1 Stability Study of a Mixed Islanded Power Network A.C. Padoan Jr., C. Nicolet, Member, IEEE, B. Kawkabani, Member, IEEE, J.-J. Simond, Member, IEEE, A. Schwery, Member, IEEE, F. Avellan Abstract—This paper presents the modeling, simulation and analysis of the dynamic behavior of a mixed power network of 2.78 GW including hydro, thermal and wind power plants. The modeling of each power plant is described. The set of parameters of the turbine speed governor of the hydroelectric power plant is determined with a specific identification procedure to achieve stable operation for different cases such as interconnected, isolated or islanded operation. The analysis of the stability of the entire mixed islanded power plant is investigated through time domain simulations for different sets of controllers parameters and for different disturbances (load rejection and turbulent wind speed profile). Index Terms—Wind park, hydroelectric power station, power network, stability, identification I. I NTRODUCTION R ENEWABLE energy sources like wind farms and hy- droelectric power plants will play in the next coming years a more important role in the production of electricity. The stability study and analysis of the transient behavior of mixed power networks involving different types of energies such as hydro, thermal and wind power, become an important challenge for planners of such complex power networks. The aim of the present study is to investigate the dynamic behavior of an islanded power network of 2.78 GW presented in Fig. 1, featuring about 17% of wind power, 35% of hydropower and 47% of thermal power. It is well-known that the governor control characteristics of the hydroelectric units have a great importance to stabilize the network, and accurate turbines models are also very important for the controllers’ parameters tuning [1] [2]. Therefore, the power system behavior is analyzed from the hydroelectric power plant point of view in order to determine the best control parameters set of the hydraulic turbine speed governor to guarantee the stability of the frequency of the mixed islanded power network. In order to determine the control parameters of the hydro- electric power plant, an identification procedure is used in the present study. It consists of the determination of a discrete transfer function of the turbine. The input and output signals This work was developped by the Laboratory for Hydraulic Machines and the Electrical Machinery Laboratory in the Swiss Federal Institute of Technology in Lausanne. A.C. Padoan Jr. is with the ALSTOM (Brazil) Ltd., Sao Paulo-SP, Brasil. A. Schwery is with the ALSTOM (Switzerland) Ltd., Hydro Generator Technology Center, CH-5242 Birr, Switzerland. C. Nicolet is with Power Vision Engineering, Chemin des Champs-Courbes, CH-1024, Ecublens, Switzerland B. Kawkabani and J-J. Simond are with the Laboratory of Electrical Machines of the Swiss Federal Institute of Technology, CH-1015 Lausanne, Switzerland. F. Avellan is with the Laboratory for Hydraulic Machines of the Swiss Federal Institute of Technology, Av. de Cour 33Bis CH - 1007 Lausanne, Switzerland. are respectively the guide vanes opening and the turbines speed. Two different input signals are employed: a pseudo random binary sequence (PRBS) is used to estimate high- frequency Bode diagram (0.5 to 20 Hz), and a square-wave signal is used to estimate low frequency Bode diagram (0 to 0.5Hz). A similar process of identification was successfully implemented for a pump-storage unit [3]. This identification procedure has been applied for different cases of operation of the network: interconnected, isolated and islanded operation. For each case of operation, the corresponding turbine transfer functions are deduced and governor parameters determined. The paper is organized as follows. First, the models of different power plants are presented in section II. The identifi- cation procedure is described in section III, and the results of identification of the governor parameters are presented in section IV. Finally, the performances of different sets of controllers’ parameters are compared in section V through time domain simulations in the case of a load rejection and turbulent wind speed profile. II. MODELING A. Power network presentation Figure 1 presents the power network comprising a 1.3GW thermal power plant, a 1.0GW hydroelectric power plant and a 480MW wind farm. Table I presents three circuit breaker configurations. Each configuration represents one mode of operation, as described below: 1) Interconnected operation: only the hydroelectric power plant is connected to an infinite network; 2) Isolated operation: the hydroelectric power plant is connected to a passive consumer load; 3) Islanded operation: the three power plants are connected to a passive consumer load. The transmission lines of 500KV have the same parameters: R =0.1Ω and L =1mH . The power network represented in Fig. 1 is simulated using the software package SIMSEN, developed for the simulation and the analysis of power sys- tems [4], [5], [6]. The modeling of the 3 power plants are presented below. H T W power plant power plant power plant Transmission line Transformer 500KV Infinite grid Passive load CB A CB B CB C 1.3GW Thermal 1GW Hydroelectric 480MW Wind farm Fig. 1. Layout of the mixed power network.