Balancing Protection Dependability And Security In Large Transmission Grids

Although main characteristics of supergrids and smart grids have not been univocally identified yet, it can be foreseen that the networks of the future will be generally characterized by more sensors, computation, ICT, monitoring and control functions. In such scenario, it is foreseeable that ordinary power system components will be improved in order to develop more complex on-line control functions. Experience with past severe cascading events showed how in many cases a selective approach to system protections could have avoided blackouts, or mitigated their impact. In this paper, a method is proposed to automatically adjust tripping zones of relays in order to balance protection system dependability versus the functional security of the system. A nonlinear programming technique is proposed for avoiding improper tripping of distance relays. The approach aims to modify the settings of the protection system (i.e. distance relays) in order to guarantee that no contingency can cause line tripping due to large swings. The tuning of protection schemes can be operated in a “preventive control fashion”, ensuring transient/voltage stability and keeping system trajectories off the tripping areas of distance relays (practical stability) with respect to selected major contingencies. Control actions (relay settings) are evaluated through the formulation and the solution of a nonlinear optimization problem. Time domain simulations associated to the optimization problem embed distance relay in the dynamic system representation. The approach treats concurrently, in the same optimization problem, multiple contingencies. In this way, it is possible to find out the best compromise between dependability and security during contingency analysis. The procedure can be applied in the extended real-time control framework of power system operation or, simply, whenever stressed conditions are experienced in the power system. Test results are provided for a test case based on a representation of the Italian power grid, along with a relevant part of the UCTE power system, during the transient that conducted to the 2003 Italian blackout. The computational burden associated is limited since basically it consists in time domain simulations of transients caused by contingencies selected after contingency screening. Since only tripping areas are modified no operative costs are associated to this security function.