Energy is the main cost in glass production, representing over 50% of the total costs of production. 80% of this cost comes directly from the fuels used to feed the furnaces that fuse the raw materials. In Vidrala this means a total amount of 3.488 GWh per year, with a tremendous cost in an upscaling energy costs scenario as the actual one. Apart from this direct cost that comes from the consumption of fossil fuels, the use of these fuels adds another cost related to CO2 emissions. All these factors made the company think about improving the efficiency of the production processes, optimizing the furnaces activity to use as little energy as possible without compromising quality.

This would mean not only a quite considerable reduction of energy consumption but also a reduction in CO2 emissions with the consequent saving in emission rights.

The solution aimed to save energy through the implementation of an in-plant process control system based on the deployment of new end-to-end control systems in the glass furnaces for the manufacturing process of the packaging using Artificial Intelligence and Machine Learning techniques. To this end, monitoring systems and IT tools were introduced to analyse the data collected to identify operational trends and propose possible solutions to the operators, so that they can take the necessary measures to adjust the parameters of the furnaces in order to reduce unnecessary fuel consumption and to prevent equipment or quality failures.

The strength of the system is that through continuous data collection and analysis of the results obtained, the system is supposed to be able to learn and propose constantly better solutions to the operators.

Other Case Studies