oil analysis - lots of silicium ?

Hello,

I have been searching for oil analysis results and interpretation, since I had oil tested which had been in the engine for only 10 miles. The oil filter had been changed as well. It looked troubled but I was not especially alarmed at first.

The sample had 32 ppm iron, 97 ppm aluminium and 149 ppm silicon. I had taken it from the bottom level of the oil volume.

So I had the oil retested, this time from the surface level. And this time they had 27 mg/kg iron, 86 mg/kg aluminium and 230 mg/kg silicon.

Considering the oil had only been in the engine for 10 miles, is this silicon level and associated wear (?) really from a bad air filter ? For 230 mg/kg, there should be some 600 mg of silicon in the oil, which seems a large quantity.

These readings have not shown on any oil results I saw, especially with this low duration of usage. Anyone can shed any light on these ?

MTIA,

Peter

Analysis 1 Analysis 2

Iron 32 ppm 27 mg/kg Lead

Reply to
peter
Loading thread data ...

Silicon can come from several sources: (1) Dust and dirt can carry silicon dioxide... (2) Engine blocks, heads, etc can be made with aluminum silicon metallurgy (Mercedes and others have used this) (3) Sodium silicate is often used in the older technology antifreezes, as an attempt to protect aluminum parts from corrosion (4) Silicone sealants and rubbers have been used in the assembly of engines. RTV (Room Temperature Vulcanizing) is one you may have heard of. (5) Silicone oils have been used as antifoams and as lubricants themselves

Engine oil analysis can be an exercise in interpretation. Some labs have error in analysis as great as or greater than the variation you are seeing.

I am not sure an analysis like this tells you very much of value. Perhaps if we knew more about your car and the fluids you use in it, it might make more sense.

Reply to
<HLS

I can't think of anything else but antifreeze leaking in from a bad head gasket (but .14% H2O rules that out) or an air leak because of a vacuum hose or lack of a PCV breather filter. I'd change the oil, install a new air filter, and retest, possibly with another lab.

Reply to
rantonrave

Reply to
Shep

i cant think of one logical reason to have your oil tested.oh wait.doctors test your blood.that must be it.the car is a living thing.or is it man is really a machine.stupid is as stupid does.have a choc-o-lat!

Reply to
oldkid

Doesn't the lab that does the analysis provide an assessment? Any of the ones I've used in the past always have, plus you would've provided them with a lot more information than you've given us here -- eg make and model, type of fuel, type of coolant, engine mileage, etc, etc.

You don't say what the interval between the two samples is. Nor should it matter where you take the sample from so long as you take it correctly -- full operating temperature, immediately after shut down.

FWIW iron and aluminium levels are probably within the normal variability of the test method. You should also be aware that at least two different test methods are used, but you forgot to tell us which one(s) applies here.

High silicon normally indicates poor air filtration. Diesels are more susceptible than gasoline engines.

It'd also be useful to know what type of filter you're using and what type of environment you're operating in. If you happen to be running a K&N in a dusty situation then high silicon levels didn't ought come as a surprise. If you're running paper with a safety element I'd be looking for an air leak after the filter at that sort of level.

Not without a lot more information.

You've also forgotten to tell us why you've had an analysis done after

10 miles... Perhaps you're digging an underground tunnel and the going is painfully slow -- depending on whether you're using a Caterpillar or a Briggs & Stratton. :)

The usual practice is to take a sample of the old oil at the normal change period.

-- John H

Reply to
John_H

Hello,

Thanks for all the replies.

The point is that the oil was in it only 10 miles, but I was surprised to see it had a troubled colour when I checked with the dipstick. I found an opportunity to do an analysis with two companies - hence the different units - but I took two samples because I did not want to run the engine again. The readings at surface and bottom level corresponded exactly.

The air filter was not replaced, but is only 8000 miles old. I live in a neighbourhood with more sand deposits, due to sand transport traffic without covers. The filter housing and engine compartment was cleaned though, to eliminate sand deposits from storage of a couple of months. There is no air leak after the filter , it is paper.

The car is a opel astra 1.4 XE 1999, has 70.000 miles and was in good condition. The interval between the samples was nil, the test methods can't be that different, results were in in a day, I suspect spectrography. Anyway, both companies said - even not knowing how many miles the oil had been inside - that they recommended immediate action because of these levels. So the figures can't fall in their normal variability...........

They showed a wear factor of 2.81, on a scale where 2 is high and 4 is the highest digit. After 10 miles...

I don't know which liquids are used because it was professional work. The low amount of coolant/water maybe doesn't exclude a leaky head gasket ? It has only run for about 16 minutes.

Please ask when specific info is needed, and thanks again for the replies

MTIA again,

Peter

"John_H" schreef > >

Reply to
peter

I don't know who he used, but the UOE's that I've seen done by Blackstone Labs do exactly that. You have to provide all that information along with the oil when you send it in.

Reply to
Steve

Hi,

I did send in this information and they determined a high wear factor. Meanwhile I contacted the lab technician, and he agreed that the silicon content is harmful for the engine, but that it can be caused by a bad air filter, in extreme sandy conditions. It is left open whether it is originating from the air filter or from accidental contamination.

I will be in touch with the lab analyst who would know the most. I'll post his advice.

Thanks,

Peter

Reply to
peter

**********If engine work had just been performed, then the oil could have taken up coloration and residues from sealant, assembly lubricant, etc. It could also have some antifreeze components in it. Oil may darken rather quickly nowadays.
*******Paper oil filters dont normally leak a lot of dust, and when they do, it goes into the intake, not primarily into the oil. Of course, it can go past the rings and into the oil but this is not the easiest path..
*******They usually use something like ICP (Inductively Coupled Plasma) spectrophotometry for the inorganics. It is fast, sensitive, and accurate if done according to exacting standards (this last item being the more frequent caveat). On a production line analytical basis, you would be lucky to reproduce answers within +/- 10%, although the technique is capable of better.
*******Garbage In Garbage Out.

******No, coolant leak into the oil might indeed explain the silica and the water content. I would change the oil and filters, run it a while longer, and repeat the analyses, if you are dedicated to this sort of technology. This will give the engine a chance to purge the old fluids, and then you will have a better idea if these data are repeatable.
Reply to
<HLS

********It can be useful in large industrial engines, turbines, etc which operate nonstop for months to years. Even so, they are serviced regularly, and the benchmark data means something for the particular installation.
********Particulate value might or might not give you a better clue but I agree that particle count is of interest. For example, sodium silicate from a coolant leak would revert to particulates once the water is sparged from the oil by heat and PCV. There might be iron particulates that indicate wear, or aluminum and silicon particulates might also indicate wear. If you want to make the correlations you need to know something of the morphology and exact composition of the particulates. That is something you easily get with electron microscopy and a SEM. Not something you will get with a production line oil analysis for $25.
******* Even Werner Heisenberg was uncertain about some things.
Reply to
<HLS

Right. Same thing goes for auto engines: if you don't do regular service and you don't get numbers on a regular basis, you don't have anything useful. But if you do service right on schedule, you have a service history, and you periodically get numbers so you have a baseline and can spot trends, it is useful.

For the average driver with a disposable car, this is not useful at all, but for fleet management it can be very useful.

Shouldn't need electron microscopy, just a simple light microscope and an area count. No more difficult than doing a blood count at your local clinic. Actual shape of the particles tells you a lot about basic composition, because, for instance, aluminum is more gummy than steel, and precipitated material tends to form round balls.

It's true that most of my experience with this stuff is with airplanes, where failing to follow maintenance guidelines gets people sent to jail. But we get particle counts, some subjective info about particles, and various metal counts (which include boron, nickel, and chrome so if there is abnormal wear you can sometimes tell what sort of alloy is wearing and guess the part from that). All done with one machine, a light microscope, and a tech who works part-time.

A lot of the reason this is done is that big reciprocating aircraft engines take a _huge_ amount of oil, and you don't want to change it if you can avoid it. The cost of the tech's time and equipment is lower than the cost of regular changes. With a car this is clearly not the case.

--scott

Reply to
Scott Dorsey

"peter" writes in article dated Fri, 04 Nov 2005 14:05:09 GMT:

Actually ppm is the same unit as mg/kg -- namely unitless 10^-6.

Could the silicon be left over from the previous batch of oil, before the change? It's impossible to change 100% of your oil. Doing it with the engine hot helps. I have heard you can clean it out pretty well by running the engine with kerosene instead of oil for 5 minutes. I haven't tried it though, and I'd try it first on a cheap car before doing it to one that I cared about.

-- spud_demon -at- thundermaker.net The above may not (yet) represent the opinions of my employer.

Reply to
Spud Demon

to see it had a troubled colour when I checked with the dipstick.

taken up coloration and residues from sealant, assembly lubricant, etc. It could also have some antifreeze components in it. Oil may darken rather quickly nowadays.

There has been only mechanical work, no sealants were used, no assemblies taken apart. The oil filter had been replaced. I never checked oil I put in myself after 10 miles, so I have no reference ... Should it have remained clear in my case ?

do, it goes into the intake, not primarily into the oil. Of course, it can go past the rings and into the oil but this is not the easiest path..

So if the major possibilities can be ruled out, it is possible it was contaminated ? I am becoming anxious ...

spectrophotometry for the inorganics. It is fast, sensitive, and accurate if done according to exacting standards (this last item being the more frequent caveat). On a production line analytical basis, you would be lucky to reproduce answers within +/- 10%, although the technique is capable of better.

Which means ? I can understand 2.81 isn't really a substantiated figure ...

the water content. I would change the oil and filters, run it a while longer, and repeat the analyses, if you are dedicated to this sort of technology. This will give the engine a chance to purge the old fluids, and then you will have a better idea if these data are repeatable.

But the aluminium content is said to be very high and might by itself justify alarm.

So should it be OR coolant OR sand ? I suspect I should clean the oil pan more thoroughly to be sure it is safe ?

Thanks again for all this advice,

Peter

Reply to
peter

That is something you easily get with electron microscopy

*****Well, you dont need it maybe, but if you run on those particles with the electron microprobe, it will tell you more in a second that you can get with an optical microscope ever.

I think that what we need is competent people to work on an industry standard for oil analyses and evaluations.

Reply to
<HLS

I've used kerosene before, but you dont run the engine with it full of kerosene.

You drain the oil, fill the engine with kerosene or similar aliphatic solvent which will not attack gaskets, etc, and let it stand. Overnight is good enough usually. Then you drain it, replace with fresh oil, run it for awhile, and then change oil and filter a last time.

You dont usually need to go to such an extreme flush.

You will have no damage if you do this correctly, and it will dissolve/disperse/ flush out all but the most stubborn goop.

Reply to
<HLS

But fresh oil with only 10 (or whatever it was) miles? There's no WAY that airborne silicates could elevate the oil silicon contentin *that* amount of time unless you were running through a dust storm with NO filter and shot rings already. Leftover from the previous fill oil? It would ahve had to be a ridiculous concentration. Antifreeze? With no moisture content? Nuh-uh.

Something just doesn't add up here...

Reply to
Steve

peter wrote: I never checked oil I put in

Absolutely not!

The oil will darken as soon as the pump circulates it as it picks up the left over oil.

The silicone numbers mean nothing without a reference point. You are chasing ghosts.

Mike

86/00 CJ7 Laredo, 33x9.5 BFG Muds, 'glass nose to tail in '00 88 Cherokee 235 BFG AT's Canadian Off Road Trips Photos: Non members can still view! Aug./05
formatting link
(More Off Road album links at bottom of the view page)
Reply to
Mike Romain

Hi Steve,

I can only say it was found in three samples with two companies. And that the particles must be micron sized, because other sizes are not being considered in the analysis.

It comes down to what's accepted as normal for this engine. I might try to call Opel and ask their opinion. I'm sure they run oil analysis on combinations of engine, oil, coolant, petrol ?

Unfortunately I cannot check the cylinder walls, unless I buy an endoscope. Any other inspections possible ? I took a look under the valve cover, did not see any particles.

I suspect verifying what's in the oil filter might help.

Am I really chasing ghosts as suggested ?

MTIA again,

Peter

Reply to
peter

Maybe. Since you only had 10 miles of running time since the last change, WHATEVER numbers you got are going to be pretty meaningless. Put 3,000 miles on it and then get an analysis and you will be more apt to see something useful.

The high count could indeed be residual synthetic oil from some previous oil change, or some coolant leakage. Or it could be dust. Get a real count and then you'll at least know whether to worry or not.

Check your air filter and if it looks dirty, change it. Come to think of it, do that every time you change your oil...

--scott

Reply to
Scott Dorsey

MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.