Smelting furnace: a technology to see and control what’s going on inside to increase throughput.

Electric smelting furnaces operating with high amperage (> 80 kA) are difficult to maintain in stable operation close to the electrical limits of their equipment (for instance in FeMn or FeSi production). 

The reasons have been identified for a long time:
- Large furnace with a low power factor (around the stability limit, 0.5)
- Inaccurate voltage measurements at the level of the furnace (downstream of the furnace transformer)

Inductive effects from the main currents are responsible for these measurement errors. 

Upstream measurement does not provide all the information necessary to fully solve the electrical equations, and high temperatures limit the possible techniques.

The state of the art technology uses the Bockman bridge to measure the voltage of each electrode relative to the metal bath. This technique relies on geometric and electrical assumptions that are poorly verified in practice, and calibration requires a perfectly balanced operation that is difficult to obtain and prove. 

Other systems exist and are based on even more arbitrary assumptions, such as the relationship between reactances.

In addition, there is no simple and reliable technology for measuring electrode length. The reactance or weight of the electrode column only gives an approximate value.

Today, there is no possibility to assess homogeneity and charge quality around each electrode.

Our self-checking device/technology measures real data without making any assumptions.

It thus provides access to precise, rich and new information on the distribution of currents inside the furnace. Now measurements are available for resistance, reactance, electrode length, homogeneity, charge quality and much more, including all electrical losses inside and outside the furnace, independence or interdependence of the reaction zones near electrodes, power lost due to the gap between the electrical neutral point and the metal voltage.

The new technology uses the accuracy of these measurements to ensure process stability based on multi-variable predictive control. Whatever the choice of control parameter, current, resistance, reactance, voltage or power, the regulation is stable and anticipates all drifts.

The technology rely on two main innovation. The first is a measuring and controlling technique who needed years and heap of dollars to develop in another application. The cost reduction of numerical technique make it now affordable for the industry. We have adapted it to smelting furnace and sized to get a macroscopic view of the furnace with 1µΩ accuracy on impedance. The electric description is precise enough to independently solve the equations and get what the currents and voltages should be. The comparison with reality give a real time error estimate. First tests gave errors of less than 1% on instantaneous value and less than 0.3% on rms value.

Depending on the level of industrial performance of the sites, the increase in furnace throughput is estimated at 10% to 30% and gives many opportunities for further improvements to the metallurgists who manage the furnaces.