This post is partially a means for me to vent a little pent up frustration as I wait for BMWNA to finally approve the inspection of my oil pan after 8 days of “pondering” what on earth might be wrong with my 11/01 M3 while ignoring the several hundred examples which have very much resembled my car and would give a really good hint about what it probably is and how to check for it to anyone with an ounce of common sense.

But it also may serve as a nice summary of the M3 bearing issue, and how it has been dealt with by BMWNA from the perspective of the customer base. It's not a very pretty story. Please feel free to use this in any manner desired - no rights retained. I would love for an executive at BMW to actually read this story, print it out, put it in front of the people who are "managing" the issue and with a cold angry gaze ask them

"is this stuff true, and is this really how we are treating our customers?"

============

===========================
M3 Rollout - the early days
===========================

Our story begins before M3's were even available to the public. There was much publicity and anticipation of the "new" e46M3 which would take over the honored title of M3 from it's predecessors the e30M3 and e36M3. In mid 2000 there had been a few press reviews and test drives, lots of photos, and intense speculation from the enthusiast community. BMW was telegraphing a "Fall 2000" rollout under a 2001 model year designation for the new car.

And then, as late 2000 approached, rumors of delays surfaced. It was all hush hush. No one in authority would comment (this is a recurring theme). But soon it became clear that the M3 would not arrive on time for some reason, and the rumor was that engines were exploding.

After a few months, some semi-official comments were made that indeed a very few motors had failed in testing, that BMW had caught the problem, stopped production, called off the rollout, and was now locating the root cause and re-starting the process.

More details emerged. The failed motors had died catastrophic deaths - broken cranks, thrown rods - major metal mayhem. After initially suspecting that the long inline 6 crank was at fault in the 8000 rpm stressed environment, outside experts were brought in, including none other than tuner Nowack, and the seeming real root cause was located: critical parts in the crank+bearing+block area had been mis-manufactured due to a programming error on the CNC machine which created them. This out of tolerance part had resulted in an oil starvation issue which then overheated the crank and led to overall catastrophic failure (is this sounding familiar yet)?

Because the out of spec parts were already deep inside the stack of motors and M3's being prepared for the rollout, a full halt had to be called on production, and the whole assembly line backed up so that motors could be inspected, removed, parts replaced, and things put right again. This pushed the originally planned ~November 2000 rollout date back 4 months to around March 2001.

===========
M3's arrive
===========

March finally came, and enthusiasts went into a frenzy. M3's were glorious, fun, and kinda raspy it turned out. They clunked a little, but still everyone loved them. Life was good.

And the cars appeared very strong. Some new owners immediately took them to the track, ran them hard, and had not a single problem. Everyone was relieved that the rollout engine health concerns seemed to have been dealt with.

Production trundled along, though it was mighty slow. After shipping around 500 cars in March for the rollout (333 demos - 1 to each dealer, plus a hundred or more customer cars), the following months saw what appeared to be lower numbers - 200 here, 300 there, barely one per average sized dealer in the US. Word was that capacity was totally limited by BMW's ability to build S54 motors at the Munich plant. Long waiting lines continued for eager enthusiasts looking for delivery.

Starting around July and August of 2001, about 6 months into the M3 production run, a couple reports from the US, and from foreign owners surfaced of blown engines. Thrown rods was the word. It was just a couple cases, and both BMW, and the enthusiast community looked at them as likely some kind of driver error - previous generation M3's had been famously easy to mis-shift due to strong gearbox synchros, and one could blow an engine by grabbing 3rd gear instead of 5th at 120 mph and spinning the motor up to 10k or 12k rpm. A few of the failed-car owners protested vehemently that nothing of the sort had happened, but they were generally dismissed. BMW was famous for making very strong engines, and the manufacturing problem from fall 2000 had long since been located and dealt with - at least this was the thinking at the time.

Of minor interest, in mid 2001, BMW ordered an oil change on the M3 - the 5w30 oil which had been shipping on all cars was ordered changed to a new special BMW-only 10w60. New cars came with the new oil, and 5w30 cars were asked to come in for a free change to 10w60. This seemed to be some kind of reaction to concerns over long term high rpm running ability, and the oil having enough high temperature capacity to hold together on such autobahn blasts.

======================
Late 2001, SMG arrives
======================

As the year 2001 progressed, the SMG gearbox was announced for the next model year, to become available in the US in November built cars. Enthusiasts eagerly awaited this new feature. M3's continued to carry huge long waiting lists - stick or SMG - everyone wanted one.

After some minor fiascos with BMWNA trying to neuter the S6 mode of the SMG gearbox, it arrived on US shores, and at around this same time M3 production gradually ramped up to around 2 to 3 times the rather meager levels which had been accomplished during mid 2001.

A few folks pointed out that one of the really nice things about the SMG was that it was incapable of a missed shift. Those few folks who'd seen blown engines - they'd have been ok with the SMG, since the computer would refuse to obey any requested gearshift that might run the engine over its 8000 rpm limit.

In October 2001, around 800 or so M3's came to the US, essentially all stick shifts.

In November 2001, the SMG option came online, and another 800 to 1000 cars arrived on US shores. Around 80% of them were ordered with the new SMG option, the remaining 20% were sticks.

In December 2001, production appeared to drop - perhaps due to holiday season time off, though there were a couple rumors of a production stop that were never confirmed. Something like 300 to 500 M3's appeared to come to the US in December 2001.

At around this time, a couple of running changes were recorded at the factory. On September 24th 2001, the part number for main crank connecting rod bearing shells for the S54 was changed to a new "red" variant. Strangely, on November 14, 2001 the shells were changed again to an older "yellow" part number which had been used early in the M3 production. At about this same time - on November 8th, 2001 service instruction 11 08 01 was issued raising the con rod bearing shell tolerance from 0.03 mm to 0.04 mm along with the advisory that if you ever went to change the con rods in an S54 you had to change the entire carefully balanced package, you could not swap out just a few of them or re-use a subset of the bolts from the old ones.

Things were still going pretty good, and as the 2002 calendar year arrived, most everyone was happy with their M3's - a few clunks, hk rattles and grinding differential notwithstanding.

===========================
Spring 2002 - fails ramp up
===========================

With production cranking along nicely, March through April of 2002 found a series of new reports of M3 engine failures. Most were thrown rods. Some appeared to have been driven hard, others didn't show much sign of it.

BMW's reaction was to aggressively go after the owners of these cars. The engine DME records peak speed, peak rpm, and elapsed time over 7800 rpm. DME's were dumped, and some showed evidence of at least brief operation in the 8100 to 8500 rpm range. As more reports of engine fail hit the Internet ( 3 in March, 5 in April, 10 in May ) BMW began a pattern of regarding any DME data over about 8300 rpm as proof that the driver was at fault, and several stick shift owners were considerably hassled about paying for their $17k new engines.
There was a lot of back and forth. After various threats about the DME data, it became clear that BMW could not really prove when these "peak rpm" events had happened, and it was also becoming clear that the nature of the engine failures didn't really make sense for a pure overrev scenario.

Two major flaws in the "mis-shift overrev" theory came to light:

1. SMG cars were getting failures, and were showing DME data in the 8400 region in a few cases, yet it's impossible for that gearbox to be mis shifted.

2. Damage to failed engines was typically isolated to one or two of the rod bearings - numbers 3, 4 and 5 being prevalent, and there almost never evidence of top end valve-train damage on the other cylinders. If the motor was massively overreved, the first thing to go is usually the valves (at around 10k rpm, not 8.4k), followed by the bearings at some higher rpm range. Fried specific bearings without any valve damage looked - _strange_.

After a lot of agony for what was becoming a couple dozen or more M3 enthusiasts, BMW inched forward to what appeared to be a policy of eventually replacing all the failed motors under warranty, even in cases where BMW wanted to claim driver fault in these 8200 or 8400 rpm readings. BMW also began to show signs of being less accusational towards SMG owners, since they clearly could not have mis shifted their car, and the rev limiter should prevent throttle induced overrevs.

BMW didn't seem to see the irony on being hostile towards stick shift owners showing 8300 rpm on their DME, while being kinder to SMG owners with the same set of failed bearings and a similar DME reading of 8300.

=============================
The Summer of Discontent 2002
=============================

Due to the easy communication of the Internet, enthusiasts began gathering together data on failed M3 motors - one in particular set up a very thorough site to collect up and organize the fail data for analysis.

In June 2002, another dozen failed engine reports surfaced, followed by a similar amount in July. Cars were failing at the rate of about 3 to 4 per week - getting close to one a day, and these were just the reports which happened to reach the Internet.

Some folks screamed in panic. Some argued for calm. BMW very very clearly said _nothing_. When comments could be extracted from BMW staff at any level, the following were the themes:

o Fails are really rare and/or people are lying about it on the Internet
o All engines which have failed were massively overreved
o M3 owners don't know how to break in and maintain their cars - we're lucky BMWNA brought the M3 to the US at all

In general, most folks could already see that these statements were factually incorrect. Many enthusiasts personally knew some of the people with failed motors, so it wasn't all lies. Ample documentation of non-overreved blown motors was available - and an intelligent person would assume BMW had access to this data. And M3 owners were well aware that as a group they were obsessive about their cars and hounded their dealers to do proper maintenance.

As July led to August and September, the patterns in the fail data became more and more clear.

Failed cars were disproportionally coming from 3 specific build months, and more exactly 7 specific build weeks which happened to come almost exactly on a 3 week delay offset from the bearing shell changes back in late 2001. This became known as the “11/01 era” (covering late 10/01, all of 11/01, and early 12/01). Cars made in the 11/01 era were around 20 times more likely to fail than either before or after that time. Massive speculation about the exact failure rate raged on the Internet. Problem being that only BMW had the actual fail stats, and they were not talking. With guesses that perhaps ½ to ¼ of all fails got reported onto the net, a general guess is that there is a background M3 bearing failure rate of somewhat less than 1%. And the 11/01 era appears to carry a failure rate more in the 10% range, and is still rising, though at a slowing rate compared to the summer failures.

More data got cooked, and more things were noticed:

1. Fails had no correlation to gearbox – it was just a cross section of whatever was being ordered for each production month. October fails were all sticks because that what was shipping. November fails were 80% SMG – same proportion they were ordered in.
2. Fails looked very much alike. Main rod bearing death, typically on cylinder 3, or perhaps 4 and 5. No signs of valvetrain damage other than that inflicted by the spun bearing.
3. Data showed no correlation to which oil was used – in fact the cars from the 5w30 era were somewhat more robust that the 10w60 cars.
4. The DME dumps of SMG cars appeared to prove that the factory rev limiter was not functioning well, at least not in some cars.
5. Most 11/01 era cars were failing right around 6000 miles
6. January 02 cars and newer saw a few fails, but not at the rate of 11/01
7. There is a hint of early data that cars newer than March 02 may not be seeing much in the way of bearing failures – BMW may have iteratively fixed the problem
8. Owners have begun to show that a simple $18.50 oil analysis can detect elevated levels of lead in the engine oil that are markers for imminent bearing failure.

And as more experiences were exchanged, BMWNA’s “policy” on M3 failures became more and more clear.

BMWNA’s formal stance is that there may be a few failed motors out there, that customers need to use the right oil, shift properly, break their motors in right, warm them up, and that each failure will be examined on a “case by case” basis. In other words, BMWNA specifically wants to reserve the right to accuse drivers of being at fault.

This despite massively accumulating evidence that examples of driver-abuse being involved in these fails is vanishingly rare. After what we guess to be several hundred cases of replaced motors, we are not aware of BMWNA ever successfully blaming the failure on the driver. They’ve attempted it some, insulted a lot of customers, stalled, forced things into litigation in a couple cases, but never once had the customer pay for the replaced motor in the end.

You can spin that positive. If you own and M3, and the motor fails, you’re likely to get a new motor under warranty in the end.

And you can spin it negative: You may very well go through a very painful and insulting process along the way to having BMW fix a motor which by all logic was flawed when it was delivered to the customer.

And the sad part here is that BMWNA is not saving a dime with all their accusations and ill will with their customers. Engines are being replaced no matter what.

The three specific areas of focus right now are the following:

A. BMWNA continues to examine spun bearing M3’s and look for opportunities to blame their customers despite massive evidence that the problem is in the motor, not the customer. This is insulting to their customers, and is not saving any money nor protecting any reputation.B. Despite several hundred examples of bearing death, and its associated symptoms, BMWNA field techs are required to examine each and every unhealthy motor before dealers are allowed to perform and service, and these techs in many cases are still leading with the assumption that despite the car having the telltale noise and being from the famous 11/01 era, failing bearings is not the most likely cause. Their marching orders appear to be to actively try and NOT find failed bearings, and if they do find them to actively look to blame anyone except BMW. This is again pointless, and inefficient. The rational move is to have a quick and easy test ready for the very large number of ongoing bearing fail cases which are cropping up, and to quickly and politely get new motors into those cars
C. BMWNA is in a communications blackout with their customers on this issue. Customers want to know what’s really going on, and believe that BMW knows more than they are saying. Customers would like to see some kind of confidence boosting warranty on the engine bearings to address owner stress and resale value issues, yet BMW is still in a damage-control mode where they won’t even discuss the problem let alone take ownership of it. This is doing serious harm to their reputation and the feelings of their most enthusiastic customers, and it doesn’t appear to be buying BMWNA anything in return for that suffering. There is obviously a problem with the bearings – everyone can see that, and continuing to let the issue fester helps no one.

As of this writing, there are 107 detailed documented cases of bearing failure in M3 engines, and probably a couple hundred more that have happened but are not documented. That number continues to rise, and both BMW and BMWNA are doing nothing pro active about it.

Starting around July and August of 2001, about 6 months into the M3 production run, a couple reports from the US, and from foreign owners surfaced of blown engines. Thrown rods was the word.

With production cranking along nicely, March through April of 2002 found a series of new reports of M3 engine failures. Most were thrown rods. Some appeared to have been driven hard, others didn't show much sign of it.

So with 9,127 miles my engine decided to self destruct and spit a rod out the drivers side of the block.
I will say that I do drive my car quite hard regularly, but I did not over-rev the engine, and it did have the approiate amount of oil in it, and I was just wondering if anyone else has had this happen to them yet, and if so, are they blowing up for one particular reason?

On Friday, I found out that a rod put a hole through the short block.

Today I was at the track here at Heartland Park in Topeka for my first D.E. event with the M3, (event sponsored by Kansas City PCA, with help from the local BMW CCA and Audi clubs).
My car had complete and catastrophic engine failure! It must have thrown a rod/piston thru' the block since all of my anti-freeze and oil blew out on the racing line, (sorry guys).

I took delivery of a 2002 M3 Cab in Nov of '01, followed the breakin period faithfully and have a total of less than 5K miles-Last week I experienced total engine failure which resulted in a hole in the block

I was driving to work this morning when CLUNK... complete loss of engine power. I was cruising along at 65 mph when it happened. I managed to get over to the shoulder and called roadside assistance. It took over an hour for the tow truck to get me but when the mechanic arrived he opened the hood and told me that there was a hole in the side of my engine block!!

A friend of mine has a '68 Pontiac GTO with 70k original miles on the original engine. The other day, he had it out for a drive. He briefly got it up to 120 mph, but slowed right away. Two or three miles later he heard a vibration. As he was pulling off to the side of the road, the vibration suddenly got worse and the steering wheel suddenly turned abruptly to the left, (counter to his steering inputs). The car spun out of control across two lanes of traffic and into the median. Fortunately, there was no other traffic close by and the car stopped in the median without crossing into oncoming traffic. The car was towed to a local repair shop where the problem was discovered to be a thrown rod. The thrown rod also perforated the power steering line, causing the car to lose steering capability and the subsequent spin.

Why am I posting this here? As my friend related his story to me, I was reminded of all the stories of engine failures on this site. In particular, I thought of those people who have said that, if their engine fails, they'll just have it towed to the dealer and have the engine replaced with a new one. As my friend described his car spinning across two lanes of the highway, I found myself thinking, "Those guys will get their engines replaced IF they survive having their engines fail!" While an engine failure in traffic can be dangerous all by itself, an engine failure can cause more damage than simply the loss of power. It can cause other, collateral damage as well. I'm sure that "steering loss" is not the only type of collateral damage that can occur. These collateral damages can affect the ability of the driver to control the car, resulting in an even more dangerous situation.

I, myself, have some experience with the dangers of a loss of control due to a mechanical failure. I purchased a "premiere edition" Mitsubishi 3000GT VR4 back in 1991. Three days old and it had a drivetrain seizure in the middle of a busy, 4-lane highway. I was fortunate enough to get out of the car before a poor, unsuspecting motorist collided with the back end of it. Fortunately, no one was hurt and the damage was repairable. Nonetheless, a loss of power or control is a very dangerous situation, no matter where, how or why it occurs.
I feel that, given the frequency and severity of the engine failures in the 11/01 and adjacent production months, and the dangers inherent in an engine failure situation, BMW should be pro-active about replacing those engines, before someone gets killed or seriously injured. In fact, other motorists besides the M3 drivers themselves are at risk. The "right thing" for BMW to do is to recall those high-risk production date engines and replace them. From a purely pragmatic viewpoint, it may seem fiscally unsound at this point. However, if someone loses a life, or is seriously injured, the lawsuit and eventual financial payout, not to mention the bad publicity, will dramatically outweigh the fiscal impact of correcting the situation in advance.

Those of you who have been on this board for a while know that I am not some "flaming" BMW hater. I have been on this board for over three years, I own two M3's and I love my cars. Over the years, I have been a big BMW supporter. However, I think that there is sufficient evidence at this point to clearly show a problem with November production engines. In fact, I feel certain that BMW has identified the problem and corrected it in subsequent production runs. If this is the case, then they should also correct the problem in the "at-risk" production run.

We've all witnessed the effects of the loss of corporate ethics, vis-à-vis Enron, Worldcom and Adelphia. While this is not an "accounting" or financial scandal for BMW it is clearly an ethical/financial issue. I think BMW should be an ethical leader in the corporate world. They should step up to the plate and do the right thing. Recalling and replacing these engines is the right thing for BMW to do.

I am sending a copy of this post to BMWNA. If others agree with me, they should send copies as well. A lot of people on this site have said that they plan to just continue to drive their 11/01 cars and, if the engine blows, they'll gladly accept a "free" replacement engine. I think we need to go further than that. I think we need to lobby BMW to do the right, ethical and, in the long run, appropriate thing.

Craig