Sei sulla pagina 1di 80

Concepts for energy & emission friendly glass melting

Evolution or Revolution in Glass Melting


Ruud Beerkens

Cairns 2011 AFPG Ruud Beerkens 1


The difficulty to observe the glass melting process….

exhaust
flame

fresh raw material batch glass


melt
bubblers
foam

Cairns 2011 AFPG Ruud Beerkens 2


Approaches
• Glass quality – dedicated laboratory tests
• Optimum melting
• Optimum fining conditions

• Energy Efficient (Container) Glass Furnace


• Low NOx emissions
• Normal glass quality
• Highest energy efficiency

• Combination Energy Efficiency, LowNOx & High Glass Quality


• Application of CFD models for furnace design & operation

Cairns 2011 AFPG Ruud Beerkens 3


Specific energy consumption NL
conventional container glass furnaces
red bar: range of theoretical low-limit based on thermodynamics for melting and heating of
o
glass (1350 C) for 100 % batch (upper range) and 100 % cullet (lower range)
orange bar: best practical limits, no wall heat losses
Energy consumption melting in GJ,
14
13
12
11
per net ton glass

10
9
8
7
6
5
4
3
2
1
0
1940 1950 1960 1970 1980 1990 2000 2010

Year

100 % cullet ultimately insulation, optimum regenerator: 2 MJ/kg


100 % normal batch, dry, ultimately insulated: 3 MJ/kg (*2.7 MJ batch preheat)
Cairns 2011 AFPG Ruud Beerkens 4
Ranking energy consumption:
Container glass furnaces database 2003 - 2010
10000
Primary energy normalized
50% cullet [MJ/metric ton] , 168 furnaces
9000
8000
Database
7000
World top Deciel (4160 MJ/Ton)
6000
5000
4000
Average value  average of 10 % most energy efficient:
3000 gives on average 21% energy savings
0 25 50 75 100 125 150 175
Ranking
Conclusion: bringing all furnaces to the top 10 % (2010) will give
on average 21 % energy savings: BEST PRACTISE
Cairns 2011 AFPG Ruud Beerkens 5
Among most optimum container glass furnaces today, without batch preheat

Cairns 2011 AFPG Ruud Beerkens 6


Lowest emissions & energy consumption

Container glass sector – best results

NOx: 500-700 mg/Nm3 air-fired (no SCR)


NOx: < 0.35 kg/ton melt (oxygen-fired)

Particulate: 80-90 mg/Nm3 (upstream filter)

Energy: 3.4 - 3.55 MJ/kg melt (50 % cullet)


3.1 MJ/kg including batch preheat, high cullet level

Cairns 2011 AFPG Ruud Beerkens 7


Typical energy (primary energy equivalent) distribution in
different glass sectors (2009)
Container Float Tableware Fibre Wool total primary energy plant
7-8.5 8-9.5 20-50 17-25 20-25 GJ/net ton glass product
100%

Batch preparation
80%
Energy consumption
distribution in plant

Glass Melting (glass furnaces)

60% Conditioning of melt (feeders)

Forming
40%
Annealing

20% Post treatment, inspection,


packaging
Buildings and diverse utilities
0%
Air Pollution Control
ss

e
ss

es

l
oo
ar
la

la

br

W
ew
G

Fi
bl
er

at

Ta
Fl
in
ta
on
C

Cairns 2011 AFPG Ruud Beerkens 8


Glass Quality Aspects

Analysis of industrial glass


melting processes today

Glass melt quality = time x


temperature x
f (mixing) x
g (chemistry)

On average melting consumes 65 % of the total energy of glass production


High quality glass melting

• Control of batch composition


• Cullet quality: no ceramics/glass ceramics > 2 mm
• Cullet colors
• Redox number & reducing power of cullet
• DWL, purity, seeds
• Complete melting of batch
• Dissolution of sand in melt
• Primary fining: bubble removal
• Secondary fining: bubble re-absorption
• Conditioning: Uniform viscosity prior to forming

Cairns 2011 AFPG Ruud Beerkens 10


Return flow Generation blisters
for batch heating from refractory
Day
hopper Refining
Bubble absorption

Conditioning of melt
Thermal homogeneity
Hot spot &
evaporation

Zone for sand


grain dissolution

Batch melting Return flow


• 40-60 minutes Spring zone from working end
• 80-90 % of net heat flux & primary fining

Cairns 2011 AFPG Ruud Beerkens 11


Critical Analysis of Conventional Glass Melt Process

Heat input into melt tank

excess air

emissivity
25 oC, 0 % Hmelt
1300 oC, 85 % Hmelt
foam
humidity
1350 oC,
600 % 200 %
100 %
100 % Hmelt
100 %
700 % 120 %
boosting
20 %

working end
return Hot spot Spring zone
return flow
flow

Cairns 2011 AFPG Ruud Beerkens 12


Residence time distribution float, container
Average τ and TV panel glass furnace
1
Container
TV-panel
0.8
Float
% per h

0.6

0.4

0.2

0
0 50 100 150 200 250 300
Residence time τ [h]

min. residence time/average residence time = 0.15-0.20


TNO Glass Group, (c) 2005

Cairns 2011 AFPG Ruud Beerkens 13


Many different trajectories (paths) in tank

Cairns 2011 AFPG Ruud Beerkens 14


Improved design features

• Optimize “Space Utilization”


• Avoid strong recirculation flows from fining zone
into melting-in zone
• Separation of melting-in and sand dissolution section
from fining area, enhance convective flows only in
melting zone but not in primary fining zone!

Cairns 2011 AFPG Ruud Beerkens 15


Example ‘ideal situation for glass melting’
‘inert refractories’

low pressure fining or


fining shelf ?

20-1200 oC 1300-1350 oC heating 1500-1600 oC 1350-1100 oC


batch melting 1550 oC fining refining &
blanket conditioning

0.75 hour 0.75 hours 1 hours 1- 2 hours 1.5 -2 hours

Selection of refractory per zone

Cairns 2011 AFPG Ruud Beerkens 16


Advanced fining & melting Rue

Submerged combustion: short residence time &


high heat flux (SCM)

Thin film melter: refractory issues


SCM

Plasma melting: very high temperatures plasma, Gonterman


- short torch lifetime high energy costs
- suitable for special glass types (low alkali)

Fining shelf: short distance for ascension

Low pressure fining: fining shaft at pressures < 0.3 bar

Centrifugal fining: short bubble removal time in fast rotating vessel


Klouzek, Prague

Cairns 2011 AFPG Ruud Beerkens 17


Batch (blanket) melting

• Residence time in batch blankets: 0.5 - 1 hour


• Floating of batch isles on top of melt
• > 80 % of energy to be transferred to melt is
absorbed by batch blanket (within 0.5 -1 hour)
• Most heat is supplied by re-circulating glass melt
from hot spot zone

Cairns 2011 AFPG Ruud Beerkens 18


JSiO2(melt) = 0.332•v1/2•x-1/2•ρ1/6•η-1/6•DSiO22/3•(C*m,SiO2-Cm,SiO2)

Cb,SiO2 • dXb/dt = C*m,SiO2 • dXb/dt + JSiO2(melt)

Melting of batch layer thickness X:


dX/dt=0.332•v1/2•x-1/2•ρ1/6•η-1/6•DSiO22/3•(C*m,SiO2-Cm,SiO2)/(CbSiO2-C*m,SiO2)

Cm,SiO2 = concentration SiO2 in melt (bulk) kg/m3


C*m,SiO2 = saturation concentration SiO2 in melt kg/m3
Cb,SiO2 = concentration SiO2 in batch kg/m3
v = glass melt velocity underneath batch m/s
x = distance from batch tip m

Batch blanket T

x 1450 oC
1300 oC v Bubbling in melt

Cairns 2011 AFPG Ruud Beerkens 19


Rapid Melter
Concept 2

batch batch

0.8 m

0.8 m 1.6 m
flow roll
bubbler bubbler
3.2 m

Cairns 2011 AFPG Ruud Beerkens 20


Sand dissolution
• Typical temperature window 1100-1350 oC
• Strong convection improves sand grain dissolution
• Separation between sand grain dissolution zone and fining zone
is essential
• Avoid temperatures > 1350 oC to minimize overlap between
fining and sand grain dissolution.

• Analysis of minimum residence time required


• Batch free time experiments
Depending on: Glass Service
- Glass composition (Na2O, CaO, MgO content)
- Sand grain sizes (size distribution)
- Convective flows
- Temperature

Cairns 2011 AFPG Ruud Beerkens 21


Effect of temperature and grain size on “batch
free time” in soda-lime-silica glass melting

batch free time

grain diameter

Cairns 2011 AFPG Ruud Beerkens 22


o
1250 1300 1350 1400 1450 1500 C
25000
Do = diameter sand
Do = 0.4 mm,
20000 -1
grad v = 0.01 s
Dissolution time
sand grain in s

Do = 0.4 mm,
15000 -1
grad v = 0.05 s

10000

5000 Do = 0.2 mm,


-1
-1 grad v = 0.05 s
Do = 0.2 mm, grad v = 0.5 s
0
1500 1550 1600 1650 1700 1750 1800 1850 1900
Temperature in K

Dissolution time of sand grains in container glass melt


as function of temperature for different convection intensities

Cairns 2011 AFPG Ruud Beerkens 23


Sand grains in molten glass and bubble formation

Cairns 2011 AFPG Ruud Beerkens 24


Sand grain dissolution in melting tank

Batch charge Primary fining area

Throat area

Cairns 2011 AFPG Ruud Beerkens 25


FINING

Just after batch melting:

50000 - 200000 seeds (D = 50-400 microns) per kg glass!!

Cairns 2011 AFPG Ruud Beerkens 26


0 to 8 mm

Cairns 2011 AFPG Ruud Beerkens 27


Fining (primary)
• Fining agents added to the batch to enhance bubble growth &
to increase the rising velocity of these bubbles
• Often used fining agent: Sodium sulphate

Fining reaction: T > TFining onset pSO2 ⋅ pO2


K =
'

[SO3 ]
Na2SO4 ⇔ Na2O + SO2 (gas) +1/2 O2 (gas)

Cm CO2 CO2
SO2
Stripping of CO2
and N2 from melt O2

Cm N2 N2

Dilution of N2 & CO2 in bubble by fining gases

Cairns 2011 AFPG Ruud Beerkens 28


100
1350 oC
1400 oC
Time required to ascend 1 meter

1450 oC
in float glass melt in hrs.

10 1500 oC
1550 oC

0.1
0 0.5 1 1.5 2 2.5 3 3.5
Diameter of gas bubble in melt in mm

Cairns 2011 AFPG Ruud Beerkens 29


400 600 800 1000 1200 1400 1500 1500 °C
2.5 I I I I I I 3
10 °C/min 4 °C/min constant

2.5
2 Foam height

specific gas volume (ml/g)


gas release (micro l/s//g)

SO2 2
1.5 CO2 / 40 peaks
SO2 1.5

1
1

0.5
0.5
CO O2
0 0
0 1000 2000 3000 4000 5000 6000 7000 8000
time (s)

Foaming and gas release from normal sulphate containing


batch without reducing agents in N2 atmosphere: flint glass
Fining at high T!
Cairns 2011 AFPG Ruud Beerkens 30
Slightly reduced batch – fining at lower T !!
Na2S + 3Na 2SO 4 + nSiO 2 →
10 1000
CO2 (Na2O)4 ⋅ (SiO2 )n + 4SO 2

Concentration CO & SO2 in


Concentration CO2 in purge

CO
8 800

purge gas [vol ppm]


SO2
gas [vol%]

6 600
Direct reaction
4 sulfate with 400
reducing
component
2 200

0 0
500 600 700 800 900 1000 1100 1200 1300 1400 1500
Temperature [°C]
Sulfur content = 0.5 mass%SO3
No oxygen evolution observed Added carbon content = 0.05 mass%

Cairns 2011 AFPG Ruud Beerkens 31


100
1350 oC

Primary fining 1400 oC

Time required to ascend 1 meter


1450 oC

in float glass melt in hrs.


10 1500 oC
1550 oC

• Bubble Growth
• Above fining onset temperature 0.1
0 0.5 1 1.5 2 2.5 3 3.5

• Bubbles should grow to D >>0.5 mm Diameter of gas bubble in melt in mm

• Removal of dissolved gases from melt (stripping)

• Fining onset temperature depends:


• sulfate content batch
• reducing agents in batch
• water vapor (oxygen firing versus air-firing)

Cairns 2011 AFPG Ruud Beerkens 32


Fining onset temperature

• EGA tests

• Bubble observation in glass melt

• Determination of gas evolution from melt &


sudden increased bubble growth

Cairns 2011 AFPG Ruud Beerkens 33


III. Fining gases and
other dissolved
gases diffuse
strongly into bubble

II. start of fining:


gases diffuse into
bubble

I. static bubble

Reaction in melt: release of fining gases


Pgases melt > pt (pt is pressure in bubble)

Cairns 2011 AFPG Ruud Beerkens 34


3 hours molten float glass batch in laboratory at 1450 oC

Cairns 2011 AFPG Ruud Beerkens 35


1500
fining onset
1400
1300
Temperature oC

1200
1100
1000
900
800
700
600
500
0 2 4 6 8
Time (hours) in melting tank

Temperature – time course of ‘most critical’ trajectory


In float glass melting tank
Cairns 2011 AFPG Ruud Beerkens 36
1500
Fining onset temperature in C 1490
o

1480
1470
1460
1450
1440
1430
1420
1410
1400
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7

Initial content SO3 in float glass melt in mass-%

Cairns 2011 AFPG Ruud Beerkens 37


Secondary fining

Second step: Secondary fining/Refining


• Dissolution of (small) remaining bubbles
• Only effective if bubble contains gases
(CO2, O2, SO2+O2) that dissolve in cooling melts
• Glass melt should be lean in dissolved gases

Cairns 2011 AFPG Ruud Beerkens 38


Proposal – for SLS glass with high melting quality

• High convective melting-in zone


(residence time 60 minutes)
• High heat transfer rate to batch
• Temperatures < 1300-1350 oC to avoid early sulfate
decomposition
• Shallow fining zone (H < 0.5 m) with extra heating
(1400-1525 oC, > 1 hour depending on H and T-Tonset)
- sufficient time > fining onset temperature for all melt
trajectories
• Slow cooling of melt between 1400-1250 oC
(secondary fining)

Cairns 2011 AFPG Ruud Beerkens 39


Energy & NOx

CFD modeling of melt & combustion

Cairns 2011 AFPG Ruud Beerkens 40


Classical Design  Energy Efficiency & primary
NOx emission reduction are main targets
Characteristics
• End-port fired
• Spacy regenerators η = 70 %
• Spacy combustion chamber
• Specific pull: 3 - 4 tons/m2⋅day
• Batch preheating
• Insulated bottom & crown (< 2 kW/m2)
• Sidewalls average loss < 4 kW/m2

Disadvantage:
• Wide residence time distribution
• Relatively high structural losses
Cairns 2011 AFPG Ruud Beerkens 41
Important TOOL: GLASS FURNACE MODELING:
Results of CFD simulation models
• Temperatures at all possible positions
• Combustion space
• Glass melt
• Refractory

• Glass melt and Combustion gas velocities


• Trajectories (particle tracing) in tank
• Thousands of different paths from charging end to throat or spout
• Redox and dissolved gases
• Redox state of melt at each position (pO2 or Fe2+/Fe3+)
• Residence time distribution
• Minimum residence time is of importance for melting process
• Glass melt quality indices per trajectory
• Trajectory with minimum melting or fining index is decisive for glass

Cairns 2011 AFPG Ruud Beerkens 42


Application of CFD models
• For furnace design (lowest energy, highest glass quality)
• Optimum depth of tank
• Position bubblers or dam or burners
• Size and design of throat
• Design combustion chamber (LOwNOx, less evaporation)
• For optimum process settings
• Optimum fuel-boosting ratio
• Temperature profile (energy distribution) & Bubbling rate
• Creation of distinct spring zone to avoid short cut
• Time-transient (time dependent) for colour or pull change
• Optimize colour change process: reduce transition time
• Time-transient for process control (r-MPC)
• Sensors give model continuous new information: model tracking
• Model continuously gives recommendation for input parameter
changes to follow optimum process path (low energy, high glass
quality, constant T)
Cairns 2011 AFPG Ruud Beerkens 43
CFD Models
• Start with basic concept
• Type of furnace
• Combustion chamber & burner port features
• Boosting, bubbling, weirs, deep refiner
• Type of doghouse construction

• Optimize design by CFD and determine glass quality and energy


efficiency by CFD model
parameters:
• Effect of glass tank depth
• Throat design
• Positions bubblers & electrodes
• Heating profile
• Length – width ratio
• Dimensions weir/dam or fining shelf
• Number and positions of burners
• Burner port angle and burner port dimensions

Cairns 2011 AFPG Ruud Beerkens 44


Design, Operation & CFD Models
• Optimize glass melt flow pattern

• Optimize heat transfer and heat transfer distribution

• Avoid short cut flows along melter bottom

• Design combustion chambers & burner ports


• Slow mixing of fuel (gas) and air (or oxygen)
• Create soot flames (improved heat transfer and low NOx)
• Avoid post-combustion in ports/regenerators
• Avoid excessive wall temperatures in superstructure

Cairns 2011 AFPG Ruud Beerkens 45


Geometry & Grid

Doghouse

Batch Boosting Electrodes


3 Burners (underport)

Cairns 2011 AFPG Ruud Beerkens 46


Cairns 2011 AFPG Ruud Beerkens 47
Temperatures in combustion chamber (scaling in oC)

Cairns 2011 AFPG Ruud Beerkens 48


NOx
End-port fired furnace horizontal cross section at level of burners

Base case

4 inch higher crown

Cairns 2011 AFPG Ruud Beerkens 49


2500
(8 % O 2, dry) end port
2000 cross fired

1500

1000
3
NOx mg/Nm

500

0
0 0.05 0.1 0.15 0.2 0.25
3
m combustion chamber space
3
per Nm natural gas input/hr

NOx emission levels depending on


relative size of combustion chamber

Cairns 2011 AFPG Ruud Beerkens 50


Found (rough) correlation between specific energy
consumption container glass furnace and relative combustion
space volume
6000 average residence time combustion gas:
τ = 5 - 10 s in combustion chamber
consumption in MJ/ton

5500
Specific energy

molten glasss

τ=±5s
5000

4500 τ = ±10 s

4000
0.05 0.1 0.15 0.2
3 3
m combustion space versus natural gas flow in Nm /hr

Cairns 2011 AFPG Ruud Beerkens 51


Particle path for minimum melting index
Melting index distribution

0.9

0.8

Fractional distribution [1/h]


0.7

0.6

0.5

0.4

0.3

0.2

0.1

t res 0

T 0 5 10 15 20 25

MID = ∫η
0
dτ Melting index / 1.000.000 [-]

Boosting?

Cairns 2011 AFPG Ruud Beerkens 52


base

electric boosting

bubbling

Cairns 2011 AFPG Ruud Beerkens 53


Characteristic flow and temperatures:

Temperature [C]

Tendency for short cut flow!


Refiner

Cairns 2011 AFPG Ruud Beerkens 54


Efficient glass melting

• High glass quality (requirement depends on glass type)


• Homogeneity
• No bubbles
• No inclusions
• Color
• High space utilization (limit recirculation)

• Furnace lifetime
• Low energy consumption
• Low emissions
• Flexibility, short transition times

Cairns 2011 AFPG Ruud Beerkens 55


Design & Operation for Conventional Glass Furnaces

Cairns 2011 AFPG Ruud Beerkens 56


Recommended design considerations
conventional glass furnaces

• Carry-over:
• Use extended doghouse with open radiation to
incoming batch
• Avoid gas velocities just above non-sintered loose
batch > 6 m/s
• Avoid very fine raw materials
• Especially for end-port fired furnaces & batch
preheating: special doghouse constructions are
necessary

Cairns 2011 AFPG Ruud Beerkens 57


Particulate (dust) emissions
Dust emissions will increase:
• Reducing flames may increase evaporation and dust emission > 50 %
• 2 x higher gas velocity gives 70 % more dust
• 50 oC higher surface T melt: gives 50 % more particulate emission
• Water vapour increases evaporation rate of Na, K, Boron
Effect of burner position above melt
model predictions U-flame furnace
20 0.3

emission [kg/ton
gas velocity above

Na2SO 4 dust
16 0.25
melt [m/s]

0.2
12
0.15
8
0.1
m/s
4 Na2SO4 g/kg glass 0.05

0 0
0.6 0.8 1 1.1 1.2
distance burner to
glass melt surface

Cairns 2011 AFPG Ruud Beerkens 58


Important dimensions !

CO
O2

CO
O2

Option throat with slope


NOx
End-port fired furnace vertical cross section at 25 % from furnace length from port

NOx scaling in mole fraction

Base case

4 inch
higher crown

Burner port Exit port (flue gas)


Lower NOx-concentration in exit

Cairns 2011 AFPG Ruud Beerkens 60


Installation of CO and O2 sensors (laser absorption)
in hot exhaust gas of glass furnace

Furnace

Regenerator

Flue
gases

Red = air cooled


Blue = water cooled
CO O2 = Flushing air

signals to air-fuel control burners


Cairns 2011 AFPG Ruud Beerkens 61
Recommended design considerations
conventional glass furnaces
• Glass melting tank
• Use CFD models to optimize Length-Width-Depth of tank
• Depth of melt depending on glass color
• Use CFD models to optimize position of bubbling zone, weir
and barrier or batch boosting
• Use modern barrier boost designs in case of boosting
• Keep well distributed batch over full width of melting tank
• Improve heating flux to melting batch
• Consider fining shelf BOC

Application of “top” burners CGM for high pulled furnaces


Cairns 2011 AFPG Ruud Beerkens 62
Principle fining shelf

Cairns 2011 AFPG Ruud Beerkens 63


Distance from surface during sulfate fining in flint glass melt:
seed starts at 1 meter depth
1.2
1723 K
1
Distance from surface [m]

1773 K
1798 K
0.8
1823 K

0.6 1873 K
1923 K
0.4 1773 K low P
fining shelf 1773 K
0.2

0
0 2 4 6 8 10
Time [hours]
Fining shelf: 40 % reduced fining time or 35 oC lower fining T
Cairns 2011 AFPG Ruud Beerkens 64
Recommended design considerations
conventional glass furnaces
• Glass melting tank
• Optimize flow patterns in melt
• Avoid glass melt bottom flow towards throat
• Avoid too short residence times
• Avoid static (dead water) zones  cat scratches
• Optimize “space utilization”
• Use CFD modeling to optimize:
• Flow patterns
• Heat transfer
• Position Spring zone & Hot Spot
• Barrier Boost system
• Temperatures & Residence times (e.g. above fining onset)
• Avoid too small weirs!
• Avoid short distances between bubbles & AZS weirs
• Optimize throat design (sloped throat) to avoid upward drilling
• Design throat to enable interim repairs
• Focus on doghouse design & batch charging (keep short blankets)

Cairns 2011 AFPG Ruud Beerkens 65


Energy flows modern end-port fired container glass furnace
40% cullet 320 tons melt/day 4.20 MJ/kg melt
A
A. Evaporation water
B
H 2.2%
B. Endothermic fusion reactions
28.0% 5.4%

C C. Heat enthalpy glass melt


G
2.3% 40.8% D. Wall losses
6.1%
D
F E. Cooling & leakage heat losses
15.2%
F. Regenerator wall losses

G. Flue gases

Cairns 2011 AFPG Ruud Beerkens 66


Measures 320 tons glass melt/day
Estimated savings MJ/ton % savings
Base case 4.192 0
Batch humidity 3.5 → 2 % 4.08 2.7
Emissivity flames 0.18 → 0.25 4.13 1.4
20 % better insulation 3.98 5.2
o
Batch preheating 300 C 3.43 18
Air excess 10 % → 5 % 4.15 1
Cullet 40 → 75 % 3.81 9
Crown 10 % higher 4.22 -0.6
Regenerator 63 → 68 % 3.991 4.8
o
Throat temperature 1325 → 1300 C 4.105 2
3
No cold air infiltration -500 Nm /hr 4.12 1.7

Table Energy Saving Potentials for Modern End-Port Fired Regenerative


Container Glass Furnace (moderate boosting 1000 kW) - gas firing

Cairns 2011 AFPG Ruud Beerkens 67


Considerations new (advanced)
furnace designs

Cairns 2011 AFPG Ruud Beerkens 68


Statement
• Most glass furnaces produce a ‘range’ of glass qualities
(small glass melt packages with different melting history)
because of different trajectories from inlet to outlet in the tank

• It is important to improve control of these trajectories or to


develop furnaces with distinct trajectories to avoid:
• glass ‘packages’ with low melting temperature history
• glass ‘packages’ with low residence time in the furnace
• glass ‘packages’ that remain very long in the furnace
This will cost extra energy and pull
• glass ‘packages’ that stay in dead water zones
• glass ‘packages’ that are contaminated

Cairns 2011 AFPG Ruud Beerkens 69


Optimum flow conditions glass melt tank
• High ratio minimum versus average residence time of glass
in tank: ‘good use of tank volume’ (space utilization)

• Design of throat
• Compromise between glass quality & energy efficiency & throat
lifetime
• straight through throat (high): return flow from refiner
 return flow blocks the forward bottom flow from the melter
• narrow, shallow throat or submerged: hardly (cold) return flow
• All parts of melt should reach fining onset temperature
• More important than residence time is fining index
• Fining shelf will increase chance of bubble removal without need of
excessive glass melt surface temperatures

Cairns 2011 AFPG Ruud Beerkens 70


Instead of non-controlled stages of the melting
process in one tank

• Segmented melter design


• Dedicated sections for different functionality
• No strong re-circulation patterns (this will increase
energy costs & gives variety in glass quality)

Cairns 2011 AFPG Ruud Beerkens 71


Process steps

• Batch melting
• Sand grain dissolution
• Primary fining
• Secondary fining
• Homogenisation

Cairns 2011 AFPG Ruud Beerkens 72


Example ‘ideal situation for glass melting’
‘inert refractories’

low pressure fining or


fining shelf ?

20-1200 oC 1300-1350 oC heating 1500-1600 oC 1350-1100 oC


batch melting 1550 oC fining refining &
blanket conditioning

0.75 hour 0.75 hours 1 hours 1- 2 hours 1.5 -2 hours

Selection of refractory per zone

Cairns 2011 AFPG Ruud Beerkens 73


Glass melter – segments for:
• Batch melting (average residence time 1-1.5 hours):
• Thin layer (< 0.05 m)
• or Very strong convection (small batch piles)
• Compacted batch (intense grain contact and fast heat conduction)
• 0-800 oC takes 2x longer than 800-1300 oC: batch preheating helps!
• Sand grain dissolution (average residence time 1.5 hours)
• Strong convection: stirring or bubbling
• 1200 < T < 1350 oC
• Narrow sand grain distribution 80-250 microns
• Fining (primary), average residence time 2-2.5 hours
• Glass level < 0.3-0.4 meter
• Bottom fining shelf > 1400-1450 oC
• Surface melt < 1500 oC *(to limit evaporation)
• Residence time about 2 hours (fining zone: 50-75 m2 for 500 TPD)
• Preconditioning of melt by helium bubbling (< 1350 oC)
• Shallow atmosphere of helium (or <<1 bar) above melt in fining zone
• Refining/ Conditioning (1325-1275 oC: 1 hour)
Cairns 2011 AFPG Ruud Beerkens 74
Example “segmented furnace & fining shelf”

Cullet pre-heater connected to Recuperative Container Glass Furnace

Cullet preheater
TE=20°C
TA= 280°C fining shelf

Recuperator
TA= 1130°C (flue gas in)
TV= 670°C (air out)

Cairns 2011 AFPG Ruud Beerkens 75


Typical aspects of industrial melting Soda-Lime-Silica Glass
T interval oC Time in hours Conditions
Batch heating 30-1300 oC 1 Thin layer, creation of batch tips
Strong convection, aggressive
melt, avoid too high temperatures
Sand grain dissolution 1100-1350 oC 2 (> 1350 oC)
No recirculation, high
temperatures, redox control and
Primary fining (sulfate) 1350-1550 oC 1.5 to 3 sufficient fining agent
Stirring, bubbling avoid glass melt
with recent refractory or glass
Homogenisation 1200-1550 oC surface contact in throat or canal
Secondary fining 1350-1250 oC 0.75 to 2 Slow cooling of melt
o
Silica > 1400 C (air firing) &
Silica > 1480 oC (oxygen firing)
tank walls: cooling avoid strong
Refractory protection convection No aggressive melts or NaOH vapours
Avoid: High glass melt surface
temperatures, high CO and H2O
> 1400 oC strong contents combustion chamber, high
Limit evaporation evaporation gas velocities above melt

Limit foam formation Control of redox, avoid excess fining

Cairns 2011 AFPG Ruud Beerkens 76


X: allow
no return flow Primary fining
1 - 2 hours
Sand grain dissolution Plug flow & shallow tank
Batch blanket heating & fusion <2 hours X 1350 → 1550 oC
heating & fusion<0.45 hr Intensive stirring &
Plug flow & 1200 → 1350 oC X: no return flow
300 → 1250 oC
most energy Secondary fining
transfer required 1 - 2 hours
Preheating of Controlled cooling
batch τbatch = 1350-1250 oC
3 – 6 hours
30 → 300 oC
Conditioning
Batch preheater (optional) 1 - 2 hours
Stirring & thermal homogenization
1300 → 1150 oC

Forming/shaping

Cairns 2011 AFPG Ruud Beerkens 77


New elements in furnace designs
• Revolution
• Segmented (compact) melters
• Physical fining techniques
• Limited number of trajectories in furnace
• New heating techniques for batch blanket & fining zone

• Evolution:
• Fining shelfs (shallow areas in tank)
• Increased combustion space sizes
• Model based control & sensors (CO, redox melt/batch)
• Improved refractory materials, especially downstream spring
zone
• Use of modelling for furnace design (tank, throat, electrodes,
bubbling, barriers/fining shelfs, combustion chamber)
Cairns 2011 AFPG Ruud Beerkens 78
Conclusions
• Most glass melting tanks show very wide residence time distribution
• Average versus minimum RTD: 0.15-0.2
• Ratio RTDav/RTDmin → 0.4: Furnace volume reduction (50 %): less
capital costs
• But batch melting process has to be changed to realize this!!
• Each process step requires its own conditions
• Temperature level, chemistry, mixing conditions
• Design per section & Control of these parameters per section
 Segmented melting unit to be developed by aid of CFD models
• New furnace design with narrow RTD will improve:
• Energy efficiency ⇒ 15-20 % extra energy savings
• Transition speed during glass change
• Conventional furnaces: Energy efficiency in glass furnaces can be improved
by about 20 % on average for container glass furnaces

Cairns 2011 AFPG Ruud Beerkens 79


Cairns 2011 AFPG Ruud Beerkens 80

Potrebbero piacerti anche