The whole OLE system wears over time.
1) the contact wire wears over time. I have in my possession a short length of original contact wire off the GEML (taken out in 2011), and it has a cross section slightly less than a semi circle (it was circular when installed in 1948/9). Contact wire does break on it's own although that is relatively rare. Incidentally the new contact wire going in on the GEML is about 75% of the cross section of what is coming out. The original was installed for 1500v DC and was necessarily thicker to deal with increased current. Incidentally (2) I'm reasonably sure that the contact wire on all running lines of the WCML was changed south of Crewe, not just the fasts.
2) the catenary (support) wire also wears. Obviously not through pantograph contact, but usually at mechanical contact points ie where it is attached to registration arms, anchors and the like. It is usually a stranded wire (as opposed to a single thick 'rod' for contact wire) and the typical failure is of some of the strands, which causes additional stress on what remains, and eventually they all go.
3) the most common thing to break is a dropper. This is the short lengths of wire that connect the catenary wire to the contact wire. These break routinely, normally at the connectors. I should think that one comes off somewhere on the network every day. Fortunately you can get away with the odd one being broken, so it's not an urgent job. Unless of course it is hanging below the contact wire and then it has to be removed ASAP (10 minute job once someone gets there and has the juice off)
4) other key components that fail are
a) insulators, particularly older designs which are susceptible to water penetration and frost shatter.
b) pulley wheels, usually due to lack of maintenance (lubrication); although these don't cause a failure themselves they can trigger something else, cause the wire to go out of alignment which then causes a pan to fall off
c) registration arms. These can occasionally break / bend badly, but that is pretty rare.
d) insulation failures. Where live components are closer to structures etc than the standard, they must be insulated. This insulating can perish over time (particularly if under a road bridge that has regular salting), which can then cause a flashover which will burn through the relevant component in a trice.
e) all the connections that hold it all together, basically nuts and bolts. The OLE system is designed to be flexible to deal with thermal expansion, wind, snow loading etc. Some of the connections are thus designed to travel, and if they get too tight (through corrosion, wear, being particularly dirty, lack of lubrication) then they can seize, and cause part of the system to lose that flexibility. This is why 'problems with the OLE' are more likely in very cold or very warm weather.
Finally the OLE support structures do wear, albeit often in unseen ways. Some obviously corrode, although that's not usually a problem. The main issue is failure of the structure foundations, particularly in places where they are 'stray' currents around from heavy industry and particularly other (DC) railways. this can cause the bolts between the concrete foundations and masts to corrode rapidly.
All of this is inspected regularly, and some components are changed on an as required basis. Often it is decided to do a 'campaign' replacement of some components on a certain stretch of line, eg all insulators. I've done that myself in a number of places.
In terms of failure of the OLE itself, as opposed to the distribution equipment (transformers, substations etc) my guess for failure causes, based on personal experience, is as follows:
20-25% caused by the train, usually pantograph failure, but sometimes other parts of the train hitting the wires
20-25% caused by 'outside parties' coming into contact with the live part of the system, thence causing failure through mechanical failure or electrical flashover. Often the 'outside party' is of avian origin, landing or flying at just the wrong place / wrong time. Pigeons seem to be particularly conductive.
20-25% caused by the failure of one or more component, broadly split between a failing of maintenance, or because of a hidden flaw in that component.
15-20% caused by extreme weather outside the design capability of the system (usually very, very hot temps, or hurricane force winds)
The rest fall into the category of no-one is really sure.
Hope this helps.
