A shortcoming of the contemporary power grid monitoring is that the system does not know its own state. Instead of taking automatic note of energy-flow disruptions, one deals with haphazard telephone reports of “no light in our house”. We propose a novel monitoring system that requires no restructuring of the power distribution network and can be applied both to the existing grids and the future “smart grids”. The proposed system is based on a network of inexpensive sensors, installed on every connecting line and communicating measured data to a central processing unit. Our approach is topological in nature, based on the connectivity aspects of the power grid embodied in Kirchhoff's current law that must be valid at every node of the network. We argue that the state of the network can be adequately characterized by specifying the RMS currents and the direction of energy flow in all connecting lines. It is essential that in this description one does not have to know the magnitude of the energy flow, only its direction. This eliminates the need to measure voltage, which would be prohibitively costly on the massive scale. In contrast, the relative phase between the current and voltage can be measured easily. Another essential point is that the instantaneous RMS currents are impractical to record and communicate, hence local averaging is required. Since Kirchhoff's law should remain valid upon averaging, the latter must be carried out at each sensor synchronously over the entire network with global synchronization provided by the GPS.
Reflections on the Future Electric Power Grid Monitoring System
