How AI/ML helped a biomass power plant maintain boiler pressure?

The Biomass that can be used for the biofuel or electricity generation are animal waste, agricultural residue, industrial waste, forest residues, solid waste, and sewage, etc. The power generation through biomass can be focused on and is never ending process as biomass is a renewable source of energy. So, improvisation and adopting advanced technologies in the biomass power plant can be considered as great initiative for increased fuel generation. The process involves certain stages that can be digitalized for better production of the biofuel.

The amount of energy or fuel generated in the Biomass power plant depends on multiple factors such as moisture content of feed, temperature, flow rate, pressure created in the boiler, etc. The process involves various stages or unit starting from feeding the biomass in combustion chamber through the conveyor. Biomass can be agricultural waste, forest waste, animal waste, industrial waste, sewage, etc. which depends on the industry. The combustion depends on the moisture content of the feed and certain other factors. After the combustion comes the monitoring of various parameters in the boiler that has some predefined values that needs to be maintained for maximum efficiency. Now, the steam generated is used for rotating the turbine, there by converting mechanical energy into electrical energy. So here the important unit and parameter to be concentrated is the boiler where the steam is generated.

Though a lot of factors such as temperature, flow rate, etc is responsible for steam generation, pressure created in the boiler plays an important role in the entire biomass plant and it must be maintained under certain limits for the efficient performance of the boiler. The increase in the pressure may lead to leakages and the decrease in pressure may affect the performance leading to low productivity. The pressure and temperature in the boiler were initially maintained manually after the occurrence of its increase or decrease represented by the sensors. The precautions taken after the occurrence in change results in fuel loss, increased probability of lower equipment life and continuous manpower to operate the plant.

The trending technology that is IIoT has helped a lot in increasing the efficiency of the plant by providing various solutions and predicting the failures before its occurrence. In case of Biomass power plant, the pressure in the boiler is maintained in a certain range to avoid any kind of failures with the help of AI and ML. The historic data of the plant and its relativity with other parameters such as temperature, flow rate, moisture content of biomass fuel, etc is studied through the AI algorithms and then the solutions are provided depending on it that predicts the pressure abnormalities and identifies the pattern.

The study of the historic data from various equipment through AI/ML algorithms has led to the predictions of pressure abnormalities, fluctuations that may occur in the boiler due to various factors. Rubus Platform is one such IIoT that uses AI/ML to identify the abnormalities and provide indication before it happens. Rubus platform is Industry 4.0 solution that provides different modules such as Trend Analysis, Rule Engine, Predictive Analysis, condition-based monitoring, and many more that can enhance the plant performance.

To be specifically concentrating on the boiler performance, we can take the live example or the use case study of the solutions provided by the Rubus platform to Kokusai Pulp & Paper Co. Ltd. Rubus has provided various modules and number of tags or the parameters that are monitored which includes the temperatures, pressure, water levels and steam generation at different levels. The Trend analysis module provided identifies the pattern and represents it either graphically or through dashboards. Rule Engine provided checks the conditions provided for certain limits in temperature and pressure and provides notification depending on the condition it satisfies. For example, here, the alerts were created 11-20 mins prior to any abnormalities that may be increase or decrease in pressure. Depending on that the operator acts accordingly.
The advantage of such prediction can be explicitly said to be saving in the fuel for steam generation, maintained pressure in boiler ultimately leading to good equipment life, less manpower required if alerts are created. All the above advantages ultimately lead to efficient plant performance and cost optimization.