Characterization, Dynamic Modeling, and Monitoring of the Degradation of Hydroelectric Production Infrastructures

Hydropower infrastructure globally faces three primary challenges: aging infrastructure, climate change, and hydro-peaking. These issues result in increased degradation rates, with a higher degree of associated unpredictability. This preliminary research aims to identify a modeling approach that would inform an optimized maintenance plan within a host organization, to aid in ensuring the availability and good operation of hydropower assets, while balancing strategic production objectives with risks. The methodology for modeling the asset degradation phenomena must leverage how degradation mechanisms evolved historically for critical assets, considering condition monitoring data over time, to recognize trends in their health state and thus optimizing maintenance interventions, minimizing production losses. The work presented in this paper describes an investigation of the existing data and its sources, including experts' feedback, within the host enterprise, a review of the literature on dynamic modeling and the monitoring of degradation mechanisms, and an evaluation of potential degradation modeling methods that could be applied to two distinct assets that were selected as case studies: the spillway gate and the alternator. It is proposed that a model based upon a Bayesian Stochastic Petri Net (BSPN) would meet the desired criteria for a degradation model for asset management, allowing for refinement and adaptation over time as more data becomes available and as variable degradation drivers continue to evolve.

Keywords: Degradation modeling, Asset management, Maintenance decision-making, Hydroelectric powerplant, Critical infrastructure.