Operation Risk Assessment Of Wind Farm Integrated System Influenced By Weather Conditions

There have been several severe blackouts in recent years that attracted much attention to system operation security. In this way, how to assess the external and internal probable factors’impact on system risk is of vital importance to grid risk prevention. As global climate changing, extreme meteorological events are more and more frequent. These events result in severe weather which could cause pole leaning or collapse, conductor breaking, insulator flashover etc. They are the main causes of system accident. With large scale wind farm development and employment, wind power fluctuation brings in a series of difficulties that intensify the potential system risk. This paper studies the weather conditions and wind power fluctuation impact on system risk.Firstly, this paper builds composite component outage models that could reflect system real-time conditions and external weather including a time-varying outage model for generators and a weather-conditioned and overload protection combined time-varying outage model for transmission lines. A two-state weather model is used to demonstrate the outage rate dependency on weather conditions. A DC power flow optimal load curtailment model is used to analyze system states. Risk levels are calculated in both system level and component level.Secondly, wind farm output feature extraction and modeling is conducted based on the massive wind farm monitoring datasets from North China Grid and the USA. Based on the multiple space-time data mining technology, the stochastic characteristic, fluctuation characteristic and correlation characteristic of wind power are analyzed. Statistics reveal that power variation is correlated with time and wind speed. In this way, this paper proposes the conditional variation index(CVI), and its fitted probability function could be used in MC sampling based on real-time joint conditions. A wind farm output variation model is proposed based on CVI, this model could show the potential deviation from power prediction as the alternative of prediction error.Then, the wind power variation risk(WPVR) index is proposed to assess wind farm’s contribution to system risk. This risk value is determined by power variation, conventional generator ramp rate and generator output limit. This index could reflect system ramp capability to coordinate with wind power positive and negative variation and the corresponding wind abandoning risk and load curtailment risk.Lastly, RTS system and Weihai city power grid in Shandong province are tested in the above models. Results show that severe weather could increase system risk value. And influenced by system network constraint, the risk for different component is different even they are on the same weather condition. Wind power positive variation results in wind abandoning risk and wind power negative variation results in load curtailment risk. WPVR values in two cases with and without generator re-scheduling are contrasted. The result is helpful for operators to refresh scheduling, to balance system risk and benefit and to cope with severe weather.