Cumulative data on S65 engine failures.

[OneMmmm]

Originally Posted by ///M Power-Belgium

More salt...

The problem is that the entire world is not on this forum,most people don't even know the existence of it !
I had also several PM's from people that are asking to stop talking of our bearing failures .
Sure they had bearing failure , but they don't post it on here !
Damn why not....?

It's only one single and simple reason and not complicated...."Thinking about the value of their car" !
I don't want to buy a car with the history of bearing failure, do you.. ?
So even when member on here and in case of bearing failure...Personal I think not even 50 % of it will post it on here !

How we should call this in two words ?...."Smart moves" ???
I've nothing against that (free country) and it's all about money and I'm ok with that , but they are not on here and not counted...

About my info I'm 99% sure,and maybe in may I hope to get the total number if it's possible .(Depends of Germany)
But at that moment you can tell the same thing , so were are we going ?
Think sometimes maybe it was better to eat my numbers , but I'm too angry for that !

And like you sad...Feel peace with my engin "for now" ! ( same here )
The big question is for how long ...
Found also something interesting for you,at least I'm curious => http://www.m3post.com/forums/showthread.php?t=1087004

You are still only speculating about failures, whereas SFP is presenting data based on documented failures in US/Canada.

I think everyone on here acknowledges that the current dataset does not capture every failure in these regions. But until someone presents valid documentation and data for the purported dozens (hundreds?) of other failures, those claims will remain nothing more than hearsay.

As such, we would need 4-5X the number of currently documented North American failures to approach a 1% failure rate. I can live with that risk... For now

[///M Power-Belgium]

Originally Posted by OneMmmm

You are still only speculating about failures, whereas SFP is presenting data based on documented failures in US/Canada.

I think everyone on here acknowledges that the current dataset does not capture every failure in these regions. But until someone presents valid documentation and data for the purported dozens (hundreds?) of other failures, those claims will remain nothing more than hearsay.

As such, we would need 4-5X the number of currently documented North American failures to approach a 1% failure rate. I can live with that risk... For now

Let's hope...for now .

[SenorFunkyPants]

Originally Posted by Z K

The numbers in the first post are very conservative. Most of the failures are not reported on the forum. Not everyone is online.

Yes there are failures not posted on M3post. But then there are also ~ 25,000 E9x M3 USA owners who are not on M3post. You can't have it both ways.
Lets say that we knew exactly how many E9x M3 owners there are on M3post and we asked everyone if they had an engine failure and everyone replied honestly. These figures would give us a pretty decent snapshot of the expected % failure rate of E9x M3s in the USA. This is what I am trying to do here.
I've already noted that we don't have exact membership figures or exact numbers of failures but the data there is gives an indication of where we are.

[SenorFunkyPants]

I have nearly completed updating the excel file with engine failures that have been reported since the last (Aug 2014) update 6 months ago.
The updated numbers and graphs should hopefully be completed later today.
There have been quite a few, 13 with sufficient details (minimum M3post username, year and mileage) two I'm awaiting email replies, and 1 supercharged.
Of these, one had a double engine change where the second engine was a used one of unknown provenance which then also failed (this second failure has not been added).
So we are looking at about 2 to 3 failures a month for M3post members.

The trends appear to follow the same patterns seen in previous updates.
The stand out one again is that 2011/2/3 cars are more likely to fail than earlier cars as a function of production numbers and mileage.
Indicating that the change that occurred around the 2011 MY (July/Aug 2010) to the newer spec bearings with a larger bearing clearance significantly reduces reliability in the S65 engines.

[Helmsman]

Originally Posted by SenorFunkyPants

I have nearly completed updating the excel file with engine failures that have been reported since the last (Aug 2014) update 6 months ago.
The updated numbers and graphs should hopefully be completed later today.
There have been quite a few, 13 with sufficient details (minimum M3post username, year and mileage) two I'm awaiting email replies, and 1 supercharged.
Of these, one had a double engine change where the second engine was a used one of unknown provenance which then also failed (this second failure has not been added).
So we are looking at about 2 to 3 failures a month for M3post members.

The trends appear to follow the same patterns seen in previous updates.
The stand out one again is that 2011/2/3 cars are more likely to fail than earlier cars as a function of production numbers and mileage.
Indicating that the change that occurred around the 2011 MY (July/Aug 2010) to the newer spec bearings with a larger bearing clearance significantly reduces reliability in the S65 engines.

Thanks a lot SFP, although I don't like the latter as having a 2011...

Will follow your updates.

Cheers

[SenorFunkyPants]

Originally Posted by Helmsman

Thanks a lot SFP, although I don't like the latter as having a 2011...

Will follow your updates.

Cheers

On the plus side (for you I guess) cars in cooler climates seem to be less susceptible to failure.

[aussiem3]

Originally Posted by SenorFunkyPants

On the plus side (for you I guess) cars in cooler climates seem to be less susceptible to failure.

But then in Australia where the temp is moderate to high throughout the year failure numbers are very small. That's what we know of ... so I take the above as an assumption rather than evidence based?

[SenorFunkyPants]

You would have to find out some idea of the number of failures and the production numbers for Aus to see what it tells you. Without these its all just guessing.
The indication for the USA is that the rate is higher in warmer states than cooler ones.

[dreamspeed]

Originally Posted by SenorFunkyPants

The stand out one again is that 2011/2/3 cars are more likely to fail than earlier cars as a function of production numbers and mileage.
Indicating that the change that occurred around the 2011 MY (July/Aug 2010) to the newer spec bearings with a larger bearing clearance significantly reduces reliability in the S65 engines.

Wait, so according to the data currently available to us...the newer "improved" bearings are actually causing MORE failures?

[chadwick]

You can't infer that warmer states lead to higher failure rates with just this data, can you? You'd need to know the total registrations by state.

CA, NY (not a warm state), TX are the highest population states, so of course there are going the more failures there, as there are simply more cars there. California will show more failures as there is likely a disproporationate number of these cars there in the first place.

People in colder states are also less likely to buy a car like this in the first place.

So while interesting, I'm not sure the failure by state is telling us anything at all, without vehicle registration information, portion of time spent in that state, or time of year of failure.

[BMRLVR]

Originally Posted by chadwick

You can't infer that warmer states lead to higher failure rates with just this data, can you? You'd need to know the total registrations by state.

CA, NY (not a warm state), TX are the highest population states, so of course there are going the more failures there, as there are simply more cars there. California will show more failures as there is likely a disproporationate number of these cars there in the first place.

People in colder states are also less likely to buy a car like this in the first place.

So while interesting, I'm not sure the failure by state is telling us anything at all, without vehicle registration information, portion of time spent in that state, or time of year of failure.

Good work on documenting these numbers SFP!

I do agree with the above poster though, 12% of the U.S. population is in CA, 8% is in TX, 6% in Florida and 6% in NY. So we have roughly 32% of the U.S. population in only 4 states. Of these 4 states 3 of them are "hot" states. I think it stands to reason these three states will have more failures than the rest.

Another thing to consider is the fact that generally speaking, sports cars are more prevalent in areas without winter not to mention they accumulate more miles since more of them are driven daily year round. In areas with winter many people park their sports cars during the cold months.

[SenorFunkyPants]

I used this map of M3post M3 owners to get a snapshot of the distribution of owners by state...not perfect but for states with a large population it is going to be fairly accurate.
https://www.zeemaps.com/map?group=1061097
One of the suggestions is that 10W60 oil significantly contributes to bearing wear when cold due to poor flow. This is extremely difficult to quantify given all the variables but, what would be the prediction for engine failures as a function of sales and location?
As I've noted before when I first looked into this I expected cold climates to have a higher engine failure rate or at least the average rate. Instead warm/hot climates seem to feature higher in the failure rate table as a function of sales per state that should be expected.
Whatever is the mechanism of S65 engine bearing failures it doesn't seem to be dominated by wear due to cold oil.
I'm in the process of adding 16 more engine failures to the xcel file which may shed a little more light on this question.
What is slightly worrying though is that 9 of these 16 are 2011/2 cars (the other 7 are 2008/9) - this is against expectation.

[SenorFunkyPants]

The latest excel file of S65 engine failures is attached for download if anyone is interested.
Contains 66 entries with clickable links to details of each fail.

[ilikebmxbikes]

great work on the data, but your sample data is very small. You may want to use the number of m3 forum members who had failures to the number of m3 forum members who dont. Then you can apply that percentage toward the total production numbers.

[SenorFunkyPants]

The above is for USA/Can cars without superchargers.
The "average mileage at engine failure in production year" green line is pretty much as expected.
The "number of cars produced in year divided by the number of failures" purple line however is not. It should follow the green line to show that newer cars are less likely to fail but clearly doesn't.
Currently of the 47 entries:
22 are 2007/8/9/10 cars from a total production of 15203 cars @ an ave miles at failure of 61,500.
25 are 2011/12/13 cars from a total production of 15053 cars @ an ave miles at failure of 25,800.
The change in rod bearing spec occurred sometime around July/Aug 2010 (probably for the 2011 MY)

[SenorFunkyPants]

Updated Excel file attached.
Two fails added.
2 to 3 fails a month seems about the going rate.

Summary (USA cars only):

[SenorFunkyPants]

This shows how the rate of failure of later cars is increasing while that of older (much higher mileage) cars is steady. This against the prediction that as mileage rises so will the failure rate.
The three dates are the total number of failures to that date




.
.

[aus]

Great work bud.
.

[SenorFunkyPants]

No new failures to add.
But I have changed one of the columns to show an approximate date of failure.
Gives a total of 20 for 2013 and 23 for 2014 - which is surprisingly constant.
I've also added if it looks like the car was tracked at some time.
14 out of 54 cars were tracked = 25%
Attached is the latest zipped excel file.

[e46e92love]

Originally Posted by OneMmmm

You are still only speculating about failures, whereas SFP is presenting data based on documented failures in US/Canada.

I think everyone on here acknowledges that the current dataset does not capture every failure in these regions. But until someone presents valid documentation and data for the purported dozens (hundreds?) of other failures, those claims will remain nothing more than hearsay.

As such, we would need 4-5X the number of currently documented North American failures to approach a 1% failure rate. I can live with that risk... For now

Common sense post.....almost shocking to find that around here.

Cheers,
e46e92

[08 E92 Jerez Black]

Is there anyone working this that is gathering specs from rod bearings pulled out of engines? I have the ones I pulled out of my 08 e92 M3 six speed with 87k Miles, and even though I would want them back (in the event I have problems with the Main bearing in the near future) I would be fine with shipping them to someone that is accepted as an expert that is already gathering the info and that BMW might take seriously. this is a street car that didn't see track time and is pretty much stock (no supercharger) the pic is showing the number one (left) and number 8 (right) upper bearings (piston side) all eight looked the same.

[08 E92 Jerez Black]

I replaced them with treated bearings and stock rod bolts (not ARP) from Bimmerworld. If I was planning on doing a supercharge I would have pulled to line bore the rods before using ARP since I have always been taught that you have to since ARP grabs so much better it will actually throw rods and mains out of round.