Oil Analysis?

[Dockhead]

Anyone do oil analysis on their main engines?

Any lab to recommend? Tips?

 

[Halo]

Years ago I ran maintenance on a large fleet of dump trucks, dozers,graders,loaders and draglines. After 3 years using oil analysis we concluded that the overall cost of the system did not prevent sufficient damage to make it worthwhile so we stopped it.
One part of the fleet was put on a shorter oil change regime and that did pay off

 

[mjcp]

One part of the fleet was put on a shorter oil change regime and that did pay off

Shorter change periods is an interesting one:

BMW M engines from the period ~2006-2013 (the S65 V8 and S85 V10 lumps) have a reputation for spinning bearings. While there were some material changes to the bearing material which helped later builds, 2010-> onwards, there did seem to be link between oil change frequency and failure: People in the US are obsessed with 3k mile changes (stock interval ~10-15k mls, based on fuel used iirc), they seemed to have more frequent failures.

The theory was that the "stock" 10W60 oil degraded to 10W50 over time (a few thousand miles), and that "thinner" degraded oil provided better lubrication as it could get to certain surfaces (like the rod bearings) more easily. Where some people were changing oil more frequently, the oil never got to the "worn" stage and thus they were inadvertently starving their engines in the that key place they were known to fail.

No evidence mechanically/scientifically, but never had an issues in 5 years ownership, 50k mls, daily abuse and standard dealer changes etc... YMMV as they say.

 

[vyv_cox]

I have quite a lot of experience with oil analysis. I always disagree with the recommendation to sample for analysis before purchase. The result is pretty meaningless. It does have its uses if used in a very organised manner, sampling at the same oil age each time and looking at trends. But for most leisure applications hours are insufficient to justify.

I did a job on generators in a bank, running constantly. A big end bearing failed about a week after a faultless analysis!

 

[superheat6k]

I have had my last and current boat engines oil tested by Finnings. But the oil does need some run hours to be effective in telling you anything, and for most an annual oil change at less than 100 hours isn't going to tell too much of a story. The reports I received in 2022 are attached.

Unfortunately for me and due to personal circumstances I have hardly had any use of the boat so I am now selling her on, this is also why I don't come on here so often. For me and very sadly the unexplainable magic of boating has gone. The Oil reports are useful for would be buyers, if they actually understand what is being reported.

 

[ducked]

Standard oil analysis only measures the tiniest wear particles, because the bigger stuff doesnt participate in the flame emmission spectroscopy usually used for elemental analysis.

Its the bigger wear debris that tends to be characteristic and predictive of failure.

This CAN be monitored, but usually isn't

 

[William_H]

Oil analysis can be a useful tool in certain circumstances. Things like helicopter gear box or engine is an example where reliability and prediction of failure is vital. ie the cost of unexpected failure is great compared to cost of regular analysis.
It is generally more useful if monitored on a regular basis from engine new. ie looking for trends rather than exact particle composition.
I can't imagine it would be cost effective in a twin engined boat where failure would not be disastrous and cost of repair not necessarily any less from knowing the failure was imminent. ol'will

 

[Jim@sea]

Years ago when I ran transport I was talking to a mechanic about the make of the next HGV I was going to buy, he said that Mercedes Truck Engines lasted longer as they had bigger sumps so the engines had more oil. A friend in the 1960's whose father ran haulage was given a Austin Healy Sprite for his 21st Birthday and immediately changed the engine to Diesel Engine Oil as with it being a Heavy Detergent would last longer plus he would change the oil every year. And after reading about the Volvo owner whose car had done over 1 million miles with the same engine it was because he changed the oil every 3000 miles.

 

[penberth3]

Years ago when I ran transport I was talking to a mechanic about the make of the next HGV I was going to buy, he said that Mercedes Truck Engines lasted longer as they had bigger sumps so the engines had more oil. A friend in the 1960's whose father ran haulage was given a Austin Healy Sprite for his 21st Birthday and immediately changed the engine to Diesel Engine Oil as with it being a Heavy Detergent would last longer plus he would change the oil every year. And after reading about the Volvo owner whose car had done over 1 million miles with the same engine it was because he changed the oil every 3000 miles.

Interesting, but that says nothing at all about oil analysis!

 

[AntarcticPilot]

My understanding is that oil analysis is useless on a one-off basis; that it needs to be done routinely at the same point in the oil change cycle so that trends can be identified and if necessary action taken. I suppose that really bad wear might show up as metal particles, but @ducked suggests that standard analyses don't pick big particles up!

 

[Ceirwan]

I used to skipper a wind farm crew transfer vessel, we had twin V10 twin turbo MTUs, about 600hp each.
Expensive engines, so we had a religious oil testing regiment based on hours of usage.

We had two major failures in my time there (the engines were never particularly reliable) and in neither case did the oil analysis pick up anything untoward. I'm not saying they are useless, but in our case they didn't work.
I doubt very much for the average boater it helps at all, change the oil with good quality stuff as per the service schedule for the engine & that's probably the best you can do.

 

[ducked]

My understanding is that oil analysis is useless on a one-off basis; that it needs to be done routinely at the same point in the oil change cycle so that trends can be identified and if necessary action taken. I suppose that really bad wear might show up as metal particles, but @ducked suggests that standard analyses don't pick big particles up!

The "classical" method of wear particle analysis (as opposed to the standard spectroscopic wear element analysis), IIRC, is known as a ferrogram (or ferrograph?) and involves flowing the oil sample over a microscope slide in a strong magnetic field.

The ferrous wear particles settle out by size, and the smear is then microscopically evaluated by a human expert. I believe this is/was quite expensive and mostly just works for ferrous particles, which bearings arent.

I imagine by now there are more automated methods of particle scanning and counting, Coulter-counter stylee, perhaps involving lasers, but I dont know anything about such recent developments.

Being long ago slightly involved in the biosciences, I would think a particle size separation could also be achieved by rate-zonal centrifugation in a sucrose density gradient, but I dunno if this has been tried in this field.

It isnt part of the standard analysis supplied by, for example, Blackstone Labs (the big name in the field) but IIRC they do mention the size of a post-centrifugation smear, which will probably be a mixture of wear particles, soot, and solid oil polymerisation products aka varnish.

IIRC the standard spectroscopic wear element analysis CAN be modified to cope with big particles by acid digesting the sample, but this is not routinely done.

I did come across a UK-based lab that provided particle-size distribution data as part of thier standard analysis at punter-friendly prices but cant remember who they were, and, from a quick look, couldnt find them ont internyet, but there are quite a lot of companies advertising that sort of thing. I dunno how they did it but would guess it involved laser scanning.

EDIT: Might have been these people

LubeWear.com

In which case the enhancement seems to be acid digestion, they have moved to a harder sell presentation than I remember, and I dont see any published prices

 

[Bran]

I have never bothered with oil analysis, not for several heavy vehicles I used to own or for my current boat and I used to have access to wear metal analysis systems. Never felt the need to analyse the oils as the engines are all in good condition and oil analysis is really only of benefit to monitor trends.
If you want to use oil analysis I would recommend you are very consistent in how you take your oil sample and after how many hours running you take the sample.
I have no experience to recommend one particular lab.
Ferrography still exists but is not so common for routine analysis.

Many independent analysis laboratories will use methods defined in an ASTM method, this is a global system that provides a defined method for this type of analysis. When looking at the wear metals in lubricant oils the method defines using ICP-OES, that is Inductively coupled plasma- optical emission spectroscopy. As mentioned previously this will only look at small particles, the plasma is not able to atomise larger particles.
The systems I worked on use X rays in a technique called XRF, that is X Ray Flourescence. This is used in many companies to monitor wear metals in applications from train engines, gearboxes to aircraft engines, generators etc, but is not used so much in the general methods described earlier.
Each analysis type has its advantages and disadvantages, often down to personal preference of the lab managers when specifying new instrumentation.
All these analysis systems can be semi automated to cope with large number of samples if needed including some online systems that monitor continuously.

Hope that gives an insight into some of the methods used. PS no AI in use here!