Reducing Heat Loss in Secondary Aluminium Casthouses: Furnaces, Feedstock and Insulation

Secondary aluminium production avoids the energy-intensive reduction of alumina required for primary metal production. Nevertheless, remelting and casting still require substantial energy, and heat can escape through furnace exhausts, cooling systems, refractory linings and material transfers. Identifying these losses helps casthouses target improvements without confusing furnace efficiency with the carbon footprint of the finished product.

This article reviews a KTH Royal Institute of Technology master’s thesis by Wenjing Wei, conducted in collaboration with SAPA Heat Transfer in Sweden using 2011 plant inventory data and on-site process investigations. These are historical case-study findings, not Firebird measurements or current industry-wide benchmarks. The insulation discussion adds practical material-selection considerations; it does not represent a Firebird retrofit tested at the SAPA plant.

1 Understanding the process and its boundaries

The casthouse route covered charge preheating, melting, holding, casting, sawing and slab homogenisation. Its feedstock combined primary aluminium ingot, internal production returns and externally sourced scrap. Unlike primary aluminium production, remelting does not require electrochemical reduction of alumina, but feedstock preparation and subsequent thermal processing remain important energy demands.

Clean internal returns can often be remelted with limited preparation. Externally purchased scrap varies in composition, cleanliness and physical form; its preparation requirements depend on the material received. Melting and alloy adjustment are followed by melt treatment and casting, with homogenisation required for relevant slab products.

Two different accounting boundaries must be kept separate. A furnace energy balance tracks energy supplied to the charge and losses from that equipment. A product carbon assessment also considers upstream materials and energy production. Results at the melting-furnace outlet cannot be treated as the footprint of a finished, homogenised slab.

Although described as cradle-to-gate, the thesis applies recycling-related allocation assumptions, including 95% recycling and a recovery burden. It also excludes some inputs and external transport. Its carbon figures are therefore method-specific, not directly transferable to a present-day product declaration.

2 Where energy leaves the casthouse

The study reported approximately 67,000 tonnes of slab production in 2011 and an energy consumption of 3,826 MJ per tonne of slab. Electricity supplied about 61% of the plant’s energy input and fuel about 39%. These figures describe the studied operation and its accounting boundary, rather than a universal target for secondary aluminium plants.

The main loss pathways were exhaust gases, cooling water, furnace surfaces, material transfers and hot process residues. Each requires a different response. Exhaust heat may be recovered for suitable preheating duties. Casting necessarily removes heat from solidifying metal, so the opportunity is to recover usable heat rather than eliminate cooling. Furnace surface losses can be addressed through lining design and maintenance, while transfer losses depend strongly on timing and distance.

The distinction matters when setting an insulation target. Improving a furnace wall cannot directly recover heat discharged through casting water or an exhaust stack. Savings should be estimated from the loss pathway being changed, not from the plant’s total energy consumption.

3 Furnace efficiency and carbon performance

Induction melting

Induction furnaces heat metal electrically and avoid combustion exhaust from the melting process. In the studied equipment, however, cooling of the inductors was a major energy-loss pathway, alongside furnace surface and residue losses. The reported comparison placed induction furnace efficiency at approximately 60–61% under the study’s calculation assumptions.

Oxy fuel melting

Oxy-fuel combustion reduces the nitrogen introduced with combustion air and therefore reduces the volume of gas that must be heated and exhausted. The SAPA installation also used electromagnetic stirring to improve heat transfer within the melt and assist melting of the charge.

The thesis reports useful melting heat of approximately 66–69% for the compared oxy-fuel furnace, increasing to 68–71% when the oxygen-vaporisation heat-recovery contribution is included. Efficiency comparisons therefore require consistent treatment of recovered heat and auxiliary equipment.

Higher thermal efficiency did not automatically mean lower carbon emissions. The study used a very low electricity emission factor of 6.5 g CO₂e/kWh for the plant’s nuclear electricity supply, whereas propane combustion generated direct emissions. This favoured induction melting in the carbon assessment.

The results do not establish a universal furnace ranking. In particular, the two AA3003+Zn cases used different primary-ingot proportions, so their footprint difference cannot be attributed to furnace type alone. A new equipment assessment should compare equivalent feedstock, metal yield, output temperature and production duty, using the actual electricity supply and fuel emission factors.

4 Feedstock changes the carbon result

The case study illustrates how upstream material burdens can substantially influence the carbon result even when the remelting operation is efficient. Its selected melting cases are summarised below.

These are historical, method-specific melting-stage results, not full finished-slab footprints. The lowest value depends on the study’s treatment of internal scrap and its low-carbon electricity assumption; it does not mean that recycling has no energy demand or environmental burden.

For casthouse operators, feedstock optimisation must balance recycled content with alloy chemistry, contamination, metal recovery and melt quality. A lower primary-ingot addition can reduce the allocated upstream burden, but a poorly controlled scrap mix can introduce other processing demands. Avoidable internal scrap generation should also be reduced, since remelting production returns consumes energy even when their incoming carbon burden is assigned as zero.

Data qualification: the thesis contains inconsistent aggregate carbon totals and contribution shares. Accordingly, its plant-wide footprint and percentage breakdown are not adopted here as verified benchmarks. The melting-case table above is retained to illustrate the interaction between feedstock, energy source and accounting assumptions.

5 Controlling heat loss through the furnace lining

Lining improvements address heat conducted through furnace walls, roofs and doors, together with defects that create local hot spots. They complement combustion control, heat recovery and production scheduling. Material selection should distinguish the working lining, which must withstand the process environment, from the backup insulation, which limits heat flow towards the casing.

Furnace walls and holding furnace backup insulation

In areas exposed to molten aluminium, the working lining must be selected for the alloy, chemical exposure and mechanical duty. Backup insulation sits behind that protective lining and should not be treated as a substitute for it.

Microporous insulation board is an option where low heat flow is needed within limited lining space. Its suitability depends on the calculated temperature at the board’s hot face, the available thickness and installation conditions. A furnace’s metal temperature alone is insufficient for selection. Joints, anchors and penetrations also need attention because they can interrupt the insulation layer or provide local heat-conduction paths.

Holding furnaces merit particular attention because heat escapes throughout the holding period. Reviewing the backup lining can help reduce this continuing loss, but burner operation, air leakage and unnecessarily long holding times remain separate improvement opportunities.

Furnace roofs and upper walls

Roof and upper-wall linings must accommodate their operating atmosphere, thermal cycling and mechanical support requirements. Ceramic fibre products can be considered for suitable non-metal-contact areas and backup duties. Standard fibre insulation should not be assumed suitable for direct molten-aluminium contact; splash, vapour and flux exposure require specific review. Manufacturer application guidance also distinguishes specialised non-wetting products from general insulation materials.

Adding insulation changes the temperature profile through the entire lining. Check the resulting temperatures of intermediate layers and anchors, rather than judging the design only by a lower predicted shell temperature.

Furnace doors and openings

A well-insulated closed door still loses heat when it remains open. Door repairs, suitable seals and reduced opening time therefore need to be considered together. Radiation losses depend on the exposed opening, the temperature of the surfaces facing it and the duration of exposure. Openings can also allow cold-air infiltration or hot-gas escape.

Reducing unnecessary charging and inspection delays can complement a lining upgrade. Fibre-based seals and door insulation must be selected for the local temperature, movement, abrasion and chemical exposure rather than specified solely by nominal temperature classification.

Assessing the benefit

Before changing the lining, record operating temperatures, shell-temperature patterns, lining construction, holding time and energy consumption. Review abnormal hot spots for possible lining damage before treating them as a routine insulation issue. Compare performance after the change under similar production conditions; lower shell temperature by itself is not a measured percentage reduction in total plant energy use.

6 Retaining recovered heat and setting priorities

The SAPA case provides a useful example of losses between operations. Material leaving preheating at approximately 196°C was reported to cool by about 160°C before charging. Recovering exhaust heat achieves limited benefit if the warmed charge then waits long enough to lose most of that energy.

Coordinate preheating with furnace availability and review unnecessary transfer delays. Assess heat-recovery opportunities against the temperature and timing of an actual heat demand. Furnace cooling systems must continue to meet their equipment requirements; heat-recovery proposals should not compromise cooling performance.

A practical improvement programme links each measure to a measured loss: lining work for surface losses, door management for opening losses, process coordination for transfer losses, and appropriate recovery equipment for exhaust or cooling-water heat. Feedstock and energy-supply choices should then be assessed separately for their effect on the product carbon footprint.

Firebird insulation support for aluminium casthouses

Firebird supplies microporous insulation boards, ceramic fibre products and insulating firebricks for suitable refractory and backup-insulation applications. Material selection should follow the duty of each furnace zone, with direct-metal-contact areas assessed separately.

Planning a melting or holding furnace lining upgrade? Send Firebird your operating temperature, existing lining build-up, available insulation thickness and target shell temperature. We can discuss suitable materials and provide technical data for your evaluation.

Website: www.firebirdref.com

Email: service@firebirdref.com