Which is the better method – don’t get fixated on the method and more on what you are trying to achieve?

You will get a better service if you describe what you are trying to achieve rather than state the suite of tests and methods you need.

I had an interesting question the other day by another lab looking to subcontract to us in that I was asked could we perform a test by a certain method. In this example it was could we measure 3 elements by XRF (X-ray Fluorescence). I asked them what they were trying to achieve and what range of concentrations were they expecting. The client explained they were looking for concentrations between 0 and 20ppm on lube oils, which was bordering on or even below the limit of detection (the point where you can distinguish between a true value and noise of the instrument data) for XRF.

I clarified why they were using XRF to achieve this when for instance ICP (Inductively Coupled Plasma) is far better at measuring low levels of a wide range of elements. The answer after a bit of digging was their customer didn’t really know why they wanted the analysis to be XRF and had always had it because the lab they first started testing with over 8 years earlier was a small lube supplier lab that only had an XRF.

The problem with dictating a method means you leave no room for improvement

I had a similar question about grease wear element analysis from a UK utilities supplier in a tender document recently too, funnily enough also asking for elemental analysis by XRF and they were very specific about the pre-processing method they wanted.

“we were not prepared to offer an inferior test method to a client”

In this case the limit of detection would have been good enough for what was needed, but trying to achieve this with a grease that contains tiny air pockets and where the penetration through the grease would be limited for the X-Rays meant the data they would have obtained would have been very poor and wouldn’t have been as good at detecting early signs of a bearing failing which was their ultimate goal. Although we could have performed this analysis to their method requirements and the preprocessing was fairly simple we decided to instead say we were not prepared to offer an inferior method to our client, especially for such critical machinery. As an alternative we offered our superior LubeWear analysis which performs elemental analysis two different ways at no additional charge, which would have been far more accurate and give a better early detection of bearing wear compared to their current method.

We asked why the customer was so specific and detailed about the pre-processing required and it turned out they had just asked for a copy of the way their current supplier does the test and turned that into the tender document of requirements. Now I appreciate this is how most tenders work as buyers want a way to remove any difference technically between suppliers by having a very rigid technical specification so ultimately they can go with the lowest price option. If you work for or worked with any large organisations in terms of buying or selling you will perhaps be familiar with this way of working. For me I find it very flawed as it assumes the incumbents way of doing things is the best and doesn’t really consider there may be something different or new developments since they started using their current lab. I appreciate this is also for ease of switching suppliers and I have even seen details of column sort orders for test results and sample information layouts in the headers of reports specified in tender documents. This to me means the customer is just looking for the same service but a cheaper price rather than a better service which might be competitively priced too. The problem with this approach of being so rigid is it also drives the price up and limits the customers options, as for instance redesigning processes for all but the incumbent lab (even down to IT costs of changing report layouts etc) will naturally mean the overall tender cost will be more expensive for the alternative labs. It also means the incumbent lab can be more confident in their pricing that the tender is inadvertently designed around them and also means they don’t have to give any added value or offer better ways of doing things.

State your problem, not the solution

I personally, whenever trying to source a new product or instrument have always given my requirements, but stated to highlight if there is a better option than I am suggesting and if the potential supplier would recommend a different approach. This has been a great approach in my experience and has highlighted faster, more accurate and automated ways of performing tests or reducing waste etc than if I had just dictated what I wanted. This is because there are constantly new developments in all manners of technology and so although I would say I’m pretty up-to-date with industry changes I would never be so arrogant to assume I know it all. Hence I am always open to better ways of doing things and when looking at your fluid analysis requirements this can help you. I appreciate some tenders do give options for deviations to be stated, but the perception for the suppliers is that it is seen as a negative if you try deviate which reduces innovation in providing better options to clients. Thankfully there are still many companies that are open in their approach to procuring lab services and benefit from having a bespoke solution tailored to their needs. Below are two case studies of where being more open to test suite options and methods really helped the clients achieve more accurate data, less unnecessary fluid changes and faster turnarounds.

Case study 1 – be open and let the lab help you – less can be more.

When working with a large international paper manufacturer during a trial with our lab before changing suppliers we had a discussion on test suites.

The customer was very open and said “we will be guided by you on what you think we need.” We walked round their production facilities, I witnessed 5 samples being taken and was shown reports for the last month from their current supplier. It turned out they were on some machines changing the oil every single month based on the sample reports coming back as serious and it was costing them a fortune.

The suite they had was identical for all their samples including gearboxes, hydraulics and compressors. The suite had ISO4406 cleanliness code on all samples including their worm gear gearboxes. This was the suite their oil supplier had suggested when starting analysing with them and although I don’t think the reason was to sell more oil as it was a very reputable lube supplier, it certainly end up in selling a lot more oil, but also putting a negative perception about the quality of it too in the customers eyes. About 80% of the time the samples came back with a high ISO4406 code on the gearboxes and the supplier had worked with the customer to move the limits up to make more pass, yet still the value was too high and they were doing repeated oil changes. They also were measuring NAS class (SAE4059) too, which they had said they had almost every sample came out at 12 and they had requested to move their limits to only flag the NAS class if greater than 12. It was not clear if this had been explained to them by their current lab, but I explained that the highest value on the NAS scale is 12 so the test was not much use if it is impossible to fail with their limits of never flagging anything at 12 or below. Equally if they were not able to filter the oil, pre-filter new oil or had any filtration on the system there was no point doing particle count on the gears at all, particularly the worm gears in which the samples normally come out looking like metallic paint so I was surprised the current labs particle counter didn’t need a huge strip down and clean after a batch of those samples.

Hence we suggested for these systems we drop the particle count test entirely (keeping it only on the hydraulics and compressors that do have filtration), monitor the wear very closely and only perform an oil change if there was wear. The customer ultimately reduced their oil consumption, reduced downtime from repeated oil changes and showed a huge benefit in waste reduction for their ISO14001 Environmental Management System audit. As of writing they have not experienced any issues and with what might appear less tests on the reports the relevance has become far better for the customer. The time spent on repeatedly changing the gear oils has been applied to an improvement in compressor lubricant water content which we have been working with the customer to reduce and which had fallen between the cracks in terms of tackling after the raft of oil changes on the gearboxes before switching to ourselves.

Case study 2 – Fuel analysis, water and bugs – there is a better way

This was a power generation customer with hundreds of bulk storage diesel fuel tanks who was having to wait several days for each analysis but needed results urgently. The problem was they has requested each diesel fuel sample to check for evidence of microbes such as fungi, bacteria, yeasts and moulds. Admittedly these are a huge issue in some systems, but the microbiology check the customer wanted as a standard test. This included culturing the sample for up to 72 hours meaning the fastest report the customer could receive would be 5 days.

The reason this had been added to the test suite was the lab they had used in the past was a microbiology lab primarily with a sideline in fuel analysis and hence the suite was favoured towards microbiology testing. The lab had also applied not too dissimilar thresholds for microbial content to the drinking water testing they also performed. This meant the customer was heavily treating their fuel tanks with biocides too or cleaning the tanks out, yet they had never had a ‘fuel bug’ related problem.

We reviewed this with the customer and our suggestions were:

  1. Target your limits for fuel bugs to reflect what is significant to your fuel system – the risk from fuel bugs is not you will get ill – the diesel would do that by itself if you were to ingest it. The problem is that fuel bugs act like any other insoluble material in the fuel and block the fuel filters with the added problem if the fuel filters are wet the microbs grow and block it further.
  2. If you have no water there can’t be any microbes – there is a reason NASA invests billions in searching for liquid water on extraterrestrial planets and that is all known forms of life rely on liquid water. Hence if the water content is very low in the sample there will not be any microbes in the sample and certainly not a level that will be significant enough to cause filter blockages. Hence the microbial content testing should be based on if there is high water present or not.
  3. Dissolved water measured by Karl Fischer (KF) titration won’t necessarily detect all the water in the sample and a secondary check is required-  When we highlighted point 2 about with no water there can’t be any microbes to the customer there were a couple examples they came back with that had less than 50ppm of water but very high microbial counts. In this case the lab they were using were working rigidly to the way you measure water in a fuel rather than looking at the problems this can cause. For instance, the KF method, which is the standard way everyone measures water in fuel needs the sample to be perfectly mixed to give an accurate result. If you have ever tried mixing diesel and water you will realise they separate very quickly even despite vigorous shaking. So to take a small sub sample that is representative is nearly impossible because the fuel is so thin the liquids separate very quickly. This means if you take from anywhere but the bottom you will pick up very little water and if you pick up from the bottom you will over-estimate the water – you are basically between a rock and a hard place. So why is this test used at all? This is because the test was designed when it was assumed you would be testing new production fuels that are very dry and only low levels of dissolved moisture would be present. Anything above these low levels and the water over-saturates the diesel and separates out meaning the sample is not homogenous. In this case that is why I have always recommend measuring both dissolved water and free water content so the customer has too angles to see if they have a water problem.
  4. Use a screening microbial test when possible – In this case when there was high dissolved or any free water we recommended performing a cyclic ATP test to look for microbial activity. This doesn’t classify into the microbe types such as bacteria and fungi, but gives an indication that they are present. This test can be performed same day and means many cultures can be avoided if not positive.
  5. Avoid heavily treating your fuel with biocides and treat the problem, not the symptoms – the microbial type is generally only useful for determining treatment of the system. With the above 4 suggestions we managed to help reduce their requirements for culturing by 92% giving a huge reduction in turnaround times. This could be reduced to nearly 100% by actually treating the cause and not the symptoms. The reason for a fuel bug is the presence of water, so by monitoring and tackling this early you can avoid ever needing to perform a microbial test on the systems. This means the turnaround time can be same or next day rather than 72 hours.

Hopefully you will have seen in these case studies that by highlighting the problem you are trying to solve the laboratory can much better aid you, rather than specifying a test suite or assuming your existing one is still the most suited for you. If you are open to discussing better ways to improve your fluid analysis testing and condition monitoring programmes then click the blue contact us button at the bottom right of this screen to get in touch.