Truth Hurts – not all analysis is the same

img 4627 1024x576 Truth Hurts not all analysis is the same
I’m sorry but it’s a fact that not all lab analysis is equal.

I was recently on LinkedIn and saw a post where a filter supplier mentioned that over 80% of failures were contamination related. They were doing an offer to their customers of a free oil sample “no questions asked” using a lab that wasn’t ours to identify the problem. I made a comment on the post that this seemed a great gesture, but pity it didn’t include LubeWear and what he thought about that. 1 hour later my comment was deleted by the poster and the connection status deleted – Clearly this struck a nerve and clearly no questions asked was being held to in the literal sense.

Now I might add this wasn’t an advert comment with links to our site or something like that, which I appreciate can annoy people jumping into a post like this without contributing to the debate. It was simply stating a truth and hopefully intended to get some debate started on the best way to catch failures. In fact it would have actually boosted his post to all my followers and potentially people would have discovered his filtration services as a result if he had kept the comment.

Since the debate was not allowed to happen on LinkedIn, I thought it was worth discussing what this analysis could have picked up Vs our analysis. I assume this was a hydraulic system they were offering analysis on so let’s start there.

What can basic oil analysis detect?

  • Wrong oil grade.
  • Wrong oil additive package if reference oil provided.
  • High particle count.
  • High water.
  • Oil degradation (depending what you get with the free analysis).
  • Potentially some low level rough wear detection – now this is being generous as standard ICP wear metal analysis used by the industry is quite honestly abysmal and in my opinion not fit for use.

So what about LubeWear analysis?

Well naturally we can do all the above, but I would like to focus on the comment by the poster that most hydraulic failures are contamination related, which is true. So let’s explore that by looking at dirt and water.

Dirty and wet oil needles in haystacks

Dirty oil or wet oil will generally mean visual contamination or a high particle count or water content. There are two causes to this, one is the system is truly dirty and wet and the other is the sample has been taken poorly or there is an external contaminant. At least 20% of all samples have some form of external contamination in the sample that isn’t really in the system and about half of the high particles / debris / water identified in samples is sampling error rather than really in the system.

So if we are trying to detect failures identifying the sample is dirty is often not enough as you need to know which ones are genuinely dirty. If you just have one machine you will likely target that one if it’s dirty or wet and check the filtration etc. However what if you have 50 machines and say only 1 service engineer available today? Their work need prioritising to triage the worst ones first. How do you know which ones genuinely need attention vs ones that could wait for a re-sample to confirm?

Ultimately this comes down to wear analysis. If you have a dirty or wet oil it will be causing wear or corrosion but unfortunately routine analysis just doesn’t pick this up most the time and certainly not early enough and without a trend. Hence LubeWear is a great tool to help identify the machines that are not only genuinely contaminated but also are being adversely affected. Meaning you can actually identify that needle in a haystack of high particle counts and dirty /wet samples.

The rebuff to my argument

As I always like healthy debate I am having to debate myself here. So a natural rebuff would be to say that if you train someone to take samples correctly you will eliminate the issue of bad samples. To this I would mostly agree, but it is impossible to eliminate ALL of these cases as I have customers who are excellent samplers taking pristinely clear oils that occasionally still get external contaminants into the sample. Ironically the post I was replying to had pictures of a clean oil sample and dirty looking oil sample next to each other, but the dirty oil sample bottle had a load of dirt engrained in the threads of the lid. This means means that dirt had been on the samplers gloves/hand and likely was an external contaminant and could be in the sample too. However let’s assume the argument is correct and every sample is taken perfectly every time.

20% is a lot to not be detecting

The problem comes in the 80% of hydraulic oil failures are contamination related. So what about the other 20%. This means approx 20% of failures will show no clues of dirt or water ingress but still fail. That’s a big percentage to be being missed at least by routine analysis that simply can’t see the true level of wear in the system.

Hence LubeWear is essential for helping catch these failures as it detects all the wear in the oil sample container. This means it detects wear far in advance of other techniques and can indeed classify it early as normal or abnormal too.

So does this mean there will never be a machine failure with your oil analysis?

I will be honest, you could still have failures no matter what you do. No condition monitoring technique can prevent 100% of failures firstly because everything wears out eventually and secondly sudden failures that don’t have enough time to generate wear first will not be detected. For example say you sample every 3 months e.g. 1st January then 1st April etc. If on February 5th you Chuck a proverbial spanner in the works of your system and it fails as a result within an hour no analysis programme will detect this. Even if support your oil analysis with a live online vibration system or something like this that would potentially pick something up like this, but depending how frequently the data is looked at and how quickly someone responds you would just be reducing the size of the failure rather than preventing it entirely.

Thankfully sudden failures like this are a rare occurrence and most failures happen gradually and more subtlety. In these cases the earlier the detection of the problem the better the outcome and hence why our LubeWear analysis can help you do this much earlier than routine elemental analysis.

If you would like to discuss the pros or cons oil analysis “with questions asked” please get in touch using the contact button below.