Non-intrusive identification of hydro power plants' dynamics using control system measurements

Sigurd Hofsmo Jakobsen (1) , Xavier Bombois (2, 3) , Kjetil Uhlen (1)

1 NTNU - Norwegian University of Science and Technology [Trondheim]
2 CNRS - Centre National de la Recherche Scientifique
3 MIS - Ampère, Département Méthodes pour l'Ingénierie des Systèmes

In this paper we propose two non-intrusive methods for identifying hydro power plant dynamics using standard system identification techniques. The purpose of the tests is to obtain transfer functions relevant for the frequency containment reserves delivered by hydro power plants. To validate the methods, we analytically demonstrate under which conditions the methods will yield consistent results. Moreover, using a simulation example we discuss the validity of the methods when one of the conditions is not met. Results using measurements from the control system of a power plant in the Nordic power system are also provided. Our proposed methods also allow us to identify a transfer function that can be used to check new requirements developed by the Nordic TSOs. The transfer functions obtained by our methods is compared with the transfer function obtained using the TSOs' methods using Monte Carlo Simulation.

Domaines: Automatique / Robotique

Sigurd Hofsmo Jakobsen, Xavier Bombois, Kjetil Uhlen. Non-intrusive identification of hydro power plants' dynamics using control system measurements. International Journal of Electrical Power & Energy Systems, 2020, ⟨10.1016/j.ijepes.2020.106180⟩. ⟨hal-02642471⟩