The MOV, metal oxide varistor degrades every time it absorbs a surge. Commonly, the MOV in a surge suppressor is designed to activate at a voltage 10-15% above the expected standard voltage - which for this case will be about 140 volts.
The standard actually is a minimum of 125% but can be higher, depending on the application.
Small surges just over 140 are fairly common, being caused commonly by motor startup and shut down, ie. air conditioners for example.. One report I've seen said that an average home will be subjected to 8-20 surges per day.
MOVs have a highly nonlinear voltage-current response curve. A slight voltage over their threshold does not conduct much power away. A large spike in voltage results in a much larger current diversion.
A true "surge" is a momentary voltage spike lasting milliseconds, not seconds. Those longer ones are over-voltage conditions that a MOV will barely touch due to their nonlinear response curve.
Each surge degrades the suppressor until its remaining capability reaches a level determined by the manufacturer and is then reports itself as failed.
Yes. When an MOV actually conducts a spike to ground the MOV develops a tiny, permanent leakage current. The leakage current depends on the size of the surge. A bunch of small ones can affect an MOV to the same extent as one large one.
As surges are conducted away the leakage current keeps increasing. Once the permanent leakage reaches the manufacturer-defined level the MOV module is considered "used up" and it triggers a light or something. It actually still works but not as effectively and can be subject to overheating and occasionally bursting into flames. (That's what the most recent standards changes address; the protective device causing the problem.)
There is a tremendous amount of surge protection built into the grid and MOVs are so cheap that a lot of electronic equipment have them built in to reduce warranty exposure but the owner just doesn't know it. Kind of makes you understand why RV companies charge hundreds for a box that contains maybe $10 retail worth of MOVs.
Think about it. There are tens of millions of homes in the US without a whole house surge protector yet their equipment is not failing regularly. It's almost a non-problem in the real world unless there is a nearby lightning strike and no consumer surge protector is going to survive that wallop anyway.
People love to talk about how they always use a surge protector strip and they've never had a failure so that must be the reason. Me, I repaired electronics down to the component level for decades and I never use one and I don't have failures either. They're trying to prove a negative to justify their expenditure.
Yes, I know a recent NEC revision is supposed to mandate whole house surge protectors but there is no retrofit requirement and those tens of millions of homes are still doing just fine. There is a whole lot of industry "input" to NEC revisions that does not necessarily improve safety in a material way and occasionally to hurt the competition.
Ray