Influence of slag properties and operating conditions on slag flow in a coal gasifier

Transcription

1 8 th International Freiberg Conference on IGCC & XtL Technologies Innovative Coal Value Chains 12-1 June 21, Cologne, Germany Influence of slag properties and operating conditions on slag flow in a coal gasifier Insoo Ye 1, Changkook Ryu 1, Bongkeun Kim 2 1 School of Mechanical Engineering, Sungkyunkwan University 2 Energy Conversion System Research Team, Doosan Heavy Industries & Construction Co., LTD.

2 Coal gasification technology and process 2 Coal gasification technology l Conversion of coal into syngas (CO, H 2 ) l Used for power, fuel and chemical production Entrained bed coal gasifiers l Typical operating conditions Coal size (μm) ~5 Particle residence time (sec) 5 1 Carbon conversion (%) ~ 99 Wall condition Slagging Slag layer formation on the gasifier wall l Deposition of molten ash onto the wall l Protection of the wall from chemical, physical and thermal damages l Ash discharge by slag flow on the surface Entrained-bed gasifiers (up) and slag layer on the wall (down)

3 Slag layer formed on the gasifier wall 3 Structure of the slag layer l Solid slag layer facing the cold wall l Liquid slag flowing downward on the surface Deposition Gasifier (T > 15 C) l T cv at viscosity of 25 Pa.s considered as the interface temperature of the two layers Molten ash Key parameters influencing the slag layer l Ash content of coal l Ash composition (SiO 2, Al 2 O 3, CaO, etc.) l Gas temperature l Reactor shape and flow pattern, etc. Understanding the slag layer behavior l Difficult to directly measure/monitor during gasifier operation Pressure vessel Water/steam tube Refractory Solid slag Liquid slag Ash Slag formation on the wall in a gasifier Coal l Numerical models required to understand its behavior

4 Existing models for slag flow 4 Seggiani (1998) model l Assumption: linear temperature profile in the slag layer l Slag viscosity: μ( T ) º μ( r) = μ( TS ) exp( - ar δ L ) l Adopted in Li et al. (29): Vertical wall Kittel et al. (29): Siemens gasifier Ni et al. (21) : Lower part of pilot scale PC gasifier Yong et al. Seggiani Yong et al. (212) model l Assumption: cubic temperature profile in the liquid slag layer l Slag viscosity: constant across the liquid slag layer l Adopted in Pressure vessel Water/steam tube Refractory Solid slag Liquid slag r Chen et al. (212) : Vertically-oriented oxy-coal combustor T Steam T R T CV Temperature profile assumed within the slag layer T S

5 Objective and methods 5 Research objective l To propose and validate a new slag flow model l To evaluate the influence of key design/operating parameters Methods l Target gasifier: PRENFLO gasifier (Entrained bed, Spain) l Part I: model development and evaluation In comparison to existing models (Seggiani, Yong et al.) l Part II: parametric analysis Gas temperature, ash deposition, slag properties Reactor geometry (bottom cone angle) Tcv l Part III: model expansion to a transient system PRENFLO Gasifier

6 Approach of the new numerical model T C Model for the liquid slag layer: direct solution of the governing equations l Solution along the direction (i) perpendicular to the wall, and marching downward (j) No assumptions required for temperature profile or slag viscosity. l Control volumes inherit the same amount of mass from above (m i,j = m i,j+1 ) l New ash deposition: a new control volume added on the surface l Programing with Excel Visual Basic for Application (VBA) Governing equations (steady states) Mass Momentum Energy T tube T R CoolantRefrac- tory tube Q steam = Q R = Q solid m = m + m out out in in deposit M = M + M + M + M out in deposit phase viscous conduction gravity H = H + ΔH + Q + H + Q T CV r r1 r 1 r i-1 r i Solid slag T 1,v 1 T i,v i Q cond r i deposit r I-1 GL T J,v J - Critical viscosity temperature - Interface temperature between solid and liquid slag layer (μ C = 25 Pa s) r I T surf m dep,h dep Q gas Δy j j Solid slag layer T cv i Liquid slag layer Deposit Deposit Deposit

7 Slag thickness and properties Slag thickness l Liquid slag thickness : summation of each cell width l Solid slag thickness : d = r from = Q Q, Q S, j = A R, j S, j R, j Slag properties l Heat capacity, Cp [kj/kg.k] l Density, ρ [kg/m 3 ] k r R, j ln( rr, j / r, l Thermal conductivity, k [W/mK] T cv -T S, j r R, j -, j j ) and å d S, j = I å i= 1 Dr C P, = XC liq P C = ( 2), å + bt - c P glass X a 1 T r = Q cond,, j S, j = C, j Q - C, j = U RC, j AC, j ( TR, j TC, j (FeO% wt. + Fe2O3% wt. + MnO% wt. ) ) i, j steam tube refractory Solid slag Liquid slag ri,j r,j r R,i r tube,j 7 Thermal diffusivity of slag, α [m 2 /s] = l Emissivity, ɛ =.83 (constant) k = a r slag C P l Viscosity, μ [Pa.s] μ = a T exp( 1 b / T )

8 Geometry and Input conditions: Reference case 8 Gasifier: PRENFLO gasifier Input and boundary conditions l Main slag composition (%wt.) Al 2 O 3 CaO Fe 2 O 3 K 2 O SiO 2 SiO 3 Misc l Tcv : 1548 K (at 25 Pa s) l Uniform ash deposition (m dep ) along the height l Boundary condition T gas (18 K) T dep (175 K) m dep (4.5 kg/s) Refractory Tube wall Water/steam Control volume k = 8 W/m K k = 43 W/m K T = 523 K Tube wall Water/steam Tube wall Refractory l sections (2 cells in each part).3 1

9 Results: Reference case 9 Liquid slag layer l Slag thickness (δ): increase in the in the downward direction At the bottom cone: rapid increase due to inclined wall (angle: 78 ) l Temperature: linear increase to the surface from Tcv on the wall (x=) l Velocity: fast flow near the surface due to low slag viscosity Heat flux (kw/m 2 ) Height (m) 8 4 x [Temperature, K] Top cone Main body Height (m) 8 Height (m) [viscosity, Pa s] Heat flux (q out ) Solid slag thickness (δ S ) Height (m) 8 4 [velocity, m/s] Bottom cone Liquid Slag Thickness (mm) Liquid Solid Slag slag Thickness thickness (mm) (mm) Liquid Slag Thickness (mm)

10 Part I: Model Comparison and Validation 1 Comparison with existing models l Seggiani: Good agreement l Yong et al.: Slag thickness under-estimated by 1% (due to constant slag viscosity) Surface temperature (K) Constant slag viscosity condition l Good agreement with Yong et al s model confirms that the under-estimation of slag thickness is by the assumption of constant viscosity Surface temperature (K) T surf T surf Height (m) 4 d L Height (m) 4 d L μ=7.12 Pa s 2 μ=7.12 Pa s This study Yong et al's model 2 This study Yong et al's model Seggiani's model Liquid slag thickness (mm) Liquid slag thickness and surface temperature Liquid slag thickness (mm)

11 Part I: Model Comparison and Validation (cont d) 11 Case of low T gas at the bottom cone (T gas =1518 K or Tcv-3K) l Liquid slag thickness (δ L ): three models give similar values l Solid slag thickness (δ S ): Seggiani s model over-estimates due to the linear temperature profile in the liquid slag (zero heat flux to the solid slag = infinite value of δ S ) This study Yong et al's model Seggiani's model 8 This study Yong et al's model Seggiani's model Temperature, K Height (m) Position in the liquid slag, r (mm) Temperature profile in the liquid layer at slag tap Solid slag thickness (mm) Solid slag layer thickness

12 Part II : Parametric analysis 12 ±1% changes from reference values Gas temperature l Largest influence on the slag flow and heat transfer Radiation (~T 4 ) dominant at high temperature Ash deposition rate l Influence on the slag behavior is not large (~ 3%) Slag properties: viscosity, thermal conductivity, emissivity l Very small influences on δ (< %) Varied Parameters δ L (%) δ S (%) Q LS (%) Gas temperature +1% (ref. 18 K) Ash deposition rate (ref. 4.5 kg/s) Slag viscosity (ref. f(t) Pa.s) Slag conductivity (ref W/mK) Slag emissivity (ref..83) 1% % % % % % % % % δ L δ S Q LS Liquid slag thickness at the slag tap Solid slag thickness at the slag tap Heat transfer rate to the solid layer from liquid slag

13 Part II : Parametric analysis Bottom cone angle 13 Influence of bo om cone angle: 78 l Surface velocity:.75 m/s.12 m/s By increase in the gravity force l Liquid slag thickness: 17 mm 13 mm l Solid slag thickness: 9 mm 53 mm l Thermal conduction rate to the solid layer strengthened and affected its thickness Velocity (m/s) Steeper angle Reference (12 (78 ) o ) o 24 o 3 o Position in the liquid slag, r (mm)

14 Part II : Parametric analysis - Tcv 14 Relative changes in δ assessed for Tcv under various conditions l Exponential increase in δ at high values of Tcv l High Tcv means Small temp. difference between the liquid slag surface and Tcv (Low heat flux to the solid slag Thick solid slag) l Little contribution of liquid slag thickness Relative magnitude of slag layer thickness (%) Influence of Tcv on the slag layer thickness At the slag tap At the end of main body Bottom cone angle = o T gas = 175 K T gas = 185 K Hollow symbols: Shifts in Tcv Solid symbols: Different ash:flux ratios Tcv (K)

15 Part III : Model expansion to a transient system 15 Example results: Instant change in gas temperature from 18 K 198 K l It requires more than 12 sec (2 min) to reach the new steady-state l For the first 1 seconds, slag thickness at the slag tap increases by faster flow of liquid slag l Then solid slag turns into liquid slag and flows down [Temperature, K] Initial condition (T gas : 18 K) sec Slag layer thickness (mm) At the slag tap [velocity, m/s] Time constant for 3.2% change Liquid slag layer Solid slag layer sec δ at slag tap 449 heat transfer to the wall Time (sec)

16 Conclusions 1 A new numerical model developed for slag flow and heat transfer l Evaluated by comparison with existing models l No assumptions required for gas temperature profile or slag viscosity Influence of key parameters on slag thickness and heat transfer to the wall l Gas temperature has a dominant influence on the slag flow and heat transfer l Steep bottom cone decreases the slag thickness l Ash deposition and slag properties are not very influential l Exponential increase of solid slag thickness with increase in Tcv Model expansion to a transient system l Slag thickness and heat transfer rate on the wall adjust very slowly (~ several minutes) to the changes in the operating conditions