxygen available for reaction with volatiles in the pyrolysis zone [24,25], and then a greater extension of partial oxidation reactions:
531
ð1Þ
Accordingly, the diagrams show a limited reduction of the contents of methane, ethylene and CnHm hydrocarbons (with n = 2–4) and a corresponding increase of that of carbon monoxide. The latter could be also related to the (much) slower reactions of steam and dry reforming:
Cn Hm þ nH2 O ! nCO þ ðn þ m=2ÞH2
ð2Þ
Cn Hm þ nCO2 ! 2nCO þ m=2H2
ð3Þ
The concentration of hydrogen shows a very limited variations for the opposite effects of H2 generated by the recalled reactions of hydrocarbons and that consumed by its own oxidation. Similarly, the CO2 concentration as a function of ER appears to have limited variations, as a consequence of the opposite effects of Boudouard reaction:
C þ CO2 ! 2CO
ð4Þ
and CO partial oxidation. The greater extension of exothermic partial oxidation reactions also generates an overall effect of reactor temperature increase, which is described by the average values of bed temperatures reported in Fig. 4. They range from the lowest value of 849 °C at ER = 0.256 to the highest value of 932 °C at ER = 0.332. As mentioned above, since the tests were carried out under thermal and chemical steady state conditions, the bed temperature is a state variable of the system, i.e. its answer to the set of operating conditions. The air preheating temperature was gradually reduced to compensate this increase, even though its role appears limited, as it is suggested by the data related to the two tests at ER = 0.256. It is noteworthy that the minimum value of ER to efficiently operate the pilot scale gasifier without additional heating was 0.25: lower values would reduce too much the carbon conversion efficiency. A single test was carried out by using a mixture of air and steam as fluidizing stream, with a steam-to-fuel ratio of 0.69 kgsteam/ kgwaste, i.e. a steam-to-carbon ratio of 1.46 kgsteam/kgC-waste. The process performance for this test greatly got worse: this is likely related to lowering of reactor temperature (in our test down to 802 °C) determined by the heat requirement of endothermic steam reforming reactions. This consequently induces a reduction of carbon conversion or the necessity of more oxygen to maintain the thermal level (which in turn implies a reduction of the heating value of the producer gas) [25]. A specific and deeper investigation is necessary to better understand this aspect. 3.2. Process performance parameters The four diagrams of Figs. 5 and 6 report the values of the main process performance parameters, as a function of equivalence ratio: – The LHV of the producer gas (excluding the contribution of tar compounds). – Its specific energy, i.e. the chemical energy of the producer gas per kg of SRF. – The Carbon Conversion Efficiency, CCE, defined as the carbon flow rate converted to gaseous products with respect to that fed to the reactor with the SRF. – The Cold Gas Efficiency, CGE, defined as the fraction of the chemical energy of the waste which is transferred to the producer gas. As expected on the basis of previous studies on biomass and waste-derived fuel [14–16,21,22,24,25], the LHVs reduce as ER
532
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Table 3 Operating conditions and experimental results of all the tests. Operating conditions Test# ER () Air/fuel (A/F) (kgair/kgfuel) Bed amount (kg) SRF flow rate (kgfuel/h) Fluidizing velocity (m/s) Air preheating temperature (°C) Process results and parameters Overall fluidizing velocitya (m/s) Bed temperature at steady state (°C) Syngas temperature at reactor exit (°C) Syngas production (volumetric basis) (m3N /kgfuel) Syngas production (mass basis) (kgsyngas/kgfuel) Syngas LHV (kJ/m3N ) Specific energy (kW h/kgfuel) CCE, CGE, -
4 0.256 1.47 146 61.4 0.70 375
6 0.255 1.54 146 58.6 0.70 520
3 0.272 1.56 146 57.65 0.71 355
5 0.302 1.73 146 51.9 0.72 329
1 0.318 1.82 146 49.4 0.73 542
2 0.332 1.90 146 47.40 0.75 273
0.77 849 807 1.62
0.79 852 780 1.73
0.78 869 857 1.72
0.79 879 801 1.91
0.82 898 822 2.04
0.79 932 815 2.00
1.99 5390
2.18 5652
2.10 5550
2.34 5160
2.56 4910
2.46 4690
2.42 0.70 0.53
2.71 0.80 0.56
2.66 0.75 0.58
2.78 0.81 0.61
2.79 0.92 0.61
2.61 0.79 0.57
Syngas composition (%) N2 CO2 CO H2 CH4 C2H4 C2H6 C2H2 C3H6 C6H6 Elutriated fines (g/kgfuel) Carbon elutriated fines (gC/kgC-fuel) Tar (g/m3N )b Naphthalene (% of identified tar species) Pyrene (% of identified tar species) HCl (mg/m3N )c
61.77 13.36 9.74 8.48 4.18 2.19 0.04 0.20 0.02 0.20 87.0 23.7 73
60.64 15.69 8.98 7.11 4.20 2.72 0.08 0.12 0.10 0.32 67.3 19.4 47
61.67 12.28 10.47 8.44 4.57 2.14 0.04 0.21 0.03 0.15 64.9 32.5 58
61.86 12.83 10.40 8.24 4.35 2.05 0.04 0.16 0.02 0.10 102.8 20.2 39
60.66 14.04 12.73 7.08 3.33 1.78 0.07 0.18 0.02 0.08 62.7 24.2 5
64.61 11.66 1058 7.68 3.78 1.27 0.02 0.18 0.00 0.21 60.7 22.7 42
29 51 10.5
38 3.0 n.a.
65 3.2 3.1
12 60 117.5
76 3.5 2.1
80 1.3 21.7
H2S (mg/m3N )c
7.4
n.a.
0.2
39.1
0.5
0.6
NH3 (mg/m3N )c
74.2
n.a.
134
39.6
2.7
18.7
a
This value was calculated taking into account the produced syngas (then assuming that it was completely generated inside the bed). This value was conservatively determined on the basis of mass balance on atomic species. The samples of gas for measurements of NH3, HCl and H2S were all taken at the reactor exit and then before the cleaning gas system, with the exclusion of test#1 for which they were taken downstream of the wet scrubber. b
c
Fig. 3. Composition of the producer gas in all the gasification tests. Fig. 4. Bed temperature and air preheating temperature in all the gasification tests.
increases. This is the result of opposite effects: the strong reduction of content of hydrocarbons having very high heating values (mainly methane, ethylene and acetylene), and the corresponding increase of carbon monoxide contents. The values reported in Fig. 5 range between a minimum of 4.7 MJ/m3N and a maximum of 5.7 MJ/m3N , and grow up to more than 7 MJ/m3N if the energy content of the tar compounds is taken into account (assuming all the tar as naphthalene) [15]. In the range of experimental conditions tested, the specific energy of the producer gas slightly increases,
from 2.4 kW h/kgwaste to 2.8 kW h/kgwaste, as a consequence of ER increasing. This is related to the corresponding slight increase of both CCE and CGE as a function of ER in the tested range, which is shown in the diagrams of Fig. 6. In particular, higher values of CCE are mainly due to the greater extension of oxidation reactions of carbon fines, being char conversion not much affected by oxygen [25]. The increase of CGE is instead related to the reduced production of condensable heavy hydrocarbons (tar) that is shown in
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
533
Fig. 7. Tar content in the producer gas at the reactor exit in all the gasification tests.
Fig. 5. Low heating value and specific energy of the obtained producer gas in all the gasification tests.
Fig. 6. Carbon conversion and cold gas efficiencies in all the gasification tests.
Fig. 7. The values of all these process parameters, together with the other results of the experimental activity, seem to suggest that the best process performance should be obtained for value of equivalence ratio in the range 0.30–0.33. 3.3. Tar concentration and characterization The amount and composition of tar contained in the producer gas is another crucial parameter of the gasification process. The
decreasing trend with ER and the values of tar concentration reported in Fig. 7 are in accordance with the results recently reported for a different SRF fed into a laboratory scale FBG reactor [15]. It is noteworthy that Fig. 7 reports the tar content in the producer gas at the reactor exit, i.e. before the syngas cleaning unit, since the aim of the investigation was providing data for designing of all the components of a commercial FBG plant, and those of the cleaning section in particular. For this reason, the tar composition has been investigated by means of the described GC–MS procedure. The results reported in Fig. 8 indicate that four main identifiable compounds form the largest fractions of these condensable hydrocarbons: naphthalene, C10H8 (that is generally between 30% and 80% of the identified species of tar, as reported in Table 3), fluorene, C13H10, phenanthrene, C14H10, and pyrene, C16H10. The first three are light PAHs (with 2 or 3 aromatic rings) that belong to the class IV of the recalled ECN taxonomy, i.e. that of tar condensing at intermediate temperatures (218 °C, 295 °C and 336 °C, respectively) even at relatively low concentration; pyrene is instead a heavy PAH (with 5 rings) that belongs to class V, i.e. that of tar condensing at relatively high-temperatures (404 °C for pyrene) at low concentrations. In cold gas applications, such as gas turbines and internal combustion engines, the producer gas requires compression before its injection, becoming saturated when the tar vapor pressure overcomes the tar saturation pressure. This means that tar classes 4 and 5 are a major cause of condensation, and then of tar plugging in pipelines and fouling in energy conversion devices [26,27]: the selectivity of treatment to remove or convert tar is then of crucial relevance. The state-of-the-art of mechanical, catalytic and thermal methods for tar reduction indicates the strong difficulties in selecting and designing a proper fuel gas cleaning system, in terms of plant complexity and availability, overall costs and efficiency [26–28]. This suggests that the final configuration of the FBG for the tested SRF must carefully take into account this aspect and it should evaluate the possibility of a ‘‘thermal gasifier’’ configuration, where the hot producer gas is directly sent to an adequate gas burner, with the additional advantage of a potentially complete exploitation of the tar heating content [12]. This option appears even more relevant taking into account that some specific energy conversion devices, such as mild combustors [29] or externally fired gas turbines [30], have recently reached a good degree of development and are now available, or close to be available, on the market. The first are characterized by an elevated temperature of reactants and low process temperature increase, in order to reduce the maximum temperature of the whole process below the value of main pollutants’ formation and to increase the stability of the combustion process [31]. The EFGTs utilize ambient air as working fluid and a gas–gas high temperature exchanger to provide heat addition: the separation of the working fluid from the combustion fumes preserves the rotating parts from fouling and plugging, while the use of the exhaust clean
534
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
Fig. 8. Tar composition at the reactor exit in all the gasification tests.
hot air from the turbine outlet as the oxidizing gas in the syngas combustion, assures high thermodynamic efficiencies [32]. These devices could be coupled to a low-medium scale FBG to obtain an efficient conversion of all the combustible components of the producer gas, including tar compounds. This option, if applied to the tested SRF, could greatly could grow up the maximum value of syngas specific energy, up to a value of 3.6 kW h/kgfuel.
of particles sampled from the bed, as obtained by means of X-ray Fluorescence (XRF) analysis. The table also reports, for each chemical element, two dimensionless parameters: the ratio Qi,fines/Qi,fuel between the mass flow rate of the element that escapes the reactor as fines and that of the same element that enters into the reactor inside the fuel, and the enrichment factor EF in the ash, defined as [33]:
EF ¼ ½ðelement concentration in ashÞ= 3.4. Bed material and entrained fines characterization
ðelement concentration in fuelÞ ½% ash in fuel=100:
Table 4 reports, for a couple of tests carried out in the same day and with the same stock of SRF, the complete characterization of inorganic fractions of the fines collected at the cyclone and that
Data reported in the table could be utilized for analysis of element partitioning during the gasification process of SRF [34], highlighting in particular that QZn,fines/QZn,fuel and QCu,fines/QCu,fuel are
Table 4 Composition of the inorganic fraction of fines collected at the cyclone and material sampled from the bed, during air-gasification tests 2 and 3, as obtained by means of X-ray Fluorescence (XRF) analysis. Element
Al Sb As Cd Ca Co Cr Fe P Mg Mn Hg Ni Pb K Cu Si Na V Zn Chloride
Units
g/100 g mg/kg mg/kg mg/kg g/100 g mg/kg mg/kg g/100 g g/100 g g/100 g g/100 g mg/kg mg/kg mg/kg g/100 g mg/kg g/100 g g/100 g mg/kg mg/kg mg/kg
SRF stock A (units)
Test#2 – cyclone Collected fines (units)
EF ()
Qi,fines/ Qi,fuel ()
Test#3 – cyclone Collected fines (units)
EF ()
Qi,fines/ Qi,fuel ()
Test#2 – bed Bed particles (units)
EF ()
Test#3 – bed Bed particles (units)
EF ()
as
1.79 50 50 50 6.25 50 129 1.16 0.246 0.672 0.0185 50 67.1 162 0.541 319 3.52 0.005 50 314 1200
11.5 118 50 50 32.9 50 370 5.23 1.39 1.77 0.0736 50 567 1700 3.13 3670 18.1 0.005 50 5610 43800
1.19 0.44 0.19 0.19 0.97 0.19 0.53 0.83 1.05 0.49 0.74 0.19 1.56 1.94 1.07 2.13 0.95 0.19 0.19 3.31
0.39 0.14 0.06 0.06 0.32 0.06 0.17 0.27 0.34 0.16 0.24 0.06 0.51 0.64 0.35 0.70 0.31 0.06 0.06 1.07
11.9 172 50 50 36.1 50 252 4.47 1.39 0.0122 0.0665 50 241 1280 3.36 4850 17.6 0.005 50 5390 41200
1.23 0.64 0.19 0.19 1.07 0.19 0.36 0.71 1.05 0.00 0.67 0.19 0.66 1.46 1.15 2.81 0.93 0.19 0.19 3.18
0.43 0.22 0.06 0.06 0.37 0.06 0.13 0.25 0.37 0.001 0.23 0.06 0.23 0.51 0.40 0.99 0.32 0.06 0.06 1.11
1.61 50 50 50 3.95 195 2090 10.7 0.264 44.2 0.123 50 2580 50 0.265 738 36.7 0.005 50 1064 89.6
0.17 0.18 0.18 0.18 0.12 0.72 2.99 1.71 0.20 12.2 1.23 0.18 7.11 0.06 0.09 0.43 1.93 0.18 0.18 0.63
2.17 50 50 50 6.18 205 2690 10.9 0.416 41.1 0.126 50 2620 50 0.238 1770 34 0.005 50 1490 136
0.22 0.18 0.18 0.18 0.18 0.76 3.85 1.74 0.31 11.3 1.26 0.18 7.22 0.06 0.08 1.03 1.79 0.18 0.18 0.88
Al2O3 Sb As Cd CaO Co Cr Fe2O3 P2O5 MgO MnO Hg Ni Pb K2O Cu SiO2 Na2O V Zn Cl
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
equal to about 1. Taking into account the confidence interval of data, mainly related to the fluctuations in feeding rates of trace elements, this means that most of zinc and copper are entrained together with elutriated fines, as well as about half of lead. This appears to be the consequence of low-boiling point of the element (as in the case of zinc) or that of the chemical compounds formed in the reducing atmosphere of the reactor (as hydroxides or sulfates in case of copper and chlorides in case of lead). The corresponding values of EF in the collected fines are larger than 1, indicating the ash enrichment for these elements. The high or very high values of EF evaluated in the bed material for Fe, Si, Cr and Mg (about 1.7, 1.9, 3 and 11.5, respectively) are instead influenced by the composition of olivine particles. Data related to iron in the elutriated fines collected at the cyclone can suggest some observations about the catalytic activity of olivine. When olivine provides a great catalytic support for tar cracking and carbonization, fines collected at the cyclone contain greatly larger quantities of iron than those entering the reactor with the fuel. Several tests [14] highlighted that part of the elemental iron of olivine, which as a consequence of catalytic action is linked with the coke layer by coordination complexes, can be detached from the particle by mechanical attrition and entrained out of the reactor in the syngas. Consequently, the values of QFe,fines/ QFe,fuel were estimated to be greatly larger than 1, typically 100 or more [14]. Data in Table 4 for the runs carried out with SRF at ER = 0.332 and ER = 0.272 indicate instead that QFe,fines/QFe,fuel is equal to about 0.25. This suggests that the catalytic action of olivine is not, or is only partially, present during air gasification of SRF, as indirectly confirmed by the absence of coke over the external surface of olivine particles in all the experimental tests. It is likely that the catalytic support to the cracking and isomerization determined by magnesium (that is 43% on mass basis of olivine) is always active, so that the heavier fragments are broken and a number of unsaturated hydrocarbons with two or three carbon atoms are formed. On the contrary, the catalytic enhancement of the dehydrogenation and carbonization determined by active sites of elemental iron is absent, and tar formation is not inhibited. There are two main possible reasons: (i) considering the very high content of Fe and Al in the SRF, the ferrous and non-ferrous metals in the ash can act as competitors of the iron oxides on the external surface of olivine particles, so avoiding their reduction to elemental Fe; (ii) the high content of oxygen in the SRF could prevent the reduction of the same iron oxides or may generate CO which, in turn, binds the iron to different kinds of carbonyls (such as Fe2(CO)9, Fe3(CO)12, and their hydrides): in both cases, iron catalytic activity is inhibited.
4. Concluding remarks The technical feasibility of air gasification of a solid recovered fuel in a bubbling fluidized bed gasifier has been investigated in a pilot scale reactor, having a maximum thermal output of 400 kW. The unit was operated under conditions of thermal and chemical steady state, for values of equivalence ratio ranging from 0.25 to 0.33. The temperature of gasification reaction zone varied in the range 850–930 °C, then it was always safely lower than the ash fusion temperature and sufficiently high to allow an efficient carbon conversion. The control of the optimal temperature can be achieved mainly by modulating the equivalence ratio and, just for a small extent, the air preheating temperature. The best process performances have been obtained by fluidizing the reactor with air and for values of equivalence ratio in the range 0.30–0.33. The obtained syngas has a LHV of about 5 MJ/m3N , that increases up to more than 7 MJ/m3N if the energy content of the tar is also taken into account. The cold gas efficiency, i.e. the
535
fraction of chemical energy of the SRF that is transferred to the syngas, can be as high as 60%. The carbon conversion efficiency, i.e. the percentage of inlet carbon that is converted to CO, CO2 and gaseous hydrocarbons, is between 0.8 and 0.9. The remaining part of carbon flow rate is present in the producer gas as condensable heavy hydrocarbons (tar) and carbonaceous fine particles escaped from the bed as entrained fines. The tests on the pilot scale reactor measured the flow rates of fines and tar at the reactor exit. These values indicate that, even under the optimal set of operating conditions, the concentrations of dusts and tars in the obtained syngas make very difficult to select and design an adequate fuel gas cleaning system and then could not allow the utilization of an internal combustion engine or a gas turbine, even because the catalytic action of olivine is only partially present during the SRF gasification process. It is likely that a proper configuration for small- or medium-size plants should directly sent the hot producer gas to a specifically designed gas burner. In this way, there will be no more necessity of a (probably complex and expensive) gas cleaning unit, even though some conditioning on the producer gas can be always operated before the producer gas combustion, and the plant will also value the heating content of tar compounds.
Acknowledgments The study was carried out with the financial support of Smaltimenti Sud s.r.l. Data reported in the paper are original calculations developed by the authors and cannot be considered as official information of Smaltimenti Sud s.r.l. The authors are greatly indebted with Mr. Onofrio Annoscia and Mr. Gianfranco De Troia of Eco-Engineering Impianti s.r.l. for their technical assistance during all the gasification tests and with Dr. Donato Santoro, which performed large part of the off-line laboratory analyses.
References [1] World Bank. What a waste_a global review of solid waste management. Urban development urban series knowledge papers 68135, 2012,
536
U. Arena, F. Di Gregorio / Fuel 117 (2014) 528–536
[15] Dunnu G, Panopoulos KD, Karellas S, Maier J, Touliou S, Koufodimos G, et al. The solid recovered fuel Stabilat: characteristics and fluidised bed gasification tests. Fuel 2012;93:273–83. [16] Di Gregorio F, Zaccariello L. Fluidized bed gasification of a packaging derived fuel: energetic, environmental and economic performances for waste-toenergy plants. Energy 2012;42:331–41. [17] Van Paasen SVB, Kiel JHA. Tar formation in a fluidized bed gasifier: impact of fuel properties and operating conditions. Report ECN-C-04-013, 2004, Petten, the Netherlands. [18] Felder RM, Rousseau RW. Elementary principles of chemical processes. 3rd ed. J. Wiley & Sons; 2000. [19] Channiwala SA, Parikh PP. A unified correlation for estimating HHV of solid, liquid and gaseous fuels. Fuel 2002;81:1051–63. [20] Kim M-R, Jang J-G, Lee S-K, Hwang BY, Lee JK. Correlation between the ash composition and melting temperature of waste incineration residue. Korean J Chem Eng 2010;27(3):1028–34. [21] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar elimination in biomass gasification processes. Biomass Bioenergy 2003;24:125–40. [22] Arena U, Di Gregorio F, Santonastasi M. A techno-economic comparison between two design configuration for a small scale, biomass-to-energy gasification based system. Chem Eng J 2010;162:580–90. [23] Kirnbauer F, Wilk V, Hofbauer H. The positive effects of bed material coating on tar reduction in a dual fluidized bed gasifier. Fuel 2012;95:553–62. [24] Mastellone ML, Arena U. Olivine as a tar removal catalyst during fluidized bed gasification of plastic waste. AIChE J 2008;54(6):1656–67.
[25] Gómez-Barea A, Ollero P, Leckner B. Optimization of char and tar conversion in fluidized bed biomass gasifiers. Fuel 2013;103:42–52. [26] Li C, Suzuki K. Tar property, analysis, reforming mechanism and model for biomass gasification – an overview. Renew Sustain Energy Rev 2009;13:594–604. [27] Anis S, Zainal ZA. Tar reduction in biomass producer gas via mechanical, catalytic and thermal methods: a review. Renew Sustain Energy Rev 2011;15:2355–77. [28] Brown D, Gassner M, Fuchino T, Maréchal F. Thermo-economic analysis for the optimal conceptual design of biomass gasification energy conversion systems. Appl Therm Eng 2009;29:2137–52. [29] Cavaliere A, de Joannon M. Mild combustion. Prog Energy Combust Sci 2004;30:329–66. [30] Cocco D, Deiana P, Cau G. Performance evaluation of small size externally fired gas turbine (EFGT) power plants integrated with direct biomass dryers. Energy 2006;31(10-11):1459–71. [31] Milani A, Saponaro A. Diluted combustion technologies. IFRF Combust J 2001;1:1–32. [32] Arena U, Di Gregorio F, Amorese C, Mastellone ML. A techno-economic comparison of fluidized bed gasification of two mixed plastic wastes. Waste Manage 2011;31:1494–504. [33] Zevenhoven R, Kilpinen P. Control of Pollutants in flue gases and fuel gases. Turku, Finland: Åbo Akademi University; 2001.