There is (in the US) apparently no definition for how joules are measured. As a result, some manufacturers measure jules in a deceptive manner, which puts honest manufacturers at a disadvantage. As a result, some good manufacturers are not providing joule ratings anymore. Instead they give surge current ratings (which are comparable).
The IEEE guide warns against comparing devices based on joule ratings unless the test method is the same. Too bad - I like joule ratings.
Certainly can happen.
In my first post I quoted an investigation by Martzloff that found only about 35J max at a plug-in suppressor with even the strongest surge that can be reasonably expected on power wires. One reason is arc-over at the service panel. The other is the impedance of branch circuit wiring. Both are mentioned in other posts. If the branch circuit is very short - outlet under the panel, the energy can be higher.
I recently bought a major brand plug-in suppressor (about $30) with ratings of 590J and 30,000A per MOV, 1770J and 90,000A total. I don't expect the suppressor will ever fail. The manufacturer apparently doesn't either - there is a connected equipment warranty.
The current ratings, 30,000A, are in some ways odd - there is no way you could get that current to the suppressor. It is higher than what will come in on the service. But the high current ratings go along with the high joule ratings.
The joule rating for a MOV is the single event energy hit the MOV can take and be at the defined end of life (but still functional). If you look at MOV ratings curves, you find that if the individual energy hits are a small fraction of the rated energy, the cumulative energy rating is far higher than the single event rating. If my 590J MOVs only see 35J or far less events, the cumulative energy rating will be far over 590J - another reason for connected equipment warranties.
High ratings are useful in the same way for service panel suppressors.
The normal failure mode for MOVs is to start to conduct at lower voltages until they conduct at normal voltages and have thermal runaway. Any suppressor you buy in the US should be listed under the appropriate UL listing - 1449. UL1449 has required, since 1998, a thermal disconnect for overheating MOVs.
For plug-in suppressors, the IEEE guide goes on at length about how the protected load can be connected across the MOVs, and be disconnected when they fail, or can be connected across the incoming line. In the former case, the load is 'protected' even if the MOV fails. Another reason why manufacturers can have protected equipment warranties.
A good idea, particularly in high risk areas. It is suggested by Martzloff in another post. Plug-in suppressors are particularly useful if the protected equipment has both power and phone/cable connections - equipment that is particularly at risk.
Note that all interconnected equipment has to be connected to the same plug-in suppressor and external wires (including phone and cable) have to go through the suppressor. I think this is not as well understood as it should be.
The max probable current on hot service wires is 10,000A. Service panel suppressors with much higher ratings are readily available.
Phones off the wall sounds odd. I would look at the phone entrance protector and length of ground wires from phone and cable entrance protectors to the ground at the power service. The entrance protector and earthing should protect the phone (using US installation practice). And look at earthing system?
All good information. Never been around a ferro-resonant transformer - interesting.