Oil Analysis lessons

[REBrueckner]

Not that anyone needs convincing about lube oil analyses, but here is a current unfolding story.
During the survey of 1976 Cummins 903 VTA's we got an oil analysis on one engine with the following SEVERE readings. We do not know the hours on the oil, but it's likely been sitting mostly unsued for three or four years. In the Chesapeake Bay area. The other engine had all normal readings:[all in parts per million]
Iron, 456
Aluminum 70
Copper 200
Sodium 453

Other readings had some high results warranting 'monitor' .

So I called the lab and spoke to an ex navy mechanic, now working for the lab on analysis of test results. Here are the notes I made:

" Recommends NOT running this engine as test results indicate SEVERE wear and existing engine damage. The wear metals indicate bearing and or bushing wear and the extremely high sodium suggest s either antifreeze and/or raw water intrusion. Heat exchanger as well as head gasket leaks could be at fault.

The standards against which these readings are based are statistical for a range of Cummins engines not specifically for Cummins VTA 903 nor specifically marine engines. The indicators in the report are based on standard deviations for all Cummins engines they have tested. [This is common, not unusual.#]
These wear metal [iron, aluminum and copper] and contaminant metals [sodium] are entirely out of normal range and running the engine for the 50 hours I mentioned back to CT is ill advised. The lab normally see potassium and sodium levels at about the same levels, as is presented in the port engine.
Total Base Number of about 6 [under fluid properties] reading is normal and indicates the acid control of the oil, the alkalinity of the oil, remains in the satisfactory range. This oil [Rotella T 15W 40] originates with a TBN of 10."

So I called back and asked: "Do you think the wear metals, the first three severe readings, and the sodium level are related?

"Sure: Without any doubt the extreme sodium reading has caused the out of normal wear readings.

What about running that engine for about five hours to get to a repair facility?
" Can you have the boat towed?' OMG!!!!

So I called the prospective buyer, a friend, and told him: If you plan to run to the repair marina, we should run on one engine after leaving dockside. We can restart the damaged engine for maneuvering for a few minutes to get dockside at the repair facility if required.

Next step: we get advice from a certified Cummins mechanic at the reapir facility.

# I reported similar baseline comparisons for my own 8V71TI oil analyses [different analysis laboratory] maybe 6 or 8 years ago in these forums. Turned out back then, and still seems the case, such engine wear and contaminant levels are proprietary.

Will post subsequent developments as they unfold.

Comments, insights welcome.

[REBrueckner]

Today 10/10 we find out from a confidential source the starboard Cummins with the bad oil analysis experienced a 'riser failure' some unknown time ago.

I explained to my friend, the prospective buyer "I never heard of water [sodium] intrusion in oil from a riser failure without the need to rebuild/repair, at a minimum, some aft cylinders, rings liners, whatever. I'm still thinking it's more likely there is some sort of current problem....head gasket or oil cooler leak....A pressure test of the cooling system via the fill cap on the heat exchanger expansion tank is a quick and easy test to get a preliminary look."

Any criticisms or alternative views?? I'm pretty sure the buyer is likely to get a certified and well recommended Cummins mechanic aboard for a detailed survey after the sellers are made aware of this mess. I suggested the buyer ask the mechanic if it's smart to get the oil changed BEFORE we do a sea trial....get rid of as much of the nasty sodium as possible.

[Dan Mapes]

I would certainly change the oil. We have been running 671TIs since 1999 with some of this. How long this engine will keep ticking is anyone's guess. I am thinking this is a gasket issue in our case. The 71's have a mess of them between the head and block.

Attachment 16340

[drburke]

Seems to me to be a straightforward distinction between what's happened in the past, possible damage done, and any ongoing issues.

You now have a baseline, so I would change out as many fluids as possible and gingerly try to get a few more hours on and retest.

My surveyor had recommended such an approach to me because of unknowns with previous maintenance.

DAN

[rsmith]

Said it before and I'll say it again. I don't have very much faith in oil analysis. There are too many variables in how the sample is taken and other factors right down to the sample bottle and how well the guy in the lab cleaned his equiptment between samples. If it's going to show anything at all sampling needs to be over a long period of time and done at the exact number of engine hours between samples. Even then how the engine is operated between samples can throw things off. If it makes you sleep better at night fine but in IMHO you might as well do a Jesse Pinkman and start chucking cash out the window.

[MikeP]

Concur 100%!

Unless they are done as part of a regular program, they are hugely overrated. In this particular case, I'd pay no attention to it at all. Normal mechanical checks/observation/oil change is all I would do and unless there is some obvious problem, start 'er up and go from there.

[GJH]

All that stuff got in that oil somehow and I doubt it was a gift from Santa. To completely ignore it would be foolish.

[krush]

A guy name Don Bentley (of Bently Nevada) came up with a new way to measure shaft position (and thus vibration) on rotating machinery like 50 years ago. Today his stuff (in one form or another) is on any piece of valuable equipment. It will tell the position of a shaft in a bearing down to the 0.0001".

The funny thing is, some machines that ran fine for 50 years all of a sudden become problem children and huge headaches after the new monitoring equipment was installed. Engineers would freak out and wonder how the equipment hasn't blown up.

Data is great, more data is better. But somebody with a brain still has to draw useful conclusions from the data. For those without the brain, ignorance (no data) is better.

[REBrueckner]

The story continues to unfold: Our final lab report on the suspect STB Cummins 903VTA 450 HP engine.

Based on previous discussions in these forums, we had a ferrography report prepared to see if small iron oxide particles from extended lack of use might be a culprit. If such small particles are scraped from cylinder walls and removed via an oil change, such could could be a minor and temporary contamination issue.

The ferrography reported overall SEVERE [RED] not surprising since the lab used the same sample.

The ferrography reports states [and includes magnified microscopic pictures] :

"Suggest flushing system. Suggest inspecting this unit for excessive wear and/or pitting. Severe red oxide (rust) contamination is likely from a heavy water contamination source - judging by the high Na content it is suspected to be a salt water contamination source but this is speculation. SEVERE amount of ferrous rubbing wear. Excessive rubbing wear results when
abrasive particles (silica/dirt and environmental contaminants) are present. Severe amount of abrasives (silica/dirt and environmental contaminants), which cause component wear. SEVERE amount of red oxide (rust) possibly resulting from external contamination. Red oxide particles are abrasive and increase component wear. Significant amount of cutting wear particles, which is the result of excessive abrasive particles (silica/dirt and environmental contaminants). Significant amount of lead/tin alloy, likely from bearing wear. Moderate amount of fatigue wear. Fatigue wear results from cyclic stress and metal fatigue. Fatigue wear is not common for diesel engines, and may indicate cam/roller rocker wear or oil pump wear. Moderate
amount of copper alloy particles."

Not a single GREEN reading resulted which is NORMAL. { Oh darn,if you get my drift!}

WEAR METALS [range is in {EDIT} microns] [I originally posted "ppm"]

Rubbing 800-1000 RED [SEVERE]
Fatigue 5-40 YELLOW [ABNORMAL]
Cutting 4-60 ORANGE [MORE ABNORMAL, not SEVERE]
Ferrous 1-60 RED
Cu Alloy 25-140 YELLOW
Pb Alloy 5-40 ORANGE

CONTAMINATION METALS
Abrasives 1-130 RED
Red oxides 1-130 RED

I've never seen such a report before so will again call the lab this week for additional feedback. And will report what they have to say. I'd like to get a separate lab analysis person and compare comments with the first.
Right now what concerns me most is the possibility of excessive bearing wear. Any interpretations, suggestions are welcome.

[tpld3]

It is at least very useful to know of high sodium in the oil as well as the high iron, whether or not oil testing is done at "scientific intervals". Knowing this information warrants further investigation to prevent possible cataclysmic failure of the engine.