Used Oil Testing

I've wanted to establish an oil analysis program for Gretchen for a while now. Her engine is quite worn and is due for a rebuild in the near future, but in the meantime I want to run the car as much as I can, while not going past the point of doing irreversible damage to the internals. An effective oil analysis program can help me with this.

The first thing I want to make clear is that I'm no oil analysis expert. In fact, as I’m writing this, I'm only just preparing to take my first sample. What I've written here is based on the small amount of research I've done so far, and I'm sharing my experience as I take that first sample. It seems an interesting enough subject to make a blog post out of...

So, what’s this all about?

Used oil analysis is, as far as I know, the best way to determine with pretty high certainty the state of internal wear in an engine (or indeed, in any mechanical system with a closed lubrication system) without taking it apart and performing a detailed inspection with precision measuring instruments. Most racers who do oil analysis (for the record, I am not a racer) use the information to help them judge when the time is right to take out and rebuild an engine before any catastrophic failures occur, avoiding a lot of expensive extra machining and saving engine parts like the crankshaft, connecting rods, block or heads. Sometimes oil analysis is used to help chase oiling system issues, strange bearing wear or clearance problems, overfuelling/mapping issues or even mysterious coolant losses. It doesn’t have to be limited to analyzing engine oil, either. Oil analysis can be used to monitor transaxle bearing, gear, and synchronizer wear, as well. Anything that uses lubricant in a closed system can be analyzed.

How does it work?

The way oil analysis probably works best, is if you first have a baseline analysis done of the lubricant you're using – send in a fresh, unused oil sample. The report will identify the types of base oils that make up the sample, properties like viscosity, and the concentrations and types of other components in the oil, like anti-wear, detergent, and anti-foaming agents. Re-establish a baseline whenever you change oil brands or weight. The SPEEDiagnostix website has a section where they've posted test results of a few new oil samples, and the oil I use is in there, so I don't plan to have a baseline sample done; I'll just send in a sample of my used oil and compare the results to the online report.

With your baseline information in hand, you have something to compare against later when you have your used oil samples tested. You will be able to tell if you’re changing the oil soon enough or waiting too long to change it – is the viscosity still in spec? Is the oil full of dirt and other contaminants? Are the detergent and anti-wear agents still there in the right amount, or have they been consumed? If you’re getting back reports that suggest the oil is still pretty fresh, maybe you can safely put more miles on it before changing it.

The report will also give you information about things that shouldn't be in the oil at all. For example, a large increase in copper or tin indicates that the bearings are wearing out. A lot of iron or chrome could indicate ring wear, if you're running iron or chrome rings. Silicon in large amounts generally indicates dirt (i.e. sand) has gotten into the oil supply, usually via the cylinders due to a torn air filter. You can see if there's coolant from a small head gasket leak or fuel in the oil from bad injectors or incorrect fuel mapping. Brass in a transaxle oil analysis report would indicate that the syncros are wearing. You get the idea. The drain plug below is not from Gretchen! It must have been a pretty bad day for someone, though.

In a nutshell, the whole idea behind analysis is comparing the latest analysis to previous baselines. You map out your engine’s wear pattern, and get a clear picture of your engine or transaxle's current health and overall life cycle. You determine the right time to perform maintenance, repairs, or rebuilds before anything gets trashed, like what probably happened in the pic above. It seems to me that if you keep at it, analysis should ultimately save you money, especially if you race or track-day your car.

Preparing for the sample

As I mentioned earlier, I had been meaning to start sampling the oil for some time, but had never actually gotten around to doing anything about it. During the last oil change (June), I had drained the oil as usual, stuffed rags around and pulled the filter as usual, and then, as usual, sucked out the remaining oil sitting in the oil filter housing using a piece of surgical tubing and a 20cc syringe. I'd thought about what a good oil sample it would make, and that I should really send out for a sample kit. A couple of days later, I finally did that and ordered a kit from SPEEDiagnostix. I also sent them an email asking if a sample sitting for a week in a syringe would be OK and would 20 CC's be enough of a sample? I must have also mentioned where I took the sample from.

The reply didn't mention the syringe, but did tell me that 20 CC’s wasn’t enough, and more importantly, that the oil filter housing was not a good location; a sample from there would cause inaccurate results. Hmmm, I hadn't considered that some locations could be better or worse than others; I would have to look into this whole oil analysis thing. The only thing to do now was pull the sample from the new batch of oil I had just put in the engine, after putting some mileage on it. I emailed again, asking how many miles the oil needed on it to register anything, and the response was about 1000 miles.

Another point the SPEEDiagnostix technician made in that first email was that I should include information with my sample on how much oil had been topped up, and also when it was added between the initial oil fill and the sample date; topping up between oil changes has a bearing on the concentration of contaminants in the total oil system. If you think about it, adding fresh oil dilutes the contaminants in the oil supply (especially when you’re using as much oil as I am – about 48 oz/1000 miles) and they need to account for it in the analysis to be accurate. I thought it was pretty impressive that they should be that thorough.

So I drove the car whenever I could, and hit 1000 miles just the other day (mid-August). Now, how was I going to get a sample? I wasn’t going to drain it from the pan at only 1000 miles. Off I went to the Home Depot to get some plastic tubing; my plan was to use the Mighty-Vac and its brake bleeding attachment to suck the sample out via the dipstick tube. Once I got there and saw what they had, I realized that the tubing might melt when you put it down the dipstick tube of a hot engine. My oil temp is usually 80-90 °C; I had done a bit of research by this time, and knew that you're supposed to take a hot sample. I started to look around for some small diameter brass tubing to use instead - no luck, and this was fortunate, because it wouldn't have fit down past the curve in the dipstick tube. When I got back home, I did some more research, and discovered what people who do this and know what they're doing actually do. Everything I had been planning to do was more or less totally wrong!

Best practices for accurate results (because there’s no point in getting inaccurate results)

Up to now, I hadn't realized that there was an "Oil Sampling Industry", or that used oil analysis was a widespread practice outside of auto racing circles. But it is! It turns out that many companies which utilize complex or expensive machinery have a regular oil testing program in place. Here are a couple of "best practices" for oil sample collection that I've come across while researching:

• Use a new sampling tube every time, never re-use sampling tubes!

You need to use a clean, new, fresh piece of sampling tubing every time you take a sample. If you re-use tubing you're contaminating the sample.

• Only the clean, new sampling tube tube may come into contact with the sample.

If anything else (pump internals, old coffee cups, beer cans, jelly jars, oil drain pans, funnels, etc) comes into contact with the sample, it will be contaminated.

• The sampling tube must not scrape and collect old, nasty debris from the sides of the dipstick tube or the oil pan.

If this happens, the sample gets contaminated. You should notice that an obvious pattern is starting to develop... The points above are referring to cross-contamination, which occurs when you have anything in the sample (dirt, foreign substances, oil from a previous sample) that is not actually part of the sample. Contamination alters the test results. And if the test is giving you false information, what's the point of testing?

• The sample should be taken from the same point every time.

In this case, it means from the same location in the oil pan, at the same height, ideally in the middle of the oil supply. This is to create consistency test to test, and therefore the maximum degree of "comparability" when you're looking at the results. I made that word up.

• The oil sample should be taken from the most turbulent part of the stream.

Stagnant oil won't give you an accurate sample, because many of the solid particle contaminants will have settled out of suspension; the oil must be fully stirred up to get it all. This is related to the next point:

• Oil should be fully up to temperature (engine run for at least 20-30 minutes) and the engine running when you take the sample. If it can't be running, the next best thing is to shut it down immediately prior to taking a sample, so that the oil is at operating temperature and evenly mixed (i.e. contaminants are as evenly distributed in the oil/sample as possible).

So the cross-contamination issue pretty much ruled out my Mity-Vac idea; it's vacuum brake bleeder parts would all be in contact with the sampled oil. Sure, I could have maybe cleaned the parts in between tests with something like brake cleaner, but that’s a pain, and I know from experience that those plastic parts really don't like brake cleaner. Clearly, the best tool to use is the special vacuum pump designed specifically for this purpose, along with a supply of clean new tubing that can withstand being immersed in hot engine oil for a few minutes without melting. Voila:

The cool thing about this style of pump besides its name (in oil analysis jargon, it’s called a "sampling thief pump"!) is that the sample tube goes right through it, and the sample bottle screws in underneath it. The only part that comes into contact with the actual oil sample is the inside of the tubing and the inside of the sample bottle. Perfect. This is the one I purchased from Amazon, it cost about $40.00 (USD).

 

 

Based on my research so far, it looks like Nylon tubing is the best/cheapest way to go. It can handle temperatures in excess of 100 degrees C, so it's no problem. BTW, if you're interested in science and chemistry, take a look at the Wikipedia page for Nylon some time, it's a pretty interesting read! I also wondered about silicone tubing, but apparently it should be avoided at all cost because the silicone will contaminate your sample. I settled on 5mm OD Nylon. It costs around $9.00 (USD) for 3 meters (Amazon.com again). The round steel ball on the 944's dipstick is right around 7mm, so this should slide right down the dipstick tube into the oil pan.

I measured the dipstick and marked off the tubing, adding about three inches, so the tubing will be sitting more or less in the middle of the oil when the mark I made is at the top of the engine dipstick tube. The other end of the nylon tube slips through the top of the pump, and you tighten the top ring onto the tube like a drill chuck.

Taking the sample

It was a beautiful afternoon; I had everything I needed to take my sample. Just over 1000 miles on the oil, and the sample pump and tubing had arrived the day before. I only needed to get the oil up to temperature and mixed thoroughly, but I ended up going for a two hour drive anyway. That’s just the way it is with this car.

Making a mess

Once back in the garage, I attached the sample bottle to the pump, and got ready to insert the sample tube down into the engine. With apologies to James Herriot, I pictured myself standing peacefully over the engine bay with the sample tube inserted a polite distance into the dipstick tube, while Gretchen calmly and happily gave up her oil sample to the little sterile sample bottle, with a gentle moooo. But that's not exactly how it went down...

Uggg, I had not given myself enough tubing between the top of the dipstick tube and the pump! Now I had to pump in an awkward spot right next to the brake fluid reservoir, and the slippery Nylon tube kept trying to come up out of the oil. I could only hold the pump at an angle, and I found out a little later that because of this, oil got all over the inside of the pump rather than flowing nicely out of the tube straight into the bottle. I would have to take the pump apart and clean it all out with alcohol.

Life Lessons learned while sampling hot oil

• Lesson 1 - Make sure the sample tube is long enough for you to be comfortable.

• Lesson 2 - Make sure the pump stays level and there is about 1/2" of the tube sticking down from the pump body so the oil flows out of the tube and into the bottle.

• Lesson 3 - Be careful what you clean the pump with - the wrong stuff can contaminate your next sample, and some solvents could damage the pump. I used 90% alcohol; any residue will evaporate away long before the next sample is taken.

I removed the sample bottle and placed it on the bench while I started to clean the pump.

• Lesson 4 - Put the cap on the sample bottle immediately after removing it from the pump!

Somehow, while disassembling and cleaning the pump, I knocked the uncapped oil sample bottle off the bench. Oil went everywhere, especially all over me! I ended up losing a pair of shoes, and a favorite pair of shorts is now relegated to garage-and-barn-only service from here on out. After quickly cleaning up the majority of the spill, I had to go in to take another sample, fast, so I would still be within the less-than-10-minutes-after-shutdown window, and of course I had to deal with all the same problems from the too-short sample tube. In the mayhem and confusion, I forgot to take any more pictures! This time though, I capped the sample bottle as soon as I unscrewed it from the pump!

So the first time out I made a couple of avoidable mistakes, ruined some clothing, and made a huge oil slick on the floor and the rubber mat in front of my bench, but at least in the end I got a good sample, and mailed it off the next day. I have to say that Gretchen is worth it. The next time I do this it should be much easier, and hopefully a lot less messy.

Tribology / Oil sampling Glossary

While I was doing research to find out more about oil sampling, I came across this surprisingly interesting glossary of terms. Here's the link in case you want to check it out: https://testoil.com/knowledge-center/glossary/

Thanks for stopping by! Stay tuned for more mechanical adventures with Gretchen. Next time, I’ll discuss the results of the oil analysis, don't miss it - sign up for email notifications! Click the green button below or enter your email address in the sidebar at the upper right.