Melting and Refining

Melting

Melting, the combination of the individual raw materials at high temperature to form a molten glass, is the central phase in the production of glass. There are numerous ways to melt glass depending on the desired product, its end use, the scale of operation, and the prevailing commercial factors. The glass formulation, raw materials, melting technique, fuel choice and furnace size will all depend on these factors (IPTS/EC, 2013. p. 40). Majority of the glass melting takes place in continuously operated melting furnaces, also called melting tanks, with discontinuous melting being used for the production of speciality glass products (Worrell et al., 2008. p.9). 

In continuous operations, the mixture of batch and cullet is continuously charged into the glass-melting furnace to compensate for the glass withdrawn and keep the glass level in the furnace constant. Combustion heating – using natural gas and fuel oil – and direct electrical heating, as well as the combination of both, are commonly used heating methods for the melting tanks. Electric boosting is commonly used in order to increase production rates and/or to increase production flexibility (Worrell et al., 2008. p.9). Currently, fiber glass production is the main user of electric melting, as it allows producing a very homogenous and high quality product (Worrell et al., 2008. p.12.

The minimum residence time of the glass melt in the furnace is a crucial parameter for ensuring glass quality and varies significantly by the type of glass produced. Normally, the higher the quality of glass produced, the longer the residence time, in order to ensure a perfect homogenisation and elimination of possible stones, bubbles, etc. which would affect the properties of the final product. The difference in residence time of the glass melt in the furnace is directly associated with the specific energy consumption; therefore, for a given capacity of the melting furnace, the type of glass produced can be associated with a significantly different energy consumption (IPTS/EC, 2013. p. 40).

Refining

During refining, also known as fining, molten glass is freed of bubbles, homogenized, and heat conditioned. Refining also takes place in the melting tank. 

Glass melting is one of the most important and energy intensive processes in the manufacturing of glass prodcuts, consuming 60 to 70% of the total energy used in glass production. While only about 40% of the energy used goes to heating the batch and converting batch constituents, approximately 30% can be lost through the furnace structure, and another 30% through the exhaust gases (US DOE, 2002. p.55). Consequently, particular emphasis needs to be placed on optimizing the melting tank. Recuperative and regenerative furnaces, which allow for heating the incoming combustion air with heat recovered from exhaust gases, are commonly used and these assist energy efficiency. Remaining heat in the exhaust gases can also be used in a waste heat recovery boiler or for preheating cullet. 

Various energy efficiency measures can be considered for existing furnaces. For furnaces to be newly bulit or to be replaced, new designs and oxy-fuel technolgoies can provide considerable savings (Worrell et al., 2008. pp.9-11). 

Melting and RefiningTechnologies & Measures

Technology or MeasureEnergy Savings PotentialCO2 Emission Reduction Potential Based on LiteratureCostsDevelopment Status
Computerized Process Control

Applications of Expert System II™ show energy savings of about 2 to 3%, improved yields of about 8%. 

The developers of Glass Max claim a 4% increas in annual production. 

The reported paybak time for Expert System II™ is around six months. 

Due to high price of electricity Melting Expert has a potential to reduce costs by reducing electricity consumption. 

In a Canadian plant Glass Max system was installed for US $150 000 and the payback time was estimated as less than one year. 

Commercial
Optimizing the Burner Position

Quantitative information not available. 

Quantitative information not available. 

Quantitative information not available. 

Increased Cullet Use

As a general rule, every 10 % of extra cullet results in a 2.5 – 3.0 % reduction in furnace energy consumption (IPTS/EC, 2013. p. 314)

Based on a study in german-speaking European countries, energy savings are estimated at around 8 MJ for every percent increase (by weight) in cullet use (Worrell et al., 2008. p.67)

The use of cullet generally results in significant cost savings as a result of the reduction in both energy and raw material requirements. Economics will depend on availability and cost of cullet.

Commercial
SORG® Flex Melter

According to SORG®, these furnaces are more energy efficient than other furnaces of similar size but quantitative savings potentials are not reported .

Commercial
Oscilating Combustion for Glass ProductionDemonstration
Improved Refractories and Insulation

According to one advanced refractory material producer, their products can reduce the fuel budget by 1 – 1.5% with upgrading the crown insulation, and by up to 4.5% when the crown insulation is newly applied (Worrell et al., 2008. p.61).

Typical energy savings of about 7% can be achieved by replacing refractory bricks in the regenerator by specially shaped fusion materials ((Worrell et al., 2008. p.61).

 

Quantitative information not available. 

Quantitative information not available. 

Commercial
Tall Crown Oxyfuel Furnaces

Two other furnaces have reached campaign lives of 7 and 5 years. The energy intensity is comparable to that of a conventional oxy-fuel furnace and has reached 3.48 GJ/ton for flint glass with 60% cullet (Worrell et al., 2008. p.67).

Not available

Not available

Commercial
Fuel Selection

For a given pull-rate and with the common technologgy, use of fuel oil is considered to be 7-8% more efficient than natural gas. 

As natural gas has a higher carbon to hydrogen ratio, for a given pull-rate its use can result in 25% less CO2 emissions as compared to fuel-oil (IPTS/EC, 2013. p. 313)

Commercial
End Fired Furnaces

Compared to cross-fired furnaces, thermal efficieny is around 10% higher (Worrell et al., 2008. p.63)

In a Croatian green soda lime plant, replacing the old melter with 150 tpd capacity and 60 m2 melting area with an end-fired melter reduced energy consumption by 25 to 30%, giving a final average energy consumption of 4 GJ/t-glass (Worrell et al., 2008. p.63).

In a UK based plant producing flint glass, replacing the old furnace with an end fired regenerative furnace and implementing a range of other measures – including enclosing doghouse, increasing and improving insulation, sealing burners, and increasing glass bath – reduced energy consumption by 12.2% (Worrell et al., 2008. p.63).

Compared to cross-fired furnaces, investment costs are around 20% lower (Worrell et al., 2008. p.63)

In the Croatian plant, installing a new end-fired melter instead of repairing the old melter provided a payback time of one year. In addition to energy savings, improvement in total tank loads across the plant contributed to savings.

At a flint glass manufacturing facility in the UK, implemented measures costed $305 000 and produced annual savings of $507 000, giving a payback time of 7 months. Majority of the savings were connected to eliminated electric and oxygen boosting. For plants without electric and oxygen boosting, the payback time would have been 16 months (Worrell et al., 2008. p.63).

Commercial
Replacing Electricity with Fuel Firing

An audit performed in a speciality glass plant in the US found that the conversion of an all-electric furnace to a combined electric melter with a gas-fired batch preheater would result in a reduction of electricity use by 9 GWh/year at additional gas use of 3.52 GWh/year (12 000 MMBtu/year) (ITP/US DOE, 2004. p.3)

At 2003 energy prices, the conversion of an all-electric furnace to a combined electric melter with a gas-fired batch preheater would have resulted in savings of $208,000/year and a payback period of 1.2 years (ITP/US DOE, 2004. p.3)

Commercial
Premix Burners

Depending on the reduction of excess air they enable, premix burners may provide energy savings as much as 11% (Worrell et al., 2008. p.60).

Commercial
Recuperative Oxy Fired Furnaces - CO System ®

The energy consumption of the CO System® is comparable to that of oxy-fuel fired systems (Worrell et al., 2008. p.66).

Commercial
Porous BurnersResearch
Recuperative Furnaces

Energy savings up to 30% compared to cold-air furnaces (Worrell et al., 2008. p.62).

Payback times are estimated between 7 to 14 months [based on a fuel cost of US $ 4.3/GJ] (Worrell et al., 2008. p.62)

Indian flag In an Indian plant a recuperation system was installed for Rs. 0.3 million, and provided savings of Rs. 0.88 million per year - resulting in a payback time of 5 months (CII, undated. p.144). 

Commercial
Top Heating Electric Furnaces

Net energy savings are estimated at 4% (Worrell et al., 2008. p.70).

The payback period was 1.3 years for the plant operating in Sweden (Worrell et al., 2008. p.70). 

Commercial
Optimizing Excess Air

Generally savings of around 10% can be achieved, depending on the current practice . (Worrell et al., 2008. p.59)

Quantitative information not available

Quantitative information not available

Commercial
Oxygen Enriched Air Staging

Energy savings arise due to reduced need for flue gas treatment to remove NOx. 

Not available

Not available

Commercial
Reducing Residence Time in the Melter

 

Quantitative information not available. 

 

Quantitative information not available. 

Research
Using Sealed Burners

Reducing the cold air by 5% will provide energy savings of 2 – 3% (Worrell et al., 2008. p.61).

A plant in the UK was able to reduce gas consumption by 1% with the installation of burner sealing rings.

A plant in the UK was able to reduce melting costs by 1.75% with the installation of burner sealing rings. The payback time was around 4 months (Worrell et al., 2008. p.61). 

Commercial
Batch and Cullet Preheating

Specific energy savings of between 10 and 20 % can be achieved (IPTS/EC, 2013. p. 319)

EU flag By installing a cullet/batch preheating heat exchanger, a packaging glass company in Netherlands was able to raise the temperature of cullet to 277 °C, which reduced electric boosting by 60 (or 90 kWh/t) and natural gas usage by 8% (or by 0.32 GJ/t) (Worrell et al., 2008. p.68).

EU flag For a 350 tonnes/day cross-fired regenerative furnace, the additional investment cost associated with the use of the preheater is about € 2.5 million, including some adaptations in the batch-charging machinery. The annual operational cost savings are about € 820 000 per year, assuming a fuel price of € 9.4 per GJ gross combustion value. The average cost savings during the furnace campaign are estimated at € 3 per tonne molten glass, calculated on the basis of 2013 energy prices (IPTS/EC, 2013. p.320).

EU flag The application of batch/cullet preheating to a 450 tonne/day furnace allows for an increase in pull capacity from 450 to 500 tonnes/day and for saving energy. The investment costs are € 3.4 million and cost savings (based on a 500 tonne/day capacity) are € 1.1 million per year. In this case, the payback time is three years. The total cost savings are equivalent to € 5 – 6 per tonne glass, partly due to the increased melting capacity of the furnace without the need of enlarging its structure (IPTS/EC, 2013. p.320).

EU flag Installation of a cullet/batch preheating heat exchanger in a a packaging glass company in Netherlands in 1996 costed $ 1.4. The project had a payback time of 2.6 years, based on natural gas prices of $4.11/GJ and electricity prices of $0.05/kWh (Worrell et al., 2008. p.68).

Commercial
Oxy-fuel Furnaces

On-site energy savings can be greater than 50 % when small, thermally inefficient furnaces are converted to oxy-fuel firing.
For a medium-sized recuperative furnace with no specialised energy saving measures, standard levels of insulation, and using only internal cullet, the energy use with oxy-fuel melting would be in the region of 20 – 50 % lower (IPTS/EC, 2013. p. 231).

Even for large efficient regenerative furnaces, savings would be between 5 and 20% (Worrell et al., 2008. p.64)

EU flag A container glass manufacturer in Germany was able to reduce energy consumption from 5.02 MJ/kg to 3.02 MJ/kg (including energy consumption for oxygen generation) and realized energy savings of 35% by installing an oxy-fuel furnace with preheater (Worrell et al., 2008. p.65). 

The capital costs of a new oxy-fuel furnace are around 20% lower compared to recuperative furnaces, and 30-40% lower compared to regenerative furnaces. The costs of the on-site oxygen plant are about 10% of the capital costs of the plant. Oxy-fuel furnaces also benefit from reduced costs for flue gas treatment (Worrell et al., 2008. p.64).

However, the necessity to install extra durable refractory material (high duty silica crowns may increase the capital costs by €300 000 to 400 000 and use of fused cast materials may increase the costs by €4–5 million (IPTS/EC, 2013. p. 236). Operation costs will increase.

Commercial
Segmented Heater

Quantitative information not available. 

Quantitative information not available. 

Research
Oxygen Bubblers

Oxygen bubblers can increase the heat exchange efficiency by 10- 15%. With the introduction of oxygen bubblers (at the design stage) a plant in Netherlands eliminated the need for electric boosters for their furnace and saved 4 million kWh of electricity per year – equal to 170 kWh/t-product (Worrell et al., 2008. p.60).

Quantitative information not available. 

Quantitative information not available. 

Commercial
Electric Melting

State of the art electric melters consume 2.81 to 2.88 GJ/t of soda-lime or sodium-borate glass (Worrell et al., 2008. p.69)

Thermal efficiency of electric furnaces are 2 to 4 times better than air-fuel-fired furnaces (IPTS/EC, 2013. p. 175)

Emission reductions will depend on the nature and efficiency by which electricity is produced and supplied to the site. 

Electric furnaces have lower capital costs, but have shorter campaign life (2 to 7 years compared to 10 to 20 years for conventional furnaces) and higher energy costs (In the EU, average electricity costs per unit of energy are 4 to 5 times the cost of fuel oil) (EPTS/EC, 2013. p. 175)

Commercial
Regenerative Furnaces

Regenerative furnaces can have thermal energy efficiency of up to 65% (US DOE, 2002. p.55)

Energy savings are correlated with the preheated combustion gas temperature. For example, with a combustion air temperature of 800 ºC gives energy saving of 35%.

Multi-pass regenerators can reduce the energy intensity of the furnace by 15% (Worrell et al., 2008. p.63).

Commercial
Adjustable Speed Drives on Combustion Air Fans

US flag In a US based plant, a potential to reduce electricity consumption by 800 000 kWh/y with the use of ASDs has been identified (Worrell et al., 2008. p.60). 

US flag

 In a US based plant, the payback time for installing ASDs was estimated to be 1.7 years. 

The savings and the payback time will depend on the operating conditions of the fan system and the size of the furnace (Worrell et al., 2008. p.60).

Commercial
High Luminosity Burners (Oxyfuel Furnaces)

Thermal efficiency increases by approximately 4% (Worrell et al., 2008. p.66).

Commercial
Waste Heat Boilers

Between 1500 kWh/h (for end-port regenerative furnaces) and 6472 kWh/h for (side-port regenerative furnaces) can be recovered in container-glass production.

Investment costs can exceed EUR 1 million with variable payback periods, depending on performance and prevailing energy prices.

Commercial
Vertically Fired Furnaces

This technology can supply more energy per m2 of melt surface area, without increasing the refractory temperatures beyond operational limits. 

Commercial
Optimized Placement of Electrodes - Electric Furnaces

 

Quantitative information not available. 

 

Quantitative information not available. 

 

Quantitative information not available. 

Commercial
Reduced Air Leakage

In a plant in the UK, sealing the doghouse in a furnace reduced energy consumption by aroun 2.5% (Worrell et al., 2008. p.59).

Depending on the natural gas prices, typical payback of reduced air leakage for large glass operators who do not use oxygen and electric boosting would be less than 1 year (Worrell et al., 2008. p.59).

Commercial
Regenerative Oxy Fuel Furnaces

Test results show that heating incoming oxygen to 104 C results in energy savings of 15% compared to traditional oxy-fuel furnaces (Worrell et al., 2008. p.66)

Not available

Not available

Commercial
Submerged Combustion Melting

Energy savings are estimated at 5-7.5% when compared to a state-of-the-art oxy-fuel furnace, and depend on the utilization of heat losses from the furnace wall (Worrell et al., 2008. p.74).

Energy savings are more than 20% as compared to oxy-gas melters (Rue, 2005. p. 7)

 

More than 55% reduction in capital costs are estimated (Rue, 2005. p. 7) 

Demonstration
Using Low-NOx Burners

The LoNOx® Melter with cullet preheating consumes 3.23 GJ/t of energy, 14.4% less than regenerative end-fired furnaces. 

Cleanfire burners are reported to reduce energy consumption by 40% and 20% compared to recuperative and regenerative furnaces, respectively. 

For one manufacturer WideFlame™ has saved 6.1% of energy over previous models. 

The energy savings that can be achieved with EsDeNOx system are up to 5% for natural gas fired furnaces and range from 2 to 7% for oil-fired furnaces (Worrell et al., 2008. p.62).

Quantitative data not available.

The payback time for EsDeNOx system is reported to be less than one year (Worrell et al., 2008. p.62).

Commercial
Raining Bed Batch and Cullet Preheater

By preheating the batch/cullet to up to 538 °C, the technology is able to recover appx. 528 MJ/t-glass.

Fuel and oxygen consumption can be reduced by up to 25% in oxyfuel furnaces (ITP/US DOE, 2007). 

With preheat temperatures of arond 450 °C, payback times between 1 –4 years is expected (Worrell et al., 2008. p. 69)

Demonstration
Oxy Fuel - Synthetic Air

Fuel savings are at least 15% compared to conventional air-fired systems (Worrell et al., 2009. p. 64).

Commercial
Plasma Melter

Tests with High Intensity Plasma Melter in fibre production demonstrated 50 – 70% improvement in energy use compared to direct firing (specific energy requirements of 4.33 GJ/t are reported for the technology, compared to 8.88 GJ/t which is industry average in the US) (GMIC, 2004. p. 167).

System costs are estimate at US $500 000 to $700 000 (GMIC, 2004. p. 167)

Research