coal to biomass (wood pellet) mill conversion to biomass (wood pellet) mill conversion presented by...
TRANSCRIPT
Coal to biomass (wood pellet) mill conversion
presented by
Corniel Zwaan
Coal Milling Projects
STAR Global Conference, Berlin 2017
Company background
• CMP is a South African company and was formed in December 1997. Celebrating 20 years
this year.
• The company consists of three departments namely:
• Design and CAD department.
• Manufacturing and refurbishing department.
• Maintenance department.
• License agreements:
• Southwestern Corporation – USA
• Merrick Feeders – USA
• Synergy – Australia
• Patented technologies:
• Version 1 to 6 rotating throats (vane wheels, port rings or nozzle rings).
• Ultra high performance static classifiers (UHPSC).
• Static modified air re-entrainment technology (SMART) for Low NOx purposes.
• High ash coal de-sanding technology to remove abrasive materials from coal.
• Coal to biomass mill conversion classifier.
Company background
• CMP’s design department started to use in 2008 as CAE software to do
computational fluid dynamic analysis (CFD) and discreet element modelling (DEM).
• Since May 2014, CMP have a full time design engineer that uses to do
CFD and DEM in house.
• 9 years using
Company background
Presentation content
• Project background
• Project outcomes
• CFD setup
▪ Geometry
▪ Computational mesh
▪ Physics
▪ Results
▪ Lagrangian Multiphase
▪ CFD compared to site tests
• How CFD helped the milling industry
• Questions
Project background
• Drax Power Station (UK) approached CMP to help convert their
Babcock 10 E vertical spindle mills from coal to biomass equivalents.
• 3960 MW Power station (6 Units)
• Replacing coal with biomass pellets as primary energy source.
• The mill is used to break up the pellet into its primary shape called a
“flake”.
• Ultimately replacing coal which is a carbon accumulator, with
biomass, a carbon neutral, to help reduce CO₂ emissions.
Project background
• The modifications imposed before our involvement had the following
issues:
• Could not get the mill throughput up (maximum load required)
• Settling of biomass material in one of the production pipes.
• The accumulative settling effect blocked the pipes causing uneven distribution to the
burners.
• Due to material settling, they continuously need to open the mill to clear the blockages.
• The mill primary air (PA) fan continuously trips on high amp-alarm.
• The mill was not able to run on auto operation.
Project background
Project outcomes
• All biomass flakes (broken pellets) must pass to the burners.
• No passing of biomass pellets to the burners.
• No reject of biomass material into the mill reject boxes
• Increased throughput to accommodate lower calorific value of biomass.
• Equal fuel distribution to the burners.
• Static classifier design.
• Use most of the existing mill components (i.e. the primary air fan)
• Easy to manufacture and maintain.
Biomass with a coal classifier?
CFD SetupGeometry
PA inlet
PA ducting
Biomass classifier
Biomass feed pipe
Loading cylinders
Loading ring
Grinding elements
Biomass build up (bed)
CMP v 4 port ring
Mill table
Plenum chamber
CFD SetupGeometry
CFD SetupComputational mesh
• 10.7 million polyhedral cells
• Volume refinement in areas with high
gradients.
• High Y+ (2 layer) prism layer to solve the
boundary layer.
• Extruded mesh on production pipe
outlet generating a separate region to
incorporate mill back pressure effects.
CFD SetupPhysics
• Steady state analysis
• Reynolds stress model (RSM) to solve turbulence
• Energy extraction field function, used in a cell set, to
simulate heat losses due to conduction and moisture
driven off.
• Extruded region on pipe outlets, set with desired
porous inertial resistances to simulate wind-box back
pressures.
Operating Data Value Units
PA flow rate 21.21 kg/s
PA inlet temp 170 ᵒC
Mill outlet temp 87 ᵒC
Barometric pressure (Ermelo)
101.325 kPa
CFD SetupResults – mill temperature
CFD SetupResults – mill velocities
CFD SetupResults – mill velocities
0 °
CFD SetupResults – mill pressure
With just replacing the existing port ring with a CMP version 4,
the system resistances came down with 30 %.
CFD SetupLagrangian Multiphase – Calibration
0.01
0.1
1
10
100
1000
0.1 1 10 100 1000 10000Dra
g C
oef
fici
ent
Reynold's Number
Drag coefficient data for a sphere vs. a square prism
SN (Sphere)
Stoke's Law
Square Prism (Experimental Data)
Square Formula
P1
P3
P2
𝐷𝑟𝑎𝑔 𝐹𝑜𝑟𝑐𝑒 = 12𝜌𝑉2 × 𝐶𝐷 × 𝐴𝑟𝑒𝑎
𝐺𝑟𝑎𝑣𝑖𝑡𝑦 𝐹𝑜𝑟𝑐𝑒 = 𝑚𝑎𝑠𝑠 × 𝑔
CFD SetupLagrangian Multiphase – particle size distribution
Height/Thickn
ess (mm)
Length =
Width (mm)
Dh (mm)
0.6 5 0.00305983
0.6 4 0.00263688
0.6 3.8 0.00254824
0.6 3.4 0.00236612
0.6 3 0.00217670
0.6 2.6 0.00197864
0.6 2.2 0.00177011
0.6 1.8 0.00154846
0.6 1.4 0.00130959
0.6 1 0.00104645
0.6 0.6 0.00074442
0.6 0.2 0.00035788
0.6 0.08 0.00019429
0.6 0.04 0.00012239
Diameter (mm) Length (mm) Dh (mm)
12 42 0.031666
12 36 0.028573
12 28 0.024166
12 20 0.01931
12 12 0.013737
6 6 0.006868
Flake to sphere
Pellet to sphere
Flake specification
CFD SetupLagrangian Multiphase – setup
Value Units
Number of flake injectors 14 #
Number of pellet injectors 6 #
Particles per injector 300 #
Density of biomass 700 kg/m³
Track time 30 Seconds
Pellet injection plane Flake injection plane Biomass material passed plane
CFD SetupLagrangian Multiphase – flakes animation
CFD SetupLagrangian Multiphase – pellets animation
CFD SetupLagrangian Multiphase – pellets and flakes animation
CFD SetupLagrangian Multiphase – flakes in pipes animation
CFD SetupLagrangian Multiphase – biomass passing
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
90.00%
100.00%
100.00 1000.00 10000.00 100000.00
Per
cen
tage
of
inje
cted
par
ticl
es p
assi
ng
to b
urn
ers,
%
Particle Size, hudraulic diameter (µm)
Particle Size Passed
Flakes passing Pellets passing
CFD SetupCFD compared to site tests
Pipe 1Pipe 2
Mill load (Tons/h) Pipe split (%)
Pipe 1 Pipe 2
42 48 % 52 %
CFD maximum load run (incident mass flux
run)
49.05 % 50.95%
• Pipe 1 and 2 temp were within 0.5 °C indicating no settling (on site test).
• System resistance (mill differential pressure) reduced with 50 % (confirmed on site).
How CFD helped the milling industry
Accurately predict Eddy currents that might lead to particle suspensionAccurately predict wear patterns
How CFD helped the milling industry
Optimise patented technology
Conclusions
• STAR-CCM+ played an integrate role in the design process of the biomass classifier.
• The ability to incorporate field functions determined from calibration experiments etc.
• STAR-CCM+ identified areas of focus to help overcome problems that will not be able to be
solved with conventional design methodologies.
• The CFD model correlated well with on site results and this ultimately led to solving the
distribution problems at Drax.
• With the help of STAR-CCM+ we managed to solve all the outcomes of this project and the
mill was able to run on auto for the first time.
• The STAR-CCM+ software is a platform where a set CFD methodology for mills can be applied
and still correlate well with conventional design methodologies and on site test results.
• Ultimately, STAR-CCM+ helped to understand mills better
• Help solve real-world engineering problems.
• Complex Multiphysics solution