Application of real-time simulation for hydropower plants monitoring
The global emission of CO2 has increased in the sector of power generation from countries outside the OECD, particularly in China which comprises two-third of the share. Meanwhile, OECD countries focus their effort in reducing their emission, where the industry area declined their emission by a quarter. Consequently, governments set up energy policies where renewable-based energies such as solar or wind energies, are highlighted. The drawback is the dependency of environment factors which generates power variations in the network leading to instabilities and a blackout in the worst case. The use of hydropower is an excellent complement to versatile renewable energy, capable of compensating these fluctuations and even be used as an energy storage. In Switzerland, hydropower is the most available resources and is constantly under development to increase capacity and energy-efficiency. Nevertheless, a lot of effort is put to extend lifetime of power generating equipment as much as possible in order to optimize the best timing for replacement or refurbishment, which requires efficient and flexible tools such as real-time simulations, which are widely used with the rapid development of computation technology. The objective of this present thesis is to study the feasibility of a multi-physics model-based real-time simulation with SIMSEN for an existing hydroelectric power plant. The concept of such system consists of using a validated model and achieves a real-time simulation taking into account boundary conditions such as water level of upstream reservoir, voltage of power network, but also set point of control system including turbine guide vane opening and the excitation system of generator. The system would enable to detect potential dysfunctions if the behavior of the power plant shows significant discrepancies in the simulation-measurement comparison during the real-time simulation. The study is divided into two main parts, which are the set-up of a test bench and the implementation of the real-time system in the power plant. The first part describes the methodology for setting-up a small-scale power unit which simulates similar behavior as a large scale one. The model includes the parameters from experimental tests used for parameters identification of synchronous machine. Two validations tests are presented: (i) a load variation (ii) sudden three-phase short circuit. The model validation is achieved comparing offline simulation results with measurement, followed by a validation in real-time with similar tests. The set-up of the test bench is concluded by a demonstration of feasibility in monitoring application by detection of a dysfunction during an opening failure of a circuit-breaker. The second part explains the modeling of a 72MW hydroelectric power plant, which includes the modeling of hydraulic components such as pipe, surge tank and Pelton turbine besides the modeling of generator. Two measurement campaigns were organized. The first consisted in collecting data in order to validate the present model. The second consisted of implementing the model running in real-time in a general-purpose computer, in the power plant and performed variations of active and reactive powers.
Programme doctoral Energie
Faculté des sciences et techniques de l'ingénieur
Groupe de scientifiques STI
Jury: Prof. Mario Paolone (président) ; Dr Basile Kawkabani (directeur de thèse) ; Prof. Farhad Rachidi-Haeri, Prof. Carlos Antonio Platero Gaona, Dr Christophe Nicolet (rapporteurs)
Public defense: 2015-11-6
Record created on 2015-10-28, modified on 2016-08-09