Around 57k miles. Here is the last page of the post I did on the details of my adventure: http://986forum.com/forums/performance-technical-chat/31883-advice-my-winter-project-3.html

Just to recap, here are some high mile UK 2.5s for sale:

153k:
Porsche Boxster 2.5 2DR + LEATHER 1999

190k:
Porsche Boxster 2.5 2DR 1999

There appear to be a reasonable number of high mile examples for sale.

Bleh. It's very hard to know what to do. Funny really. A month ago I was pretty confident it would just keep on going. As soon as I start thinking about getting the clutch done, the paranoia creeps back in!

I'm at 100K with my '01S. My IMSB shows no sign that its near the ends of its working life. My clutch has seen a lot of miles, but isn't slipping yet or making noises. While I don't what to spend the $2,200 US to replace these parts, I plan to do so in the next few months for two reasons.

First, I plan to keep the car for another 10 years so longevity is important to me. Second, and most important, I've owned my car for 10 years. I know how it's been maintained and what has been repaired. So I'd prefer to stay with a car I know intimately than either spend a lot of time looking for a well documented alternative or go with an unknown.

So that's my risk calculus. I'm willing to spend the $800 US extra to install the LN bearing when replacing the clutch because that's a small price to pay when amortized over 5 to 10 years or when avoiding buying an older car with a unknown history. And that's really the heart of my risk profile.

BTW: if you visit the following site, you'll see some data on LN bearing failures: dual row = 0% and single row = 0.1%. IMS 101 | IMS Retrofit. See third paragraph from the bottom.

As today has worn on, I am beginning to come round to this way of thinking.

I know my car well, it already has plenty of newish parts on it (water pump, rads, coils etc).

But I do find the admission that some of the single-row retrofit bearings have failed pretty worrying. What has caused them to fail? I thought the point of the hallowed LN bearing design removes the known failure modes.

Obviously no mechanical part is perfect. But to me "about half a dozen" sounds like a fair few failures for an aftermarket part that can take the engine with it when it goes.

I also find it hard to believe the 99.9% thing if they have had six failures. That means they are claiming to have at least 6,000 single row bearings running in the field. That sounds like far too many to me.

As a data point I chose to do clutch/IMS at 96K miles when my trans had to be replaced due to a failed pinion bearing. The 1st gen IMS bearing still had both seals intact and there was no visible rust, corrosion or play in the bearing at all. The clutch plate was just off the rivets and the FW was far out of spec. All were replaced. A Boxster will still run with an out-of-spec flywheel but it will never feel right. I have never overheated this car, it has short oil change intervals, and I live in a very low humidity environment which all contribute to IMS longevity as we understand it.

I do plan to keep the car because she is a known entity and I don't suffer from a bad case of "the wants". To me it is a lot more than simple dollars and cents. If it were only about dollars and cents none of us would be driving a Porsche. This is my weekend toy, joy ride, and track rat. I spend far more on tires each year than a clutch/IMS job. It was a no-brainer for me to get it done while we were already in there.

I'll add couple more thoughts to the discussion and stop.

A well respected Porsche dealer in my area installs the LN bearing as its preventive maintenance measure and not the Porsche bearing. That's saying something.

At least LN has put numbers on the table which Porsche apparently hasn't done.

I'll take LN's 0.1% over Porsche guesstimated 2% to 5% any day of the week.

Humidity as an IMS bearing failure factor?

What is the thinking there?

If that's the case, surely driving style will have much more impact? If a car just does infrequnet short journeys, moisture in the oil won't be boiled / burned off. If it regularly gets nice long runs, it won't really matter what the environment is like as the engine temp will be nice and hot for extended periods, burning off the moisture.

Also, if humidity was a material factor, I think there would be fairly clear anecdotal evidence of higher failure rates in the UK than a lot of other countries. But I just don't see that. Hardly anyone has had an IMS failure on Boxanet since I started using it nearly three years ago.

I do get your thinking, Topless, and I'm not a million miles away from it myself. It's just a pretty big investment in my current financial circumstances, so it's a very tough decision.

If I was sure the upgraded bearing was a dead cert not to fail, it would make the decision a bit easier, I think...

Honestly, I'm not sure it does. It's not like we fully understand either exactly why some factory bearings fail and others don't or if the LN design is unquestionably advantageous and / or if it has its own issues.

To my mind, we have very little proven data. All we know for sure is that the factory bearing definitely fails. I'd say the jury is still out on the LN kit to at least some degree.

And I'm not totally sure LN has put numbers on the table when its numbers are apparently saying it has sold 6,000 single-row bearings, which seems implausible to me.

On a related note, i have to say that IMS Guardian thing strikes as pretty preposterous. Obviously that a whole separate topic. But the IMS subject is tricky and the lack of hard data has a tendency to have people making assumptions that may not be true.