Your melting furnace is running great. The melt rate is good. Maintenance on the flue and the doors is a bit high, but it doesn’t affect you too much. However, energy costs are high. You’ve checked the air-fuel ratio and furnace pressure, but nothing seems to help. You may have a furnace that is over-fired.


Many aluminum melting and holding furnaces are designed with a burner fire rate that is higher than it needs to be. Furnace designers want to make sure the furnaces will melt at the stated rate. Yet this higher fire rate does not necessarily increase the melt rate. Reducing the fire rate to optimum doesn’t cost anything, but it can save you 5-15% on energy. While this paper discusses aluminum melting and holding furnaces, the same principles can apply to other furnace types.


Firing Density

Furnace designers increase the fire rate on furnaces as a safety factor to meet the specified melt rate. After all, the purpose of a melting furnace is to melt metal and melt it quickly. However, there is an optimum burner fire rate to melt metal. The metal surface only accepts heat so fast. Many factors go into this heat transfer including the residence time of the products of combustion (POC), insulation of the bath surface with metal oxides (dross) and metal stirring transferring the heat away from the bath surface. Too low of a firing density means a slow melt rate. A higher firing density improves melt rate up to a point. After an optimum level is reached, the extra heat is lost up the flue. Symptoms of over-firing include high flue temperatures, high maintenance costs on the flue and stingers around the door.

    Firing density is defined as the fire rate (BTU/hour) divided by the bath surface  area (feet2) exposed to POC. It does not include exterior bath surfaces such as wells. For example:


•   Firing rate = 24 MM BTU/hour

•   Interior bath length = 18 feet

•   Interior bath width = 14 feet

•   Firing density = 24,000,000 BTU/hour /(18 feet x 14 feet) = 95,238 BTU/hour/feet2


    In the 1970s, a firing density of 100,000 BTU/hour/feet2 was normal for furnace design. In 1982, a paper[1] mentioned that a firing density of 80,000 BTU/hour/feet2 was found to be more energy efficient while still obtaining a good melt rate. In 2006, a study using computer modeling and a prototype furnace concluded a value of 66,000 BTU/hour/feet2 was the optimum point.[2]

    This study found several benefits from a reduced fire rate. Reduced heat loss from the flue gases was the primary benefit. The lower fire rate also increased POC residence time to allow increased heat transfer to the load and furnace. Of course, these two items are interrelated. When residence time is low, heat is not transferred to the metal and walls, so flue temperatures are higher and flue heat loss is high. Finally, modeling showed that controlled soot production optimized radiative heat transfer.

    In the study, firing density was varied while both melt time in hours and specific energy (BTU/pound) was monitored (Fig. 1). It was found that the drop in melt time leveled off at a value of 66,000 BTU/hour/feet2. Even an increase of 35% in the firing density only decreased melt time by another 7%. However, the higher firing density meant energy use increased by 26%.

    While modeling and laboratory studies are good, real-world results are better. A reduction in firing density has been made in multiple furnaces at different plants. In-plant testing on nine different furnaces showed 5-15% energy savings without changing the melt rate (Fig. 2). Another paper states a 14% reduction in flue energy loss using this technique.[3] Another test showed 10% reduction in energy use with no reduction in melt rate. Figure 3 shows energy use plotted versus production for that test.

    In a 2010 study, over 100 reverberatory furnaces with multiple manufacturers were analyzed.[4] It was found that 80% of the melting furnaces were designed to run over the recommendation of 66,000 BTU/hour/feet2. The majority of holding furnaces were also found to exceed the recommended rate. While this means the chances are your furnace is over-fired, it also means that many furnaces run at the recommended fire rate and perform well.

    The steps to making such a change are:

1. Calculate the “ideal” fire rate. This is [bath area (feet2) x 66,000] / 1,000,000 = fire rate (MMBTU).

2. Make the change to the burner fire rate in steps with several weeks between each step to make sure improvement is seen without affecting melt rate.

3. If possible, drop the fire rate below the calculated value.

4. Once a point is reached where melting is affected, move the fire rate back up a step.


Holding Furnaces

A high fire rate is not needed for holding furnaces. It has been said that the high fire rate should be exactly the amount required to keep the bath at temperature and no more. To leave some capability to raise the bath temperature, however, a higher fire rate is normally used. Varied values have been found in literature and practice, but a value of 20,000-40,000 BTU/hour/feet2 is appropriate in most cases.


Adjustable Fire Rates

For many furnaces, the metal surface is the simple bath area. But batch furnaces may require a higher fire rate since the metal surface is not a flat bath but actually a mound of complex shapes in the form of solid charged metal. In furnaces such as a round-top, values of 100,000 BTU/hour/feet2 have been used as the recommended starting firing density.

    As the mound of charged metal begins to melt down, the surface area decreases and eventually changes to a flat bath. Lower fire rates are appropriate as the surface area decreases. At the same time, the temperature differential between the flame and the bath surface lowers. On both batch and continuous furnaces, dross begins to form on the surface, which increases insulation. All of this means a lower fire rate is needed since heat cannot be absorbed as quickly.

    This can be done in one of three ways. Some companies have programmed the burner fire rate to drop over time. This is sometimes done in steps using a user-developed program.[5] Others drop the fire rate a certain percentage when the flue temperature reaches a setpoint.

    The last method is to use cascade logic. When the furnace is cool after a charge, bath temperature controls the burner. As the furnace heats up (and the charge melts down), the flue temperature rises. Cascade logic begins to lower the fire rate as the flue reaches a certain setpoint. While cascade logic is used in many furnaces (sometimes with roof temperatures as the second input), the temperature is set high in many furnaces so that the firing rate rarely lowers. Adjusting this setpoint lower and using an appropriate PID setting will allow the fire rate to gradually decrease as the charge melts. Testing showed a savings of 2-3% with this change.[6]



Lower firing rate is a simple and low-cost method to reduce energy costs in the cast house. While a quick calculation highlights furnaces that may be over-fired, the best method is running a test. Burner velocities, burner placement, furnace configurations, refractory designs and the stirring process will all alter the optimum level. This technique has been used to obtain 5-15% energy savings in multiple furnaces. IH


For more information:  Cindy Belt is an energy-management consultant located in Pensacola, Florida.  She can be contacted by e-mail: and found at LinkedIn:



1. John A. Marino, “High Efficiency Aluminum Melting, European Experience and Background,” presented at the IHEA Energy Seminar, 1982

2. C. Belt, B. Golchert, P. King, R. Peterson, J. Tessandori, “Industrial Application of DOE Energy Savings Technologies to Aluminum Melting,” Light Metals 2006, Travis J. Galloway Ed., TMS, 2006, pp. 881-885

3. M. Roy, V. Goutiere, C. Dupuis, “Implementation of a Global Casthouse Furnace Energy Efficiency Program at Rio Tinto Alcan,” Light Metals 2010, J. Johnson Ed., TMS, 2010, pp. 669-673

4. C. Belt, R. Peterson, and D. Bequette, “Five Low Cost Methods to Improve Energy Efficiency on Reverberatory Furnaces,” Energy Technology 2010: Conservation, Greenhouse Gas Reduction and Management, Alternative Energy Sources, Neale R. Neelameggham Ed., TMS, 2010, pp. 71-79

5. J. Wang, Y. Zhou, H. Yan, J. Zhou, “Computational Analysis of Thermal Process of a Regenerative Aluminum Melting Furnace”, Light Metals 2014, J. Grandfield Ed., TMS, 2014

6. C. Belt, “Energy Efficiency Tests in Aluminum Combination Melting and Holding Furnaces,” Light Metals 2004, Alton T. Tabereaux Ed., TMS, 2004, pp. 613-617