In reported incidents, different types of voltage instability have occurred and each has presented in a quite distinct manner. To better acquaint power engineers with the different types, two incidents are described as each commenced with a type of system voltage instability and culminated with radial voltage instability. In one case, the loss of a heavily loaded generator commenced as ''classic'' system voltage instability. In this incident operators were able to shed load, but not fast enough to prevent the loss of local generators so that the situation reverted to radial voltage instability. In the second incident there was no initiating disturbance but extremely hot weather created a very high level of air conditioning load. This created a situation of ''slow'' system voltage instability where the system remained intact 12 minutes after the last shunt capacitors had been switched in. A more serious collapse was avoided when radial voltage instability led to an exponentially increasing voltage reduction at three 500kV substations. The quickly altering voltages and currents triggered impedance protection on their 500kV lines interrupting 2.8 million people supplied from the three substations. In neither of these two incidents were operators fully aware of the situation. The analysis of the incidents provided in this paper can give important operational insights. In particular it shows the value of timely load shedding and the need to allow a 20% margin from the radial voltage stability limit.
