Beruflich Dokumente
Kultur Dokumente
h i g h l i g h t s
< High heat-exchange rate in uidised bed is appealing for coupling a Stirling engine.
< A mathematical model of a coupled system newly developed.
< The integrated system is of interest for micro-scale co-generation (1 kW) from renewable fuels.
< A PID control strategy largely enhances the system performances.
a r t i c l e i n f o a b s t r a c t
Article history: The present article deals with the integration of a uidized bed combustor and a Stirling engine for
Received 4 May 2012 cogeneration purposes. An experimental study was carried out, proving the ability of the bed to exchange
Accepted 3 November 2012 heat at high rate with an immersed coil that realistically emulates the heat exchanger of a small Stirling
Available online 12 November 2012
engine. The heat transfer coefcient attains values up to 280 W m2 K1. No dirtying of the immersed
surface occurred during a combustion test of biomass.
Keywords:
The paper also reports on a newly developed mathematical model of a uidized bed combustor
Fluidized bed
coupled with a Stirling engine for co-generation purposes. It consists of four fundamental blocks
Stirling engine
Biomass
describing i) the heat transfer, ii) the uidized bed combustion, iii) the heat recovery, and iv) the Stirling
Combustion engine. The model produces as relevant outputs the bed temperature, the mechanical power and the
Heat transfer efciency of the Stirling engine, at changing the operating conditions and geometrical parameters of the
system. A slow dynamic response is predicted, that it is signicantly improved by adopting an efcient
control strategy. On the whole, the model results indicate that the proposed integrated system is of
interest for micro-scale cogeneration from biomass fuels.
2012 Elsevier Ltd. All rights reserved.
1. Introduction normally in the range 100e200 kWel [1] and the thermal efciency
is lower than 20% [2]. Furthermore, capital costs are high at small
The biomass fuels have intrinsic CO2 neutral impact, indepen- scale applications. The cogeneration can be also accomplished by
dently of the way their chemical energy is transformed into heat or coupling a gasication/pyrolysis step with an internal combustion
power via thermal or biological processes. In contrast, techno- engine or fuel cell, but this option results more complex due to the
economic constraints limit the viability of biomass utilization for required purication steps of the syngas [3]. Recently, an increasing
power generation. attention is paid to the application of micro-scale cogeneration
From a technical point of view, although the combustion of from different fuels, thanks to the diffusion of even more dense
biomass fuels is relatively easy to be carried out in both xed and grids for electricity distribution and picking up [4]. In this scenario,
uidized bed devices, the process implementation at small scale is the development of reliable and cheap methods for cogeneration
not straightforward and efcient for power generation. For from biomass fuels in a range 1e100 kWel is appealing, represent-
instance, the minimal size of organic Rankine cycle (ORC) is ing a still open challenge.
The external combustion Stirling engine [5] has been conceived
and developed since the 19th century, although having never
reached a high level of penetration into the energy and transport
Abbreviations: PID, proportional-integrative-derivative control; FB, uidized
bed; SE, Stirling engine. sector. Main advantages are the smoothness, the exibility toward
E-mail address: miccio@irc.cnr.it. the external heat source and the high thermodynamic efciency.
1359-4311/$ e see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.applthermaleng.2012.11.004
F. Miccio / Applied Thermal Engineering 52 (2013) 46e53 47
2.1. Equipment
Ta, Tb, c, U, e, Q
FB heat
transfer
Fig. 4. Ideal blocks conceived for the mathematical model of the FB-SE integrated
system.
Fig. 3. Heat transfer coefcient versus the uidization velocity for ne and coarse bed
materials (P 4 bar abs.).
" #
0:5
dp Umf rg d3p rg rs rg g
2
33:7 0:0408$ 2
33:7 (5)
m m
dTb
Mb cpb AU ra cpa Ta Tb Sl Kt T0 Tb hSSE TSE Tb
Fig. 5. Heat transfer coefcient in uidized bed as function of the particle size and dt
temperature (computed values). Ff hc DHc
(6)
[18]. Accordingly, an overall combustion efciency hc in the bed is
where Mb is mass of the bed and A is the cross section. The terms at
introduced taking into account possible losses due to the fast
the right side, listed in the order as they appear, are: the enthalpy of
release of volatile matters and imperfect mixing between
the uidizing gas stream AUra, the heat dispersed by conduction at
combustible species [19]. The efciency hc is reasonably dependent
the lateral wall Sl, the heat transferred to the Stirling engine
on the bed temperature Tb, the uidization velocity ratio U/Umf and
throughout the surface SSE, and the power of combustion
the excess air ratio e. Eq. (4) provides the relationship adopted by
Q FfhcDHc.
the model among the cited variables: an asymptotic dependence of
hc on Tb and e, accounting for the increased reaction rate, and
3.3. Heat recovery
a linear dependence on U/Umf, accounting for better mixing. The
biomass properties, namely nature, moisture, composition, particle
The heat recovery block consists of a heat exchanger down-
size, are accounted for by the given parameters p1, p2 and p3 of
stream the combustion chamber, which makes possible the
Eq. (4).
combustion air pre-heating by heat transfer with the hot ue gases.
!1 For instance, we can assume that nned tubes are transversally
U T e
hc 1 p1 Tb exp b exp (4) inserted in the zone of the column above the bed (freeboard),
Umf p2 p3 forming an air-to-air cross-ow exchanger. In this region the
residual concentration of entrained bed solids would prevent the
formation of deposits on the external surface of the exchanger [12].
The simplied equation of heat transfer (Eq. (7)) is derived
assuming that the heat exchanger has very fast dynamics with
respect to the uidised bed, leading to absence of accumulation
terms.
Tb Ta Tex T0
Fa cpa Ta T0 Fg cpg Tb Tex Y (7)
2
where T0 and Ta are the temperatures of the air before and after
heating, and Tex is the temperature of ue gases after the exchanger.
The average temperature difference (TbTa TexT0)/2 is the
driving force, Y is a global coefcient of the heat exchanger, Fa and Ff
the mass ow rates of the air and ue gas.
in order to produce engine data-sheets as those reported in Fig. 7, Combustor size m 0.3
showing the dependence of the mechanical power WSE (A) and Bed height m 0.4
efciency hSE (B) on the angular speed u for different upper Bed particle size m 0.3
Bed voidage (at minimum uidization) e 0.5
temperatures. By interpolation of computed data, the 3rd order Particle density kg/m3 2600
tting equations (Eqs. (8) and (9)) are obtained. Particle specic heat J/kg/K 800
Air ow rate kg/h 50
h i Fuel heating value MJ/kg 17,000
WSE Wn b33 1 n3 Tu3 b22 1 n2 Tu2 b1 1 n1 Tu Air for combustion kg/kg 6.24
Fuel parameter p1 K1 20
(8) Fuel parameter p2 K 90
Fuel parameter p3 e 0.8
h Coefcient of the heat exchanger W/K 7.5
hSE 100 q33 1 d3 Tu3 q22 1 d2 Tu2 q1 1 d1 Tu Heat transfer coefcient (gas) W/m2/K 40
i Number of SE modules e 4
SE displacement L 0.388
d0 T 9
SE surface (1 module) m2 0.18
Heat capacity of SE J/K 5000
It is also assumed that the dynamics of the Stirling engine is very Proportional constant kg/s2/K 0.1
fast with respect to that of the uidized bed, thus its working Integrative constant kg/s/K 0.0002
conditions (h and W) can be computed on the basis of the current Derivative constant kg/s3/K 5.0
values of the bed.
The upper temperature TSE is calculated via the differential
energy balance (Eq. (10)) for the SE block: 3.5. Automated control
dTSE W A control block is taken into account for assuring stable working
CSE hSSE TSE Tb SE (10)
dt hSE conditions or improving the dynamic response of the system. It is
assumed that in presence of a temperature gap, the fuel ow rate is
where CSE is the heat capacity of the engine and SSE the exposed adjusted by acting on the nal control element (e.g. the frequency
surface. of a rotating screw feeder). The controller strategy is single-loop
feedback PID [22], where the controlled variable is the bed
temperature and the manipulated variable is the fuel ow rate. The
controlling equation (Eq. (11)) links together the bed temperature
Tb, the set-point temperature Tset and the fuel ow rate Ff.
Zt
dFf dTset Tb
G1 Tset Tb G2 G3 Tset Tb dw
dt dt
0
(11)
Fig. 8. Time proles of the bed temperature (A) and fuel ow rate (B) without any
Fig. 7. Mechanical power (A) and efciency (B) of a Stirling engine as function of the Stirling engine connected to the uidized bed: comparison between control and no-
angular speed and upper temperature. control operation.
F. Miccio / Applied Thermal Engineering 52 (2013) 46e53 51
60
50
eff. SE
40
efficiency, %
eff.t
30
20
10
0
0 10 20 30 40 50 60 70
speed, Hz
Fig. 9. Time proles of the bed and SE temperatures under control operation (NSE 4,
time of engine start 2000 s). Fig. 11. SE efciency and total efciency versus the engine speed (control operation).
50
Tb - TSE,C
40 6
30
20 4
10
0 2
0 10 20 30 40 50 60 70
speed, Hz
Fig. 10. Temperature difference between bed and Stirling engine and mechanical Fig. 12. Time required for achieving steady working regime after engine start-up
power versus the engine speed (control operation). versus SE number and engine speed (control operation).
52 F. Miccio / Applied Thermal Engineering 52 (2013) 46e53
assumed for setting steady working conditions. A rather long time ni parameters for SE power, K1
in the order of 1000 s is required for achieving steady conditions, p1 parameter of combustion efciency, K1
the thermal inertia of the uidized bed being the limiting factor for p2 parameter of combustion efciency, K
rapid changes. It also appears that the system dynamics is much p3 parameter of combustion efciency, e
slower when the number of Stirling engines or the angular speed qi parameters for SE efciency, s
increase, corresponding to conditions of higher thermal power to P pressure, bar
be transferred from the uidized bed to the engines. Q thermal power, W
S surface, m2
4. Conclusions t time, s
T temperature, K
Experiments were carried out for measuring the heat transfer U uidization velocity, m s1
coefcient in uidized bed under conditions similar to those of FB-SE Umf minimum uidization velocity, m s1
integrated system. The heat transfer coefcient attains a maximum Y global coefcient of the heat exchanger, W K1
value of around 280 W m2 K1. This value is in agreement with the W mechanical power, W
theoretical estimates and results much more high than heat transfer
coefcients in convective exchangers, in the typical range 20e Greek symbols
50 W m2 K1. The obtained values can be reasonably extrapolated d bubble fraction, e
to a real FB-SE system. No deposits were formed on the tube surface DHc combustion enthalpy, J kg1
by fouling or condensation after the combustion of a biomass, hc combustion efciency, e
proving the reliability of the proposed solution. hSE mechanical efciency of the engine, %
An integrated model of uidized bed combustion coupled with ht total efciency, %
a Stirling engine was developed. The model provides as relevant q time, s
outputs the bed temperature, the mechanical power and the ef- m viscosity, Pa s
ciency of the Stirling engine, at changing the operating conditions r density, kg s1
and geometrical parameters. u angular speed, s1
Realistic values of the mechanical power generated by the
engine are in the range 1.0e5.0 kW for a 20 kWth uidized bed Subscripts
combustor. These results indicate that the proposed integrated a air
system is of interest for micro-scale co-generation from renewable b bed
fuels in connection with smart grids. However the slow dynamics of ex exchanger
the system addresses toward applications characterized by rather f fuel
steady request of mechanical/electrical power. g gas
The automated control at changing the operation mode of the l lateral side
system was taken into account by choosing as manipulated variable n nominal value
the fuel ow rate. On the basis of the adopted variables and p particle
parameters, a prompt response is predicted when the load of the r radiative
system is suddenly changed. set set point
The results reported in this paper open perspectives for the w wall
development and operation of an experimental prototype for 0 initial value
micro-generation to be fed with biomass fuels. This option is
greatly appealing for appliances in rural and developing areas. In References
this concern, an experimental prototype of the integrated system is
under construction at CNR for proving the proposed option and [1] F. Vleza, J.J. Segoviab, M.C. Martnb, G. Antolna, F. Chejnec, A. Quijanoa,
A technical, economical and market review of organic Rankine cycles for the
validating the model.
conversion of low-grade heat for power generation, Renewable and Sustain-
able Energy Reviews 16 (2012) 4175e4189.
Nomenclature [2] A. Duvia, A. Guercio, C. Rossi, Technical and economic aspects of Biomass
fulled CHP plants based on ORC turbogenerators feeding existing district
heating networks, in: Proc. 17th European Biomass Conference, Hamburg
A section, m2 Germany, 2009, pp. 2030e2037.
bi parameters for SE power, s [3] P. McKendry, Energy production from biomass (Part 3): gasication technol-
cp specic heat, J kg K1 ogies, Bioresource Technology 83 (2002) 55e63.
[4] P. Jarventausta, S. Repo, A. Rautiainen, J. Partanen, Smart grid power system
C heat capacity, J K1 control in distributed generation environment, Annual Reviews in Control 34
di parameters for SE efciency, K1 (2010) 277e286.
dp particle size, m [5] F. Walker, Stirling Engines, Clarendon press, Oxford, UK, 1980.
[6] M. Kuosa, J. Kaikko, L. Koskelainen, The impact of heat exchanger fouling on
Dt tube diameter, m the optimum operation and maintenance of the Stirling engine, Applied
e excess air ratio, e Thermal Engineering 27 (2007) 1671e1676.
F mass ow rate, kg s1 [7] T. Li, D. Tang, Z. Li, J. Du, T. Zhou, Y. Jia, Development and test of a Stirling
gravity acceleration, m s2
engine driven by waste gases for the micro-CHP system, Applied Thermal
g Engineering 33e34 (2012) 119e123.
G1 proportional control constant, kg s2 K1 [8] P. Basu, Combustion and Gasication in Fluidized Beds, CRC Press, Taylor &
G2 derivative control constant, kg s1 K1 Francis, Boca Raton, USA, 2006.
G1 integral control constant, kg s3 K1 [9] J. Werther, M. Snger, E.U. Hartg, T. Ogada, Z. Siagi, Combustion of agricultural
residues, Progress in Energy and Combustion Science 26 (2000) 1e27.
h heat transfer coefcient, W m2 K1 [10] A. Cammarota, R. Chirone, M. Miccio, R. Solimene, M. Urciuolo, Dual-fuel
kew average thermal conductivity, W m1 K1 uidized bed combustor prototype for residential heating: steady-state and
Kt conduction coefcient of the insulation layer, W m2 K1 dynamic behavior, in: Proc. 20th Int. Conference on Fluidized Bed Combus-
tion, Xian (CN), May 17e20, 2009, pp. 441e447.
Lt tube length, m [11] F. Miccio, G. Ruoppolo, M. Urciolo, Valorization of a difcult biogenic residue
M mass, kg (olive husks) as energy source in uidized bed, Accepted for presentation at
F. Miccio / Applied Thermal Engineering 52 (2013) 46e53 53
Venice 2012 4th Int. Symposium on Energy from Biomass and Waste, session [17] D. Kunii, O. Levenspiel, Heat Transfer between Fluidized Bed and Surfaces, Ch.
C6, November 2012. 13th, In: Fludization Engineering, Butterworth-Heinemann, Boston, 1991.
[12] D.G. Klaren, E.F. Boer de, Self-cleaning uidised bed heat exchangers for [18] F. Scala, P. Salatino, Modelling uidized bed combustion of high-volatile solid
severely fouling liquids and their impact on process design, in: Dr. Jovan fuels, Chemical Engineering Science 57 (2002) 1175e1196.
Mitrovic (Ed.), Heat Exchangers e Basics Design Applications (2012), ISBN [19] R. Chirone, F. Miccio, F. Scala, On the relevance of axial and transversal fuel
978-953-51-0278-6. segregation during the FB combustion of a biomass, Energy and Fuels 18
[13] M. Gaderer, G. Gallmetzer, H. Spliethoff, Biomass red hot air gas turbine with (2004) 1108e1117.
uidized bed combustion, Applied Thermal Engineering 30 (2010) 1594e1600. [20] C.Y. Wen, Y.H. Yu, Mechanics of uidization, Chemical Engineering Progress
[14] F. Di Natale, A. Lancia, R. Nigro, Surface-to-bed heat transfer in uidised beds Symposium 62 (1966) 100e111.
of ne particles, Powder Technology 195 (2009) 135e142. [21] F. Normani, Stirling Engine Design Program (2009) (accessed on 02.05.12),
[15] R.H. Perry, D.W. Green, J.O. Maloney, Perrys Chemical Engineers Handbook, newenergydirection.com/blog/2009/03/stirling-engine-design-program.
seventh ed., Mc Graw-Hill, New York, 1997 (Chapter 27). [22] G. Stephanopoulos, Chemical Process Control: An Introduction to Theory
[16] J. Zhu, R. Grace, C.J. Lim, Tube wear in gas uidised beds e I. Experimental and Practice, Prentice-Hall International, Englewood Cliffs/New Jersey USA,
ndings, Chemical Engineering Science 45 (1990) 1003e1115. 1984.