It can be a phase to phase short circuit. My knowledge in electrical engineering has become rusty, but here are my thoughts:
Surge protectors works by shorting itself when the voltage becomes too high. They short themselves to lower the voltage on the line, protecting other equipments on the line. Common types are metal oxide varistor, where metal oxide will melt and become conductive upon high voltage, and gas discharge tubes, where gas becomes ionized hence conductive upon high voltage. Surge protectors have voltage/power rating which indicates the highest amount of power they can dissipate. If the spike voltage is too high, or the voltage spike lasted too long (“long” is not only time, it’s about the shape of the waveform too, RMS basically), the surge protector will blow out and other equipments on the line may be damaged. The expectation is that the main breaker on the line will trip before surge protectors blow out.
In a phase to phase short, two phases are connected together, which means two sinusoidal waves with 120° phase shift are superimposed onto each other. The end result will be voltage dropping, but in the transient event, there could be some power spikes, as the line and load aren’t perfect resistors. The reactance component (capacitance and inductance) can cause voltage spikes a few times higher than the main voltage when the main voltage changes rapidly. These voltage spikes likely triggered the surge protectors. Current on these two faulted lines also increases substantially as phase to phase short is essentially a short circuit, the phase angle between the two phases means that voltage on these two lines rarely equal to each other.
Another change is that since the 3 phases are no longer balanced, lots of current will flow through the neutral wire, slightly rising the voltage on neutral wire. This can damage some equipments too.
There are line differential relays in the substations to monitor phase to phase faults by comparing the current in each phase at different locations on the line, but it won’t react immediately when the short happens, as it usually need a few cycles to determine what exactly happened. By the time the relay decides to trip the circuit breaker, the voltage spikes may have already caused some damage to the surge protectors and equipments down the line. Short circuit calculations are often done using energy instead of voltage to account for the fact that main circuit breaker can not trip immediately upon fault, and it’s really the energy, not the voltage, that damages equipments.
Edit: Looks like a real expert has chimed in just above me
