Power Efficiency Tracking of Induction Furnaces

Induction melting is a very energy intensive and dynamic process. Therefore, improving the power efficiency is critical to reduce the cost of production. Saving of just a percent of energy can reduce the power bill considerably. There are several factors that contribute to power efficiency and hence difficult to control these parameters. The melting process and the furnace characteristics are unique to each melt and vary with the melt material and its physical properties. Therefore, developing a maximum power transfer method for induction melting is necessary. We have collected real time data from a 3KHz, 40KW coreless induction furnace for several heats. We have considered three different materials viz., Cast- iron, SG Iron and Wrought Iron to tabulate the data to be input to the process. The focus here is to show that the frequency of the system varies with the change in weight of melt instead of a predefined resonating frequency. This variation in frequency increases the melting time and therefore the power consumption. If frequency variations can be reduced, then the power consumption can be reduced. The paper shows that there is scope to improve performance of induction furnace by using a machine learning technique.