Adaptive remedial action schemes for transient instability

Yi Zhang

doi:10.7273/000005722

Remedial action schemes (RAS) are designed to avoid wide spread outages after a severe contingency in the power system. RAS have also been installed to increase the operational transfer capability (OTC) with the restrictions on the transmission expansion in the power systems. In current practice, these schemes are predetermined based on extensive repeated time domain simulations. This dissertation presents an adaptive method for RAS computation. Given a two-area system model and a mode of disturbance (MOD) that the system will separate after a severe fault on the ties, the presented method separately calculates differential potential energy (DPE) and residual kinetic energy (RKE) with respect to an intermediary case. A sufficient RAS is determined by comparing the RKE and the DPE. Because the DPE can be reused for the new unstable cases, the repeated time domain simulation can be avoided in RAS calculation. This method is adaptive to different operating points and fault locations; it is also can change with different fault clearing times and/or RAS initiation time. This method presents a way to reduce the time domain simulations in the RAS off-line calculation. This method also enables us to establish an adaptive RAS using fast on-line computation. To support the adaptive scheme, a new transient energy calculation method is proposed based on the curve of equivalent power versus equivalent angle (EPEA curve) of the power system, which is derived from the concept of corrected kinetic energy. It is particularly suitable for transient energy calculation in situations with varying mechanical power, such as generation rejection. This dissertation also proposes a concept using the transient instability detection as the trigger of the adaptive RAS. With Phasor Measurement Units (PMU) measurements as input, a decision tree technique is applied to detect system transient instability. MOD information is included in the classifications of the decision trees so that it can be applied for the MOD specific RAS. The effects of the time delay and the synchronization of the PMU sampling, which are independent of the occurrence of the system events, are investigated. Also studied are the effects of different locations of PMU and different input features of the decision trees, such as phasor angle and magnitude, pre-fault transfer power, and discrete system events. IEEE 39-bus and WECC 179-bus systems are used to illustrate and evaluate the proposed methods.

Adaptive remedial action schemes for transient instability

Files and links (1)

Abstract

Metrics

Details