Oil Analysis Benchmark

[REBrueckner]

If you think oil analyses are quick, simple and obvious, and that lab results are consistent and specific to your engine(s), get ready for a rude awakening. I got some real surprises.

If anyone can post Detroit Diesel wear standards here I'd love to see them.


In 1999 I had an oil analysis report from Ana Laboratries Inc in Bellmawr (That's how they spell it), NJ on my 8V71TI's. At the time they had about 1500 hours total time.

Here is what they showed as "acceptable levels of typical wear metals", in parts per millon. (These are NOT likely Detroit specific standards, but were the standards on my Detroit diesel report.)

Aluminum 0-15 bearings,bushings,valve guides
Copper 0-30
Lead 0-20 not applicable for gasoline engines
Tin 0-20
Silver 0-5

Nickel 0-10 valves
Iron varies greatly
Aluminum 0-25 pistons
Chromium 0-10 or 12
Molybedium 0-15 ring/liner (cylinder)
Iron 0-145


"Acceptable Level of Oil Contamination"

Sodium variable oil additive,coolant,environment
Boron variable oil additive, coolant
Silicon variable oil anti foam agent,anti freeze
Silicon 0-25 dust,dirt or other abrasive
Water % by vol 0.1-0.2
Solids (Carbon,soot) 0-1.5% blowby product
Fuel % by vol 0-2% leak


Both engines were reported as "satisfactory" with 75 engine hours on the oil.

Right after the above analysis, the 8V71TI engines were disassembled and checked. Inspection showed port engine cylinders were scored, so I had liners,pistons,rings, replaced while the engine was disassembled. Iron and Chromium and iron readings at that time were LOWER for the port engine than starboard. (and within "standards".) So the scoring was not reflected in the analysis results.

Now here comes the interesting part.
I just had reports done (November 2006 oil samples) with 60 hours on the oil. (Same engines, about 2500 hours total run time, about a 1,000 more than in 1999.) These were from H.O. Penn CAT ,Newington Ct. On the back or the analyses it says "Possible sources of high readings for Caterpillar Product" which makes me suspicious.

The only standard on the new report is fuel dilution: 0 to 4% is "negative". (Over 4% is positive...boy, that seems incredibly high.) And that's twice what the other lab says!!!! And their descriptions of engine wear components have such phrases as "molybdenum....certain engines piston rings.." Is this report Detroit specific?

On the 1999 report, phosphorous and zinc are shown under "wear protection", and magnesium and calcium as "detergent", but on the new report are listed as "wear elements". Phosphorous is not even listed in the 2006 report. So I'm not even sure if a higher reading is better or not. This time around, my readings were generally higher than earlier.

For my engines, the new report showed everything was normal except stbd engine Chromium at 5 (PPM) was reported as "elevated". I thought "oh shucks" until I looked at the old report: Back in 1999, and about a thousand hours of running earlier, Chrome was reported at 12 and NORMAL!!! Port was 8 at that time.

Here are some other interesting comparisons: The current starboard engine iron and chromium readings (ring and liner wear) are lower than in 1999!! And currenty starboard engine iron is lower than the port...not what I would have expected!!! Chrome is higher.

What I also don't understand: the port engine (with newer liners/cyliners/rings) currently shows some sulfur, oxidation and soot readings ("under oil contamination", apparently within limits, while the "old" starboard engine shows NO readings!!! Both engines have RACOR CCV's for about five years. The "old" stbd engine smokes a bit more during warm up than the port;both exhausts appear clear at cruise. so I would have guessed it might be the one with a bit of soot,etc.

And of course you have to wonder how/if reports REALLY take engine running time on the oil sample is taken into account. The earlier report (at 75 hours) said "oil ok for continued use" while the new report (at 60 hours)suggested changing oil and filters even though one engine showed NO contamination readings and normal viscosity and the other was in limits. The new report suggests oil samples at 250 hours for transmissions and engines.

Any comments welcomed. This was my first experience with repeated oil samples and hope it will be of use to others.

 

[SKYCHENEY]

On those lab papers that you send in with the samples did it ask for the engine make and model? What I'm getting at here is did they know what the samples were from so that they could relate them to the proper limit charts?

 

[REBrueckner]

Yes, the engine make is correctly printed on each report as I submitted,and so is the oil brand, weight and hours on the oil.

In fact, on the other recent thread about oil sample analysis, I was just noticing his analysis apparently had no oil contamination results...no sulfur, oxidation, soot, % allowable readings which my recent one did. My old one showed oil contaminanats under five categories: Sodium, Boron, Silicon in PPM, water and solids via % allowable.

And Sodium and Silicon, Na and Si, are shown in my recent reports under "wear elements" while the old report showed them as "oil contaminants" while the old report shows ' " "solids" (carbon, soot)' by % also referred to as "sludge"....Maybe this is equivalent to "soot "plus " oxidation" on the more recent report.

So here are three labs with at least some analyses category differences ...maybe all minor, maybe not, I don't know...but they make thorough year vs year analyses tougher.

I guess one lesson is to stick with the same lab although there is no guarantee their categories will remain constant over time.

 

[REBrueckner]

One of the labs I used is online at http://www.analaboratories.com/commonrootie/oiltests.asp.

They only show normal 40 wt oil viscosity as 12.5 to 16.3 when tested at 100 degrees.

A description of a number of tests is also provided.

 

[REBrueckner]

At the above website, they have a lot of good information. For techno geeks its worth a read.

For example, one description says Low calcium,Low magnesium,Low zinc,Low phosphorous IS a problem...so higher seems better as I guessed in my original post...

 

[REBrueckner]

Here is another gem from the ANA website:
"..Sample Proration.

Oil samples should be taken at regular intervals. The accumulation of contaminants such as wear metals and soot are time dependent. That is, the longer the oil is in the engine, the higher the levels of contaminants will be. If sampling is too irregular, an adjustment to the data is necessary before it can be interpreted. The data needs to be prorated relative to oil mileage or operating time. Generally, no adjustment is necessary if the samples are taken within 10% of the sampling average..."

So my two samples, one at 75 hours, one at 60 hours, should be prorated before matching. I assume that means a 25% increase in the last readings would then be comparable to the old readings 75-60/60 is 25%.

But this raises other questions, such as whether standards for a given engine are given at some fixed hours of use interval...we still need some fixed standard of comparison....
I know less now than when I started!!!!

 

[Boss Lady]

A sampling program is properly conducted for the life of the engine at regular service intervals. Oil analysis is a tool that can help detect early failure potential. If you maintain a log for a particular engine you can totalize the parts per million and determine by volume a relative amount of disolved metals. A standard paper clip if dessolved into a quart of oil will give you a 500 ppm reading. Using that as a reference then you can begin to understand what some of these numbers mean. If over the life of an engine you are seeing 150 ppm iron or other metal at every oil sample then just total it up and you can see about how much has worn off and is contained in the oil. Another issue is how long the oil is used and the ability of the testing facility to detect very minute quantities. You also need to calculate the total amount of oil, say for instance that you have a 150 ppm reading for 5 quarts that would actually represent 5 x 150= 750 or about a paper clip and a half's worth of iron. Multiply that times 20 quarts and now you have a "LOT" of wear material in the oil. 20 x 150= 3000 or 6 paper clips! If this material is rings or pistons you are toast pretty soon. The other dilema is where a particlular wear material originates, is it normal wear evenly distributed over all like wear items or is it localized from say an individual failing camshaft lobe? Oil analysis is a very good tool if you understand what you are looking at, if the information is not charted properly it will mean nothing. Also an individual sample has little meaning, for this to become a valuable tool you need to develop a reference so you can see where the engine is heading.

 

[Boss Lady]

I forgot to mention that current oil analysis methods are for 1 micron particles or smaller, gross wear products are not measured, these are usually captured in your oil filter and need to be analyzed seperately. The labs will tell you that the micron test is relative to the gross wear tests, as a rule this is fairly reliable but not always true. Hence the gentleman stated that his micron tests did not reflect the scoured cylinders. Check your filters too.

 

[REBrueckner]

Here's a brief summary of useful tips from the ANA laboratories site on oil analyses.

"...combustion soot makes its way into the engine oil past the piston rings due to blow-by, engine timing, and operation. The quantity of soot in the engine oil is related to the engine operation, not the oil...

"...water contamination may be detected effectively with a crackle test. This test is conducted either by dropping a small quantity of oil in a heated aluminum pan, or by immersing a hot electric soldering iron in a sample of the oil. A crackling noise indicates the presence of water....

"... Nitration products (Ni) are formed during the fuel combustion process in internal combustion engines....These products are highly acidic...Nitration indicates excessive "blow-by" from cylinder walls and/or compression rings...


"...An oil should be changed when the TBN drops to 1/3 of its new oil value. If high sulfur fuel (greater than 0.5% sulfur) is used, the oil should be changed at 1/2 of new oil TBN. TBN should never be below 2.0. TBN and viscosity are the most important parameters used to establish the condition of the oil.... These must be monitored for optimizing oil drain intervals

(So why don't I see a TBN number on my oil analsis report??)

"...Warning limits should be specific to an engine or a family of engines within a fleet. The use of general warning limits published for single sample analysis are often used by oil analysis laboratories and are inappropriate for fleet monitoring.

"...Oil samples should be taken at regular intervals. The accumulation of contaminants such as wear metals and soot are time dependent. That is, the longer the oil is in the engine, the higher the levels of contaminants will be. If sampling is too irregular, an adjustment to the data is necessary before it can be interpreted. The data needs to be prorated relative to oil mileage or operating time. Generally, no adjustment is necessary if the samples are taken within 10% of the sampling average.

"...The most common problem encountered in interpreting used oil analysis is not missing an impending engine failure, but reaction to false alarms -- unnecessary actions taken based on a single oil analysis. Many times an engine is unnecessarily disassembled, only to find that no engine problem exists. For this reason, under no circumstances should an engine be disassembled based on a single oil analysis unless there is mechanical substantiation.

"...The most common problem encountered in interpreting used oil analysis is not missing an impending engine failure, but reaction to false alarms -- unnecessary actions taken based on a single oil analysis. Many times an engine is unnecessarily disassembled, only to find that no engine problem exists. For this reason, under no circumstances should an engine be disassembled based on a single oil analysis unless there is mechanical substantiation.

"..When a critical level oil analysis is received a resample at 1/2 the interval that the critical sample was taken or at 1/2 the normal sampling interval is recommended. If the critical level was prompted by fuel, coolant, dirt, soot, or other identified contaminant, a correction should be made as soon as possible. Reduced oil drains may be needed until the correction is made. Remember to prorate the test result of the 1/2 interval sample to obtain the correct action level..."