Carbon Build-up Issue: SURVEY RESULTS - The Unofficial BMW M5 Messageboard (m5board.com)

wilsodh · 13th December 2005, 21:36

Dear board members,

I compiled the data to date from the "Poll on Carbon Build-up Issue" thread and present the results below. I originally posted the "Poll on Carbon Build-up Issue" thread in late August, so these data represent approximately three and a half months exposure to the message board.

The results are discussed in three sections:
I. Top-line numbers (the basics)
II. Data compilation & break-out analysis
III. Graphs

Unfortunately, the limited functions of the WYSIWYG editor limit me from making easy-to-read tables and from pasting the graphs directly into the post (at least, I haven't figured out a way), so you'll have to open the attachment to see the data. Also, I originally compiled the data in Microsoft Excel, but was disappointed to find that I couldn't attach an Excel file to this thread (wrong file format). So I pasted bit map versions into a Word document, which is attached below. If you would like the original Excel file to play around with the numbers, feel free to ask and I'll send it to you through a private e-mail.

At the end is some discussion and recommendations.

-------------------------------------------------
I. Top-line numbers (the basics)

Number of responses: 62
Number of respondents with carbon build-up problem: 22
Percent respondents with carbon build-up problem: 35%
Avg mileage of cars w/carbon problem: 67k (median 61k)
Avg mileage of cars w/out carbon problem: 45k (median 38k)
Non-US cars with carbon problem: 0

Mileage range of cars with carbon problem: 9,000 - 130,000
Build/MY range with carbon problem: 9/99 ('00) - 10/01 ('02)
Oil consumption range with carbon problem: negligible - 1L/700 mi

II. Data compilation & break-out analysis

Please open the attached Word document (below). I compiled the data onto two Excel sheets, bit maps of which are depicted in pages 1 and 2 of the attached Word document.

Page 1: Tally of the information contained in the Poll on Carbon Build-up Issue. Notes:
Links directly to the individual respondent were pasted into the Excel spreadsheet. Clicking on the name takes you to the individual's public profile on the M5board.

Respondents reporting a carbon build-up problem are highlighted in yellow.

Non-US respondants are highlighted w/ blue-green shading.

Page 2: Break-out analysis.
I broke out the data into three categories: 1) mileage; 2) oil consumption; and 3) pre ring-change (<3/00 build) and post ring-change (>2/00 build). I did the pre- and post- piston ring change for both mileage and oil consumption. Post ring-change data are of interest because these engines are identical to those present in '01-'03 cars. Notes:
I pulled non-US cars out of the analysis because it appears these cars are not programmed to monitor secondary air flow. It seems likely that non-US cars would have just a much carbon as US cars, but it is of no consequence to owners when the SES light is not tripped and cold-start emissions are not tested.

The average oil consumption in cars with a carbon problem is clearly higher than the average oil consumption in cars without a problem (although the ranges overlap to a high degree). This does not necessarily mean that issue is due to oil consumption, because the data are confounded by the mileage variable.

I attempted to tease apart the confounded oil and mileage variables with the analysis on pre- and post- ring change. Pre ring-change is highlighted in orange, and post ring-change is highlighted in green.

Interestingly, although the oil consumption for >2/00 cars is approximately one-half the oil consumption for the earlier build cars, the mileage at which the carbon problem appears is approximately the same (~60-70k).

Even after pulling out the two extreme low mileage cars (9,000 & 17,000 mi) from the ">2/00" data set, the finding of comparable mileage between pre- and post ring design is not changed (ie, the milage is approximately the same between the two groups).

The above observation suggests that oil consumption may not be the driving factor in the carbon build-up. If it were the driving factor, we would expect to see higher mileage reported by the ">2/00" group; that is, with cars that consume less oil, it should take more miles to bring on the carbon problem. This does not appear to be the case.

Since the ">2/00" group is representative of ALL later model year engine designs, these data suggest that the carbon build-up problem is a problem facing all of us in the US... not just high oil consumers.

It may that the reason for the larger number of <3/00 cars having a problem than >2/00 cars is simply because more of the older cars have met the higher mileage threshold at which the problem becomes more prevalent.

III. Graphs

Please see page 3 of the attached Word document (below). I compiled distribution plots for both mileage (Figure 1) and oil consumption (Figure 2).

Figure 1: This bar graph depicts the distribution of mileage for both cars with and without the carbon problem. Notes:
Cars without a carbon problem are distributed more to the lower mileage, while cars with a carbon problem are distributed throughout a large range of mileage. The extreme low-lying data point is BKH's car. He clarified in PM's that his situation is that he had a new engine (long block) installed in his '00. At 9,000 miles, it started throwing the SES light ('secondary air flow'). The 17,000 mile car with the carbon problem is the '01 from Hawaii. The range of occurance of the problem is striking, and suggests that there might be a hardware issue (eg, shut-off valve?) at work to pre-dispose some cars to the problem very early on.

We should be careful to interpret these data in terms of the distribution of mileage for ALL cars, which we don't have. For example, the larger incidence of cars with a problem at 50-70k miles than >100k miles may simply reflect a greater number of total cars in the former mileage category rather than a 'risk profile' per se.

Figure 2: This bar graph depicts the distribution of oil consumption for both cars with and without the carbon problem. Notes:
Again we need to be careful not to interpret these graphs as 'risk profiles'. Just because higher oil consumption looks to be associated with the carbon problem, this data may simply reflect that there are a higher number of early build cars (higher oil consumers due to the ring design) that have the higher mileage needed to precipitate the problem.

Discussion/recommendations

I went into this analysis expecting to see evidence that higher oil consumers are at greater risk for the carbon build up problem. Like everyone else, I keep seeing '00's and higher oil consumption typically being the victims of the carbon problem. The break-out analysis highlighted in blue above argues differently. Why don't we see higher mileage being required for a carbon build-up problem among the lower oil consuming group? This was surprising, and provides evidence that the carbon build-up problem is going to be appearing in all model year M5's in the US as the mileage distributions move up with time.

I did not note anything else of interest in terms of precipitating factors for the carbon build up (ie, driving style, type of oil/gas).

There are some limitations with the data. Obviously the statistical sampling is not the strongest. We don't know the total number of US board members who viewed the "Poll on Carbon Build-up Issue" thread in the last three and a half months. It seems reasonable to think that if a lurking board member viewed the thread but did not respond, he probably did not have the carbon build-up problem. The number of US board members viewing the thread would be of interest in determining the absolute incidence. From the known number of total US e39 M5 car owners, we could extrapolate an estimate of the total number of times BMW has seen this issue in their cars.

Recommendations. If we follow the JCFox/pixel example, having a list of 22 people with the carbon problem should be useful in a formal communication to BMWNA. I suspect our group of 'plaintiffs' came from a relatively small sampling of total US e39 M5 owners (many or most of whom are not active on message boards or were not active during the last three and a half months). My suggestion would be to pare down the list to respondents reporting the problem. The information on type of oil and gasoline used by owners reporting the carbon problem might even be useful in usurping attempts by BMW to pin the blame on owners; ie, that that the carbon build-up problem is "owner maintenance" related. The letter to BMWNA might be written "on behalf of" US M5 owners of the M5board, with the list provided as documentation of a representative group based on our limited survey.

This is obviously something for further discussion. Please chime in, and be sure to let me know if you have any questions about the analysis.

(Note: I will post this as both a separate thread and in the "FAQ on the Carbon Build-up" thread to keep everything in one place.)

Cheers,

ALEV8 · 14th December 2005, 04:39

I have 2 questions.

1. Can we prevent this by having the dealer take a look to see if our M5s are developing this problem? What would they do?

2. Should we have the piston rings changed as a preventative measure?

I'll research the issues but now I getting concerned, and if the dealer wants a ton of cash to diagnose this, then we really need to pressure BMWNA..................

KBK · 14th December 2005, 05:10

Dealing with the issue in almost any form imaginable would be cheaper than replacing the piston rings.

wilsodh · 14th December 2005, 05:29

Quote:

Originally Posted by ALEV8

I have 2 questions.

1. Can we prevent this by having the dealer take a look to see if our M5s are developing this problem? What would they do?

2. Should we have the piston rings changed as a preventative measure?

I'll research the issues but now I getting concerned, and if the dealer wants a ton of cash to diagnose this, then we really need to pressure BMWNA..................

1. Personally, I'm not optimistic we will get much useful help on this from our dealers. I'm not sure if a 'diagnostic' parameter or procedure is available to check for progression of carbon build up prior to the SES light being tripped. That said, I agree whole heartedly that we should press BMWNA on this: in particular, what can we do to prevent this issue? I wrote BMWNA a letter today asking this very question, as I think Greg did last week. There may be nothing that can be done, but I'm interested to see how BMWNA handles this type of inquiry. I specifically asked that the letter be elevated to someone who could give a meaningful response.

2. No! I don't see any clear evidence from the Survey that the earlier ring design makes the car more prone to carbon build up. Rather, the data indicate that the carbon problem is observed in both higher oil consumers (early ring design) and lower oil consumers (later ring design) at about the same mileage (mean ~70,000 mi). If high oil consumption is a driving factor in the carbon problem, we would expect to see carbon build-up occuring at lower mileage among the higher oil consuming group vs. the lower oil consuming group. Perhaps with a larger data set we might start to discern differences, but the data we have today suggests that these differences might not be great if they exist.

My suggestion is to write a letter to BMWNA asking about this. The more letters the better. Who knows... we might get lucky and hear from someone at BMWNA who can address the issue meaningfully.

Dave

Philip · 14th December 2005, 07:06

Dave,

Thank you for the time and effort you have put into this, and the well written summary of results. Very well done.

Because I only put about 2K miles on the car per year, I don't see myself entering this fight. If I were to, I'd start by sending this information to BMW's legal department using certified USMail to put them on notice and request a formal response, and contact legal companies that have handled class action lawsuits against BMW in the past. I would also determine if this issue may be covered under state and/or federal emissions laws, and see if a California state agency can get involved (CA is tough on emissions issues).

I understand that that this may be a U.S. only problem due to emissions programming. But, are you sure that there are no equipment differences with respect to this function that is throwing this error? And, can the mechanism/sensor that is detecting this error be defeated/disabled?

You can bet that BMW is aware of this problem.

Regards, Philip

Beamer_usa · 14th December 2005, 07:08

Unfortunately BMWNA will push this onto the individual dealerships, and in turn dealerships will give zero customer satisfaction. I'm sorry for being so negative but with past experiences I've had with my local dealership it has been a nightmare. Don't get me wrong I'm sure there are excellent delaerships out there, but I got stuck with the lemon

tranck · 14th December 2005, 12:52

Quote:

Originally Posted by ALEV8

I have 2 questions.

1. Can we prevent this by having the dealer take a look to see if our M5s are developing this problem? What would they do?

2. Should we have the piston rings changed as a preventative measure?

I'll research the issues but now I getting concerned, and if the dealer wants a ton of cash to diagnose this, then we really need to pressure BMWNA..................

Concerning the piston rings, when my engine was apart, I asked them to check the rings and although my tech said that they looked good, he also said that he could not JUST change the rings because they were specifically made/matched for my pistons.

I HAD the heads replaced earlier due to the carbon-buildup issue. One more thing, even though I had the heads done, when they went over the engine, my tech said that the tops of my pistons also had alot of carbon buildup on them.

An 11/99 build - who knows.

Sometimes I get nauseated thinking of how much $ I've spent on this freakin' car.

Looks and runs good (NOW) though. Tomorrow - again, who knows.

Salty · 15th December 2005, 08:52

Quote:

Originally Posted by wilsodh

Dear board members,

I compiled the data to date from the "Poll on Carbon Build-up Issue" thread and present the results below. I originally posted the "Poll on Carbon Build-up Issue" thread in late August, so these data represent approximately three and a half months exposure to the message board.

The results are discussed in three sections:
I. Top-line numbers (the basics)
II. Data compilation & break-out analysis
III. Graphs

...

Thanks Dave from me too.
A lot of work obviously went into compiling this data.

I don't pretend to understand the secondary air flow (SAF) system in the M5, but I think it is clear that piston ring design and oil consumption have nothing to do with the SAF carbon problem.
Piston top carbon deposits are a related, but separate, issue.

I personally think that the SAF carbon problem is most likely affected by erratic driving, especially idling and stop&go traffic. The engine stays cooler longer and the rich mixture is not optimized for these conditions. More unburned hydrocarbons are going through the exhaust valves under these conditions. JMHO.
I know this issue has been extensively discussed and the likely culprit is a weak/failed shut off valve and/or weak air pump.

The best way to determine this would be if our cars had an operating time monitor, similar to the "hour meter" on most commercial trucks.
This would keep track of the total time that the engine is running, not just the mileage.
Does the OBC hold this information somewhere?

Thanks & regards,
Salty.

Salty · 15th December 2005, 08:55

A quick afterthought...

I wonder if it is possible to have those little metal SAF feeder tubes teflon coated?
That surely must help with reducing the ability of carbon to form and clog them up.
Anybody?

Salty.

macfly · 15th December 2005, 10:11

but why not use the air pump to flush the system for a few seconds every few minutes, and for 10 seconds on turning off the ignition, before the carbon gas cools and solidifies?