Termosun and Pervasive optimize biomass combustion for industrial boilers

Pervasive Technologies , a company specializing in the development of image recognition solutions using Artificial Intelligence (AI) for different industrial sectors, together with Termosun and Imae, in collaboration with Schneider , have unified resources and knowledge in an innovative research project for the optimization of the combustion of biomass and related by-products in industrial boilers through the application of Artificial Intelligence (AI) and other disruptive technologies such as Machine Learning and Big Data.

The worrying increase in the cost of fossil gas as a result of the current war in Ukraine, as well as the increase in the price of carbon credits, are causing many industries in Spain to begin the process of converting their thermal plants , replacing obsolete traditional gas boilers with new boilers that use renewable biofuels, such as forest biomass or agro-industrial by-products.

As we can see, there is a growing demand for biomass boilers in the food, automotive, chemical, and other industries , but Termosun has observed that operators and maintenance personnel at industrial facilities do not have sufficient skills for the optimal management of the boilers, so they do not take full advantage of their energy efficiency, nor do they manage to reduce polluting emissions as could be achieved with current equipment and technology.

The operation and maintenance required for biomass boilers is more complex than that of gas boilers , and this evolution must be accompanied by advanced data acquisition and management technologies, as well as more sophisticated control systems, to prevent boiler efficiency from depending solely on the operator. This ensures efficiency and increases the boiler's lifespan, while also protecting the environment by preventing excessive biomass consumption or malfunctions that could lead to atmospheric emissions.

The 3BD project, Biomass Boiler Big Data , was created to improve current algorithm models using Artificial Intelligence . Machine Learning, through the creation of digital twins and algorithms, allows for the identification of patterns in massive amounts of Big Data. This enables the development of predictions that allow for the digitization of boiler operation for optimal combustion, efficiency, and minimal emissions.

Under the concept of a toolbox, the project combines tools that are integrated into the boiler at different levels of scalability and licensing, the basic ones being:

• Continuous measurement of the parameters that occur in the different physical-chemical states during the combustion process
• Image capture of combustion grill inside oven
• Set of oxygen and temperature probes in exhaust gas line
• Massive data acquisition and interpretation platform
• Digital training model and data affinity and self-learning
• Operating parameter correction interface

The tools are arranged in different layers, starting with layer 0, also known as field hardware, and progressively scale up to the acquisition, interpretation, iteration and alteration of operating parameters, culminating in the corresponding report to the technical support service.