Preventing Wide Area Blackouts in Transmission Systems: A New Approach for Intentional Controlled Islanding using Power Flow Tracing

Abstract

A novel method to reduce the impact of wide area blackouts in transmission networks is presented. Millions of customers are affected each year due to blackouts. Splitting a transmission system into smaller islands could significantly reduce the effect of these blackouts. Large blackouts are typically a result of cascading faults which propagate throughout a network where Intentional Controlled Islanding (ICI) has the advantage of containing faults to smaller regions and stop them cascading further.
Existing methodologies for ICI are typically calculated offline and will form pre-determined islands which can often lead to excessive splits. This thesis developed an ICI approach based on real time information which will calculate an islanding solution quickly in order to provide a ‘just-in-time’ strategy. The advantage of this method is that the island solution is designed based on the current operating point, but well also be designed for the particular disturbance location and hence will avoid unnecessary islanding. The new method will use a power flow tracing technique to find a boundary around a disturbance which forms the island that will be cut. The tracing method required only power flow information and so, can be computed quite quickly.
The action of islanding itself can be a significant disturbance, therefore any islanding solution should aim to add as little stress as possible to the system. While methods which minimise the power imbalance and total power disrupted due to splitting are well documented, there has been little study into the effect islanding would have on voltage. There a new approach to consider the effects that islanding will have on the voltage stability of the system is developed.
The ICI method is based on forming an island specific to a disturbance. If the location of a source is known along with information that a blackout is imminent, the methodology will find the best island in which to contain that disturbance. This is a slightly different approach to existing methods which will form islands independent of disturbance location knowledge. An area of influence is found around a node using power flow tracing, which consists of the strongly connected elements to the disturbance. Therefore, low power flows can be disconnected. This area of influence forms the island that will be disconnected, leaving the rest of the system intact. Hence minimising the number of islands formed. Finally the methodology is compared to the existing methods to show that the new tool developed in this thesis can find better solutions and that a new way of thinking about power system ICI can be put forward.