One of the most important parameters of a power system is the system frequency. The frequency can vary over a small range due to generation-load mismatches. That is, when the power consumed by loads and overall losses is greater than the generated power, the system operating frequency will decrease, resulting in a situation known as the underfrequency condition. The main reasons can be either the loss of any one or more generators in the generating side stations or if there is a sudden increase in the power consumed during certain peak hours of a day. Indeed, if the loads in a system are suddenly lost or disconnected from the system, it also leads to a condition called as the overfrequency condition. This condition is characterised by greater input power than the consumed power by the loads. The rest of the loads in the system will absorb the extra power and the generator inertia, leading to an increase in the system frequency. In both the cases, the system frequency changes from the safer frequency range, leading to tripping off of the substation and further collapsing of the entire system. In this paper, the author is proposing a novel approach, by which employing a smart meter, which estimates the system frequency change caused by overload or underload condition and comparing it with a preset frequency value set by the Load Despatching Authority, the line responsible for the frequency change is being isolated from rest of the power system or being given only the demanded power. The preset frequency is determined by analysing the parameters of electric power-based on a load forecasting. The method is employed with an embedded system along with an algorithm and Network Synchronisation with central server. © 2014 IEEE.