IEC 61850 based predictive power management system for industrial networks

Abstract

The industrial power distribution network, of late, has become complex as utility network due to multiple generating units operating in parallel and distributed loads. Stability of the power system is critical for the efficient operation of a process plant. Loss of generation is one of the severe contingencies in the plant network, which may result in black-out of the sources due to the cascaded tripping of sources. Different techniques of load shedding are used like under-frequency based, rate of change of frequency based, contingency based etc. The paper discusses about a model that is predictive. The load shedding scheme is based on power balance with a back-up frequency and/or header steam pressure. Industrial distribution network have multiple sub-network islands based on the source and tie-feeder status and are dynamic in nature. The model includes dynamic sub-network determination matrix. The model considers the Spinning reserve power available in the machines in the power balance calculation to avoid over-shedding of the loads. The paper also discusses about the various load shedding techniques and the importance of the fast load shedding. The proposed model is reliable, fast, flexible and effective that results in maximum availability of the power with minimum plant disturbance. The model uses the ethernet based network for faster communication and would be best suited for the management of networks of large process plants of oil and gas, cement, steel, paper and chemical industries.