Model Development for Power System Analysis with a Substantial Wind Energy Capacity Installed in the Nordic grid

The worldwide development of wind power installations now includes planning and construction of large-scale wind farms ranging in magnitudes of 1000 MW and more. As part of the planning and design of such systems, it is well established that the transient and dynamic stability of the electrical power system needs to be studied. Modelling work of the electrical behaviour of wind turbines and wind farms as well as model validation by measurements have been important parts of this project work. The models have been used to study dynamic phenomena during normal operation and fault occasions in the electric system. Fault Ride Through (FRT) measurements have been carried out on new wind parks connected to Estonian power grid and in all of them FRT tests were made. In several wind parks the tests were not successful and the tests will be repeated. In Finland measurements have carried out in 6 MW Högsåra wind farm. Measurement results from real scale tests of frequency control were presented in the project, where different amount of fluctuating wind power were injected into a power system. The allowed limit for system power fluctuation was shown, which is given by the acceptable variation on grid frequency. Results show a strong dependency on the wind turbine operational states as well as on wind speed conditions. Nevertheless it is important to consider the underproduction power and the associated large recovery period that follow a Variable Speed Wind Turbines overproduction operation. Current has work has been carried out to integrate marked and network models. This has been done either directly in the optimization algorithms of the marked model, such as for Samlast, Samnet or PSST, or by providing algorithms and functionality for converting the result of the marked model into a full detailed power flow description. For the PSST (Power System Simulation Tool) , which optimizes power flow by minimizing the cost of generation and assumes a perfect market with nodal pricing, focus has been on interpreting the result establishing time series of area prices as well as including power losses in the model. By cooperation within the Nordic countries the existing knowledge has been spread, new knowledge has been created and the results have been transferred to utilities. Over 35 journal or conference publications and five PhD-theses have been presented. Two more PhD theses are on the way next during 2011. Two Nordic Wind Power Conferences have been organized during the project period.