Bo Leckner 1

COMBUSTION SCALING OF FBC

Bo LecknerDepartment of Energy and Environment, Chalmers University

of Technology, Göteborg, Sweden

Pal Szentannai and Franz WinterTechnical University, Vienna, Austria

59th Technical Meeting, IEA-FBC Czenstachowa 2009

This presentation is taken partly from a chapter by the same authors in M. Lackner Ed. Scale-up in Combustion, Verlag ProcessEng

Engineering GMbH, ISBN: 978-3-902655-04-2

Bo Leckner 2

METHODOLOGIES OF SCALING

Scaling is carried out by

• Formulation and solution of the fundamental equations describingthe processes together with their boundary conditions

• Using these equations and boundaty conditions to formulatedimensionless criteria Z1, Z2 …

Two procesesses I and II are similar when the dimensionless criteriaare equal

1, 1,

2, 2,

I II

I II

Z ZZ Z

=

=

Bo Leckner 3

ON THE SCALING OF FB BOILERS

Scaling from one size to another can be done in manyways and for many purposes:

• Fluid dynamic scaling (from cold to hot conditions)

• Combustion scaling (from lab reactor to boiler)

• Boiler scaling (scale-up of boilers)

Bo Leckner 4

FLUID DYNAMIC SCALINGBased on the dimensionless form of the fluid dynamic

equations, including contituitive relationships:

220 00

0

, , , , , , ,s p fs s

f s

u d u Lu G geometry PSDgL L u

ρ ρρ ϕρ µ µ ρ

20 0

0

, , , , , ,s s

f mf s

u u G geometry PSDgL u u

ρ ϕρ ρ

20 0

0

, , , , ,s

mf s

u u G geometry PSDgL u u

ϕρ

Simplifications

Further simplifications

5 Numbers

4 Numbers

3 Numbers

Bo Leckner 5

FLUID DYNAMIC SCALING—EXAMPLE: 1/9th cold model

The beds

The particles

Pressure drops Pressure fluctuations

Bo Leckner 6

COMBUSTION SCALING I: Simplifications

In combustion scaling most operation parameters are similar in Plant I and Plant II

• TI = TII Temperature• λI = λII Total excess-air ratio• φI = φII Primary-zone stoichiometry• Same fuel• Same bed material• uI ≈ uII Almost same fluidisation velocities

Bo Leckner 7

COMBUSTION SCALING: FLUID DYNAMICS

0

0 0

(1 )( ) 1 ( )s s ts

s s

G u u c Lu u

ρ ε ερ ρ

− −= ≈ − =

With the similar operation criteria in Plant I and II, most of the fluid dynamic criteria are similar. Only the geometry(height L or width D) and Gs/(ρuo) remain.

The latter criterion is similar to the concentration cs at the top of the furnace (L):

ss

Pcg Lρ∆

=

On an average

/II II IIP L P L∆ = ∆which gives

Bo Leckner 8

COMBUSTION SCALING: COMBUSTION

, 0 0 0/( )i v iDa R x c u=

( ) ( ) i,vi

1div div grad DaPei iC C= +U

( ) ( )i i,hPe div div grad Dai iC C= +U

where Da are the first and second Damköhler numbers or in generalDa=transport time/reaction time

Drying, devolatilisaton and char combustion are to be compared with mixing of the fuel in the furnace. The dimensionless species transport equation

gives the dimensionless criteria

2, 0 0/( )i h i iDa R x D c=

Bo Leckner 9

COMBUSTION SCALING: HORIZONTAL

• Transport (mixing) time

• Devolatilisation time

• Char combustion time

21

adev pa dτ =

2

4c p

charc ox

dM ShDcρ

τ =

2 /(2 )disp hx Dτ =

FUEL

FUEL

AIR

X

Bo Leckner 10

COMBUSTION SCALING: VERTICAL

, 0 0 0 , ,/( ) / /i v i I o I II o IIDa R x c u L u L u= => =

since R and c are already equal and xo=L.

L/u is gas residence time (transport time).

Particle residence time is nL/up, where n is the circulationnumber and the particle velocity up=uo-ut.

1/(1 )n η= −

( ),50 P

1

1 /m

pd dη =

+

and the cyclon efficiency

Bo Leckner 11

INFLUENCE OF RESIDENCE TIME

Temperature 850 oC

Oxygen concentration 5%

dp50 =27 µm

L=30 m

Bo Leckner 12

CYCLONE SCALING

The cyclone is influenced by a great number of parameters

• particle properties• solids as a whole (e.g. concentration)• gas properties• cyclone configuration• force field

But with some simplification (low-loaded cyclones) for similargeometries the efficiency can be expressed as

250

( ) (Re, )

18

p

p in

d f Stk

d uStk

D

η

ρµ

=

∆=

Bo Leckner 13

CYCLONE SCALING

Above a certain Re the efficiency scaling is even more simplified

( ) ( )pd f Stkη =

Bo Leckner 14

CYCLONE SCALING: HIGH-LOADED CYCLONES

, ,50 , ,50

1

( ) ( )(1 )

No L p o L p

i iio o

c d c dm

c cη η

=

= − + ∑

According to one of Muschelknautz’ formulations, the cyclone efficiency of a cyclone with inlet loading co is

where co,L is a limiting loading, separating out the ”saltation” part from the ”inner vortex” part. Both parts depend on Stokes number, and for scaling

,( ) ( , / )p o L od f Stk c cη =

Bo Leckner 15

THE IMPACT OF THE CYCLON

2,

2,

I r I cycl I

II r II cycl II

P uA u D

P uA u D

At the same combustion conditions the fuel powers of combustors I an II are

2

2I I

II II

P DP D

=

2 250, 50,

9 9p I in p II in

I II

d u d uD D

ρ ρµ µ

∆ ∆=

250,

250,

p I I

p II II

d Dd D

=

4

50,

50,

p I I

p II II

d Pd P

⎛ ⎞=⎜ ⎟⎜ ⎟

⎝ ⎠

Stokes number scaling

The cyclon efficiency is stronglyproportional to the size of equipment

Bo Leckner 16

COMPARISON BETWEEN TWO PLANTS USED FOR COMBUSTION SCALING

Bo Leckner 17

Bo Leckner 18

EXAMPLE: (Knöbig et al.) COMPARISON OF EMISSIONS FROM LAB RIG AND BOILER

CARBON MONOXIDE NITRIC OXIDE

Bo Leckner 19

EXAMPLE (Alliston and Wu): DESULPHURIZATION

0

0.5

1

1.5

2

2.5

3

3.5

4

0 1 2 3 4 5

DISTANCE FROM FRONT WALL,m

SO2

REL

EASE

D/A

VER

AG

0

1

2

3

4

5

6

7

8

9

O2

CO

NC

ENTR

ATI

ON

, %

Overall O2O ll

Top of primary zone, O2

SO2 Released/Average SO2

In comparisons between desulphurisation with the same coal and limestone in a test rig and boilers, the test rig always showed the best performance!

Bo Leckner 20

CLASSICAL BED SCALING

vdCU DadZ

=

In the Z=z/L direction the species transport equation is

and in dimensional formdcu Rdz

=

For the two-phase bed this can be written

0( ) ( )bb mf b gg b b be b e

dcu u R a k c cdz

ε ε ε− = − −

(1 )(1 ) ( )ee mf e b c gs b b be b e

dcu R a k c cdz

ε ε ε σ ε= − − + −

( )

(1 ) (1 )(1 ) ( )

bb b gg b e

ee e b c gs b e

dC Da NTU C CdZ

dC Da NTU C CdZ

ε β ε

ε β ε ε σ

= − −

− = − − + −

or in dimensionless form

Bo Leckner 21

CLASSICAL BED: DIMENSIONLESS CRITERIA

0 0= / ;mfu u uβ −

/ ;gs gs e oDa K C L u=

/ ;gg gg b oDa K C L u=•Gas-gas Damköhler number

•Gas-solid Damköhler number

•Bed Number of Transfer Units

•Gas-flow partition number0/ ;b b beNTU a k L uε=

A most simplified result

( )gsz L

gs

Da NTUC exp

Da NTU= = −+

Bo Leckner 22

CFB BOILER SCALE-UP

Capacity of CFBCs worldwide (2005)

0

200

400

600

800

1000

1975 1980 1985 1990 1995 2000 2005 2010Year

Cap

acity

[MW

th]

Duisburg IGermany

Robert son Texas-New Mexico

Gardanne France

Poludniowy Koncern Energet yczny (PKE)Poland

Diagram from M. Hupa 2005

Bo Leckner 23

MODULAR SCALE-UP (Foster Wheeler)

Bo Leckner 24

ALLOCATION OF HEAT TRANSFER SURFACE

Size L

1/L

Bo Leckner 25

BOTTOM REGIONS FROM VARIOUS MANUFACTURERS

Bo Leckner 26

Net capacity of thermal power plants in EU-25 by fuel and age 2006

Bo Leckner 27

Comparison between conceptual designs

10 m wide to allow penetration of secondary air, pant-leg in Alstom. Less than 50 m high, and as wide as needed for the power of 2.5-4 MWth/m2 cross-section area

Size

External heat exchangers (No size limitation)

Internal walls+INTREX (10-20 MWth each), 8 consisting of two in series

Bed cooling except walls

68Separators

600/620604/621Steam temperature,oC

270 (header outlet)315 (design pressure)Steam pressure bar

450-600800Size MWe

Alstom [26]Foster Wheeler [25]Item

Bo Leckner 28

CONCLUSIONS• Strict scaling is impossible.

• Fluid dynamical scaling is useful to represent the flowpatterns of hot large-scale boilers in cold small-scaletest-rigs.

• Combustion scaling is possible if reactors with similaroperation data are compared. Often, the horzontaldimension cannot be scaled, but the vertical can.

• The influence of cyclone operation on test results has not been stated in literature

• Boiler scaling utilises building-stones from smaller scalesthat are duplicated to obtain larger scales.

Bo Leckner 29

REFERENCES• L. R. Glicksman, M.R. Hyre, P.A. Farrell, Dynamic similarity in fluidization, Int. J.

Multiphase Flow 20 Suppl., 331-386 (1994). • G. Damköhler, Einflüsse der Strömung, Diffusion und des Wärmeüberganges auf

die Leistung von Reaktionsöfen, Zeitschrift der Elektrochemie 42, 846-862 (1936).• WPM. van Swaaij, Chemical reactors, in Fluidization, Second Edition, Eds. JF

Davidson, R. Clift, D. Harrison, AP, London (1985), pp. 595-629.• S.J. Goidich, O. Sippu, A.C. Bose, Integration of ultra-supercritical OTU and CFB

boiler technologies, Proceedings, 19th International Conference on Fluidized Bed Combustion, Ed. F. Winter, Vienna, May 2006.

• G.N. Stamatelopoulos, J. Seeber, R. Skowyra, Advancement in CFB technology: a combination of excellent environmental performance and high efficiency, Proceedings, 18th International Conference on Fluidized Bed Combustion, Ed. D.W. Geiling, ASME, Paper FBC2005-78081 (2005).

• F. Johnsson, A. Vrager, B. Leckner, Solids flow pattern in the exit region of a CFB-furnace - influence of exit geometry, 15th Int. Conference on Fluidized Bed Combustion, Savannah, 1999.

• B. Leckner, P Szentannai, F. Winter, Scale-up of fluidized-bed combustion in M. Lackner Ed. Scale-up in Combustion, Verlag ProcessEng Engineering GMbH, ISBN: 978-3-902655-04-2

• A.C. Hoffmann, L.E. Stein, Gas Cyclones and Swirl Tubes, Springer, Berlin, 2008.