Process and device for rapid torrefaction of biomass

Abstract

A process for controlling a unit for torrefaction of biomass particles including: measurement of a mean diameter of the biomass particles; as a function of the mean diameter, calculation of a maximum torrefaction temperature for which the ratio of a characteristic time of the torrefaction to a characteristic time of the heat transfer at the level of the particles is equal to a minimum value; adjusting the torrefaction temperature to a value less than or equal to the maximum torrefaction temperature; adjusting the torrefaction time to a value such that the final yield by weight of the torrefaction is equal to a predefined target value.

Claims

1. A process for torrefying biomass particles or lignocellulosic particles t.sub.r T.sub.g in a torrefaction reactor, said process comprising: measuring a mean diameter d.sub.p of the biomass particles by a particle-diameter measurement device; receiving the mean diameter d.sub.p of the biomass particles at a computational device; according to the mean diameter d.sub.p of the biomass particles, computing by the computational device a maximum torrefaction temperature T.sub.g.sup.max for which a ratio H of a characteristic torrefaction reaction time to a characteristic heat transfer time in the biomass particles is equal to a predefined minimum value H.sup.min; setting a heater to maintain a torrefaction temperature T.sub.g of a fluid in said torrefaction reactor to a value less than or equal to said maximum torrefaction temperature T.sub.g.sup.max; setting a torrefaction time t.sub.r in said torrefaction reactor to a value such that a final mass yield R of the torrefaction is equal to a predefined target value; and torrefying the biomass particles in the fluid at the torrefaction temperature T.sub.g in the torrefaction reactor, during a time equal to the torrefaction time t.sub.r.

2. The process according to claim 1, wherein the torrefaction temperature T.sub.g is set equal to said maximum torrefaction temperature T.sub.g.sup.max.

3. The process according to claim 1, wherein said predefined minimum value H.sup.min of said ratio H is greater than or equal to 2, or greater than or equal to 5, or greater than or equal to 10.

4. The process according to claim 1, wherein said predefined target value of the final mass yield R of the torrefaction is greater than or equal to 60%, or 70%, or 80%.

5. The process according to claim 1, wherein said torrefaction temperature T.sub.g is within a range of 300° C.-400° C., or a range of 300° C.-350° C., or a range of 325° C.-350° C.

6. The process according to claim 1, wherein said torrefaction time t.sub.r is less than 15 minutes, or less than 10 minutes, or less than 5 minutes.

7. The process according to claim 1, wherein said maximum torrefaction temperature T.sub.g.sup.max is determined according to said minimum value H.sup.min of said ratio using the formula: $H^{\min }=\frac{1/k_t}{\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p^2}{36{\lambda }_{\mathrm{eff},p}}}$ wherein: k.sub.t is a reaction rate constant for the torrefaction reaction, considered to be a first-order reaction: ρ.sub.p is the volumetric mass density of the biomass; Cp.sub.p is the specific heat of the biomass; and λ.sub.eff,p is the effective thermal conductivity of the biomass.

8. The process according to claim 1, wherein said maximum torrefaction temperature T.sub.g.sup.max is determined according to said minimum value H.sup.min of said ratio using the formula: $H^{\min }=\frac{1/k_t}{\begin{array}{c}\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p^2}{36{\lambda }_{\mathrm{eff},p}}+\\ \min \left(\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p}{6h_{\mathrm{conv}}};\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p}{6{\omega }_p\sigma \left(T_g^{\max }+T_p\right)\left(T_g^{\max 2}+T_p^2\right)}\right)\end{array}}$ wherein: k.sub.t is a reaction rate constant for the torrefaction reaction, considered to be a first-order reaction; ρ.sub.g is the volumetric mass density of the biomass; Cp.sub.p is the specific heat of the biomass; λ.sub.eff,p is the effective heat conductivity of the biomass; ω.sub.p is the emissivity of the biomass; h.sub.conv is the external heat transfer coefficient between the biomass and said fluid; σ is the Stefan-Boltzmann constant; and T.sub.p is the initial temperature of the biomass particles before torrefaction.

9. The process according to claim 8, wherein said external heat transfer coefficient h.sub.conv is determined using the Ranz-Marshall correlation, by the formula: $h_{\mathrm{conv}}=\frac{{\lambda }_g}{d_p}\left(2+0.6{\mathrm{Re}}_p^{1/2}.Math.{\mathrm{Pr}}^{1/2}\right)$ wherein: Re.sub.p is the Reynolds number determined using the formula: ${\mathrm{Re}}_p=\frac{{\rho }_g\times V_g\times d_p}{{\mu }_g};$ Pr is the Prandtl number determined using the formula: $\mathrm{Pr}=\frac{{\mu }_g\times {\mathrm{Cp}}_g}{{\lambda }_g};$ wherein: ρ.sub.g is the volumetric mass density of the fluid; μ.sub.g is the viscosity of the fluid; λ.sub.g is the thermal conductivity of the fluid; Cp.sub.g is the specific heat of the fluid; and V.sub.g is a velocity of the fluid relative to the biomass particles.

10. The process according to claim 1, wherein the torrefaction time t.sub.r is determined according to the torrefaction temperature T.sub.g and the predefined target value of the final mass yield R of the torrefaction, by predetermined experimental correlation data.

11. The process according to claim 1, wherein the mean diameter d.sub.p of the biomass particles is less than 40 mm, or less than 20 mm, or less than 10 mm.

12. The process according to claim 1, wherein the biomass consists of wood.

13. The process according to claim 1, wherein said fluid is an inert gas, or dinitrogen N.sub.2.

14. A system for torrefying biomass particles or lignocellulosic particles t.sub.r T.sub.g, the system comprising: a torrefaction reactor; a heater for heating a fluid in the torrefaction reactor; and a control device including a particle-diameter measurement device configured to measure a mean diameter d.sub.p of the biomass particles and a computational device configured to receive the mean diameter d.sub.p of the biomass particles from the particle-diameter measurement device, wherein said computational device is configured to compute, according to the previously measured mean diameter d.sub.p of the biomass particles, a maximum torrefaction temperature T.sub.g.sup.max for which the ratio H of a characteristic torrefaction reaction time to a characteristic heat transfer time in the biomass particles is equal to a predefined minimum value H.sup.min; said control device is configured to set the heater to maintain the fluid in the torrefaction reactor at a torrefaction temperature T.sub.g less than or equal to said maximum torrefaction temperature T.sub.g.sup.max; said control device is configured to torrefy the biomass particles in said fluid during a torrefaction time t.sub.r; and said computational device is configured to compute the torrefaction time t.sub.r to be a value such that a final mass yield R of the torrefaction of the biomass particles is equal to a predefined value.

15. The system according to claim 14, wherein the control device is configured to preset said minimum value H.sup.min of said ratio H and/or to preset the final mass yield R of the torrefaction.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) The invention will be understood more clearly and further details, advantages and features thereof will emerge on reading the following non-limiting description with reference to the appended figures wherein:

(2) FIG. 1 is a partial schematic longitudinal sectional view of a torrefaction unit according to one preferred embodiment of the invention;

(3) FIG. 1a is a view on a larger scale of the detail la in FIG. 1;

(4) FIG. 2 is a graph illustrating the relationship between the maximum torrefaction temperature T.sub.g.sup.max and the diameter d.sub.p of the biomass particles for three values of the ratio H;

(5) FIGS. 3 and 4 are graphs illustrating, for two biomass particle diameter values d.sub.p, the progression of the final mass yield R of the torrefaction as a function of the torrefaction time t.sub.r, for three torrefaction temperature values T.sub.g;

(6) FIG. 5 is a partial schematic side view of a device for the experimental estimation of torrefaction homogeneity in torrefied biomass particles;

(7) FIG. 6 is a schematic top view of a torrefied biomass particle, illustrating a measuring principle used by means of the device in FIG. 5.

(8) Throughout these figures, identical references may denote identical or similar elements.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

(9) FIG. 1 represents a laboratory torrefaction unit 10, intended for the torrefaction of solid biomass particles, for example wood, and comprising a tubular reactor 12, along with a heating element 14 formed from resistors encompassing the reactor 12 and capable to reaching up to approximately 1000° C.

(10) The reactor 12 is of the double-jacketed type, and thus comprises an outer tube 16 made of quartz and an inner tube 18 also made of quartz and extending inside the outer tube 16 from a first 20 of the ends thereof to a median part thereof, wherein the inner tube 18 has an open end 22. The opposite end 24 of the inner tube projects beyond the outer tube 16 and is connected to an analysis device 26. The first end 20 of the outer tube 16 is closed around the inner tube 18. The outer tube 16 comprises a first part 27a comprising the first end 20 of the tube 16 and open at the opposite end thereof, along with a second removable part 27b comprising the opposite end 28 of the tube 16. The second part 27b of the tube 16 is suitable for being connected to the first part 27a of the tube by means of a coupling 27 as illustrated in FIG. 1.

(11) The outer tube 16 comprises, in the vicinity of each of the opposite ends 20 and 28 thereof, a coupling 30, 32 to a pressurised gas source 33, such as dinitrogen N2. Each coupling 30, 32 is connected to monitoring means 34 and means 35 for measuring the gas flow rate entering the reactor 12.

(12) The second end 28 of the outer tube 16 is provided with an orifice for the passage of a rod 36 having at the end thereof arranged inside the reactor 12, a sample-holder 40 equipped with a thermocouple 42. The rod 36 is movable in translation along the longitudinal direction, between a retracted position wherein the sample-holder 40 is near the second end 28 of the outer tube 16 and is situated outside the hole of the heating element 14, and an extended position wherein the sample-holder is removed from the second end 28 of the outer tube 16 and is situated in the hole of the heating element 14.

(13) In the example illustrated, the length of the outer tube 16 is approximately 1290 mm and that of the inner tube 18 is approximately 850 mm. Furthermore, the internal diameter of the outer tube 16 is approximately 70 mm and that of the inner tube 18 is approximately 55 mm. The longitudinal span of the heating element 14 is approximately 620 mm.

(14) The torrefaction unit 10 further comprises a control device 44, represented very schematically in FIG. 1, and connected, at the input, to the flow rate measuring means 35, to the thermocouple 42, to temperature measuring means integrated in the heating element 14, and to the analysis device 26, and, at the output, to the flow rate monitoring means 34 and to the heating element 14.

(15) The control device 44 comprises a device 46 for measuring the mean diameter d.sub.p of the biomass particles. This measuring device may be of any conventional type, operating for example using optical measurements and image processing algorithms, or using mechanical sensors. This device may alternatively adopt the form of an input unit for entering a value of the diameter previously measured by a human operator.

(16) The control device 44 also comprises a computing unit 48 designed to determine, according to the mean diameter d.sub.p and the initial temperature T.sub.p measured for the biomass particles, a maximum torrefaction temperature T.sub.g.sup.max for which the ratio H of a characteristic torrefaction reaction time to a characteristic heat transfer time in the biomass particles is equal to a predefined minimum value H.sup.min.

(17) The ratio H is defined using the formula:

(18) $H=\frac{1/k_t}{\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p^2}{36{\lambda }_{\mathrm{eff},p}}+\min \left(\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p}{6h_{\mathrm{conv}}};\frac{{\rho }_p{\mathrm{Cp}}_p{d}_p}{6{\omega }_p\sigma \left(T_g+T_p\right)\left(T_g^{{}^2}+T_p^2\right)}\right)}$

(19) wherein: k.sub.t is a reaction rate constant for the torrefaction reaction, considered to be a first-order reaction; ρ.sub.p is the volumetric mass density of the biomass particles; Cp.sub.p is the specific heat of these particles; λ.sub.eff,p is the effective thermal conductivity of these particles, given by the formula: λ.sub.eff,p=ε.sub.pλ.sub.g+(1−ε.sub.p)λ.sub.p where ε.sub.p is the porosity of the biomass particles and λ.sub.p is the thermal conductivity of these particles; ω.sub.p is the emissivity of these particles; h.sub.conv is the external heat transfer coefficient between these particles and the inert gas; σ is the Stefan-Boltzmann constant equal to 5.67.10.sup.−8 W.Math.m.sup.−2.Math.K.sup.−4.

(20) It should be noted that the numerator of the above formula corresponds to the characteristic torrefaction reaction time whereas the denominator corresponds to the characteristic heat transfer time in the biomass particles. The first term of this denominator corresponds to a characteristic heat transfer time by conduction and radiation in the biomass particles whereas the second term corresponds to the minimum characteristic external heat transfer times by convection and radiation respectively.

(21) The computing unit 48 determines the reaction rate constant k.sub.t on the basis of prerecorded data for one or a plurality of types of biomass to be torrefied, as will emerge more clearly hereinafter.

(22) The computing unit 48 determines the external heat transfer coefficient h.sub.conv on the basis of the Ranz-Marshall correlation, using the formula:

(23) $h_{\mathrm{conv}}=\frac{{\lambda }_g}{d_p}\left(2+0.6{\mathrm{Re}}_p^{1/2}.Math.{\mathrm{Pr}}^{1/2}\right)$

(24) wherein: Re.sub.p is the Reynolds number determined using the formula:

(25) ${\mathrm{Re}}_p=\frac{{\rho }_g\times V_g\times d_p}{{\mu }_g};$ Pr is the Prandtl number determined using the formula:

(26) $\mathrm{Pr}=\frac{{\mu }_g\times {\mathrm{Cp}}_g}{{\lambda }_g};$

(27) and where: ρ.sub.g is the volumetric mass density of the inert gas; μ.sub.g is the viscosity of this gas; λ.sub.g is the thermal conductivity of this gas; Cp.sub.g is the specific heat of this gas; V.sub.g is a velocity of this gas relative to the biomass particles.

(28) The velocity V.sub.g is determined using the formula:
V.sub.g=D.sub.g/S.sub.18

(29) wherein D.sub.g is the total gas flow rate injected via the couplings 30 and 32, which may be for example in the region of 1 L/min, and S.sub.18 is the cross-section of the inner tube 18.

(30) The control device 44 further comprises a monitoring unit 50 designed to set the torrefaction temperature T.sub.g to a value less than or, preferably, equal to the maximum torrefaction temperature T.sub.g.sup.max determined by the computing unit 48.

(31) Furthermore, said computing unit 48 is designed to compute a torrefaction time t.sub.r for which, given the torrefaction temperature T.sub.g, the final mass yield R of the torrefaction is equal to a predefined value.

(32) For this purpose, the computing unit 48 comprises a memory wherein experimental data giving the final mass yield R corresponding to various triplets {torrefaction temperature T.sub.g; torrefaction time t.sub.r; mean diameter d.sub.p of the biomass particles} are recorded. The computing unit is thus designed to determine the torrefaction time t.sub.r by extrapolation on the basis of these pre-recorded data.

(33) The monitoring unit 50 is designed to discontinue the torrefaction treatment in the reactor 12 after the torrefaction time t.sub.r determined by the computing unit 48.

(34) The control device 44 further comprises an input unit 52 for entering the minimum value H.sup.min of the above-mentioned ratio H and the final mass yield R sought for torrefaction.

(35) Preferably, the input unit further allows the entry of the type of biomass to be torrefied, in which case the computing unit comprises a memory wherein the values of the parameters relating to various types of biomass, i.e. the reaction rate constant k.sub.t, the volumetric mass density ρ.sub.p, the specific heat Cp.sub.p, the effective thermal conductivity λ.sub.eff,p, and the emissivity ω.sub.p, are recorded. Alternatively, the input unit may be designed to enable direct entry of the respective values of these parameters or of parameters suitable for determining same.

(36) The same applies for the parameters relating to the inert gas, i.e. the volumetric mass density ρ.sub.g thereof, the viscosity μ.sub.g thereof, the thermal conductivity λ.sub.g thereof, and the specific heat Cp.sub.g thereof.

(37) The torrefaction unit 10 described above is a unit intended to be used for experimental purposes in a laboratory, for small volumes of biomass. However, the above description may be readily adapted by those skilled in the art to any type of torrefaction unit, particularly for industrial purposes, for the torrefaction of large volumes of biomass.

(38) The torrefaction unit 10 described above may be used as follows for the torrefaction of a sample 54 of biomass particles.

(39) Firstly, the sample 54 may optionally be weighed after drying to constant mass.

(40) After starting up the control device 44, an operator enters, by means of the input device 52, the type of biomass to be torrefied, the minimum value H.sup.min of the above-mentioned ratio H, and the value of the final mass yield R sought for torrefaction.

(41) The computing unit then determines the maximum torrefaction temperature T.sub.g.sup.max compatible with these values, along with the corresponding torrefaction time t.sub.r.

(42) The control device 44 then causes the start-up of the heating element 14 and the temperature regulation thereof using temperature measurement means integrated in this heating element.

(43) Furthermore, the second part 27b of the outer tube 16 being detached from the first part 27a thereof, the operator arranges the sample 54 on the sample-holder 40 which is then situated outside the reactor 12. The operator then connects the second part 27b of the outer tube 16 to the first part 27a thereof, and arranges the sample-holder 40 on the side of the second end 28 of the outer tube, outside the region encompassed by the heating element 14, by holding the rod 36 in the retracted position thereof.

(44) The control device 44 controls the flow rate monitoring means 34 and the inert gas source 33 such that pressurised gas constantly enters via the couplings 30 and 32 of the outer tube 16 of the reactor 12. A corresponding gas flow is thus discharged to the analysis device 26, so as to drain the reactor 12.

(45) The gas injected into the reactor 12 via the coupling 30 near the first end 20 of the outer tube 16 is heated between the outer tube 16 and the inner tube 18 before reaching a median region of the reactor 12 encompassed by the heating element 14. The gas injected into the reactor 12 via the other coupling 32 is suitable for immersing the sample 54 so as to maintain same at the initial temperature thereof before the start of torrefaction and thus prevent the heat treatment of the sample 54 from being initiated in an uncontrolled manner. It should be noted that the gas flow rate injected via this other coupling 32 is preferably equal to one-third of the total gas flow rate injected into the reactor 12 via the two couplings 30 and 32. The control device 44 regulates the heating element 14 such that the temperature of the gas inside the reactor 12, determined by the temperature measuring means integrated in the heating element 14, remains constantly equal to torrefaction temperature T.sub.g which may be less than, or preferably equal to the maximum torrefaction temperature T.sub.g.sup.max determined previously.

(46) Once the analysis device 26 indicates to the control device 44 that the dioxygen level present in the reactor 12 is sufficiently low, such that this reactor is filled with an inert gaseous medium suitable for torrefaction, and if the gas temperature has reached the torrefaction temperature T.sub.g, the rod 36 is moved in translation to the extended position thereof illustrated in FIG. 1, either manually by a human operator, or automatically by means of a robotic device provided for this purpose (not shown in FIG. 1).

(47) The temperature regulation of the heating element 14 may as such be based on a temperature measurement made by the thermocouple 42 arranged as close as possible to the sample 40 for maximum precision. The temperature measurement may obviously be made by any other means within the scope of the present invention.

(48) After a time equal to the torrefaction time t.sub.r determined previously, the rod 36 is moved to the retracted position thereof, either manually by the human operator, or automatically by means of the above-mentioned robotic device.

(49) One particular application of the torrefaction process described above to the torrefaction of beech particles will now be described.

(50) For this material, the values of the parameters required for computing the ratio H are for example: ρ.sub.p=710 Kg.Math.m.sup.−3; Cp.sub.p=1522 J.Math.Kg.sup.−1K.sup.−1; λ.sub.p=0.112 W.Math.m.sup.−1K.sup.−1; ε.sub.p=0.7; ω.sub.p=0.9.

(51) Furthermore, the reaction rate constant k.sub.t is determined according to the studies by Di Blasi and Branca presented in the article Di Blasi, C. and C. Branca (2001): “Kinetics of Primary Product Formation from Wood Pyrolysis.”, Industrial & Engineering Chemistry Research 40(23): 5547-5556.

(52) This reaction rate constant k.sub.t is thus estimated using the formula:
k.sub.t=4.38.10.sup.9e.sup.−141200/R.Math.T.sup.g

(53) The gas used being dinitrogen N2, the values of the corresponding parameters are well known to those skilled in the art. It is noted that the thermal conductivity of this gas is obtained by the following formula:

(54) 0${\lambda }_g=\frac{{\mu }_g}{M_g}\left[1.3\left({\mathrm{Cp}}_g{M}_g-R\right)+14644-\frac{2928.8}{T_g/{\mathrm{Tc}}_g}\right]$

(55) wherein: M.sub.g is the molar mass of the gas equal to 0.028 kg/mol; R is the ideal gas constant equal to 8.3145 J/(mol.Math.K); Tc.sub.g is the critical temperature of the gas (−146° C.); Cp.sub.g is the specific heat of the gas obtained using the formula:

(56) ${\mathrm{Cp}}_g=\frac{6.5+0.001T_g}{M_g}*4.18$

(57) FIG. 2 represents three curves illustrating the relationship between the torrefaction temperature T.sub.g and the mean diameter d.sub.p of the biomass particles such that the ratio H calculated using the above formula is respectively equal to 2, 5 and 10, in the case of beech particles torrefied in dinitrogen N2, these particles being initially at a temperature T.sub.p=25° C.

(58) This type of graph illustrates the principle used in the control device 44 described above consisting of determining the maximum torrefaction temperature T.sub.g.sup.max guaranteeing a predefined minimum degree of homogeneity H.sup.min, according to the mean diameter d.sub.p of the biomass particles.

(59) FIGS. 3 and 4 illustrate data pre-recorded in the control device 44, for determining by extrapolation the torrefaction time t.sub.r according to the torrefaction temperature T.sub.g and the final mass yield R sought (on dry basis).

(60) For information, these data were obtained experimentally by means of the torrefaction unit 10 described above, by manually adjusting the torrefaction temperature T.sub.g and the torrefaction time t.sub.r during various tests, and each time measuring the mass yield R obtained at the end of torrefaction.

(61) FIG. 3 relates to experiments conducted with particles having a mean diameter equal to 10 mm, whereas FIG. 4 relates to experiments conducted with particles having a mean diameter equal to 5 mm. In both cases, three torrefaction temperature values T.sub.g were tested, i.e. 300° C., 325° C. and 350° C.

(62) These experiments made it possible to validate the formula used for calculating the reaction rate constant k.sub.t, and thus the characteristic torrefaction reaction time, based on the kinetic model of Di Blasi and Branca as explained above. Indeed, the difference between the theoretical yield provided by this model and that measured experimentally is on average 4% for the 5 mm diameter particles, and 8% on average for the 10 mm particles.

(63) Furthermore, a second thermocouple was used during these experiments. This thermocouple was arranged inside a biomass particle. A comparison between the temperature inside the particle measured using this thermocouple and the temperature of inert gas measured using the thermocouple 42 described above made it possible to determine experimentally the particle heating time. These measurements made it possible to validate the calculation of the characteristic heat transfer time explained above since the difference between this characteristic time and that measured experimentally represents on average 10% of the time measured experimentally for the 5 mm particles and on average 7% of the time measured experimentally for the 10 mm particles.

(64) Furthermore, the graphs in FIGS. 3 and 4 demonstrate that the torrefaction can be conducted rapidly, in less than 9 minutes for both particle sizes tested, with a loss of mass of less than 30%.

(65) Obviously, data based on a larger number of experiments may be recorded in the control device 44 so as to increase the precision of the calculation of the torrefaction time t.sub.r.

(66) Following experiments such as those described above, the torrefaction homogeneity in the biomass particles may be checked using measurements and compared to the results provided by the calculation described above.

(67) FIG. 5 illustrates an analysis device for conducting such measurements.

(68) The principle of this analysis device is based on the penetration depth measurements, in the torrefied beech particles, of a punch subjected to a given stress.

(69) The torrefied beech particles are previously cut so as to enable a measurement on a plane surface, in the core and on the periphery of the sample.

(70) This sample 54′ obtained in this way is kept stable on a metal part 110 by bonding. Firstly, the punch 112 is brought to the surface of the sample; a pre-stress, measured by a force sensor 114, is applied as a reference for the analyses, and a greater stress is applied and the resulting penetration depth of the punch 112 is measured using the comparator.

(71) This measurement is repeated at various points 115 along a median axis 116 of the sample (FIG. 6), i.e. “in the core”, and on the periphery, 2 mm from the edge, so as to compare the effect of the heat treatment on the surface and in the core of the sample. A plurality of measurements may advantageously be made along the same axis to obtain a mean penetration depth value, and avoid heterogeneity of the sample material. These measurements should not be made on end zones 118 of the sample due to edge effects.

(72) As a general rule, a particular application of the torrefaction devices and processes according to the invention to beech particles has been described above, but it is obvious that the invention may be applied to any type of biomass particles.

(73) In particular, the invention may be applied to any types of wood, to any hardwoods, such as beech, but also to softwoods. Indeed, the latter have a lower reactivity than beech, and are thus suitable for obtaining high values of the ratio H representing the homogeneity of the torrefaction reaction more easily.