how common are the transaxle failures?

Definitely not #2. Luscious Garage | Blog | Gen 2 Prius (2004-2009) Transmission Failure, P0AA6, P0A92, P0A7A has a mention. Perhaps dig around on their blog?

As I posted at Transaxle Failure: no ICE, no codes, no go. | PriusChat, what was found there is a bit worrisome.

 

3. Failures are very rare, and for some reason only happen to first time posters who never reply, so we have little data to give you.

The failures I have rumors of, are all Motor/Generator failures. I think (but have so little data I hesitate to even call it a theory) that not changing the ATF WS in the transaxle lets metal shaving short out the M/Gs.

 

In a word, not common.

There is plenty of circumstancial evidence to support this.
1. Reports from Taxi cab companies indicate Prius transmisions making over 200K miles in hot dusty climates, and over 300K miles in colder cleaner enviornments is the norm. There are individual accounts of prius cabs of going over 800k miles without major failures.
2. There are two long running threads of people on Priuschat who have gotten over 200k miles, and over 299,999 miles have reported very few major failures.
3. My own experience shows that after little maintenance (two drain and fills after 220k miles) on a car that has had severe useage of daily major climbs and descents with a lot of "B" useage, and over 5 years in a hot dusty climate, and after 270k miles I have had no major repairs. My first drain and fill at 220K miles showed dark and gritty drained oil. Toyota experts now recommend changing Transmission oil every 60K miles in hot dusty areas.

 

Really? When I follow that link I get a pathetic forum with a total of 4 posts, ALL OF THEM SPAM, and nothing at all about hybrid cars? Has something gone wrong with the link?

 

Strange then, it almost looks like that whole website has been hijacked or something jacek. If you click on that "main forum" there's now 5 posts, and again all spam!

 

There seem to be two types of failures in gen 2 transaxles. The more common type is associated with the motors shorting out as the amount of metal shavings increases and the resistance of the trans fluid decreases.

This seems to be more common with repeated high torque conditions, for instance San Francisco cabs seem to see it at ~200k-300k+ miles. Even in this case it only destroys the trans after the cab owner ignores/gets around the problem by resetting the ECU/error codes, which leads to more arcing that eventually destroys one of the bearings. At that point the whole transmission fails.

The other one, which appears to be relatively uncommon, appears to be due to assembly errors made from the factory.

 

How are shavings shorting out the windings? They are covered in very strong conformal coating.
These are not bare windings.

 

Where can I find info on this what you claim common failure? What does metal shavings have to do with shorting out the motors?
Please explain.

 

There are 2 types of failure on Gen II cvt's. The common one is the windings overheat from super hi torque demand on the electric motor. Think constant hills, low soc on hybrid battery & overheated Inverter coolant. The windings overheat and the conformal coating that covers the windings fails and allows voltage to leak out into the case. Gen I's would not crowbar under this condition just throw a dtc. Car was driveable until the leak became big enough to draw current and then the Inverter would either shut down or fail. Gen II's are programmed to shut off if any ground fault current is measured. Toyota considered this high voltage safety.
The other common way is a bearing would fail either in the motor section or in the differential area. Usually due to lack of fluid maintenance.

 

It's just my WAG concerning what Lucious Garage describes here. According to that article they've seen various DTCs associated with winding deterioration, probably from high temperatures and greater metal wear associated with driving a cab versus the type of driving most other drivers see.
Cab drivers in particular tend to bypass these DTCs by resetting the ECU. The garage also sees bearing failures, especially of a specific needle bearing. I'm speculating that the two aren't independent, and that arcing from winding degradation/lower resistance can eventually causing pitting and failure of bearings in the case.