Gasification Of Two Biomass Fuels In Bubbling Fluidized Bed

Paper No. FBC99-0014

Gasification of Two Biomass Fuels in Bubbling Fluidized Bed

Proceedings of the 15th International Conference on Fluidized Bed Combustion May 16 - 19, 1999 Savannah, Georgia

Copyright ©1999 by ASME

Gasification of two Biomass Fuels in Bubbling Fluidized Bed F. Miccio* Istituto Ricerche sulla Combustione – Consiglio Nazionale delle Ricerche IRC-CNR via Metastasio 17, 80125 Napoli, Italy - Tel +39-81-5935379 - Fax +39-81-5931567

O. Moersch, H. Spliethoff, K.R.G. Hein Institut für Verfahrenstechnik und Dampfkesselwesen IVD-Universität Stuttgart Pfaffenwaldring 23 - 70550 Stuttgart, Germany - Tel. +49-711+6853565 - Fax. +49-711+6853491

leading to a finer elutriable particulate, since the ABSTRACT residual fuel particle after the devolatilization The gasification of two biomass fuels, conserves a good mechanical resistance. beech wood and a dry granular sewage sludge from the Swiss Combi process, has been experimentally studied. The gasification was car- INTRODUCTION The properties of the fuel particle play an ried out in a laboratory scale ABFB facility, operated at steady state. The fuels have both a high important part during the conversion of fossil content of volatile matters but differ principally and renewable fuels in bubbling fluidized beds. for their ash content, which is approximately 1% Segregation phenomena can be ascribed to highand 56% in beech and sewage sludge, respec- volatile fuels because of the propensity of partitively. The attention was focused on the pres- cles in course of devolatilization to move up to ence of particulate and tar in the producer gas the bed surface [Fiorentino, 1997] with a velocwhich affect the process efficiency and give ity depending principally, among various panegative drawbacks in the utilization in motors rameters, on particle density. The conversion of or turbines for power generation. The influence the residual char particle is strongly affected by of operating variables (i.e. process temperature topological and mechanical properties, the diffuand equivalence ratio) on the gasification per- sion of reactants and products as well as the formances was explored. Results show that the patterns of the associated comminution being composition of the producer gas is quite inde- strictly related to such properties [Bhatia and pendent of whatever fuel is gasified. As far as Perlmutter, 1981, Chirone et al., 1991]. Highsewage sludge is concerned, process perform- ash fuels can be subjected to catalytic effects ances are poorer and steady state operation is [Smoot and Smith, 1985]; consequently, the difficult, because of the high elutriation rate of burnout time of particles drastically changes. fines, the continuous increasing of the bed height The metal inclusions in ashes undergo vaporizaand the interaction between ash and bed materi- tion and successive condensation depending on als. The tar yield was always high for both fuels. their volatility; the presence of certain species in Unexpectedly, the gasification of a blend of two ashes could favor the sorption of such metals on fuels gave a minimum tar yield, probably as- the particle surface instead of as aerosols [Cenni cribed to a catalytic effect. A difference was et al., 1998]. Ashes could accumulate in the bed, found comparing the comminution behavior of sinterize or interact with inert materials [Anbeech wood and dry sewage sludge. The former thony, 1995], leading to the generation of agundergoes fragmentation with the generation of gregates or mineral deposits on the bed particle relatively large, elutriable fragments; the latter surface and to the modification of the bed fluidis principally subjected to mechanical abrasion, dynamic. _____________________________________________ * Author to whom correspondence should be addressed

The gasification represents a more efficient and attractive process for power generation from biomass with respect to the combustion. In both environments, the particle devolatilization occurs under quasi similar conditions, because a reducing atmosphere due to pyrolysis vapors involves the particle. Consequently, the knowledge about phenomena occurring during devolatilization (e.g. the fuel segregation during this stage or the primary fragmentation) can be extended from combustion to gasification. On the contrary, since the residence time of the char particle is much higher for gasification than for combustion, it can be expected that particle properties during the char conversion stage act differently, depending on whether combustion or gasification is carried out (e.g. comminution phenomena could be emphasized under gasification). The present work deals with the comparison of the gasification behavior in a bubbling fluidized bed of two biomass reference fuels, which have strong differences as far as ash content, ash composition and topological properties of the fuel particle are concerned. The goal of the paper is to ascertain that the composition of the producer gas, at the same operating conditions, does not change significantly with the fuel flare gas analysis tar sampling freeboard 135 mm ID

cyclone (trace heated)

ceramic candle filter (trace heated)

2.8 m

freeboard probe

air 5 heating zones (3,8 kW each)

screw feeder scale fluidization column 108 mm ID air

bed material withdrawal

electrical preheater (3,6 kW) gas distribution plate

Figure 1 - Experimental facility

choice, when using biomass fuels having a similar composition on dry ash-free basis. This means that the fate of volatiles, which represents the major part of the combustible matters, and the occurrence of homogeneous reactions are quite independent of the properties of char particles. It can also be expected that differences are found as far as the condensed phase behavior (i.e. ash evolution, comminution phenomena, mechanism of chemical reaction and diffusion in the char) is concerned. The comparison was pursued by focusing the attention on the quality of the producer gas, the ash behavior, the tar yield and the carbon elutriation, during gasification of single biomass fuels and their blend. EXPERIMENTAL The experimental facility used to carry out gasification tests is sketched in Fig. 1. Steady state operation is made possible by means of external heating by electrical furnaces. The reaction chamber consists of a bubbling fluidized bed with an internal diameter of 108 mm and a 125 mm ID freeboard section. The reactor is equipped with accessories and instrumentation to perform measurements of temperature, pressure, gas, tar and particulate concentrations during steady-state experiments of gasification. A more detailed description of the facility, the experimental technique and the procedure used for the laboratory analyses can be found elsewhere [Miccio et al., 1998]. The tar concentration was measured by means of a quasi on-line analyzer, designed and developed at the IVD, University of Stuttgart [Moersch et al., 1998], which allows measurements every two minutes. Two fuels were used during gasification tests: a beech wood (BW) and a dry granular sewage sludge (DSS). Table I shows physical and chemical properties of such fuels. The beech wood originated from local forests; after a predrying stage, it was chopped to the desired size by means of a cutting mill. Beech wood is a typical ligneous biomass, with high-volatile and low-ash content and a residual moisture ranging between 16 and 22%. The dry sewage sludge was obtained from the Swiss Combi process by mechanical and thermal treatment of a raw sew-

Table I - Physical and chemical properties of biomass fuels Fuel: Density, kg m3 Low heating value, MJ kg-1 Particle size range, mm Particle size distribution Average diameter, mm 0.06 0.16 0.35 0.50 0.90 1.50 3.00 5.00 Proximate analysis Moisture, % Volatiles (dry basis), % Ash (dry basis), % Fixed carbon (dry basis), % Ultimate analysis (dry basis), % Carbon Nitrogen Hydrogen Sulfur Ash Oxygen (by diff.) Ash analysis (ash basis), % Sodium Potassium Calcium Magnesium Aluminium Silicium Iron

Beech wood 510 18.38 0-5

Dry sewage sludge 1830 7.85 0-5

Cumulat. distr., % 1.10 2.33 5.53 9.32 17.08 48.08 96.24 100.00

Cumulat. Distr., % 0 0 0 0 0 4.75 89.97 100.00

16.0 - 22.0 84.90 1.03 14.07

4.68 41.61 56.11 2.28

50.40 0.26 7.21 0.00 1.03 41.10

20.42 2.35 3.18 0.61 56.11 17.33

age sludge. The final products are rather spherical granules (Fig. 2A), gray in color and uniform in size. The dry sewage sludge is dust-free, not smelling and storable for a long period. Differently from beech wood, it presents a high ash content (56.11 %) and, consequently, the calorific value evaluated on mass basis is low (7.85 MJ/kg). The granules of dry sewage sludge have a fairly good mechanical resistance; the particle breaks if subjected to a compression stress - hand operated by means of a pencil approximately equal to 10 N and fragments can be rather easily pulverized. For dry sewage sludge, Fig. 2 reports SEM pictures of a raw granule (A), a char particle (B) and an ash particle after complete burn-

0.561 1.121 10.657 0.860 4.052 19.071 5.235

off (C), the latter two being produced by fast heating up to 800 °C in a muffle furnace. It appears that there is not a significant change of shape and volume of the particle upon its conversion. Furthermore, the char particle (Fig. 2B) shows an increased internal porosity with respect to the raw granule (Fig. 2A); the porosity disappears in the exhausted particle (Fig 2C) because of ash fusion and re-arrangement during char conversion. The granulometric distributions of both fuels are reported in Table I. The top particle size was 5 mm in both cases; fines (i.e. lower than 1.0 mm) were retrieved only for beech wood. A blend was prepared by mixing equal masses of both fuels. The feeding of such a

A

B

velocity Umf in air at the temperature of 800 °C was 11 cm/s and the expanded bed height ranged between 80 and 90 cm. The main operating variables of experiments were the process temperature T and the equivalence ratio λ, defined as the mass ratio between the actual flow rate of the air and the stoichiometric flow rate required for the fuel combustion. The fluidization air was preheated up to 500 °C. The reactor worked at fixed gas velocities in the freeboard (i.e. 0.38 m/s). Consequently, the undisturbed fluidization velocity U was in the range 30 - 40 cm s-1 corresponding to the beginning of the slugging regime for the given bed diameter and sand size. RESULTS AND DISCUSSION

C

Figure 2: SEM pictures of a raw granule (A), a char particle (B) and an ash particle after complete burn-off (C) for dry sewage sludge blend in the reactor was proved to be sufficiently steady and reliable. The bed material was quartz sand with a narrow size range (0.40 - 0.60 mm) belonging to the B group of Geldart classification [Kunii and Levenspiel, 1991]. The minimum fluidization

Interaction between fuel ashes and bed material Upon gasification of beech wood there was no evidence of accumulating ashes in the bed or their interactions with quartz sand, up to a temperature of 900 °C. On the contrary, numerous ash skeletons were retrieved in discharged materials, and sand particles turned their color, from gray to red-brown, when dry sewage sludge was gasified. An appreciable increasing of the bed height was also noted during DSS gasification. The presence of ash skeletons in discharged materials confirms that fuel particles do not undergo fragmentation phenomena during stages of devolatilization and char conversion, as already observed for the particles converted in the muffle furnace (Fig. 2). Since DSS granules have a residual mechanical resistance upon their conversion and the fluidization velocity is kept at relatively low values, the mechanical abrasion operated by bed materials on the particle surface, which is consumed by a shrinking particle mechanism, is not able to give a generation rate of elutriable fines comparable to the mass rate of fuel ashes. Consequently, ashes accumulate in the bed leading to the dilution of the original quartz sand with particles distributed over a wide range (from 150 to 2000 µm).

Table II - Results of AAS and SEM analyses for sand materials Element

Atomic Adsorption Spectrometry Fresh sand g/kg

Si Ca Fe Al K Mg Na Zn Cu

280.000 1.500 2.215 8.400 12.900 0.170 0.830 0.007 0.016

Fatigued sand g/kg 311.000 15.000 9.223 12.100 9.000 1.010 1.310 0.200 0.083

Laboratory analyses were performed on samples taken from the bed in order to give a confirmation of the deposition of elements from ashes over particle surface. Table II reports the results of off-line analyses carried out by means of atomic adsorption spectroscopy (AAS) and surface analysis via scanning electron microscope (SEM). The comparison between fresh and fatigued sand (i.e. after approximately 10 hours of operation with dry sewage sludge at 800 °C) shows that fatigued sand was particularly enriched of some elements which are present in sewage sludge ashes, as reported in Tab. I (Ca, Fe and Mg). Their deposition occurred along the particle surface, since the measured value of the area covered by Si in fatigued sand drastically falls down in SEM analysis. Of course, the large deposition of iron oxides (∆ = 317 % and 2171 % for AAS and SEM, respectively) is responsible of the observed change in sand color. Aggregates with a size up to 40 mm were also retrieved in the bed after experiments carried out at a temperature of 900 °C; such aggregates have a dark color and a porous structure with a good mechanical resistance. Their formation must be ascribed to the fusion of ashes and/or sintering of ash and bed materials, because of the higher temperature during such tests. However, it is reasonable to suppose that the aggregates were produced at the end of the gasification experiment, when the fuel feeding is switched off and the carbon in the bed is burned,

∆ % 11 900 316 44 -30 494 58 2757 419

Surfacial Analysis Fresh sand g/kg

Fatigued sand g/kg

∆ %

944.000

406.300

-57

3.400 31.000 10.000 3.600

77.200 107.300 14.500 31.600

2171 246 45 778

following the experimental procedure for the determination of the bed carbon load [Miccio et al., 1998]. In such a situation the bed temperature increases over the set-point (up to 980 °C), the cooling of the bed in the actual reactor being not possible, and the temperature of char particles increases with respect to bed temperature. Producer gas Table III reports the average results of gasification tests operated using both fuels and their blend (50% by weight). The concentration of principal gaseous species on dry basis, the tar concentration in the producer gas, the carbon elutriation rates at the reactor exit (EC) and in the freeboard (EC,fr) and the carbon load in the bed are reported. The carbon elutriation rate is normalized with respect to the total carbon feeding rate. The quality of the producer gas (i.e. concentration of combustible species) is improved in all cases by decreasing the equivalence ratio and increasing the process temperature. The comparison between the results of the two fuels shows that the hydrogen concentration attained with dry sewage sludge is higher than that with beech wood, except in the case λ=0.35 and T=800 °C. The result can be explained taking into account that the ratio between volatiles and fixed carbon is higher for DSS than for BW (18.25 and 6.03, respectively). As a consequence, the role of homogeneous reactions is more important during dry sewage sludge gasifi-

Table III - Experiments of biomass gasification Operating conditions 1 2 Equivalence Process ratio temperature

Results 3 CO2

4 CO

-

°C

%

%

0.15 0.25 0.35 0.25

800 800 800 900

17.9 18.8 19.2 16.4

18.6 14.5 11.4 17.7

0.15 0.25 0.35 0.25

800 800 800 900

16.0 16.7 17.5 11.0

13.5 11.5 9.6 20.0

0.25 0.25

800 900

17.3 13.5

13.5 19.0

5 H2

8 9 10 11 Carbon Carbon Carbon Carbon elutriation elutriation conversion load exit freeboard in freeboard mg/m3 % % % % % g Beech wood 14.8 6.2 14000 5.5 15.0 9.5 81 12.8 4.0 8200 3.8 6.2 2.4 37 10.8 3.0 4500 2.9 2.9 0.0 17 15.3 3.8 4970 2.2 3.6 1.4 19 Dry sewage sludge 19.7 2.2 11400 6.4 8.1 1.7 176 14.7 2.5 6860 5.4 7.9 2.5 104 9.0 2.9 2620 4.7 8.2 3.5 57 21.0 2.7 3750 3.9 5.0 1.1 81 Beech wood - dry sewage sludge blend 14.7 3.8 5540 5.3 6.4 1.1 81 19.0 2.5 3050 3.0 4.0 1.0 44

cation, leading to increased formation of hydrogen by cracking and partial oxidation of gaseous hydrocarbons [Littlewood, 1977]. During gasification of the biomass blend, the gas composition is always at an intermediate level with respect to BW and DSS gasification operated at the same conditions. The tar concentration in the producer gas is high for both fuels and drastically decreases with the equivalence ratio (Fig. 3) and the temperature (Fig. 4). It is worth to note that dry sewage sludge gasification results in a lower tar yield, in spite of the higher ratio between volatiles and fixed carbon. Among various variables influencing tar genesis, the different organic structure of the fuel and the ash content could be considered. The former influences the nature of vapors from the fuel pyrolysis resulting in the generation of more refractory species when beech wood is pyrolyzed; the latter modifies the particle temperature and the duration of the devolatilization stage, which are known to influence the tar yield [Moersch, 1998]. The tar concentration with the biomass blend is unexpectedly lower than the values of BW and DSS at both the temperatures of 800 and 900 °C (Fig. 4). The experimental finding could be attributed to a catalytic effect [Bridgewater, 1995] operated by DSS ash, where Ca is present in a

6 CH4

7 tar

large amount and Mg in traces, leading to enhance the cracking of tar species in the regions of the bed and freeboard. However, this effect is not able to drastically reduce the tar concentration and principally operates on tar species from beech wood pyrolysis, because the tar abatement is less marked in the gasification of DSS alone. Particle conversion Figures 5 and 6 report the dependence on the equivalence ratio of the carbon elutriation rate and bed carbon load, respectively. Dry sewage sludge shows higher values of such variables, if compared to beech wood. The difference is more pronounced in the case of the bed carbon load, which always increases at least by 100 %, when switching from BW to DSS gasification. Furthermore, the results are consistent with the experimental findings of Donsì et al. [1978] for coal combustion, which show the increasing of the carbon elutriation rate with the bed carbon load. The different fragmentation behavior of the two fuels could provide an interpretation of the large difference in the bed carbon load. A ligneous biomass, similar to beech wood, has been characterized by Chirone et al. [1997], which demonstrated the occurrence of primary and percolative fragmentation, during

15000 beech wood dry sewage sludge

Tar concentration, mg m-3

12500

10000

7500

5000

2500

0 0.1

0.15

0.2

0.25

0.3

0.35

0.4

Equivalence ratio Figure 3: Tar concentration versus the equivalence ratio (T=800 °C) combustion, with the consequent generation of smaller fragments. DSS granules do not undergo fragmentation, conserving their original size and shape during conversion. The larger the particle size, the higher the resistance to mass diffusion, the lower is the global kinetics of char gasification. It is also known from literature [Smooth and Smith, 1985] that the reactivity of carbonaceous particles appreciably decreases at a late state of their conversion, because carbon is diluted in the ashes and the internal diffusion of gaseous reactants is more difficult. Similarly, the high ash content of DSS represents an obstacle to the gas diffusion; as a consequence, the intrinsic kinetics of the particle conversion is lower than for low-ash fuels. In conclusion, the bed carbon load increases, a longer time being required for the complete conversion of fuel particles in the bed. Figure 7 reports the size distribution measured by means of a Malvern equipment of powders collected at the cyclone (Fig. 7A) and the ceramic filter (Fig. 7B). It clearly appears that the size of elutriated fines is shifted toward smaller diameters, passing from beech wood to dry sewage sludge. Furthermore, during the

beech wood gasification particles with a size larger than the upper detection limit of the apparatus (i.e. 564 µm) are collected at the cyclone. These particles must be very porous and light in order to be elutriable. Again, results reported in Fig. 7 confirm the main difference in the evolution of the fuel particle during its conversion, as far as comminution phenomena are concerned. The beech wood undergoes fragmentation (primary, secondary and percolative) because of the absence of a coherent ash skeleton and the increasing porosity with the particle conversion. Elutriable particles of relatively large size are produced in the bed via fragmentation and successively separated by the cyclone. On the contrary, the fines generation during dry sewage sludge gasification is principally caused by the mechanical abrasion along the external surface of the particle, leading to the formation of fines with smaller sizes. Results of particle sampling by means of the freeboard probe are also reported in Table III in terms of the percent carbon elutriation rate EC,F (col. 9) and percent carbon conversion in the freeboard (col. 10). The latter is the difference between columns 9 and 8. The residence

9000 T=800°C T=900°C

Tar concentration, mg m -3

8000 7000 6000 5000 4000 3000 2000 1000 0 beech wood

blend

dry sewage sludge

Fuel Figure 4: Tar concentration as a function of the fuel type (λ=0.25) time of gas in the freeboard is approximately equal to 4 s for each test. Accordingly to measurements at reactor exit, EC,F decreases with an increasing of the equivalence ratio, during gasification of beech wood at 800 °C. The situation is reversed in experiments with dry sewage sludge gasified at the same temperature: EC,F assumes a rather constant value (approximately 8%), whatever λ may be. The carbon conversion in the freeboard strongly decreases for BW tests (from 9.5 to 0 %) and smoothly increases for DSS tests (from 1.7 to 3.4 %), with an increasing of the equivalence ratio. The difference in the comminution behavior could partially explain the different trend of carbon conversion in the freeboard for biomass fuels under investigation. Since elutriated fines from beech wood have a larger size dp (Fig. 7), the particle terminal velocity Ut in the freeboard could not be negligible with respect to the gas velocity Ug,F, leading to increase the residence time of particles in the freeboard. The terminal velocity (Eq. 1) has been evaluated following the procedure proposed by Kunii and Levenspiel (1991).

Ut =

(

)

4d p ρ p − ρ g g 3ρ g C D

(1)

In the equation ρp and ρg are the densities of the particle and gas, g is the gravity acceleration and CD is a function or the particle Reynolds number. In turn, the particle residence time in the freeboard tp,F is calculated by Eq. 2, being LF the length of the freeboard: LF t p, F = (2) U g ,F − U t Assuming for BW char an apparent density of 300 kg m-3, from a comparison with data available in literature for a similar ligneous biomass [Masi et al., 1997], and average particle sizes of 800, 400 and 200 µm, the terminal velocity at 800 °C is equal to 1.76, 0.88 and 0.15 m s-1, respectively. It clearly appears that the terminal velocity is smaller than the gas velocity in the freeboard (i.e. 0.38 m/s) only in the last case. It is worth to note that the assumed value of ρp is referred to an unconverted biomass char and the irregular shape of char particles is not taken into consideration, leading to the overestimation of the terminal velocity.

8

Carbon elutriation rate, %

beech wood dry sewage sludge 6

4

2

0 0.1

0.15

0.2

0.25

0.3

0.35

0.4

Equivalence ratio Figure 5: Carbon elutriation rate versus the equivalence ratio (T=800 °C) Anyway, the finding that the terminal velocity is comparable with the gas velocity, leads to an appreciable increasing of the residence time in the freeboard, via Eq. 2. Furthermore, it is reasonable to suppose that the highest generation rate of large fragments is obtained at λ=0.15, i.e. under unfavorable conditions of gasification. Differently from beech wood, particles elutriated during DSS gasification are smaller than 150 µm (Fig. 7), corresponding to low values of the terminal velocity (e.g. Ut = 0.08 m/s at dp=80 mm and rp=1000 kg m-3) and the residence time of the particles is practically equal to that of the gas. A further effort to explain the difference of freeboard carbon conversion could be done looking at the carbon content of elutriated fines. This latter is much lower for dry sewage sludge (< 5%) than for beech wood (approximately 50%), because of the huge ash content in char particles of DSS. Again, elutriated fines which leave the bed are less reactive for dry sewage sludge than for beech wood [Smooth and Smith, 1985]. The cooperation of the above mentioned factors, longer residence time and higher char reactivity, leads to an enhancement of the car-

bon conversion in the freeboard during beech wood gasification with respect to dry sewage sludge, in particular under unfavorable conditions of gasification. Finally, as far as results of the particle conversion are concerned, the blend fuel shows an intermediate behavior with respect to the single fuels (Tab. 3). The results confirms the absence of mutual effects or interactions between the particles of different fuels (e.g. catalysis by ashes), which can expected only if the fuel mixing is more intense and at the particle scale.

250 beech wood dry sewage sludge

Bed carbon load, g

200

150

100

50

0 0.1

0.15

0.2

0.25

0.3

0.35

0.4

Equivalence ratio Figure 6: Bed carbon load versus the equivalence ratio (T=800 °C) CONCLUSIONS The gasification of two biomass fuels, beech wood and dry sewage sludge, was carried out rather smoothly in a bubbling fluidized bed reactor. The quality of the producer gas is quite independent of the fuel, achieving concentrations of combustible species which allow the utilization for power production. The feeding of a blend of such fuels was also tested, giving a producer gas with intermediate properties with respect to that of single fuels. For dry sewage sludge strong interactions with bed materials were noted: accumulation of ashes in the bed, deposition of mineral matters on bed particles and sintering of ash and sand with formation of relatively large aggregates. The tar yield was always high, whatever fuel was gasified. Tar concentration for dry sewage sludge was lower than for beech wood, confirming that tar genesis is sensitive to particle properties and to the chemical nature of the organic components. Unexpectedly, the gasification of the blend gave a minimum tar yield. The results could be explained by supposing a catalytic effect of sewage sludge ashes on tar species

from beech wood pyrolysis. The phenomenon represents an interesting aspect for further studies. Carbon conversion and elutriation occurred simultaneously during the gasification. Again, particle properties influence the comminution behavior: for beech wood the generation of coarse elutriable particles was prevalently operated by fragmentation paths. On the contrary, dry sewage sludge was principally subjected to mechanical attrition, the granule having a coherent ash skeleton. As a consequence, size and shape of elutriated particles change significantly. Measurements of particulate concentration and chemical analyses of the samples obtained in the freeboard and at reactor exit showed that the carbon conversion in the freeboard is relevant, under certain conditions, for beech wood. On the contrary, for dry sewage sludge, because of the smaller residence time of elutriated particles and the lower char reactivity, the carbon conversion in the freeboard is less sensitive to the equivalence ratio.

ACKNOWLEDGMENTS The author F. Miccio gratefully acknowledges the Alexander Von Humboldt Stiftung, Bonn (Germany) for the granting of a research fellowship at the IVD, University of Stuttgart, in which framework this research was carried out. The authors are indebted to Ms. G. Lotsch, Ms. S. Mayer and Ms. A. Wöll of IVD for chemical analysis of samples. Ms. C. Zucchini of IRCCNR is also acknowledged for SEM elaboration.

REFERENCES Anthony, E.J. (1995), "Fluidized Bed Combustion of Alternative Solid Fuels; Status, Successes and Problems of the Technology", Prog. Energy Combust. Sci., Vol. 21, 239 Bhatia, S.K. and Perlmutter, D.D. (1981), "A Random Pore Model for Fluid-Solid Reactions: Diffusion and Transport Effects", AIChE J., Vol. 27, p. 247 Bridgewater, A.V. (1995), "The Technical and Economical Feasibility of Biomass Gasification for Power Generation", Fuel, Vol. 74, p. 631

10 9

beech wood dry sewage sludge

A

Size distribution, %

8 7 6 5 4 3 Cyclone

2 1 0 10 1 9

10

100 beech wood dry sewage sludge

B

8 Size distribution, %

1000

7 6 5 4 3 Filter

2 1 0 1

10

Particle size, µm

100

1000

Figure 7: Size distribution of collected fines at the cyclone (A) and the filter (B) (λ=0.15 and T=800°C)

Cenni, R., Frandsen, F., Gerhardt, T., Spliethoff, H., Hein, K.R.G. (1998), "Study on Metal Species Partitioning along the Flue Gas Path in Combustion of Bituminous Coal and Dry Sewage Sludge", Proc. of 1998 Int. Conf. on Incineration & Thermal Treatment Technologies, May 11-15 Salt Lake City, Utah, p. 689 Chirone, R., Greco, G., Salatino, P. and Scala, F. et al. (1997), "The Relevance of Comminution Phenomena in the Fluidized Bed Combustion of a Biomass (Robinia Pseudoacacia)", Proc. of 14th ASME FBC Conference, Vol. 1, p. 145 Chirone, R., Massimilla, L., and Salatino, P. (1991), "Comminution of Carbons in Fluidized Bed Combustion", Prog. Energy Combust. Sci., Vol. 17, 297 Donsì, G., Massimilla, L., Russo, G. and Stecconi, P. (1978), "Carbon Load and Gas Composition in a Fluidized Bed Combustor", Proc. 17th Int. Symp. on Comb., Leeds, p. 205 Fiorentino, M., Marzocchella, A. and Salatino, P. (1997), "Segregation of Fuel Particles and Volatile Matter During Devolatilization in a Fluidized Bed Reactor: II Experimental", Chem. Eng. Sc., Vol. 52, p. 1909 Kunii, D. and Levenspiel, O. (1991), Fluidization Engineering 2nd Ed., Butterworth-Heinemann, Boston, MA Littlewood, K. (1977), "Gasification: Theory and Application", Prog. Energy Combust. Sci., Vol. 3, 35 Masi S., Salatino, P. and Senneca, O. (1997), "Combustion Rates of Chars from High Volatile Fuels for FBC Application", Proc. of 14th ASME FBC Conference, Vol. 1, p. 135 Miccio, F., Moersch, O., Spliethoff, H., Hein, K.R.G. (1998), "Generation and Conversion of Carbonaceous Fine Particles during Bubbling Fluidised Bed Gasification of a Biomass Fuel", submitted to Fuel Moersch, O., Spliethoff, H., Hein, K.R.G. (1998), "Untersuchung der Moeglichkeiten zur Minderung des Teergehalts bei der Wirbelschichtvergasung von Biomasse", DBU Contract Report Moersch, O., Spliethoff, H., Hein, K.R.G. (1998), "Quasi Continuous Tar Quantification with a New Online Analyzing Method", Proc. of 10th European Conf. and Technology Exhibition, Würzburg, 8-11 June 1998, Vol. 1, p. 1638 Smoot, L.D. and Smith, P.J. (1985), Coal Combustion and Gasification, Plenum Press, New York NY