Data and signal analysis in a biomass electricity generation plant Diagnostics and optimisation
Client brief
The plant in question, monitored with various sensors and measurements, recorded a decrease in energy production on an annual basis, which is critical in terms of return on investment. The data automatically acquired by the system are, in part, used for the control of the system itself, but in large part neither used nor analysed. The question that led to the commencement of an internal research project was whether it was possible, from the data acquired during four years of operation, to identify possible causes of the lower energy production.
Project challenge
Addressing the issue required an extensive analysis of the plant and the data sources, the involvement of an interdisciplinary team, and continuous interaction with the plant operator. Furthermore, it was essential to base the activities on a scientific research method, with bibliographic studies, experiments, identification of data management and analysis methodologies and signal analysis.
Execution phases
Study of plant functionality and management procedures
Organization and cleaning of data, signal pre-processing
Research correlations and identification of events
Plant shutdown assessments starting from the data
Problem identification on data and sensors
Estimation of non-measurable variables
Identification of KPIs
Design of an interactive interface for data visualisation
Achieved results
Management and plant data acquisition problems were identified. A prototype interactive dashboard was created to visualise data and calculated Key Performance Indexes (KPIs) in order to better monitor the criticality and efficiency of individual plant components. The intervention team made recommendations on how to optimise plant management and monitoring (measurement reliability, optimisation of maintenance to maximise plant start-up), and tools were developed that would also support the plant manufacturer to increase the number of plants monitored and thus the statistical significance of the data.
Efficiency derived from measurements and estimates
Interactive dashboard page for monitoring