MODELLING WIND FARM WITH FREQUENCY RESPONSE FOR POWER SYSTEM DYNAMIC STUDIES
Master thesis
Date
2015Metadata
Show full item recordCollections
- Institutt for elkraftteknikk [2401]
Abstract
This dissertation presents an analytical study on modelling wind farm with frequency responsefor power system dynamic studies. Recent trends indicate that wind energy penetration in thepower system will keep on increasing. This study models a permanent magnet synchronousgenerator fully rated converter based wind farm, since this is one of the state of art technologiesthat services this growing demand. Due to the variability in the wind resource, a high demandwill be placed on the frequency stability of the power system. This necessitates the requirementof providing both inertial and governor support by the wind farm to the power system. Highwind penetration will have a significant impact on the power system stability. These impactsneeds to be investigated by conducting power system dynamic studies.The main challenge is in identifying an appropriate level of complexity of the models torepresent power system electro-mechanical dynamics, while keeping the models as simple aspossible, to reduce the computational requirements. Taking this into consideration, the maincontribution of the modelling work, is identifying a full order model and a reduced order modelof a wind farm with frequency response. The dissertation presents the dynamic models of themain components of a wind turbine and shows how the component models are combined togenerate a full order and reduced order model.The wind farm is interfaced to the utility grid by a back to back voltage source converter.Therefore selection of robust control structures and identification of control parameters constitutesan important study element. Both, generator converter controller and grid converter controlleris given special attention. Auxiliary control loop for implementing frequency response isidentified and integrated to the models.The power system under study is the Kundur s two area network. One synchronous generatoris substituted by the wind farm and contributes to 25% of the system capacity. The dynamicsystem study investigates aspects of both frequency and small signal response.The full order and reduced order model dynamic response, for a variable wind speed sequence,shows a high degree of correspondence, both with and without frequency control. Thereforethe results of this work indicate that a reduced order model is sufficient to model powersystem electro-mechanical dynamics without significant loss in accuracy. The full order and reduced order model small signal response, for constant wind speed andwithout frequency control shows a high degree of correspondence. However the full order andreduced order model response shows deviation, when the wind farm is supported with frequencycontrol. Therefore the full order model provides a more accurate small signal responsefor wind farms with frequency control. The studied network reveals a poorly damped inter areamode. Substitution of a synchronous generator by the wind farm, increases the damping of thismode, decreasing inter area oscillations and enhancing system small signal stability.The simulations of an over frequency event, in a system with a wind farm supporting frequencycontrol, clearly shows superior frequency response, in comparison to a wind farm withoutfrequency control. The salient performance is reflected in rate of change of frequency, temporarymaximum frequency, steady state frequency and settling time of the frequency response.