thanks for the spreadsheet but I haven't a clue how to interpret the data. Someone want to clue me in?

 

Just some guesses on my part, but it looks like the first 18 columns or so represent the trace metals found in the oil but I'm not sure what measurement standard it represents. The "Time Oil" column is probably the number of miles driven on the test sample and the "Time Equipment" is probably the total miles/KM on the car.

 

no I understand that but what trace metals are bad and in what concentration are they considered unsafe? Sorry if I was unclear...

 

Sorry, should have explained. The numbers in each column reflect the amount measured in parts per million in the sample. Go to http://www.blackstone-labs.com/report-explanation.php and click on Gas/Diesel. An example report should pop up, and these numbers come from that big middle section. You should be able to click on the element names to see what may be sources of them.

 

Thank you for providing the data. With all due respect it doesn't seem possible to "prove" anything. In order to do so one has to establish a hypothesis, design an experiment or method of data generation, control for extraneous variables, biases, confounding and do statistical analysis (including an understanding if sufficient power is present in order to substantiate or refute the hypothesis). These concepts also pertain to retrospective analyses such as this. Again, without sounding anal if one doesn't do the above one is just deducing a misconstrued result.

 

I'm not claiming one can reject some hypothesis at a certain confidence level using this data. That's not the point. My point is just to show those who claim that there is one and only one oil for the S62 that there are observations out there where the engine did not blow up after using something other than, say, TWS. Besides, we've already had tons of oil threads where someone does prove a certain oil is THE oil. hiha

 

I have "started" analyzing the data. So far I have found the following:

There is NOT much variance in most of the minerals (or metals, or whatever), except for 4 of them.

They are:

Molybdenium
Boron
Calcium
Magnesium

Comparing just BMW 5w-30 to ALL other 10w-60 oils (including BMW), I found the following:

Mo = 10w-60 leaves about 1/8th of that of BMW 5w-30 (8 times as much for 5w-30).
Ca = 10w-60 leaves about 65% of that of BMW 5w-30.
Bo = 10w-60 leaves amost 2 x as much as BMW 5w-30 (half as much as 10w-60).
Mg (Magnesium) = 10w-60 leaves OVER 7.5 x as much as 5w-30. Just scan down the "Mg" column ... you will see (or sort on that column, better yet).
BMW & other 5w-30 oils ... 4 to 252ppm. 10w-60 ... 344 to 797ppm.

No conclusions, but I have seen enough.

Thanks, Nabio.

UPDATE: I am NOT a metallurgist.

 

I guess what I'd say (and trust me I'm not trying to be argumentative), it's impossible to know the driving conditions of all these vehicles, what oils were previously in the cars before the 7 quarts of stated oil. etc. Additionally there is no follow-up so actually you really can't say that any of these S62's did or did not blow up using the stated oil. For the most part the only help oil analysis will allow is for one individual to take his/her car through repeated analysis using different oils or different conditions and same oil and drawing conclusions about the variable one is varying. ...food for thought.
With regards to the 4 elements mentioned above, does anyone know what the "acceptable" levels of these are?

 

You want to concentrate on WEAR METALS- not stabilizers in the different oils.

I would develop a way to look at Lead, Chromium, Iron, Copper Tin and Nickel...then 'normalize' those by the miles on the fill ( so PPM/1000 miles driven)...then sort by oil type- average. See if the different oil types yield any difference in wear metals.

 

I played with the spreadsheet a bit. Deleted any rows that were missing data, and deleted any oil types that had less than 4 samples

I then calculated the value in each column divided by the miles on that sample, then averaged each column by oil type, then multiplied that value by 5000 miles, a typical change interval.



<table style="border-collapse: collapse; width: 761pt;" width="1014" border="0" cellpadding="0" cellspacing="0"><col style="width: 114pt;" width="152"> <col style="width: 30pt;" width="40" span="6"> <col style="width: 35pt;" width="47"> <col style="width: 30pt;" width="40" span="5"> <col style="width: 35pt;" width="47"> <col style="width: 30pt;" width="40" span="2"> <col style="width: 40pt;" width="53"> <col style="width: 35pt;" width="47"> <col style="width: 41pt;" width="54" span="2"> <col style="width: 30pt;" width="40"> <tbody><tr style="height: 12.75pt;" height="17"> <td class="xl67" style="height: 12.75pt; width: 114pt;" width="152" height="17">Oil_Type</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Al</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Cr</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Fe</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Cu</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Pb</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Sn</td> <td class="xl67" style="border-left: medium none; width: 35pt;" width="47">Mo</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Ni</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Mn</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Ag</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Ti</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">K</td> <td class="xl67" style="border-left: medium none; width: 35pt;" width="47">B</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Si</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Na</td> <td class="xl67" style="border-left: medium none; width: 40pt;" width="53">Ca</td> <td class="xl67" style="border-left: medium none; width: 35pt;" width="47">Mg</td> <td class="xl67" style="border-left: medium none; width: 41pt;" width="54">P</td> <td class="xl67" style="border-left: medium none; width: 41pt;" width="54">Zn</td> <td class="xl67" style="border-left: medium none; width: 30pt;" width="40">Ba</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">BMW Synthetic 5W/30</td> <td class="xl65" style="border-left: medium none;" align="right">4.85</td> <td class="xl65" style="border-left: medium none;" align="right">0.14</td> <td class="xl65" style="border-left: medium none;" align="right">12.63</td> <td class="xl65" style="border-left: medium none;" align="right">4.03</td> <td class="xl65" style="border-left: medium none;" align="right">4.26</td> <td class="xl65" style="border-left: medium none;" align="right">1.24</td> <td class="xl65" style="border-left: medium none;" align="right">114.07</td> <td class="xl65" style="border-left: medium none;" align="right">0.52</td> <td class="xl65" style="border-left: medium none;" align="right">0.87</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0.89</td> <td class="xl65" style="border-left: medium none;" align="right">42.61</td> <td class="xl65" style="border-left: medium none;" align="right">5.23</td> <td class="xl65" style="border-left: medium none;" align="right">7.07</td> <td class="xl65" style="border-left: medium none;" align="right">2237.4</td> <td class="xl65" style="border-left: medium none;" align="right">58.75</td> <td class="xl65" style="border-left: medium none;" align="right">774.91</td> <td class="xl65" style="border-left: medium none;" align="right">914.9</td> <td class="xl65" style="border-left: medium none;" align="right">0.15</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">BMW Synthetic 10W/60</td> <td class="xl65" style="border-left: medium none;" align="right">6.58</td> <td class="xl65" style="border-left: medium none;" align="right">0.1</td> <td class="xl65" style="border-left: medium none;" align="right">13.38</td> <td class="xl65" style="border-left: medium none;" align="right">6.14</td> <td class="xl65" style="border-left: medium none;" align="right">4.56</td> <td class="xl65" style="border-left: medium none;" align="right">0.96</td> <td class="xl65" style="border-left: medium none;" align="right">11.87</td> <td class="xl65" style="border-left: medium none;" align="right">0.36</td> <td class="xl65" style="border-left: medium none;" align="right">1.15</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">4</td> <td class="xl65" style="border-left: medium none;" align="right">104.11</td> <td class="xl65" style="border-left: medium none;" align="right">7.12</td> <td class="xl65" style="border-left: medium none;" align="right">6.18</td> <td class="xl65" style="border-left: medium none;" align="right">2180.2</td> <td class="xl65" style="border-left: medium none;" align="right">754.26</td> <td class="xl65" style="border-left: medium none;" align="right">1150.83</td> <td class="xl65" style="border-left: medium none;" align="right">1345.14</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">Castrol 10W/60</td> <td class="xl65" style="border-left: medium none;" align="right">3.88</td> <td class="xl65" style="border-left: medium none;" align="right">0.15</td> <td class="xl65" style="border-left: medium none;" align="right">7.27</td> <td class="xl65" style="border-left: medium none;" align="right">4.96</td> <td class="xl65" style="border-left: medium none;" align="right">3.54</td> <td class="xl65" style="border-left: medium none;" align="right">0.89</td> <td class="xl65" style="border-left: medium none;" align="right">38.87</td> <td class="xl65" style="border-left: medium none;" align="right">0.33</td> <td class="xl65" style="border-left: medium none;" align="right">0.7</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">1.43</td> <td class="xl65" style="border-left: medium none;" align="right">60.26</td> <td class="xl65" style="border-left: medium none;" align="right">4.95</td> <td class="xl65" style="border-left: medium none;" align="right">3.42</td> <td class="xl65" style="border-left: medium none;" align="right">1599.9</td> <td class="xl65" style="border-left: medium none;" align="right">490.12</td> <td class="xl65" style="border-left: medium none;" align="right">754.32</td> <td class="xl65" style="border-left: medium none;" align="right">904.34</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">Castrol 5W/30</td> <td class="xl65" style="border-left: medium none;" align="right">5.2</td> <td class="xl65" style="border-left: medium none;" align="right">0.18</td> <td class="xl65" style="border-left: medium none;" align="right">17.33</td> <td class="xl65" style="border-left: medium none;" align="right">4.96</td> <td class="xl65" style="border-left: medium none;" align="right">3.84</td> <td class="xl65" style="border-left: medium none;" align="right">0.9</td> <td class="xl65" style="border-left: medium none;" align="right">180.9</td> <td class="xl65" style="border-left: medium none;" align="right">0.06</td> <td class="xl65" style="border-left: medium none;" align="right">0.57</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">1.64</td> <td class="xl65" style="border-left: medium none;" align="right">64.69</td> <td class="xl65" style="border-left: medium none;" align="right">13.05</td> <td class="xl65" style="border-left: medium none;" align="right">10</td> <td class="xl65" style="border-left: medium none;" align="right">3328.1</td> <td class="xl65" style="border-left: medium none;" align="right">41.29</td> <td class="xl65" style="border-left: medium none;" align="right">1161.62</td> <td class="xl65" style="border-left: medium none;" align="right">1368.61</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">Castrol TWS 10W/60</td> <td class="xl65" style="border-left: medium none;" align="right">6.63</td> <td class="xl65" style="border-left: medium none;" align="right">0.25</td> <td class="xl65" style="border-left: medium none;" align="right">12.55</td> <td class="xl65" style="border-left: medium none;" align="right">5.4</td> <td class="xl65" style="border-left: medium none;" align="right">5.36</td> <td class="xl65" style="border-left: medium none;" align="right">1.01</td> <td class="xl65" style="border-left: medium none;" align="right">16.1</td> <td class="xl65" style="border-left: medium none;" align="right">0.28</td> <td class="xl65" style="border-left: medium none;" align="right">1.98</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0.02</td> <td class="xl65" style="border-left: medium none;" align="right">1.94</td> <td class="xl65" style="border-left: medium none;" align="right">121.97</td> <td class="xl65" style="border-left: medium none;" align="right">6.73</td> <td class="xl65" style="border-left: medium none;" align="right">3.28</td> <td class="xl65" style="border-left: medium none;" align="right">2288.3</td> <td class="xl65" style="border-left: medium none;" align="right">741.92</td> <td class="xl65" style="border-left: medium none;" align="right">1165.38</td> <td class="xl65" style="border-left: medium none;" align="right">1378.63</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">Castrol TWS 5W/30</td> <td class="xl65" style="border-left: medium none;" align="right">7.2</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">14.25</td> <td class="xl65" style="border-left: medium none;" align="right">5.84</td> <td class="xl65" style="border-left: medium none;" align="right">2.77</td> <td class="xl65" style="border-left: medium none;" align="right">1.1</td> <td class="xl65" style="border-left: medium none;" align="right">242.62</td> <td class="xl65" style="border-left: medium none;" align="right">0.92</td> <td class="xl65" style="border-left: medium none;" align="right">1.64</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">2.09</td> <td class="xl65" style="border-left: medium none;" align="right">81.03</td> <td class="xl65" style="border-left: medium none;" align="right">35.54</td> <td class="xl65" style="border-left: medium none;" align="right">18.94</td> <td class="xl65" style="border-left: medium none;" align="right">3686.9</td> <td class="xl65" style="border-left: medium none;" align="right">25.35</td> <td class="xl65" style="border-left: medium none;" align="right">1383.53</td> <td class="xl65" style="border-left: medium none;" align="right">1734.61</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl65" style="height: 12.75pt;" height="17">Mobil 1 0W/40</td> <td class="xl65" style="border-left: medium none;" align="right">5.55</td> <td class="xl65" style="border-left: medium none;" align="right">0.17</td> <td class="xl65" style="border-left: medium none;" align="right">23.64</td> <td class="xl65" style="border-left: medium none;" align="right">4.84</td> <td class="xl65" style="border-left: medium none;" align="right">4.32</td> <td class="xl65" style="border-left: medium none;" align="right">1.19</td> <td class="xl65" style="border-left: medium none;" align="right">103.66</td> <td class="xl65" style="border-left: medium none;" align="right">0.27</td> <td class="xl65" style="border-left: medium none;" align="right">12.18</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> <td class="xl65" style="border-left: medium none;" align="right">1.98</td> <td class="xl65" style="border-left: medium none;" align="right">164</td> <td class="xl65" style="border-left: medium none;" align="right">5.53</td> <td class="xl65" style="border-left: medium none;" align="right">8.61</td> <td class="xl65" style="border-left: medium none;" align="right">3539.7</td> <td class="xl65" style="border-left: medium none;" align="right">60.09</td> <td class="xl65" style="border-left: medium none;" align="right">1090.58</td> <td class="xl65" style="border-left: medium none;" align="right">1279.61</td> <td class="xl65" style="border-left: medium none;" align="right">0</td> </tr> <tr style="height: 12.75pt;" height="17"> <td class="xl66" style="height: 12.75pt;" height="17">Mobil 1 15W/50</td> <td class="xl66" style="border-left: medium none;" align="right">4.87</td> <td class="xl66" style="border-left: medium none;" align="right">0.09</td> <td class="xl66" style="border-left: medium none;" align="right">10.16</td> <td class="xl66" style="border-left: medium none;" align="right">5.84</td> <td class="xl66" style="border-left: medium none;" align="right">3.51</td> <td class="xl66" style="border-left: medium none;" align="right">2.05</td> <td class="xl66" style="border-left: medium none;" align="right">91.81</td> <td class="xl66" style="border-left: medium none;" align="right">0.81</td> <td class="xl66" style="border-left: medium none;" align="right">0.14</td> <td class="xl66" style="border-left: medium none;" align="right">0</td> <td class="xl66" style="border-left: medium none;" align="right">0</td> <td class="xl66" style="border-left: medium none;" align="right">1.28</td> <td class="xl66" style="border-left: medium none;" align="right">90.24</td> <td class="xl66" style="border-left: medium none;" align="right">4.75</td> <td class="xl66" style="border-left: medium none;" align="right">5.95</td> <td class="xl66" style="border-left: medium none;" align="right">2967.1</td> <td class="xl66" style="border-left: medium none;" align="right">62.62</td> <td class="xl66" style="border-left: medium none;" align="right">1109.08</td> <td class="xl66" style="border-left: medium none;" align="right">1318.19</td> <td class="xl66" style="border-left: medium none;" align="right">0</td> </tr> </tbody></table>

 

I then calculated the value in each column divided by the miles on that sample, then averaged each column by oil type, then multiplied that value by 5000 miles, a typical change interval.

How did you normalize for mileage exactly? And how did you group the oil types? I'm thinking of getting this in a format that can be easily compared but I'm looking for some hints as to what oils are actually the same - though have different titles. E.g. Castrol TWS 10/60 vs. Castrol TWS 10/60 Syn(tec). I have it narrowed down prety well now but a litte reassurance and I can come up with a little graph that plots oil type against ppm; mileage against ppm, etc.

Thanks


PS - Im bored at work and have Excel open 9 to 5 anyway!

 

John,

I would guess that "BMW Synthetic 10W/60", "Castrol 10W/60", and "Castrol TWS 10W/60" are all referring to the same oil.

Chuck

 

BMW 10W/60 and Castrol TWS 10W/60 are probably the same, the numbers look close, but Castrol 10W/60 could be normal Castrol, the numbers look a bit different, so I left them separate.

I have no idea what Castrol TWS 5W/30 would be, unless people are mixing TWS and 5W/30 half and half

 

Wonder what the TWS 5W30 refers to?!?!?

 

It is somewhat annoying that some didn't report their oil properly, but an interesting figure in that table above is the difference between the 10W60 and the rest in iron content. FWIW that figure is half that of most of the others in the column.

 

so which oil preformed the best?

 

There's no way to tell from the data presented. For example, to make any attempt at a conclusion, you would need to know a few other parameters that they measure but aren't represented here, 100C viscosity and fuel dilution %, plus you'd then need to know the duty cycle and ambient environment of the car in question (cold starts from what initial temperature, rpm vs load vs time for the duration of the oil interval, etc). For example, if the set of 10W60 samples includes a reasonable portion that were driven and operated in very cold climates, that could skew one's interpretation of the data. Conversely, if the 0W40 Mobil 1 samples include mostly cars driven as such, attempting to compare such data to BMW 5W30 which has a broad spectrum of operating conditions would be potentially futile.

Ponder attempting to compare data from a Minnesota car driven year-round with 10W60, parked outside, etc, with a garaged Miami car using the same oil. Well, all of these operating environments are intermixed in this data, and making a broad brush assumption that "they all are the same" with respect to key operating environment data, oil dilution and its effects, etc, isn't going to be helpful.

In net, I do not think you can gain much from just looking at this data...at least not much that would stand up under professional engineer scrutiny. By trying to do so, one is attempting to use this data for what it was not intended to be used for.

Chuck

 

Thanks for running the number xr4tic and thanks to Nabio for getting the info in the first place.

I'm not sure which I find most disturbing: the fact that there are more than 4 people running "Castrol TWS 5W30" or the fact that "Castrol 10W60" and "Castrol TWS 10W60" had such widely different results.

In either case, the important note here is that UOA's are only useful in the context of your own engine. My UOA and your UOA should not be compared in terms of raw numbers - that is not what a UOA is for - because 2 different engines may have different wear points and characteristics. Perhaps the folks assembling my engine didn't do quite as good a job fitting the bearing sleeves and I've got excess wear. Or maybe my piston rings are looser and aren't scraping the cylinder walls. In either case, the wear characteristics of my engine may or may not match yours, so comparing UOA's across vehicles is not a wise idea. Multiple oils for the same engine, by contrast, makes sense...

d-

 

from What is Oil Analysis?

Table I. Engine problems predicted with oil analysis.
Indicator Acceptable Levels Engine Problem What to Check
Silicon (Si) and
Aluminum (Al) 10 to 30 ppm Dirt ingestion Air intake system, oil filter plugging, oil filler cap and breather, valve covers, oil supply
Iron (Fe) 100 to 200 ppm Wear of cylinder liner, valve and gear train, oil pump, rust in system Excessive oil consumption, abnormal engine noise,performance problems, oil pressure, abnormal operating temperatures, stuck/broken piston rings
Chromium (CR) 10 to 30 ppm Piston ring wear Excessive oil blow-by and oil consumption, oil degradation
Copper (CU) 10 to 50 ppm Bearings and bushings wear, oil cooler passivating,radiator corrosion Coolant in engine oil, abnormal noise when operating at near stall speed
Lead (Pb)* 40 to 100 ppm Bearing corrosion Extended oil change intervals
Copper (CU) and
Lead (Pb)* 10 to 50 ppm Bearing lining wear Oil pressure, abnormal engine noise, dirt being ingested in air intake, fuel dilution, extended oil drain intervals
Aluminum (Al) 10 to 30 ppm Piston and piston thrust bearing wear Blow-by gases, oil consumption, power loss, abnormal engine noise
Silver and
Tin 2 to 5 ppm
10 to 30 ppm Wear of bearings Excessive oil consumption, abnormal engine noise, loss in oil pressure
Viscosity Change Lack of lubrication Fuel dilution, blow-by gases, oil oxidation, carburetor choke, ignition timing, injectors, injector pump, oil pressure
Water/Anti-freeze Coolant leak or condensation Coolant supply, gasket sealed, hose connection, oil filler cap and breather

 

from What is Oil Analysis?

Table I. Engine problems predicted with oil analysis....

Seems like that table might be a bit engine specific, don't you think? I'm going to go out on a limb and say that an iron block and aluminum block engine just might have different wear indicators. Sure, the crank is still typically cast, but everything else could be one of several metals, so unless you find an engine specific table, those are simply guidelines.

As an example, how do I know if my cylinder walls are wearing? According to the table, I look at Fe (iron), but unless I'm mistaken, there is no iron in our cylinder walls, so I'm pretty sure that isn't it.

It is an interesting table, but not for us without significant custom tweaking by someone who has disassembled our engines and determined what material each part is made of.

d-

 

<o:smarttagtype namespaceuri="urn:schemas-microsoft-com<img src=" ..="" ubb="" redface.gif="" border="0" alt="" title="Embarrassment" smilieid="3" class="inlineimg"></o:smarttagtype> Well, this is my first real post on the board, other than the sign up sheet thread, so please go easy... I know it is an oil thread, too, but I promise to present something reasoned. I’m a new BMW E39 M5 owner in <st1lace><st1:city>Seattle</st1:city>, <st1:state>WA</st1:state></st1lace>. It is the car I’ve always wanted, but also one that really intimidated in terms of the potential for expensive mechanical problems. Heart finally won out over head and here I am with a 2001.
<o> </o>
I have followed the oil and rod bearing threads on this site with some interest. It seems a motor replacement on these cars can approach the value of the vehicle, so I’m curious how one might be able to spot a problem developing. It also seems clear that if one can catch bearing wear early, a $2K to 4K repair might preempt engine failure. Then of course, there is the debate about oil, BMW 5w30 vs. Castrol TWS 10w60. When this thread was posted with actual oil analysis results, I wanted to try to make some numerical sense of both topics.
<o> </o>
THE OILS<o></o>
<o> </o>
The Blackstone data is not always very specific on the type of oil. So, looking at the oil type column and the additive elements (Ca, Mg, P, Zn, Ba), I tried to isolate out those samples that were clearly BMW 5w30 Synthetic or Castrol TWS 10w60. The sample population I came up with for each looked like this:
<o> </o>
BMW 5w30 Synthetic
22 samples
Mean mileage on vehicle: 50,461 miles, standard deviation 28,128 miles
Mean mileage on oil: 6,769 mi, std dev 2,834 miles
<o> </o>
Castrol TWS 10w60
67 samples
Mean mileage on vehicle: 58,118 miles, standard deviation 22,344 miles
Mean mileage on oil: 4,179 mi, std dev 2,248 miles
<o> </o>
So, the TWS folks change their oil more often. Let’s come back to the mean and standard deviation stuff in the next post.

 

--- continued ---
<o> </o>
ENGINE BEARINGS<o></o>
<o> </o>
From the information I’ve been able to gather in other threads, it appears the M5 uses Clevite rod bearings. From the Mahle website, the Clevite bearing is “a copper-lead alloy layer on a precision steel back and finally, a precision electroplated white metal ‘babbitt’ surface layer.” White metal or Babitt metal is traditionally composed of copper, tin, lead and antimony. I surmise that the bearing wear metals one should look for in the standard oil analysis results are most significantly copper, lead and tin, with perhaps iron to capture the steel shell component. Here are the numbers:
<o> </o>
BMW 5w30 Synthetic Samples
---
Copper (Cu): mean 4.6 ppm, stdev 1.9 ppm
Lead (Pb): mean 5.5, stdev 3.9
Tin (Sn): mean 1.7, stdev 1.6
Iron (Fe): mean 14.5, stdev 11.4
<o> </o>
Castrol TWS 10w60
---
Copper (Cu): mean 4.6, stdev 3.3
Lead (Pb): mean 4.3, stdev 7.2
Tin (Sn): mean 0.7, stdev 0.9
Iron (Fe): mean 10.0, stdev 5.6
<o> </o>
<o> </o>
Standard deviation is a widely used measure of variability in a normally distributed population around an average value (mean). One standard deviation from the mean captures approximately 68% of a population, two standard deviations captures approximately 95%. In basic terms, if your oil is showing a number more than one standard deviation off the mean, you are outside of what 68% of other owners are reading. If you are two standard deviations out, you are getting numbers outside of what 95% of your fellow M5 owners get.
<o> </o>
My gut feel looking at these numbers and the individual oil samples (there are some pretty odd outliers that play with the averages) is that you might want to keep an eye on your oil analysis results if you see Copper (Cu) over 7 or so, Lead (Pb) above 9, and Tin (Sn) over 2. Because of the spurious results that pop up in the tests, I don’t think I’d infer anything from a single oil analysis. You’d want to look at several over time.

 

--- continued ---
<o> </o>
CYLINDER BORES<o></o>
<o> </o>
The literature I’ve found online states that the S62 engine in the E39 uses Alusil cylinder bores, which are (as the name implies) mainly aluminum and silicon mixed 70% to 30%, respectively. Here are those numbers for each oil:
<o> </o>
BMW 5w30 Synthetic Samples
---
Aluminum (Al): mean 5.8 ppm, stdev 2.0 ppm

Silicon (Si): mean 6.0, stdev 1.5
<o> </o>

Castrol TWS 10w60
---
Aluminum (Al): mean 5.0, stdev 1.4
Silicon (Si): mean 5.0, stdev 1.2
<o> </o>
<o> </o>
Silicon can come from a variety of other sources, so aluminum is probably the one that matters in terms of bore wear. Perhaps an Aluminum (Al) number above 7 or 8 ppm would be good to monitor.

 

--- continued ---
<o> </o>
FINAL THOUGHTS<o></o>
<o> </o>
These oil analysis data leads me to believe that the 5w30 Syn and TWS are both suitable oils. I’ll admit that I came into this leaning toward TWS, not so sure now. I also think that some of the other oils, Mobil 1 0w40 as an example, probably work very well, too. If you look at the data, it appears there are a couple owners using M1 0w40 and getting very favorable results, comparably. Not sure that oil analysis is going to be a great indicator of bearing wear, but perhaps the wear metal numbers can supplement other observations.
<o> </o>
That’s my amateurish first shot at oil analysis and posting on the M5Board. Would welcome your thoughts.

 

All I can say is WOW!

 

Tarmac,

Your initial posting on this subject was excellent and alot of your insights were what I'd call very close to spot on.

Single spikes, and single metal spikes, should be watched but with a unworried mind. For example: copper can show a dramatic spike with no actual problems present (my 2008 Subaru Legacy GT showed a copper spike of 38 PPM on a 3500 mile Used Oil Analysis). It was simply the oil cooler (copper) oxidizing.

However, certain metal spikes and certain metal spikes in combination should raise concerns. If aluminum spikes, you've almost certainly got cylinder scuffing. This is super rare; but, if your rod bearings are wearing out, you might get a little more piston ring contact on the cylinder.

Lead will just about never spike unless you've got bearing wear. Exceptions are running certain fuel/oil additives or race gas. Otherwise, a spike in lead is a very concerning sign.

Silicon is a real ***** when it spikes. If it's by itself, you most likely have poor air filtration (dirt getting by) or an intake leak. If it spikes in combination with aluminum, cylinder scuffing would be further reinforced.

Iron is a bastard in that there are lots of parts it can come off of. Heck, you can get what's called a "particle streak" in that a single speck of iron comes out of a part and dissolves into the oil. A one-time event though. Iron almost always increases with mileage on the oil, and will spike higher with certain oils just because of their chemistry (i.e. redline will have iron spikes, especially during the first 3 runs of it). A spike of 20ppm Iron looks very alarming, but can mean nothing.

Tin is like lead: you shouldn't see it in more than 2-4ppm on any given UOA. If it start spiking, watch out.

...

Doug Baker said it well: it's incredibly hard to compare UOAs. However, getting a cross-section of all S62 UOAs is fantastic. It shows a few things: 5w30 isn't destroying engines. The S62 can run a variety of oils with no real ill effects. It gives a baseline for expected wear results on YOUR UOA.

Some things to remember though: a used oil analysis is a snapshot of the dissolved metals in an engine's oil. If one car is consuming 1 quart per 600 miles and ran through 10 quarts in 6000 miles, and then changed the oil, then their snapshot is going to be interpreted far differently than the guy who ran 7500 miles and only consumed 2 quarts.

Driving conditions can contribute, but it's important to remember that all is not what it seems. The guy who does alot of WOT in 6th gear at 1500 rpms may be incurring alot more bearing wear than the guy who spends all his days between 4000-6000 rpms. Air and oil filtration are critical.

I ramble on. I'd just like to commend whoever ran Mobil One 15w50 in their s62. It's a fantastic oil that has really awesome cold flow properties. I'm going to go back to see how many UOAs of that there were....also, why no Redline 5w40? That's like the creme de la creme of serious street-driven motorsports oils.

 

Some things to remember though: a used oil analysis is a snapshot of the dissolved metals in an engine's oil. If one car is consuming 1 quart per 600 miles and ran through 10 quarts in 6000 miles, and then changed the oil, then their snapshot is going to be interpreted far differently than the guy who ran 7500 miles and only consumed 2 quarts.

The provenance of each oil sample tested is what is missing as you point out here and I did up above in my post. I would imagine that without some means to normalize (can it even be done?) the data with respect to the oil consumption over the test period, that trying to compare different engine's data will have a high degree of variability, confidence, in the comparison. I change my oil every 5k miles or so, and use less than 1qt (typically) add between changes. Some other cars in the sample might have 3-5 times that amount of oil added per mile driven. I'm also curious how fuel dilution levels measured correlate (or not) with other wear data.

Chuck

 

I just pulled my first UOA from Blackstone, and my car is on line 18, test date 11/04/2004, in that file. Looks like I have averaged a bit under 6k miles/year since then on the car. I'm changing the oil this weekend, so I guess I'll send in another sample to be tested.



tarmac, JoeFromPA,

Thanks for your posts and insights into the data. Any further thoughts about the variability of oil add between intervals, fuel dilution stats (which aren't shown in the spreadsheet) and how that might effect results, etc?

Chuck

 

tarmac,

When you calculated your mean/std dev data, did you normalize the wear metal data for the Time_oil (i.e. so your data would show ppm/Time_oil = ppm/mile)? It seems like that would make the most sense to try to do any comparisons. Perhaps coming up with some combination of Time_oil and Time_equipment to normalize the wear data against would be insightful too?

 

CSB - In a large enough sample size, the variations in driving styles, fuel dilution, and car condition will start to normalize and acceptable medians will emerge. I personally think that's a large enough sample size right now, but we don't have no many individual cars are represented there either.

Honestly, UOAs are not that great for tracking wear. In almost ANY NORMAL ENGINE, you will see what I'll call "noise"...meaning metals will move up or down within even a 10ppm range, and won't mean anything. Why?

because in any normal engine, there is NO metal on metal contact. If there was, you wouldn't be able to get 100,000 miles out of an engine without cracking it open. The oil film keeps the parts seperated and transmits the force of one rotating part to the next rotating part without direct contact.

What you see in UOAs are those moments in time where:

1. Metal on metal contact occured and left some micro-damage on a surface, and the metal then dissolved into the oil.
2. The oil had a heat/pressure/chemical reaction with the metal it's protecting and leached a small amount of the metal into itself.

Every once in awhile, you'll see actual metal wear occuring. This will show up in a huge spike in AL, PB, FE, Tin, etc. I mean like 50-400ppm of a metal that usually sees something like 2-10ppm.

This is usually caused by:

1. A huge amount of dirt getting into the oil due to poor filtration or leaks in the systems.
2. Inadequate lubrication for some reason
3. A coolant leak into the oil which causes the oil to be able to maintain a film between parts.

What you want to identify in these oil results is: did an oil run for a certain period of time (in miles) show any signs of excessive wear that was consistently shown among that oil in S62s.

As far as I can tell, the answer is no: S62s can run anywhere from a 5w30 to a 10w60 without showing (through UOA) any signs within 5k miles that the oil is not sufficiently protecting the engine.

In regards to the one known S62 problem (that seems to occur very randomly and somewhat rarely): rod bearings. I'm starting to become curious if instead of an oil problem, it's a problem with individual builds and rod-balancing. I.e. Is there a problem with specific builds having some sort of odd rod balancing issue causing, at some point, the rods to start contacting the bearing surface more frequently.

Because as far as I can tell (based upon the limited information I know), it has nothing to do with oil choice and proper maintenance.

My .02

 

What you want to identify in these oil results is: did an oil run for a certain period of time (in miles) show any signs of excessive wear that was consistently shown among that oil in S62s.

As far as I can tell, the answer is no: S62s can run anywhere from a 5w30 to a 10w60 without showing (through UOA) any signs within 5k miles that the oil is not sufficiently protecting the engine.

In regards to the one known S62 problem (that seems to occur very randomly and somewhat rarely): rod bearings. I'm starting to become curious if instead of an oil problem, it's a problem with individual builds and rod-balancing. I.e. Is there a problem with specific builds having some sort of odd rod balancing issue causing, at some point, the rods to start contacting the bearing surface more frequently.

Because as far as I can tell (based upon the limited information I know), it has nothing to do with oil choice and proper maintenance.

My .02

Yeah, that makes sense based on what's been observed. Some on the board here over the years have tried to pin rod bearing issues to oil used, and so far no data to support that "feeling" has emerged.

Thanks for the insights on wear metals/causes too. Great stuff.

Regards,
Chuck

 

Joe,
<o> </o>
Thank you very much for the insight on the wear metals and overall guidance on how to look at the oil analysis data. You asked about the Mobil 1:
<o> </o>
0w40 Mobil 1<o></o>
13 samples
Mean mileage on vehicle: 58,382 miles, standard deviation 17,933 miles
Mean mileage on oil: 3,987 mi, std dev 1,185 miles
---
Copper (Cu): mean 3.6 , stdev 0.9
Lead (Pb): mean 3.5, stdev 3.3
Tin (Sn): mean 0.8, stdev 1.2
Iron (Fe): mean 18.2, stdev 10.2
---
Aluminum (Al): mean 4.2, stdev 1.7<o></o>
Silicon (Si): mean 4.2, stdev 1.1
<o> </o>
<o> </o>
15w50 Mobil 1<o></o>
15 samples
Mean mileage on vehicle: 77,293 miles, standard deviation 27,398 miles
Mean mileage on oil: 5,349 mi, std dev 2,040 miles
---
Copper (Cu): mean 5.7, stdev 2.4
Lead (Pb): mean 3.1, stdev 1.2
Tin (Sn): mean 1.4, stdev 1.9
Iron (Fe): mean 9.5, stdev 3.1
---
Aluminum (Al): mean 4.1, stdev 1.1
Silicon (Si): mean 4.1, stdev 1.0
<o> </o>
<o> </o>
The 0w40 users changed oil more often and had fewer miles on their vehicles than the 15w50 users. Again, the results for both oils look acceptable. Iron (Fe) is a very puzzling element in all of the test results, there are often spikes in that number. The 0w40 specifically had a couple of 40ppm iron spikes that affected the averages, otherwise the 0w40 tests were consistently around 11 to 16 ppm Fe.
<o> </o>
Chuck,
<o> </o>
If the Blackstone data is representative of the M5 owner population, I think that the influence of driving style, recent mechanical work, fuel dilution, etc should be captured in the variability of the data around the mean. When I have a chance, I'll try plotting the wear metal readings against mileage and time on oil. Would be interested to hear what you get back on your next UOA and how it compares to your last.

 

Iron (Fe) is a very puzzling element in all of the test results, there are often spikes in that number.

I'm assuming ring wear would be the primary cause of a higher than average iron result. Does anyone have educated opinions as to what would be the root cause of high iron readings if not the rings? Oil pump maybe? Solenoids valves that control oil movement? I suppose anything non-Al in the motor. I suppose if there is a bearing issue in progress then some iron particles off the journals could make their way into the sample?

Would it be safe to assume that if a sample was high (significantly above average) in Fe and Pb that it might possibly be a bearing issue in progress (or a one-time contact issue that might result in something further in the future)?

 

Between this and the "2002 E39 M5 oil" thread I feel like I'm experiencing deja vu.

The analysis being performed is mathematically correct, but is not statistically reliable. It is flawed is several aspects and you are drawing conclusions that may not be valid since your sample size is too small to be statistically relevant (15 points out of X thousand M's), is not randomly selected (it is self-selected), and is not drawn from the total population (results are US only at best). If you haven't yet read "How to Lie with Statistics" I would strongly recommend you do so - it is a very easy read and is pretty entertaining. Not only will it explain the sentence above in more detail, but it will give you a much different view about the numbers and stats you see presented around you in everyday life...

Just to be clear, I'm not saying that the numbers that have been calculated are wrong, I'm simply saying that we need to be VERY careful about how much faith we put in their accuracy and therefore the conclusions we draw.

d-

 

Doug,

If I remember correctly, I stated this earlier: what this data shows us is that there's no huge difference in dissolved metals between using a 5w30 or 10w60 as determined by this type of analysis. In other words: your engine isn't making significantly more metal-on-metal contact with 5w30 or 10w60 under whatever variety of conditions is captured here.

I totally agree, not alot of samples and very little controls. But 194 UOAs on this engine is a good sample size overall. You can paint certain broad strokes with a sample of that size and with the knowledge that they were all tested by the same laboratory.

Just my take. Go m1 15w50!

 

If I remember correctly, I stated this earlier: what this data shows us is that there's no huge difference in dissolved metals between using a 5w30 or 10w60 as determined by this type of analysis. In other words: your engine isn't making significantly more metal-on-metal contact with 5w30 or 10w60 under whatever variety of conditions is captured here.

The comment wasn't aimed at anyone specifically, it was just a caution about using data without knowing its origin or how it relates to the total population.

Just my take. Go m1 15w50!

I'm going to ignore that