ATF change, ATF lab report, ESSO no longer recommended by BMW or GM
 

20,000 mile update: transmission is shifting perfectly well with no issues.

My 2001 3.0i X5 has 83k miles on it, and I decided to change the transmission fluid, even though BMW says the fluid is “lifetime”. BMW does not define what “lifetime” means, but they don’t give warranties past 100k miles. Think about that.

My primary drivers for changing the fluid were that 1) GM uses this exact transmission (A5S390R) in other cars and recommends 100k mile ATF changes 2) ATF just does not last forever (what does?) and 3) I want to get at least 200k miles out of this truck.

While trying to decide what fluid to use, I thought about using the originally recommended Texaco ETL 8072B (also by ESSO?), but, both GM and BMW say that this fluid should not be used, and instead BMW says that their part 83220397114 should be used and GM says that Dexron VI should be used. Presumably, 83220397114 is Dexron VI.

Question: I’ve seen data that says that 8072B is Dexron III, and I’ve seen people say that it is a full synthetic, but I’ve never seen any documentation. Can anyone verify that 8072B is really a synthetic (Group III or higher)?

With both GM and BMW recommending Dexron VI, I did not feel the need to go with the 8072B fluid, but, since I did not find anyone online that had switched their A5S390R to Dexron VI, I decided to go with Redline D4ATF. It seems to be the highest quality (group V base) ATF out there where I can find plenty of people using it without problems in their A5S390R. (yes, I do know that it is not certified Dexron III or Dexron VI)

If I could have found several people that went with Dexron VI in their A5S390R, I probably would have gone with Valvoline’s full synthetic Dexron VI (does anyone know what base stock that is?). Judging by the fact that Valvoline claims a full synthetic Dexron VI and pretty much everyone else says their Dexron VI is a blend, I have to assume that Dexron VI does not automatically equal full synthetic (as some people seem to claim).

In any case, the fluid change went very easily. I drained the fluid, dropped the pan, changed the filter, put everything back together and filled it per the BMW procedures. It took about 6.5 quarts. Even though it is a different tranny, the fluid and pan looked about like this, which, while the oil was very black, the plan was cleaner than I was expecting. Two days later, I did another quick fluid drain and fill (no filter change) to get at most of the rest of the old oil. It took six quarts the second time (no pan drop or fitler change). One thing to watch out for: part of the gasket of the original filter stuck up in the mounting hole and I had to reach in and pull it out. It was not a big deal, but it would have been if I had left it in there.

I sent the oil from the first change into Blackstone for analysis. Here are their comments: “Judging by what we found in this sample we don't think your BMW's transmission fluid has been changed before. All wear is above average but copper was the most out of line. Iron, copper, and tin all come from the clutch plates, while aluminum is from the torque converter. Silicon is likely from sealers used when assembling the transmission. Insolubles at this level may show sludge in the system. We think a few short oil runs should clean up your transmission. Try 10,000-15,000 miles on the next oil. The TAN read 3.6, showing the oil was very acidic.”

Here are the interesting numbers from the report:

Aluminum 90
Iron 192
Copper 417
Tin 16
Silicon 51
Insolubles 0.3%
TAN 3.6

Here is a link to my full report: D67388.pdf - Windows Live

None of this is very good, especially the TAN. I think that shows that BMW’s lifetime fill is really intended to mean 100k. I do not think my transmission was going to go another 100k with numbers like this at 80k. I suppose mine could be a fluke, but I doubt it.

The transmission was working fine before the change, and it is working fine now. Maybe it shifts a little quicker and smoother, but not much.

Thanks for sharing your experience!

Great notes. Personally I would have just gone with what BMW currently specs for the car, but to each his own. Great idea to send it to blackstone.

Good post, and thanks for the information on the lab report.

A couple of comments from my experience with oil and hydraulic fluid sampling:

The lab report, while interesting, becomes much more valuable if you can plot a trend line from multiple samples. Isolated numbers, by themselves, don't tell very much. For the total acid number, you could take a sample of the original factory oil (from the dealer) and see what the change has been. For the other items listed, they weren't zero in new fluid and it helps to see when they climb, when they stabilize, and when they spike. You want to change it before they spike, essentially. If the aluminum got to 80 during break-in, for example, being at 90 now isn't an issue. If it was 5 for the entire life of the transmission and went to 90 last month, then now is exactly the time to change the fluid.

Did you do any large particle analysis? The dissolved solids info is fine, but I would look for a particle count as a key indicator.

My other comment is about the comparison with the GM transmission. I have no doubt that the mechanicals of the transmission are the same as the GM product (or ZF, for the ZF-built transmissions). What BMW does do is add their own control package, and electronics. Those electronics and sensors are programmed to provide different shift points, and different shift characteristics than the GM electronics (which is why BMW puts them in there). The way in which a transmission shifts has a large effect on the wear and heat resulting from a shift. Essentially, a firmer shift reduces the time the clutch packs are slipping, reduces heat, and so on. All of that extends fluid life. A soft, slow, shift that slips the clutche plates more results in faster wear of the clutch plate material and disks, and much more heat (which breaks down the fluid much more quickly). While the transmission durability is likely similar between the GM and BMW versions, the above means that not everything can be extrapolated. That doesn't mean that you shouldn't change your fluid if you want to, but it means that the GM recommendation for 100,000 is simply one more data point, and not an absolute truth.

I was very glad to see that after you changed the fluid, you did a second change. If you are going to change it, it makes sense to get more of the old fluid out and if you got 50% of the old fluid out each time, after two changes you are at a 75% change, considering the amount of fluid that stays in the transmission/torque converter when you drain the pan.

Thanks again for your post.

JCL,
Is there a way to perform a 100% fluid change at one time.

If you drain the transmission, torque converter, lines, and cooler, then yes. A sure way of doing it is to remove the transmission. Not worth it IMO.

I understand that, but given that GM made the transmission internals, and they use that same package in their cars, it does give what they say a lot of weight. I don't think that a reprogrammed shift pattern is going to take the fluid from lasting 100k miles to lasting forever. Whatever you think about that, the high TAN alone from my fluid says it was time to change, despite what BMW says about the fluid not needing to be changed.

Here is a link to my full report: D67388.pdf - Windows Live

I looked into this. The torque converter does not have a drain, so you can't easily get the fluid out. The only options involve opening up the ATF cooler lines and swapping out the fluid while the car is running (many shops have a system that will push fresh fluid back in while it takes out the old fluid via the cooler lines), but I've hear of lots of bad results from those things, so, I don't think there is an easy way to get all of the fluid out in one shot.

However, doing the second fluid change, just drain and fill only, was very easy and took less than 30 minutes.

John,
Thank you for the excellent info. This assured me that I did the right thing when I'd changed mine out at 50K miles when everyone else on the planet told me to leave it alone. I even had dealers that looked at me funny and refused to do the change.

Thanks!

BMW currently specs Dexron VI, and I would have gone with that if I could have found a few people that had already switched their A5S390R to Dexron VI with good results, but, since I could not find anyone that had, I decided to go with Redline D4 ATF, since I have seen many people use it in their in their A5S390R with good results.

?? got the breakdown of the elements and everything huh,:bustingup Thats just way to deep for me to worry about.

I'm with you. '01 at 47k and going to change this summer.

Thinking about diff fliuds too. What do you guys think. JCL, I think you said before not to change the Transfer case, just diff?

Tambalong, I agree. After you turn it on one of two things happens:
1. Car go.
2 Car no go.

Thanks for the full report. A baseline number would be really helpful as well, if anyone out there has a low-mileage Blackstone lab report on the factory fill fluid.

In the 'good old days' (prior to electronic shifting controls) we dealt with automatic transmission fluid overheating by installing 'shift kits' from B&M and other performance parts outlets. They were essentially revised check valves and springs for the transmission valve body, which reprogrammed the shift points for higher revs, and firmer shifts. They were good for trailer towing as well as performance V8 applications in muscle cars. They reduced fluid temperature, and the likelihood of clutch damage. It is not the shift point (in rpm) that matters as much as the shift characteristic, since clutch plates wear when the shift is happening, and not between shifts. That is so true, that for some applications engineers commonly measure transmission life by the number of shifts, not the number of hours or miles. So, IMO, changing shift characteristics has a significant effect on fluid life.

I fully agree that it doesn't make the fluid last forever. Just like the famous story of the pack of gazelles in Africa that is pursued by the lion, however, you don't have to be the fastest gazelle. You just can't be the slowest. Applying that to transmission life, my point is that it doesn't have to last forever. It just has to exceed the useful life of the rest of the vehicle, whatever that may be deemed to be.

The Dexron specs (which have been improved since your vehicle was produced) list various oxidation and stability tests. The TAN indicator varies by transmission manufacturer, but some published numbers I have seen are that TAN cannot increase by 3.0 or 3.25 over initial baseline numbers, whatever that baseline number is. I don't know what the initial TAN number is for your fluid, but there are other oils out there on the market in the 1.5 range. By that calculation, your fluid was about halfway to the change limit. All of this is an estimate, but I wouldn't say that your fluid was dangerously high. It was trending towards a replacement point, absolutely.

Interesting topic, thanks for the post.

I think that the differentials and transfer case do not have the same issues around the risk of changing as the transmission. They are straightforward drain and fill applications. If you are going to change the differential fluid, you may as well change the transfer case fluid at the same time. I think of them as 100,000 km or 60,000 mile change points, but that is just my opinion.

I would ensure that you use factory OEM fluid in all of them, as BMW does not publish good (universal) specs for aftermarket fluids and there is nothing wrong with the factory fluids, and little to be gained by playing roulette with aftermarket blends. That also covers the risk of not using the correct fluid for the clutches in the x-drive transfer case.

I did the transfer case and diff about a year ago:

http://www.xoutpost.com/x5-e53-forum/...-diff-oil.html

http://www.xoutpost.com/x5-e53-forum/...que-specs.html

:thumbup:

FWIW, I'm at least 15K miles with 2 drain bolt changes on the transmission, with the 3rd being a full pan drop/filter change.

Transfer case & diffys were done around 15K miles ago as well.

1st drain bolt was using OEM Esso.
2nd and Pan Drop was using Amsoil ATF.

Amsoil ATF in the TC and Severe Gear 75W-90 in the diffs for me.

The only fluid I had to mess around with was the PS, and I settled in on M1 for this app. The viscosity of the other 2 were slightly lighter and definitely did change steering feel characteristics.

Anyone replacing their drain bolts with ones that have magnets in them?

I don't think it makes much of a difference, but I did one for the oil pan:

http://www.xoutpost.com/x5-e53-forum/...oil-diffs.html

The ATF pan already has a magnet in it.

:thumbup:

You sure that Honda one will fit? There isn't some BMW reseller that sells them anymore?

John -
Curious do you do trending UOA's on your oil ?
I'd love to compare ....

That would be good to know that kind of info. There is a guy with a great Z Coupe website that does. Check it out:
Bill's Web Space:2007 BMW M Coupe (E86)

Nope. I just did a one time sample.

Even though I don't have a baseline, my copper was so high, I would say that I was at or beyond the replacement point.

John,

I spoke with a very knowledgable / trusted BMW independent and he said the lifetime ATF BMW uses is complete BS. In the 2001-2003 models the ZF tranny used is a weak point with tranny life of 100-150k. Unless BMW has reinvented the tranny and not told anyone; the lifetime oil bit is a marketing ruse. The metals grind and the shavings need to be emptied with the oil. I guess BMW wagers that 4-5K repair once every 125k miles or so is not enough to cause owners to riot.

Too bad it's not so easy to take a fluid sample from the transmission or transfer box, etc. to get it tested periodically. Ironically the fluid that is easiest to access for testing sample purposes is also the easiest to change.

Agree the copper is high, but if it was from a thrust washer or clutch plate I would have expected the zinc or tin to be high as well. Since it is so high by itself, my guess would be that it is from the cooler, and thus benign. Just an opinion.

There are various sampling kits available if you want to take regular samples. There are tap ports that can be installed, and extraction pumps that are simple to use.

There seems to be a perception that fluid sampling is about the fluid. While the TBN and TAN numbers do give some visibility into the state of the fluid, I have always seen oil sampling as more useful for telling a story about the component in question, much more than the oil itself. By trending wear metals over time, the fluid analysis can be used to catch a failing component prior to failure, and thus support a repair-before-failure maintenance program. In heavy equipment (high $$$) or aircraft applications (safety) there is good reason to do this. In an automotive application, are we really expecting to catch a failing transfer case prior to failure? There are a lot of other indicators that it will be failing, and I suggest that most of us would let it go close to failure in any case. I recall one specific customer who struggled with justifying his oil sampling program, as he looked back at several years worth of samples on 20+ pieces of equipment and concluded that he had in fact rarely been willing to take a machine out of service based solely on the oil sampling results. He saw a benefit in being able to predict when his downtime would occur, but didn't realize a reduction in his maintenance cost.

Oil samples are in and of themselves interesting, but they don't give us much guidance as to how to maintain our vehicles, without trend lines and correlations to failures.

Some good info. In regard to my comment though, my point was that given the almost ridiculous notion of "lifetime fill" I think people would be more apt to change their ATF (and similar fluids) if they were provided with some kind of objective analysis as to the wear rate of the components in question as determine by their fluid, doubly so if those fluid were easily accessible. I know I would.

They have objective analysis on component failures, from their warranty system. They know how long things last, on average. They know how much it costs to fix them when they fail.

Wear rate prediction from fluid analysis is complicated by the complexity of each system. Look at the oil sample posted above. It will be widely assumed that high copper is a problem (it is dissolved in the fluid and not in chunks in this case, unless an LPA showed a high particle count, so I am not sure that it is such a problem, but that is an aside). The copper reading says very little about the fluid, and not much about the transmission. There are alloy thrust washers, so if there were high levels of tin or zinc, it could be a problem. The copper level may just be telling us quite a bit about the cooler. Would anyone put the oil cooler on their PM list, even though this one sample data point suggests it? Probably not. We like to think the transmissions are the problem. And is the cooler really a problem, or just a source of leached copper?

We use scheduled oil analysis in my industry for tracking the source of engine problems, before they become catastrophic failures. Trace metals designed into different bearings/bushings/etc provide signposts to a component that is about to fail, and save searching through the engine for the problem item. That is worthwhile because the cost of downtime in these cases (heavy equipment) is much higher than the cost of analysis. With a vehicle that is as replaceable as a personal car, it is hard to justify the expense. You are more likely to invest in electronic systems that flag the failure after it has happened in order to reduce the diagnostic time, and not even trying to prevent the failure (which is exactly what the car manufacturers have done with the engine electronics).

We have to go back to the fact that oil analysis isn't really about the oil. That is why there is one TAN result, and 12 or 20 or more element tests. It is about the component or system. Pretty much the same concept as a blood pressure reading not being a cause to change your blood, but rather an indicator of a system problem. I am not a doctor, but the analogy seemed apt. All you medical practicioners, please feel free to correct me.

fwiw, trending is also only as good as the sampling procedure as well as the lab itself. Different labs have different mechanisms of testing, which skews results if you flip flop between labs.

John, I'm confused. I did drain/refill on my 02 3.0 without dropped pan and filter change, only pumped 5.0 Quarts in. What was the temp of transmission fluid when you had your done? I did it at 36 degree celsius

I did not have a tempature read on the tranny; I just waited until it felt warm, but not hot, and then topped it off.

Temperature of the fluid affects the level of the fluid. I did an experiment when I had an Audi Quattro with a ZF 4HP18 trans. I filled a glass bottle with ATF that I drained from my Audi and mark the level with tape at room temp (approx 75F) Since it was winter I left the bottle outside overnight and took a reading of the temp of the ATF (28F) and noted that the fluid level had dropped by quite a bit.

The point of this story is that the trans has to be cold when you re-fill it or the final level will not be accurate, especially if the ATF fluid in the Torque Converter is at the engine operating temp.

When I do a fluid change I jack the car up the day before and drain the fluid. I let the car sit overnight and refill the next day.

That is not correct. The fluid has to be at a specific temperature to check the level, not cold. Otherwise you will have the wrong level. See the maintenance procedure.

There's a range that the ATF has to be I believe it between 30C to 50C, not a "specific temp" like 38C.

The point of my post is if you drove the car around and then perform the trans drain and refill, since a lot of the ATF is still in the torque converter the ATF temp is well beyond the "range" that is in the procedure. Running the car with a cold engine and trans during the fill procedure will eventually get you to that "range" that is outlined in the procedure.