Development of risk assessment model for biopower plant boiler using Bayesian network

• Main Authors: Alaw, Fares Ahmed, Sulaiman, Nurul Sa’aadah Binti
• Format: Conference Proceeding
• Language: English
• Subjects: Alternative energy sources; Bayesian analysis; Biogas; Biomass burning; Biomass energy; Boilers; Clean energy; Conditional probability; Cooling systems; Electric power generation; Fuels; Occupational health; Occupational safety; Oxygen; Power plants; Risk assessment; Risk based inspection; Risk management
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0188281

The growing popularity of the biomass power generation process is connected to its label of green and cheap renewable energy. The enhancement of empty fruit bunch (EFB) is increasing following its usage for power generation; however, the assessment of potential occupational health and safety (OH&S) considerations associated with the usage of biomass fuels in combustion-based generation remains restricted. One of the main pieces of equipment in fuel-fired power plants is the boiler for the biomass power plant. Unified system management concepts and methods that may accurately assess equipment risk and save operating costs pursue security and economy through risk-based inspections. EFB biopower boilers do, however, pose a number of major concerns and health and safety problems. As EFB has around 67% moisture, it needs to be dried thoroughly before it can be used as fuel in boilers. This study used the Bayesian network (BN), an inextricably well-known method for illuminating uncertain and complicated systems, to establish a condition and safety connection framework for biomass power production boilers. The conditional probability values were elicited to determine important factors that must be considered in efficient boiler operation. A case study from FTJ Biogas Power Plant in Jengka, Pahang was applied. Deaerator, economizer, fuel preparation, and water-cooling system are the biopower boilers’ primary risk areas. The network was built using the HUGIN programme. The states and input values for a CPT (Conditional Probability Table) of each node were specified after that. To pinpoint the informal linkages responsible for the system failure based on time, prediction, diagnosis, and sensitivity analyses of the Bayesian network were conducted. The results obtained from the proposed Bayesian Networks model indicated oxygen attack as the most dangerous risk that needs to be considered. Oxygen in water produces pitting that is highly severe because of its localized nature. Furthermore, the results will change simultaneously with each change in the percentage of probability. Thus, this framework may support decision-makers to decide when and where to take preventive or mitigate measures in the risk management process for biomass power plants.