Submerged Combustion Melting The Next Generation Melting System

1
Submerged Combustion Melting The Next Generation
Melting System

• David Rue
• Gas Technology Institute
• 66th Glass Problems Conference
• U. Of Illinois, Champaign, IL
• Oct. 26, 2005

2
Submerged Combustion Melting Principle

• Air-fuel or oxygen-fuel mixture is fired directly
  into a pool of hot melt
• intense combustion
• direct contact heat transfer - combustion
  products bubble through the melt
• reduced NOx formation
• reduced CO and unburned hydrocarbon emissions
• High rate of heat transfer and rapid mass
  transfer
• High thermal efficiency
• Reduced melter size

3
Submerged Combustion Melting Features

• Melting and mixing in a single device
• No external device needs to contact the melt
• Short residence time from forced convective
  heating and mixing
• Melter is simple, robust, and reliable
• Small size low capital cost
• SCM is easily started and stopped in a few hours
• No hot repair work required
• Compatible with other segmented melting process
  steps
• Charging
• Conditioning
• heat recovery

4
GTI and GI SCM History

• Gas Institute (Ukraine) developed SCM for mixed
  nuclear waste vitrification and industrial
  melting not deployed
• Process simplified and commercial, air-fired
  units operating more than 10 years for other
  applications
• two 3-ton/h rockwool SCM units in Kiev, Ukraine
• three 3-ton/h rockwool SCM units in Byarosa,
  Belarus
• One SCM cement aggregate unit in Norilsk Russia
• GTI licensed SCM for applications outside former
  Soviet Union
• GTI has patents and background IP in melting,
  submerged firing, and heat recovery
• 500-lb/h SCM unit fabricated and operated at GTI
• Multiple melts including basalt and sodium
  silicate
• First use of oxy-gas burners

5
SCM 3 ton/h Mineral Wool SCM in Belarus
SCM Furnace
Loading Feed Hopper
SCM Interior
6
From Melt to Mineral Wool
Molten Slag Channel
Blow Chamber
4 Wheel Fiber Spinner
Product Fiber Mat
7
SCM Advantages

• Energy savings gt20 vs. oxy-gas melters
• gt55 capital cost reduction
• Compact with very little refractory 80
  refractory reduction
• Melt area is 15 of tank melter area (0.6
  ft2/ton/day)
• Reduced emissions
• NOx gt50 below oxy-gas melters
• CO and unburned hydrocarbons reduced gt20
• Rapid switching of melt composition
• Short residence time - rapid heat transfer
• Reliable, proven melting technology
• Feed flexibility lowers batch and feeder cost
• Mates with conditioning and heat recovery steps
• Excellent redox and color control

8
Approaches to Glass Melting

• Single tank
• Compromise simple and reliable, but non-optimized
  approach
• Holding furnaces, fining, and conditioning are
  needed after the melter for many glass products
• Staged or segmented
• Melting, mixing, refining, conditioning, heat
  recovery, etc. are optimized as needed for the
  glass product
• highly flexible with many potential process
  advantages
• Requires eloquent design for reliability and to
  avoid over-complexity and high capital cost

9
NGMS Project Underway at GTI

• Demonstrate melting and homogenization stage of
  low capital cost, energy efficient NGMS process
  for all industrially produced glass
• Sponsors
• DOE
• NYSERDA
• Gas industry
• Consortium actively supporting development and
  commercialization of SCM fro NGMS
• Corning Incorporated - PPG Industries, Inc.
• Johns Manville - Schott North America
• Owens Corning

10
Batch-Scale SCM at GTI
11
Lab-Scale SCM

• Industry batch melted to glass
• Full glass range melted
• Low-temp. soda-lime glass
• High-temp hard LCD glass
• Borosilicate glass
• Scrap reinforcing fiberglass
• Batch feed
• Continuous discharge
• Evaluation of glass product before pilot SCM
  fabrication
• Components scaled for 0.5-1.0 ton/h pilot SCM
• Product glass is fully melted and homogeneous

12
Special Tap Designed for Glass Melts
Soda-Lime Glass
13
Scrap Fiberglass Melt Sampling
14
Pilot-Scale SCM Unit

• Objective continuous feed and discharge made
  easier with
• Larger capacity melter (0.5-1.0 ton/h)\
• Demonstrated platinum discharge tap
• Most components are in place and tested
• Melter, burners, cooling water chiller needed
• Added instrumentation into data acquisition
  system
• Multiple burners spaced to create
• Uniform temperature profile
• Desired mixing and residence time distributions
• Elimination of poor mixing zones in corners and
  along walls
• Flexibility built into the unit
• Changeable burner patterns
• Provisions for two or more discharge locations
• Provisions for two feed locations

15
Glass Quality Varies Dramatically
Acceptable Bubble Count

• Lower-cost glass making must have BOTH
• High intensity melting
• Rapid refining
• Quality varies over 5 orders of magnitude
• SCM alone
• Makes fully melted homogeneous glass
• Only makes lowest quality glass
• SCM works well with all external refining methods

Glass Market Seeds/Oz Relative Seed Quality
LCD Display 10x better than TV panel glass
TV Panel 10x better than float glass
Float/Flat 1,000 - 10,000x better than container glass
Textile Fiber 100x better than container glass
Tableware lt 2 10x better than container glass
Lighting Glass 25 2x better than container glass
Container 10-20 10x better than funnel glass
TV Funnel 200 2x better than wool insulation fiberglass
Insulation Fiber 400
16
NGMS (SCM AND Rapid Refining)

• If refining is slow, the capital cost benefits of
  low-cost, high-intensity melting are lost
• Potential refining approaches include

Sonic Lab-scale tested Potentially low cost and very rapid, easily installed
Helium (inert gas) Entering commercial trials Helium cost is acceptable, not usable on all glasses
Thin film Limited commercial use Particularly good SCM match, bubbles are large and no CO2
Reduced pressure Limited commercial use Requires good control, hardware is expensive
centrifugal Pilot-scale tested Complex hardware, potentially very rapid
17
Toward Commercialization

• Already completed
• SCM concept
• Pilot-scale oncept validation, including
  combustion system
• Initial commercial use for low-quality products
  (mineral wool, aggregate)
• Current activities through 2006
• Lab-scale melting of full range of industrial
  glass and fiberglass scrap
• Batch feed and continuous discharge using oxy-gas
  burners
• CFD and physical modeling of SCM
• Design, fabrication, and operation of continuous
  0.5-1.0 ton/h pilot-scale SCM
• Preparations for first industrial demo-scale SCM
• Design, construction of first commercial SCM
  making abrasives from steel industry waste and
  cullet (northern IN)

18
Next Steps

• 2006-2008
• Plan first glass industry plant demo-scale 1-4
  ton/h SCM
• Fiberglass or scrap fiberglass
• Test unit not replacing existing melter
• Rapid conditioning work to develop NGMS for all
  industrial glass compositions
• 2009
• Demo-scale SCM and NGMS units in consortium
  member plants
• Initial replacement of current melters with NGMS
• 2012
• Fully developed and commercially demonstrated
  NGMS using SCM
• Licensing of NGMS to non-consortium member glass
  companies

19
Commercialization Pathway

• Expected order of market entry
• Scrap fiberglass
• Fiberglass
• Specialty glass (pressed and blown)
• Specialty glass (optical fiber, LCD, etc.)
• Container glass
• Flat glass
• Consortium agreement lays out company access
  priority to the NGMS technology
• Consortium member companies
• GMIC member companies
• Non-GMIC glass companies