Selective load reduction in power grids in order to minimise the effects of cascade failures

Cascading failure in power grids might lead to a catastrophic black out in power systems. One resolution to deter the power grid from failure is load shedding. There are different types of load shedding in the literature which mostly cut off some loads from the grid to preserve the rest of the network's connectivity and functionality. However sometimes it is not the case and disconnecting some feeders from the power grid is impractical due to their vitality for the economy or the society. In this paper, as soon as a transmission line gets overloaded due to any rise in overall load, since its breakdown can trigger a cascade of failures, we propose a method to rank the loads to be reduced in order to prevent that particular line from failure. After the loads are ranked, the top ranked load will be a candidate to get reduced. If this action wasn't enough or feasible, the second top ranked load is chosen and so on. Even a set of loads with different percentage of load reduction can be picked. As the pre-failure data, we apply the results from load flow analysis or the last working state of the power grid. These data is applied to calculate the specific transmission line's sensitivity to changes in different loads in power grids. The results show that this method is much more efficient when the classic methods suffer from divergence and setting malfunctioning.