Role of Clustering in the Probabilistic Evaluation of TTC in Power Systems Including Wind Power Generation

Transfer capability is used to schedule future transactions between areas and arrange commercial activities in multi-area power systems. Probabilistic evaluation can provide more information about transfer capability than deterministic approaches, potentially leading to more efficient operation of the transmission network. Probabilistic methods can also model generating units with variable outputs such as wind farms (WFs). Fluctuating nature of WF creates different impact on transfer capability compared to conventional generators. In this paper, three methods are proposed to compute transfer capability in multi-area power systems containing WF. In all the methods, Monte Carlo simulation is used to assess probabilistic variation of transfer capability. The methods differ in the manner of handling output of WF and load. In the first approach, system load data and power output of WF are directly used in the Monte Carlo simulation. In the second approach clustered power output of WF and system load data are used. In the third one, after clustering of input data, Monte Carlo simulation is run separately for each of the clusters. Incorporating clustering into Monte Carlo simulation can accelerate convergence speed, reduce variance and sometimes provide additional useful information. IEEE-RTS is used to demonstrate the effectiveness of the proposed approaches.