Getting the most out of operational data in energetics

Both energy and environment fields are of top priority in the current and strategic European Union policies. These sectors imply complex approaches and consistent investment in order to keep pace with depleting resources, converting to renewables and reduce the environmental footprint of anthropic activities. Operating power production facilities leads to huge amounts of installation data. The paper presents a possible alternative to large investments to improve the efficiencies of existing installations: namely the use of datamining techniques for analysing existing operational data. Data-mining is emerging as a valuable concept based upon detailed analysis of operational records with advanced mathematical/statistical tools. Outcomes are, e.g.: evaluating the consistency of measurements, identifying new hardware needs, internal/external benchmarking, correlation of consumptions with power/ heat production, pointing out causes of high environmental impact, maintenance scheduling, stock optimization, analysing scenarios for future development and decision support for local managers, etc. The paper details the data-mining
carried out at the TERMICA – Suceava power plant for the period 2007-2012.
The analysis was carried out by a multidisciplinary, complex team (engineers of
various specialities, environmentalists, economists, mathematicians, IT
experts). Dta mining is much less expensive than rushing to investment in new
equipment. Participation of Company’s top management is fundamental, this
being the driving force and motivation source for experts and operators. The
approch presented is self learning.