A tropical aquarium is keep at the same temperature. Unless somebody would disconnect it to save energy or the termostat[s] would fall, shouldn't have experienced a lot of changes in their volume by this.
Not so much. A big mass of water act as a temperature buffer and air pumps and filters are mixing the water column all the time. It takes several hours to cool.
The event does not seemed truly random to me. If it was, the probability of breaking in the 90% of the time when there was people around would be much higher than breaking in the 10% small time interval without people. Extreme luck is rare. Looking for an external trigger seems appropriate
But the only surface really accessible to people was the escalator, and now we know that it didn't broke. The outer surface is mounted over the escalator door, can't be scratched purposely without taking a lot of troubles. Can't be shoot without leaving evident marks of a crime
If we assume that the thermostat didn't broke and that material fatigue is not an optimum explanation (it was checked and maintained in 2020), then temperature differences seem a good candidate. It broke in winter, in the hour where day/night differences should be maximum. The outer temperatures that night were -10C if I'm not wrong.
Maybe the insurance companies could sue the government if this was indeed the case. I would really like to see the bad policy makers go to jail or get hit with huge fines.
The aquarium was likely very constant (otherwise the fish would die, they tend to be extremely sensitive to variations, especially rate of change), but the surrounding air likely was not.
Those fishes can stand an interval of temperature if changed slowly. A main question to check would be if the thermostat was deliberately lowered by the owner in the previous weeks to save energy