2002 University of Porto, Faculty of Engineering (FEUP)

Size: px

Start display at page:

Download "2002 University of Porto, Faculty of Engineering (FEUP)"

Sophia Ross
5 years ago
Views:

1 Holberg H, Ahtila P. Drying phenoenon in a fixed bed under the bio fuel ulti stage drying. In: Oliveira A, Afonso C, Riffat S, editors. Proceedings of the st International Conference on Sustainable Energy Technologies, Porto, Portugal; 2 4 June, 22. p. EES University of Porto, Faculty of Engineering (FEUP Reprinted with perission.

2 SET 22 st International Conference on Sustainable Energy Technologies 2-4 June 22, Porto, Portugal Paper EES2. DRYING PHENOMENON IN A FIXED BED UNDER THE BIO FUEL MULTI STAGE DRYING Henrik Holberg and Pekka Ahtila Helsinki University of Technology P.O.Box 44, 25 HUT, Helsinki, Finland henrik.holberg@hut.fi, pekka.ahtila@hut.fi ABSTRACT Drying of solid particles is generally divided into three periods: a short initial heating period, a constant drying rate period, and a falling drying rate perio Drying in a fixed under various drying conditions was experientally studie The ain goals of the easureents were to calculate the heat transfer coefficient based on the easured values, and to evaluate the validity of the constant drying rate perio The exact duration of the constant drying period concerning the entire bed is short. The calculated heat transfer coefficients see to be soewhat reasonable. NOMENCLATURE A surface ( 2 c p specific heat (J/kgK D eff effective diaeter of particle ( h specific enthalpy (J/kg l v vaporisation heat of water (J/kg Le Lewis nuber (- & ass flow rate (kg/s M olar ass (kg/ol p o atospheric pressure (=.3 kpa p v partial pressure of vapour (Pa p' v saturation partial pressure (Pa R gas constant (=8,34 J/olK t,t teperature ( o C, (K u oisture of saple (kg/kg V Volue ( 3 x absolute oisture of air (kg/kg a Greek letters α heat transfer coefficient (W/ 2 K ρ density (kg/ 3 τ tie (s Subscript initial a air a dry air dry ass e evaporation in inlet out outlet s particle surface v vapour w water wb wet bulb INTRODUCTION In Finland bio fuel represents approxiately 9 % of the total priary fuel consued in forest industry. The ost coon bio fuels are bark, forest residues, and different kinds of waste woo The oisture content of bio fuel typically varies between 5 and 63 w-% (water per total ass, considerably reducing the power production of the power plant. Copared to copletely dry fuel the theoretical decrease in power production is 2-2 % for the initial oistures 5-63 w-%, respectively. A ulti stage drying syste consists of several drying stages between which the sae air flow is heate Figure illustrates in a Mollier diagra the change of the air oisture in ulti stage drying having three drying stages copared to one stage drying. The change of air oisture is the sae in both cases but the axiu air teperatures are lower in ulti stage drying than in one stage drying. Because of the lower drying teperatures, it is possible to utilise lower teperature energy sources such as secondary heat flows in the heating of the drying air than in one stage drying. This increases the electricity production of the cobined heat and power plant. H3 D3 H2 D2 H D Moisture (kg/kga Saturation cuve Enthalpy (kj/kga Teperature( C Figure - The change of air oisture in ulti stage drying with three drying stages copared to one stage drying. D = drying period, H = heating perio The broken line represents one stage drying.

3 Drying phases of solid particles are usually divided into three periods: a short initial heating period, a constant drying rate period, and a falling drying rate period /2/. During the constant drying rate period, the particle teperature and oisture distributions are assued unifor, and the heat and ass transfer is in equilibriu at the particle surface: & l = αa t t, ( e v ( a s where t s is the surface teperature. During the constant drying rate period, the surface teperature reains constant, and it is theoretically the sae as the wet bulb teperature of air. The theoretical wet bulb teperature can be derived fro the Equation ( based on the analogy of the heat and ass transfer // : t t wb M = ρc v p po Le RT n lv ln p o po p, p ( t v v wb (2 calculated using a theoretical constant drying rate odel. Wood particles of different sizes were dried in a fixed bed reactor under drying conditions corresponding as near as possible to real process conditions. TEST RIG The flow sheet of the test rig with easureent points is presented in figure 2. The drying reactor consisted of a glass tube and an aluiniu shutter grate through which the drying air was led into the glass tube. The inner diaeter of the glass tube was, height 4 and wall thickness 5. The diaeter of the shutter grate holes was,5. The air ass flow was adjusted to a desired one by a ass flow control unit. If all particles in the fixed bed have reached the local wet bulb teperature, the teperature difference in Equation ( can be replaced as a logarithic ean value teperature: Dry air Daper Vapor Reactor Bed H T a T s (tain t s (taout t s2 Δ tln =, (3 (tain t s In (t t aout where t ain is inlet air teperature, t aout outlet air teperature, t s surface teperature in the botto part of the bed, and t s2 surface teperature in the upper part of the be During the falling drying rate period, the particle surface teperature rises, and the evaporation rate decreases. The drying rate decreases as a result of internal heat and ass transfer resistances inside the particle. Particularly below the fiber saturation point, which is around 3 % (water per dry basis for Finnish wood, the drying rate significantly decreases. To iprove the heating value of the bio fuel with reasonable investent costs, it is not necessary to dry it below fiber saturation point. If the initial oisture of the bio fuel is 6 w-% (water per total ass the relative iproveent of the heating value (MJ/kg copared to initial oisture is 8 % and 2,5 % for the final oistures 25 w-% and w- %, respectively. The drying tie is, however, considerably longer if the final oisture is w-% instead of 25 w-% If the average final oisture of the fuel particles is around 25-3 w-%, one can speculate about the accuracy of the constant drying rate odel used to calculate the evaporation rate fro wood particles. These easureents have two ain goals: to define the value of the external heat transfer coefficient between particles and air based on the easured data, and to evaluate the validity of the constant drying rate odel by coparing real drying results to results s2 Pre heater Moist air H By pass Figure 2 - Flow sheet of test rig with easureent points. H = huidity, T a = air teperature, T s = particle surface teperature, = ss flow control The air teperatures before and after the bed were easured using therocouples. There were four teperature easureent points after the bed to take into account the horizontal teperature gradient of the outlet air due to unequal drying rates in the radial direction. The surface teperature of the particle was deterined by fixing a therocouple on the particle surface with studs. The particles with therocouples were roughly placed in the botto, id, and upper parts of the be The absolute huidity (g/kg a of the air before and after the bed was easured using huidity transitters. INITIAL VALUES Although real fuel particles have no regular shape, it is approxiately possible to convert particles of different shapes to a spherical shape by using an effective diaeter. The effective diaeter is defined as follows/2/: 6V D eff = (4 A where A is the total evaporation surface of the particle and V the volue of the particle. T a

4 Four different effective diaeters 5,, 5, and 2 calculated using the equation (4 were used in the tests. All particles were of regular shape; they were ideal spruce particles specifically ade for the laboratory tests. The bed heights were chosen that the particles coposed a proper bed in a reactor but the outlet air was not yet saturated after the be The real diensions of the particles and the bed heights used in the tests are presented in table. To equalise the oisture distribution of the test particles they were kept in water for at least one week before the tests. The oisture of the particles was around 4-45 w-% before the wetting; after the wetting period, the average oisture varied fro 55 w-% to 65 w-% depending on the particle size and the duration of the wetting perio Sall particles were usually oister than the large ones. Table - The real diensions of the particles and the corresponding bed heights D eff length width thickness bed height 5 2, , Two different inlet air teperatures, 7 and 2 o C were use With inlet teperature 7 o C, the air was ostly dry but a series of tests with particle size was carried out by oisturizing the air to around 8 g/kg a before the drying chaber. With inlet teperature 2 o C the inlet oisture content was around g/kg a. Air velocities used in the tests were,3 /s,,45 /s (only with inlet teperature 7 o C,,6 /s,,9 /s and,2 /s. All velocities were calculated per free sectional area of the grate. As air density becoes saller when the air teperature rises, the ass flows were saller with inlet teperature 2 o C. RESULTS Figure 3 shows how the easured teperatures and oistures change during one hour of drying when particle size is, air velocity is,45 /s, inlet air teperature 7 o C and inlet air oisture g/kg a. It can be seen there is a clear constant drying rate period but its duration is quite short, only a few hundred seconds. The surface teperature in the botto part of the bed rises relatively rapidly quite soon after the beginning of the drying while the surface teperatures in the id and upper part of the bed reain constant longer. The oisture content of the outlet air also decreases quite soon after the beginning of the drying and accordingly the teperature of the outlet air increases. Regardless of the drying conditions and the particle sizes, the changes of the easured values were relatively unifor when the Figure (3 is scaled by dividing the drying tie by the drying tie of each test. Drying conditions and particle size have a considerable effect on the absolute drying tie. For exaple, with an inlet air teperature of 2 o C and air velocity,6 /s, the drying ties to oisture content 3 % (water per dry basis are: 84 s (u o =2,, 284 s (u o =,74, 389 s (u o =,7, 448 s (u o =,38 for particle sizes 5,, 5, and 2, respectively. t (C, x (g/kg tie (s Figure 3 - Changes of the easured teperatures and oisture contents during the drying session. Particle size, air velocity,45 /s and the initial oisture content of the saple,38. teperature of the inlet air, 2 surface teperature in botto part of the bed, 3 teperature of the outlet air, 4 surface teperature in iddle part of the bed, 5 surface teperature in upper part of the bed, 6 oisture content of the outlet air. Heat transfer coefficients were calculated based on the easureent data by applying Equations ( and (3. All teperatures required to define the logarithic ean value teperature are obtained fro the easureent data. The evaporation rate of the bed (kg/s can be calculated as follows: & = & ( x x (5 e a out in The nuber of particles in the bed was anually calculate Because of the regular shape of the particles, it was possible to precisely calculate the external evaporation surface. The teperatures and oisture contents obtained fro the easureent data are the ean values for the period when the entire bed is siultaneously or alost siultaneously in the period of constant drying rate. In soe cases (ainly with axiu air velocities the constant drying period in the botto part was so short that the particles in the upper part had not yet reached the wet bulb teperature which slightly skews the principle of the definition of the heat transfer coefficient. Nevertheless, the heat transfer coefficients have been calculated for each air velocity based on Equations ( and (3. Figure 4 shows an exaple of the period when the whole bed is siultaneously in the period of constant drying.

5 The calculated heat transfer coefficients based on the easureent data are presented in Figures 5 and 6 as a function of air velocity per free sectional area of the grate. t (C, x (g/kg tie (s THE EVALUATION OF THE VALIDITY OF THE CONSTANT DRYING RATE MODEL The validity of the constant drying rate odel was evaluated by coparing the theoretical change of the saple oisture to actual change of the saple oisture. To calculate the theoretical drying rate of the bed, it was divided into equal drying layers according to Figure 7. Evaporation rate is calculated for each layer based on the constant drying rate odel. The evaporation rate of the entire bed is the su of the evaporation rates of the layers. z n tn t n out, x n out t n in = t n- out, x n in = x n- out Figure 4 - Period of constant drying rate. Inlet air teperature 7 o C, inlet air oisture g/kg da, air velocity,45 /s and particle size. surface teperature in the botto, 2 outlet air teperature, 3 surface teperature in the iddle, 4 surface teperature in the upper, 5 absolute oisture content of the air z 2 z t 2 t t out = t2 in, x out = x2 in t in = t, x 2 in = x o heat transfer coefficient (W/2K Figure 5 - Calculated heat transfer coefficients for inlet air teperature 7 o C. Values on the * line are the calculated heat transfer coefficients with oist air. heat transfer coefficient (W/2K velocity (/s.5.5 velocity (/s 5 * Figure 6 - Calculated heat transfer coefficient for inlet air teperature 2 o C. Figure 7 - The principle of the odel for calculating the theoretical evaporation rate of the be The following assuptions were used in the calculations: The evaporation surface is equal in each layer. The surface teperature of the particles in each layer is the sae as the theoretical wet bulb teperature of the inlet air Surface teperatures reain constant during the entire drying period Drying conditions are constant in each layer. They change only between the layers (see Fig. 7 Drying rate does not change in the radial direction of the bed Drying process is adiabatic All particles have an equal initial oisture The heat transfer coefficient is the sae in each layer A short initial heating period is neglected With these assuptions, the evaporation rate in each layer can be calculated using Equation (. As the drying process is assued to be adiabatic, the air teperature after the layer can be calculated by applying the energy balance of the adiabatic oisturising: & ahin & ec pwt s = & a + h (6 where h is an enthalphy of the air per kilogras of the dry air. The enthalphy of the wet air (kj/kg a at the particular teperature ( o C in equation (6 is expressed as follows: out h = c pa t + x(c pv t + 25 (7

6 The reference point in Equation (7 is water at zero degrees. The outlet oisture of the air after the layer can be written as follows: x out & e = xin + (8 & a The air teperature in Equation ( is the ean value of the inlet and outlet air teperature of the layer. When the evaporation rate in Equation (8 is expressed by eans of Equation ( and the Equation (8 is placed in the energy balance of the adiabatic oisturising, where the enthalpy is expressed by the Equation (7, the only unknown ter in the equation (6 is the outlet teperature of the air which can be solve When the outlet teperature of the air after the layer is known the evaporation rate and the outlet oisture of the air can be calculated using Equations ( and (8. The calculated outlet teperature and oisture are then the new inlet values of the air in the next layer and so on. As long as the constant drying rate assuption is valid in the entire bed the theoretical outlet oisture of the saple can be calculated as follows: u out = u Δτ n i= αa ei ( t ai t si (9 where u o is the initial oisture of the saple, Δτ the duration of the constant drying rate period, and n the nuber of layers. Since the oisture in the first drying layer has decreased to zero there are nine drying layers left and the inlet air teperature into the second drying layer is the sae as in the first layer, and so on. It is, of course, not correct to apply the constant drying rate odel particularly below the fiber saturation point. Because the ai of this study is to evaluate the accuracy of the constant drying odel, the decrease in the evaporation rate is not taken into account in the calculations. The theoretical evaporation rate of the bed erely decreases when the nuber of the dry layers increases in the be The actual outlet oisture of the saple based on the easureent data was calculated as follows: u out = u Δ& τ a j= ( Δx j + Δx 2 j ( where Δτ is the tie step, Δx the difference between outlet, and inlet air oisture and the nuber of scans. The wet bulb teperature of the air was calculated using the Equation (2. The densities of dry air and vapour in Equation (2 were calculated by applying the equation of ideal gas. The pressure of the saturated vapour as a function of teperature can be approxiately calculated using the correlation//: 78, (t-99,64 p v ' = p exp( ( t + 23 where t is the wet bulb teperature in Celsius Degrees. The diffusion coefficient between water and air needed to define the Lewis nuber depended on the teperature according to proportion D T 3/2 /4/. The value of the exponent n in Equation (2 was,4.the bed was divided into ten layers in all cases by assuing the evaporation surface is equal in each layer. Figures 8 and 9 show how air velocity and particle size affect the validity of the constant drying rate odel Theoretical saple oisture (kg/kg.3 /s.6 /s.9/s Figure 8 - The effect of air velocity on the validity of the constant drying rate odel for particle size 5 and inlet air teperature 2 o C. Moisture content in x- axis is the theoretical oisture of the saple calculated by the odel and oisture content in y-axis is the actual oisture of the saple deterined by the easureent Theoretical saple oisture (kg/kg Figure 9 - The effect of particle size on the validity of the constant drying rate odel with air velocity,3 /s and inlet teperature 2 o C. Moisture content in x-axis is the theoretical oisture of the saple calculated by the odel and oisture content in y-axis is the actual oisture of the saple deterined by the easureent. We can see in Figure 8 that air velocity has soe effect on the validity of the constant drying rate odel but it is quite sall. Figure presents how valid the constant drying rate odel is for all particle sizes and teperatures used in the tests when the effect of the air velocity is ignore Figure shows a ean value of the curves in Figure. Regardless of the particle size and

7 drying conditions we can approxiately deterine the accuracy of the constant drying rate odel as a function of the saple oisture in Figure Theoretical saple oisture (kg/kg 5(7C 5(2C (7C (2C *(7 5(7C 5(2C 2(7C 2(2C Figure - The effect of particle size and teperature on the validity of the constant drying odel Theoretical saple oisture (kg/kg Figure - The average accuracy of the constant drying odel. CONCLUSIONS Drying of wood particles was experientally studied in a fixed bed under various drying conditions. Heat transfer coefficients were calculated based on the easured values. According to the results, the heat transfer coefficients see to behave relatively logically as a function of velocity and particle size when the factors causing deviations are observed in the results. Factors ainly causing deviations are: the easureent of surface teperatures, the easureent of the outlet air teperature and the variation of effective evaporation surface. The deviations of the easured values and the calculated heat transfer coefficients were estiated by conducting five tests for the sae initial values. Particle size was, air velocity,45 /s, inlet air teperature 7 o C and inlet air oisture g/kg a. The variations of the easured values were the following: surface teperature in the botto 26,3-28,2 o C, surface teperature in the upper 2,6-22, o C, outlet air teperature 27,2-29,5 o C and evaporation rate,384 -,4 g/ 2 s. The variation of the calculated heat transfer coefficients is 48, W/ 2 K. The ean value is 5, W/ 2 K. It is iportant to reeber the easured teperatures and oistures are not necessarily exact values. The values of the heat transfer coefficients see to be higher when the inlet air teperature is 7 o C. This at least partly results fro the saller ass flow the air had when the inlet air teperature was 2 o C. Siilarly, The heat conduction fro the reactor little iproves the absolute values of the heat transfer coefficients. The heat conduction was not wanted to eliinate as there is also heat conduction in a real application. According to easureent data, the exact duration when the entire bed is in the period of constant drying rate is relatively short. On the basis of the easured surface teperatures, the particles in the botto are quite soon in the period of the falling drying rate while the constant drying rate period is valid longer in the iddle and upper part of the bed (see Fig. 3. Air velocity and particle size have soe influence on the validity of the constant drying rate odel. When particle size and air velocity are sall (in these easureents 5 and under,6 /s the constant drying odel can be used with reasonable accuracy to average final oisture,6 -,7 kg/kg. When particle size and air velocity increase the accuracy of the odel is slightly worse. Inlet air teperature under these drying conditions has no discernible effect on the validity of the odel. It is reasonable to conclude that saple oisture is the ost influential factor affecting the validity of constant drying rate odel. Since the oisture content has decreased below,8 -,9 kg/kg (see. Fig. the accuracy of the odel deteriorates in ost cases regardless of the drying conditions and particle size. It is, however, possible to copare different drying conditions by using the constant drying rate odel because the error copared to actual oisture content sees to be in the order of the sae in all cas ACKNOWLEDGMENTS The study has been financed by the Technology Developent Centre of Finland (TeKes, Vapo, Stora Enso, and Pohjolan Voia. The authors gratefully acknowledge those who have contributed to this work. REFERENCES [] Lapinen M.J (996. Cheical Therodynaics in Energy Engineering ( in Finnish. Publications of laboratory of applied therodynaics. Finlan [2] Saastaoinen J. J, Ipola R.K (994. Drying of solid particles in hot gases. Proceedings of the 9th International Drying Sypnosiu. Australia. [3] Spets J-P. (2. A new Multi Stage Drying Syste. st Nordic Drying Conference. Norway. [4] Ryti Henrik. (98. Hand Book of Technical Science (in Finnish. Finlan [5] Mujudar A, S. (995. Hand Book of industial drying. Canada. [6] The www-pages of Finnish Forest Industries.

I. Concepts and Definitions. I. Concepts and Definitions

I. Concepts and Definitions. I. Concepts and Definitions F. Properties of a syste (we use the to calculate changes in energy) 1. A property is a characteristic of a syste that can be given a nuerical value without considering the history of the syste. Exaples