We all know cleaner oil is good, but did you know there is such a thing as too clean to the point it prevents the oil working properly, strips it of its protective properties, hides impending problems and to top it all off makes it potentially at risk of explosions in certain situations. Wow, who would have thought a bit of filtration could cause all that?
So before you go hiring a bomb disposal unit to change your oil filters going forward, we need to explain why all these things can occur and how to mitigate them.
There is a lot to get your head round. Clean oil just feels right, doesn’t it? Surely if it’s cleaner it must be healthier and lubricate better than dirty oil? Well, in truth it depends. Dirty oil is a major contributor to machinery failures and in hydraulic systems is the cause of about 80% of failures. So before I get a load of strongly worded letters from filter companies I want to first confirm I am not advocating having dirty oil or indeed saying never to filter your oil. However, what I am saying is there are diminishing benefits the cleaner you go and once you hit a certain particle size the benefits start to be outweighed by the downsides. To illustrate the point I have written a short fictional story to explain the problem first and then we can go into some real world very true case studies.
The two cures that hid a larger problem
In a quaint suburban neighbourhood lived two men, Robert and Thomas, bound by an invisible thread of fate. Robert was a diligent family man, but for months he’d been grappling with terrible headaches. Dismissing it as stress, he found solace in over-the-counter paracetamol. As time wore on, the headaches persisted, and so did his reliance on painkillers.
Just a few doors down, Thomas, a mechanical enthusiast, had a vintage car that was his pride and joy. He prided himself on a unique filtration system he’d installed, convinced that the finer the filtration, the purer the engine oil, and thus the better the performance. Over time, he escalated the fineness of his filters, believing he was prolonging the life of his cherished vehicle.
Then, one fateful afternoon, both men faced unforeseen tragedies. Robert collapsed while mowing his lawn, rushed to the hospital, only to be diagnosed with an advanced brain tumour. He passed away within two days. Almost simultaneously, Thomas’ car engine seized, grinding to an irreversible halt.
An inquiry into both situations revealed chilling parallels. Had Robert sought medical expertise earlier, imaging would’ve detected his tumour in time for intervention. Similarly, had Thomas consulted a professional to analyse the oil in his car engine, they would’ve discovered that his obsessive filtration was removing essential additives, leading to increased wear and tear.
Thus, the tragic tale of Robert and Thomas became a sombre lesson for the community. They learned that prevention is better than cure; it’s not just about alleviating symptoms or optimising superficially. One must understand the root cause, whether it’s in the complexities of the human body or the intricacies of a machine, to truly ensure longevity and well-being.
So following that rather sombre tail, what have we learnt? Well firstly if you are suffering long term headaches seek medical advice. However, beyond that we learnt that trying to fix a problem you believe is there without actually knowing it is or realising what it may be masking the real problem and more dangerous.
Making it worse
The issue with lube oil filtration is everyone talks about how far can they go. Pore size, beta ratios, dirt holding capacity are all the main sales tools of a filtration company. After all their job is to sell filters and I have seen many lube analysis trainers scaring customers into believing their oil is much dirtier than it should be.
For instance most people will be familiar with the ISO4406 cleanliness system to measure the cleanliness of an oil. A hydraulic oil new is usually about 20/18/15 on this scale and most filter companies try to clean it up to say 18/16/13, which is 4 times cleaner than new oil and very clean indeed.
Now in an excavator in a dusty mine or quarry that is exceptionally difficult to achieve and so normally I would argue if you can in those harsh and dirty environments get a sample that is anywhere near new oil cleanliness you are doing a good job.
However I have seen situations where the manufacturer has set thresholds of 18/16/13 as red critical alarm limits and suggesting oil and filter changes on construction machinery. This likely leads to unneeded downtime and also risks introducing more contamination by opening the system up to change the oil, filters or both.
This is not just true of dirty environments. A very famous injection moulding manufacturer has cleanliness requirements of 14/12/10. That is between 32 and 64 times cleaner than new oil and that’s the top end of what’s allowed. I’m calling it out, that limit is ridiculous and unrealistic, and I personally advise to take that as a target rather than beat yourself up if you don’t achieve it. After all, aircraft, formula one and space agency cleanliness requirements are not as strict as this.
So why do manufacturers set such unrealistic cleanliness targets?
Well if you think about it, if you are a manufacturer providing a warranty and I said “how much dirt can I put in your machinery oil and you still pay out on the warranty?” Your answer would likely be “None”. After all you have precisely engineered these components perhaps with years of research into their design, you don’t want some idiot pouring a load of dirt and sand into the oil now do you?
So that is how it’s set, it’s set at a threshold that means the system is exceptionally clean. Often these thresholds go to committees which on the whole will be more cautious at each stage as it becomes the final document, not wanting their names associated with some financial disaster of warranty claims.
Another way of limits being set is with research studies. Many bearing manufacturers will for instance setup test systems and deliberately pour sand into systems etc to see what happens. The same is true of additive manufacturers too, as after all, they have to account for the loosest nut of the system and that is unfortunately often the operators of the machine. I have seen studies and research showing for instance in gear systems traces of dirt causing microscopic damage to components. So that is why some training organisations say an industrial gearbox oil should be 18/16/13 as above that the research started showing damage on the test bed. They may have a point. However, the fact these oils are often smaller systems and not filtered and the new oils come in around 22/20/17 from the barrel or about 4 times dirtier than new hydraulic oil you have to put this into perspective by risking introducing contaminants when you change the oil. It also brings up the topic of overwhelming the customer with red flags
Not seeing failures for the reds
I see this problem a lot when clients move to us from other labs. The limits they have been told to use give all red flags, which means how do you tell an ‘about to fail’ machine from another red report saying it’s dirty etc. In every site there will be some machines in better conditions than others. Even if some have a poor baseline it’s always best to target the very worst condition systems first and then gradually move to the better condition machines. Otherwise you get what I call ‘sampling apathy’. Another example of that coming up.
Limits that are poorly thought out
Gas engine wear metal limits are often given by manufacturers in per hour limits. Often as per 100 or per 1000 hours. E.g. 10ppm per thousand hours. The concept is good as you don’t want to penalise an oil that’s been in use a long time by an arbitrary limit when it’s just high for the number of hours. However, the problem is these limits have a major flaw. If the machine oil hours is only 100 hours then the limit is 1ppm and 50 hours 0.5ppm. Hence I have seen situations customers have changed their oil because of a cautionary high TAN to then get on the next sample a serious chromium because the limits are per hour and a sample at say 120 hours gives a maximum of 1.2ppm. This is preposterous and why I don’t blame people for ignoring the OEM limits at very low hours when it gives such false alarms. Those that follow them without question end up with a customer with all red flagged samples. The problem is they start to become desensitised to reds and then miss genuine real issues. I have had customers I have had to completely re-educate that red means do something with our reports because with their previous lab only 4% of samples were flagged normal. Clearly that suggests something is up with the limits.
So as you can see limits sometimes need to be used as an ideal target rather than good vs bad cut offs. Limits need to be based on what is practical too. So what are the purposes of limits?
What are limits for?
Limits as I like to use them should be ultimately to help machinery operators flag problems and improve machinery condition for longer term reliability. We have already covered the problem of overwhelming operators with too many reds as well as setting limits that risk you introducing more contamination into the system by the oil change. So this brings the case that we need to have limits that are more application specific rather than blanket rules. This is why I often suggest doing stats analysis based on customers data rather than blanket taking OEM or even using customers limits. So now you understand the problem with limit and how they can be used to panic people into potentially over maintaining an asset, but I started this article talking about over filtration specifically being a problem. So can over filtration as in the story actually cause problems? Let’s discuss this now.
Can over filtration cause problems?
In short yes, or this article would be a bit pointless. The main problems come down to 3 main causes.
1. Hiding a problem
This is something I have always thought is seldom discussed. If you are filtering the oil so heavily then the wear materials from a potential misalignment or work hardened bearing never appear in the oil analysis as you are stripping them from the oil so quickly. This means they can’t enter the sample bottle and so can’t be detected by the lab. This does not mean you can’t do the filtration but does mean you need to do filter analysis too to investigate problems as you are collecting the material in the filters rather than the oil. I have posted this point on LinkedIn a few times and every single time I say it I get controversy to suggest such a thing. However if you do cut open one of these filters you do find higher concentrations of everything including wear metals as that’s what a filter is supposed to do. The problem the larger more abnormal wear material ends up in the filter and hence is not detected.
2. Static discharge
This one is not solely a problem with the filter but the oils too. Modern oils tend to be very good insulators because they are so pure and this means that they tend to build up static discharge over time. The cleaner the oil the more of an insulator it becomes. Additionally, the finer the pore size the more surface area contact of oil with the filter occurs, increasing friction and hence static. So what this means is you will have an increased static discharge issue the cleaner the oil becomes and also the finer you filter the more static discharge will build up. You are still going to filter the oil, but the benefits of cleaner oil diminish the finer you go and the static discharge issues increase. A poorly managed static discharging system with no mitigation in design or use of anti static filters can lead to some potentially explosive situations in the very worst systems.
3. Removing additives
Every filter companies will tell you their filters don’t remove additives. This is untrue and EVERY filter will remove additives. The finer the filter the more additive will be stripped from the oil.
So why do they say this, well they say they don’t remove dissolved materials but additives are not dissolved truly. They are just well dispersed at best. Modern formulations of oil are oxidation stable but as a result poor solvents and so are awful at holding onto their additives and drop them at a moments notice, so in reality once you drop under 1 micron you are basically just stripping additive in most systems.
I have a case study I like to use but I will naturally anonymise the filter company name. This is because I do not wish to embarrass the company concerned as the week I write this there is much controversy over a single car tech YouTube reviewer in USA causing a company’s stock to plummet and near bankrupting the company. My reach may not be that great with LearnOilAnalysis but best to avoid it anyway.
The story of Tom
No his or her name is not Tom, but let’s call them Tom for this example to explain. Tom came into filtration quite late in his career and developed a product for varnish removal way before the other filter companies were looking at this. This product had one of the smallest pore sizes on the market and dirt didn’t stand a chance against this fantastic filter. Tom was such a good salesman every one of his customers was impressed and I spoke to a few who were all converts to his philosophy on filtration. Tom started mentioning he was extracting such early varnish it hadn’t even darkened yet and referred to it as colourless varnish. Unfortunately this material was not varnish and was actually oil additive being stripped by the filter. I first saw this when a new oil lost 75% of its RULER value after installing the filter in 5 months when working for another lab. Tom was not very happy at this finding and actually was one of the a very few customers that actually left the lab I was working at because he didn’t like the fact we had linked the over filtration to the additive drop. The oil company equally didn’t like to admit that their oil additives were not fully dissolved so the problem was never really investigated further.
So was Tom wrong to filter so fine. Well if you want to remove varnish then very fine filtration is often a good way to do this, but varnish is basically sludged oil additive that has been spent and often polymerised. So you need to go pretty small to remove it. However if you go so small you remove not only the polymer but the original smaller additive molecules too. Many in the industry have now adopted selling filtration with supplementary additives when going this low to replenish some of the additives stripped. However this adds an extra level of cost on what is potentially a very expensive system anyway.
Tom still is very popular and has expanded his business. He now sells similar products to remove silicon contaminants in engine oils, but again I have seen additive levels stripped such as the TBN drop as a result. This does lead to earlier oil changes as a result and potentially reduced engine protection. With engine formulations supplementary additives are not viable as they require very strict OEM approvals and mixing additives is extremely difficult. Hence much research needs to be done on the effect of significantly stripped additive packages on long term engine health. Yes there are benefits to filtration but customers should be aware that there are downsides too.
Every filter does remove additives over time, but the laws of nature mean the smaller you go the more “good stuff” you remove compared to the “bad stuff”. Speaking of bad stuff, abnormal wear is bad but removing it although will help reduce abrasive wear from the metal itself does hide the fact a wear problem is occurring and like the tragic story above you could have a sudden failure with no warning. Why? Because you are removing the evidence with filtration. Hence the moral of the story is if you do filter go for what is practical not to the nth degree and don’t forget the importance of using both oil AND filter analysis in your condition monitoring strategy.
Conclusion
Should you filter your oil? Yes absolutely as it is something that can benefit your oil. However if you are going for ultra fine filtration you need to understand that not everything is perfect and you need to take extra accommodations including more regular oil testing and even filter testing to ensure your filtration is not taking things too far.
If you want to learn more about oil analysis and how to use it best with your filtration strategy then click the contact us button to reach out.
