This study evaluates the impact of decentralisation on the reliability of electricity networks, particularly under stressed conditions. By applying four strategies to add decentralised generators to the grid, the impact on network reliability has been assessed, where the blackout impact has been defined as the product of the relative blackout size and the relative blackout frequency. The general approach taken to decentralise the network is to replace the aggregated generation capacity at an existing node with three new nodes representing the total generation capacity of multiple decentralised generators. Two different networks have been used: a reduced and aggregated version of the electricity network of Great Britain (GB) and the IEEE 39 network, and each of them has been assessed for decentralisation based on conventional energy sources and for decentralisation based on intermittent renewable energy sources. The results suggest that adding significant amounts of DGs, especially if it is intermittent, can seriously reduce network reliability; however, various approaches regarding the decentralisation strategy and management of the resulting network can mitigate the negative effects.