Oil Analysis - good read

[a64pilot]

You don’t have to get so defensive, it’s not personal.
I never said it’s useless, I’ve said it repeatedly it’s excellent to determine the condition of the oil, just maybe not the engine.
So if your wanting to extend drain intervals or whatever, it’s an excellent way to get there. How else are you going to determine if the oil is breaking down etc?

[Steve DAntonio]

Originally Posted by mvweebles

Thanks so much Steve.

Question : any guidance on how to mitigate risk of an owner changing oil immediately prior to sea trial? Best I can think of is long sea trial 4+ hours. But even that isnt too long. Will the lab be able to calibrate based on that few hours? Other thoughts?

Peter

Recent oil changes are annoying, however, you can't wash away coolant or salt contamination in one oil change, so things like that will still show up. Wear metals are a bit trickier in low hours oil. However, and this is extremely critical, make absolutely certain the sample is drawn properly, this video will show you how https://vimeo.com/158187739 And, make absolutely certain the correct informaiton regarding the unit hours (hours on the engine, gear, equipment) and lube time (number of hours on th eoil), is accurate, even if it's only 4 hours on the latter. Thos numbers drive the threshold levels for contaminants. My most frequent complaint about oil analysis is sloppy sampling technique, and failure to provide all the required information with the sample, and confirming it is correct when you get the report. Mechanics who fill these out are notorious for being lazy and either making up numbers of leaving them blank.

Yes, if you suspect the oil has been recently changed, get as long and hard a sea trial as you can.

[mvweebles]

Originally Posted by a64pilot

You don’t have to get so defensive, it’s not personal.
I never said it’s useless, I’ve said it repeatedly it’s excellent to determine the condition of the oil, just maybe not the engine.
So if your wanting to extend drain intervals or whatever, it’s an excellent way to get there. How else are you going to determine if the oil is breaking down etc?

I'm sure there are exceptions, but I know of no one who owns an engine of less than about 700hp (say, up to 10 gal of oil) who uses oil analysis to figure out when to change their oil. They use it to figure out if there are harmful contaminants that are a harbinger of advanced wear - just as Dannc did. This is the recommendation of many industry experts such as Steve D, Nigel Calder, and research by BoatUS.

Owners in this class typically change their oil every 100-250 hours. That's long before oil breaks down.

Here's a great video by a guy with a 1m subscribers who does all sorts of comparison tests. In this one, he compares 70-year old Quaker State to modern oil. The advances in oil are remarkable (FYI - the guy has a ton of videos, most of which are worthy).

https://www.youtube.com/watch?v=-zHlxeu_yuM

Peter

[Steve DAntonio]

Originally Posted by dannc

I will start with the iffy one first. My truck is almost 19 years old and I have owned it since new. At one time, I was buying JD oil in 5 gallon buckets because they have a very good 0Wx40 oil at a very good price. By the book, the oil is supposed to be changed every 5,000 miles. I change about 15,000 miles. I can hear the gasps! The screams! The horror. Well UOA, shows that when I am changing at 15K miles I am throwing away perfectly good oil even though I ran it three times more than the manuals specifies. Furthermore, my UOA test results are BETTER than the universal averages for my engine.

Later,
Dan

I firmly believe we throw away a lot of good oil in this country by changing oil based on the odometer, hour meter or calendar. Just as you should change a fuel filter based on filter vacuum, for those who generate a lot of hours or miles, you should change oil based on how it is wearing and contaminant levels. In a properly running engine, the factor that most often drives a condition based oil change is the base level. This is an oil's ability to neutralize acid. Base states at somewhere between 8 and 12 (this is a good reason to take a reference sample, an analysis of new oil so you know where you are starting), and the rule of thumb is to change when it gets to two. I have clients running Deere and Scania engines that routinely go 450 hrs before a change, and at that point the oil still has life in it. If you are going with that extended drain interval, performing analysis ever 100 or 150 hrs makes sense.

Deere guidelines call for oil replacement every 6 months regardless of hours. To settle arguments, I have analyzed 6 month oil with 25 hours on it and it is absolutely fine, that change interval is just wasteful and likely designed to drive service revenue.

Now, if you put 150 hrs a year on your engine, just change is seasonally.

[Shaneesprit]

Find your local caterpillar dealer, those huge machines have 100s of litres of oil, so they always sample the oil rather than change it by time. As it saves money and downtime.
You buy the kit and it comes with a bottle and postage label.
Always fill the bottle ( warm oil ) tip the oil out and fill it again and send that in the envelope that comes with the kit, all costs are normally included in the kit purchase

[a64pilot]

Originally Posted by mvweebles

I'm sure there are exceptions, but I know of no one who owns an engine of less than about 700hp (say, up to 10 gal of oil) who uses oil analysis to figure out when to change their oil. They use it to figure out if there are harmful contaminants that are a harbinger of advanced wear - just as Dannc did. This is the recommendation of many industry experts such as Steve D, Nigel Calder, and research by BoatUS.

Owners in this class typically change their oil every 100-250 hours. That's long before oil breaks down.

Here's a great video by a guy with a 1m subscribers who does all sorts of comparison tests. In this one, he compares 70-year old Quaker State to modern oil. The advances in oil are remarkable (FYI - the guy has a ton of videos, most of which are worthy).

Peter

You said you know of no one who uses UOA to determine oil change intervals unless an engine of 700+ HP and say 10 gls of oil, then you reference Dan and his truck.
So Dan’s truck is 700 HP+ and over 10 gls of oil?
Quite a few do actually, it’s like anything else, people get so interested into a subject that they do things that are actually probably not necessary. Noting wrong with that, to many it’s sort of a hobby, go over to Bob is the oil guy .com and you read about many doing UOA’s on engines that hold 5 qts and schedule oil changes biased on them.
We have a lot of that here where people argue about things that for most are beyond the point of diminishing returns.

From an oil change interval probably the best thing going has been going for decades, it’s called GMOLS or General Motors Oil Life System, the way that works is every time the oil change minder is reset it resets with a set number, then for every cold start some numbers are removed, and then for hours spent at low RPM, short trips etc more numbers are removed, numbers are always being subtracted with highway miles probably subtracting them the least, when the number reaches zero, the oil change reminder light is lit.
So the mileage or time intervals between oil changes can vary widely, based on how the vehicle was operated.
Been around since 1988 and has probably saved millions of gallons of oil, yet no other manufacturer that I am aware of has adopted such a system
https://www.machinerylubrication.com...-of-oil-change

[Shaneesprit]

Originally Posted by a64pilot

If you guys feel an oil analysis will spot upcoming problems, knock yourselves out.
I’ve never seen it, and I maintained fleets of helicopters that had regular oil analysis done on the components, the nose gearboxes for example the sample interval was every 25 hours.
Also all our trucks were having oil analysis’s done, to determine oil change intervals, for that it’s excellent, an oil analysis is excellent for determining the condition of the oil, that is after all what it is, but for determining the conditioning of an engine? It’s next to worthless.

The theory is that when an engine begins to “make metal” an oil analysis will catch it in time as you will see a spike in a certain type of metal, they type of metal will tell you the source of the metal, bearing material and rings are of course very different metals.

The guy who’s engine threw a rod, would someone please tell me how an oil analysis would have predicted that?
About the only way is if the bearing slowly wore out before it spun, then maybe the higher metal content of the oil would have someone tear an engine apart to investigate.

You know how many engines I’ve seen taken apart or removed from service based on an oil analysis? Exactly zero, no oil analysis report will ever say, engine failure imminent, remove engine from service, none.
What they will say is change oil, resample at 25 hours and send in another oil sample, cause I’ve seen that many, many times.

If you really want to spot impending engine failure, instrument your engine, know it’s oil pressure and temp, buy an oil filter cutter and learn how to inspect the filter media, cause if an engine is actually making metal, where do you think it ends up?

You use your cutter to remove the can from the filter, you look for metal, if any is found you test it with a magnet to see if it’s ferrous metal or not, and you cut the media, wash it with mineral spirits and filter the mineral spirits to collect the metal.

But, and this is the big BUT, engines rarely make metal for very long before they go bang, and always when they do you get other indicators, like low oil pressure and high oil temps.

Since you guys like Blackstone labs and they are a very good lab, I’ll link to their how to page. Now it’s pretty aircraft engine specific, but it can also of course be used for any engine the principles are the same.
https://www.blackstone-labs.com/wp-c...Newsletter.pdf

Just noticed Blackstone reprinted John of Sacramento’s sky ranch’s article, but that if anything adds legitimacy to it in my opinion.

Every single large machine I know of in the world ( admittedly on the ground I don’t know about flying machines ) but everything from a digger to a train relies 100% on oil analysis, and oil is never changed unless the oil sample says to change the oil !
So oil analysis works, because they are all run on costs, and lots of issues are discovered by analysis including bad fuel, oil contamination with fuel or coolants. I don’t know what you do in the air but on the ground it works, so Marine use the same engines as the ground based - so oil sampling works.
On a funny note - I don’t sample my oil as for me it’s cheaper to just change it as I have 5 litre not 500 litre of oil in my engine.
Would it predict a rod failure, well that depends on the cause was - if it was oil or water in the oil then yes.
In fact I remember an engine oil that was sampled, less than a week after it was changed and the normal response is email, but in this case a phone call to stop the engine, massive change in sample- it turned out Pegasus 705 had been used, which is a perfectly good oil, however this engine requires Pegasus 1005 and the engineer had made a mistake - engine saved!

[a64pilot]

Originally Posted by Shaneesprit

Every single large machine I know of in the world ( admittedly on the ground I don’t know about flying machines ) but everything from a digger to a train relies 100% on oil analysis, and oil is never changed unless the oil sample says to change the oil !
So oil analysis works, because they are all run on costs, and lots of issues are discovered by analysis including bad fuel, oil contamination with fuel or coolants. I don’t know what you do in the air but on the ground it works, so Marine use the same engines as the ground based - so oil sampling works.
On a funny note - I don’t sample my oil as for me it’s cheaper to just change it as I have 5 litre not 500 litre of oil in my engine.
Would it predict a rod failure, well that depends on the cause was - if it was oil or water in the oil then yes.
In fact I remember an engine oil that was sampled, less than a week after it was changed and the normal response is email, but in this case a phone call to stop the engine, massive change in sample- it turned out Pegasus 705 had been used, which is a perfectly good oil, however this engine requires Pegasus 1005 and the engineer had made a mistake - engine saved!

Want you are saying agrees with what I have been saying, that oil analysis is excellent in determining the condition of the oil, viscosity, acid level, how soot loaded it is etc. as I said every US ARMY ground vehicle I know of that is a piston engine uses UOA to determine when to change the oil, been that way since the 1970’s.

Where I differ based upon my experience is whether or not that UOA will monitor engine condition, my experience says no, it won’t.
OK, so you get a spike in chrome, which maybe means rings, what are your actions? How about Iron?
I’ve never heard of anyone who based on an oil analysis removed an engine or component from service, and I therefore have never heard of an engine or component that was removed from serviced based on an oil analysis that if kept in service would have resulted in failure.

Just about every thread posted here about buying a boat, has someone post “be sure you get an oil analysis, that way you know the engine is good”
That is what I am saying may not be true.

[boatpoker]

Oil_Analysis_on_Diesel_Boats

Plus sample analysis reports.

Pay special attention to a post I copied into the end of the article from a Trawler Forum member. You pays you money and takes your chances.

[mickt243]

if oil analysis was so critical to extending engine life and predicting/preventin component failures, why dont manufacturers mandate it during warranty periods - it would seem to be an easy "out" for them to deny warranty claims based on lack of oil analysis data if that data actually proved anything.

[mickt243]

Originally Posted by boatpoker

Oil_Analysis_on_Diesel_Boats

Plus sample analysis reports.

Pay special attention to a post I copied into the end of the article from a Trawler Forum member. You pays you money and takes your chances.

Originally Posted by Boatpoker's attachment - the important bit at the end he seems to be refering to

In engines and other systems of the size that are addressed by this forum, I believe that frequent oil changes provide absolutely the most cost effective maintenance possible, sampling provides very little for the money in real terms, the money is better spent on the oil

I dont think any of these "small" engines are adequately equipped to draw a sample in the controlled manner required (as standard) and whilst taking an oil sample from the sump or wherever may provide some added interest (and certainly in itself it does no harm) I dont actually think it does much good, it's akin to the car enthusiast polishing his car every week, it doesnt hurt it but neither does it improve any measurable parameter of the vehicle.

Value of the analysis, just like you would not expect an MD to give you a (medical) diagnosis based solely on your pulse rate, anyone who is giving a diagnosis based solely on an oil analysis is kidding you. Together with other information spectographic analysis can help with a diagnosis, but it is unlikely and it is certainly beyond the scope of "normal" mechanics. As part of a test and trial process a lead mechanic should be able to draw some conclusion which an oil analysis may reinforce but he would have probably got there anyway with the other tools at his disposal

Of the thousands of reports I have dealt with I can't remember one which saved us from a failure, although I can remember a few instances where this may have been the case had the results been returned and reviewed quickly enough. In these few cases they merely helped us in a forensic sense to understand the origin of the failure, sometimes this is not as simple as it may sound when all you have is a large pile of scrap !!

this excerpt from Boatpokers link, from someone claiming 50 years experience on the subject, telling us that its not a particularly valuable tool for assessing engine /parts condition, instead he say to listen to a competent mechanic.....




just like my MD with my current blood pressure issues - he has me sampling daily, not every other year as I might bother to check.... so unless you are going to do an oil sample after every run, how can it possibly tell you "part x is about to fail, change it out now" before it fails (especially for the people sampling at hundreds of hours intervals)

[mvweebles]

When I originally posted the Steve D article, I figured it would just be a good read. I know there are natural shortcomings to episodic tests (especially when evaluating a potential purchase boat), but I figured adding data points is a good thing. I am surprised by the vehemence of the "don't bother" crowd. I had no idea this is just another holy war - sail vs power, single vs twin, cat vs mono, compost head vs marine head. You can understand my surprise: it's a $40 test!

I guess we agree a test provides data. Question is whether that data can be translated into useful information. For me, it does. It's not as easy to derive useful information as it looks, but that's life - not all data comes packaged in a nice ribbon and bow. To each their own.

[a64pilot]

Yes, back to it can’t hurt, and may help, it’s data and if done correctly it’s good data. The argument is how valid is that data? I maintain the data that determines the oils condition to be excellent, the data used to determine the condition of the engine can be questionable, in my opinion.
There is some technique to how to sample though to get good test results, it’s not just slop some oil in the bottle and send it off.
All I am trying to say is it’s not what many people want it to be, and that is a definitive test that will tell whether or not an engine is in good condition.
For example the be sure to get an oil sample if buying a boat that way you know if the engine is good or not, I don’t think that’s true, I have seen excellent sample results from a engine that is slap worn out, and high metals in nearly new ones.
And for some seemingly bizarre reason it’s not uncommon for a perfectly good engine to throw some high numbers out of the blue and then have them disappear later. Something’s wear accelerated of course and then returned to normal, maybe a burr or tiny defect on a part?
The real question is what do you do if say iron or chromium or any other metal shoots up? The oil labs I’ve worked with the answer was always submit another sample in 25 hours, if it was still high guess what? Resubmit another sample in 25 hours, exceedingly rare (once) but I have gotten the drain, flush and resubmit a sample, but that was only on a transmission that was making so much metal the oil looked like it had glitter in it and we were getting chip lights of course. You could see the metal in the sample, at the lab only a drop or so is burnt to determine the spectroscopic results, some is used for viscosity tests etc and it was then all filtered through what amounts to a coffee filter and washed with solvent and what was left in the filter inspected, so they had to see metal, and it was magnetic which is always a bad thing on a gear box.

I think it depends on if $40 is money or not to you, for some it is. I do however whole heartedly agree with the person who’s post Boatpoker copied and that’s the $40 is better spent on more frequent oil changes, if $40 is money to you.
As a mechanic who has overhauled many engines, I can tell as soon as I pull the valve cover and engine that has had frequent oil changes from one that’s been serviced IAW the manufacturers recommended interval.
Back in the day I could tell you who ran Quaker State and or Penzoil just by removing the filler cap too

One thing that hasn’t been brought up is that maybe showing a history of oil analysis may make a boat worth more come the day you sell it, I bet it probably will.