Suzanne Herman | Blackstone Laboratories

TBNs & TANs: Part 2

In our last newsletter, we did a deep dive on the science behind the Total Base Number (TBN) and Total Acid Number (TAN), and what information we can (and can’t) glean from these tests. Go back and check that one out if you missed it. For this article, we went into the lab for a more hands-on approach, and encountered a few surprises along the way. Let’s roll up our sleeves and start experimenting!

How Does Heat Affect the TBN?
We know a variety of factors impact the acidity of engine oils, but we wanted to isolate just one variable to measure its impact. Heat was the obvious choice, for two reasons. First, overheating is a common problem, and we often find increases in wear, viscosity, and insolubles after even brief overheats, so clearly heat causes the oil to change. Second, heat is pretty easy to replicate in a laboratory setting. So we broke out the hot plate and got to work!

We took three samples of Rotella T 15W/40 and tested oil at three different temperatures:
-212°F (normal operating temp)
-302°F (upper end of most oil temp gauges)
-392°F (a temperature your oil hopefully never sees, at least not for long!)

After reaching the target temperature, a portion of each sample was removed at 2-hour intervals throughout the 8-hour work day. It would have been nice to continue past 8 hours, since that’s really not very long in the oil analysis world, but fear of burning down the lab kept us in line. And once we took the first 392°F sample (see Image 1), our fears were vindicated.

The most obvious change we noticed was color. Image 2 shows that heat alone caused the oil to darken over time, and you can even see a color change between 302°F and 392°F, in Figure 3 below. So something is definitely happening as the oil gets hotter, but does it affect the TBN?

Once all the samples were collected, we tested the TBN, TAN, and viscosity, with the results shown in Figure 1. Eight hours at normal operating temperature had virtually no impact, and even at a relatively high 302°F, the TBN only dropped slightly over time. At 392°F, however, the TBN took a nosedive. By the 4-hour mark, it was already down to 1.8 — low enough that we wouldn’t suggest running the oil any longer — and it kept going down from there.

Even though this is a small sample size, the data clearly shows that high temperatures do cause the TBN to drop more rapidly, and the effect is more pronounced as time passes. Heat is far from the only factor impacting active additives out in the real world, where the oil also has to deal with the negative effects of friction, contamination, and fuel blow-by, just to name a few. But it doesn’t seem farfetched to think some of the damage from overheating could be linked to a diminished capacity to neutralize acids, so the TBN might be worth tracking in those instances when oil temps end up off the charts.

Surprising Finds
It’s worth noting that the TBN fell without throwing the viscosity out of whack — even the 8-hour sample at 392°F, the only one that ended up obviously thicker, was still within the normal range for 15W/40 (12.7 to 15.8 cSt at 212°F). We typically associate a thicker viscosity with “heat damaged” oil, but except in extreme circumstances, it looks like that viscosity increase is likely due to other factors, like high friction causing a breakdown of viscosity-improving additives, rather than just the heat itself.

Another surprise was that the 392°F samples also had a noticeably lower TAN reading. If you remember from the previous newsletter, the TBN tends to go down the longer the oil is used, and the TAN tends to increase, so at first this seems like a really weird result. What’s going on here?

Well, it’s important to remember that the TBN and TAN tests technically measure two different things. The TAN is a measure of how acidic the oil is, while the TBN measures the oil’s capacity for neutralizing acids. It appears that the excess heat has caused a chemical reaction that caused the oil to become less acidic while at the same time reducing the oil’s capacity to neutralize acids, likely by damaging the TBN-boosting additives or causing them to fall out of suspension. If this were a real used oil sample, acids would also be building up due to the other factors in the engine we mentioned above (contamination, blow-by, etc.), causing the TAN to increase. But since this was a controlled experiment where heat was the only variable, that didn’t happen here.

A Colorful Surprise
Before we end: a mystery! While we researching this article, we got curious about Aeroshell 100 Mineral oil. In theory, this type of oil shouldn’t have any additives at all (the normal additives in automotive oil can cause catastrophic detonation in aircraft engines), so we wanted to confirm that this oil would start out with a 0.0 TAN (it did) and a low TBN (also true at 0.0).

What we weren’t expecting was for the titration solution to turn bright purple! Usually samples stay evenly light yellow throughout testing, but once the basic solution (potassium hydroxide and isopropyl alcohol – the “Bruce” from Part 1) was added to this sample, the color drastically changed. A few minutes later, it began changing back to yellow and you can see that process beginning in Image 4. This made us even more curious, so we found a bottle of Aeroshell W100 (which does have some ashless-dispersant additives) and measured its TAN and TBN as well. They were 0.0 and 0.2, respectively. And lo and behold, this sample took on the same cheery magenta once the basic solution was added.

We aren’t sure why this happened, but we have a guess. Phenolphthalein is a pH indicator used in some other sorts of acid-base titration. It turns purple in basic solution, in the same way litmus paper turns blue. Phenolphthalein won’t show up in our spectral exam, since it’s made of the same elements as oil (C20H14O4), but maybe it or another pH-sensitive substance is present in these oils? We’re not sure! If you know the answer, we’d love to hear from you at bstone@blackstone-labs.com.

Answering the question! So, back to the main question — Do you need a TBN or TAN? The results of our experiments suggest that these tests can be helpful, especially if your engine oil got much hotter than it should. If you’re not having a problem with temps though, the main reason to get a TBN is what we discussed in Part 1 — seeing if the oil can be run longer.
In either case, the TBN or TAN readings provide additional data points, but they don’t replace the information on wear levels and physical properties found in the standard oil tests. Whether you’re concerned about heat damage or just looking to run a few thousand miles longer on the next oil, we look at all of the data to answer your questions and give you a complete picture of the engine’s health.

TBN/TAN: Do You Need One?

“Do I need a TBN?” It’s a question that comes up a lot. The TBN is a test we do on engine oil, while the TAN is meant for transmissions and other gear lubes or hydraulic oils. These tests are widely discussed on internet forums, where facts and misconceptions can be hard to distinguish. So let’s dig into the science behind them!

What is a TBN or TAN?

The Total Acid Number and Total Base Number are ways to determine how acidic oil has become (TAN), or how effectively it can neutralize the acids that form from combustion and other factors (TBN). An increasing TAN indicates more acidity, while a decreasing TBN shows an oil’s acid-neutralizing additives are being used up. You may remember the pH scale from science class. pH is a more familiar measure of acidity in everyday life. So why don’t we use pH on oil?

pH stands for “potential of hydrogen” and measures the flow of hydrogen ions in a water-based solution. pH doesn’t apply to oil because these ions can’t flow through oil – it’s a poor conductor. That’s why oil is used as an insulator for transformers and other applications that call for interrupting the flow of electrical current. Fortunately, we can use titration to get around this obstacle. Titration is used to determine the concentration (in this case, the acidity level) of an unknown solution (oil) by exposing it to measured quantities of a known solution (acid or base).

Running the tests

We start by mixing one gram of oil with a happy blend of toluene, chloroform, isopropyl alcohol, and a splash of H₂O. (Kids, don’t try this at home!) The solvent breaks down the oil into a solution that is a better conductor, so we can measure the pH. The next step differs slightly for the TAN or TBN.

For the TAN, we add a cocktail of chemicals – let’s call it Bruce – to the toluene solution, a little at a time. This continues until the pH reaches 11. The lab techs then use an equation to calculate the TAN from the amount of Bruce added to the oil-toluene blend.

The TBN follows a similar methodology, except the solution added is more acidic – more of a Boris than a Bruce.

The end goal for the TBN titration is a pH of 3, and as with the TAN, the lab people are doing some math to transform the amount of Boris added to the oil into your TBN number.

Why get a TBN?

As the oil circulates through the harsh environment of a hot, running engine, combustion causes acids to form. These acids can cause increasing wear and corrosion. To prevent this, the oil manufacturers add detergent additives to the oil, which help it buffer those acids and stabilize the oil’s pH. The higher the TBN, the better your oil can resist becoming acidic. That’s the main reason to check the TBN. It’s a helpful data point if you want to extend your oil change interval beyond manufacturer recommendations.

The TAN does essentially the same thing, but we use the TAN on oils that don’t have detergent additives (like hydraulic oil and ATF). Some industrial equipment manufacturers will set standards for when to change the oil based on the TAN.

Which oil has the highest TBN?

The TBN is mainly based on the amounts of calcium and magnesium (detergent additives) in the oil. Oils with more of those additives typically have a higher starting TBN, and those with less will rank lower on the list. Is more better? Not necessarily (and we’ll get into that a little later). Meanwhile, Figure 1 lists a slew of virgin engine oils and their average starting TBNs, from highest to lowest. The progression isn’t perfectly consistent, because we don’t test for every conceivable substance the oil manufacturers might include that determines the TBN. chart showing the TBN for various types of oil

As you probably know, the TBN drops pretty fast when you start using the oil. Then it levels out and drops more slowly, the longer the oil is run. Figures 2 and 3 show two types of Mobil and how the TBN tends to fall as acidic substances start to “use up” the detergent additives. That’s what they’re there for, and we consider any TBN over 1.0 sufficient, while a TBN of 2.0 or greater is ideal when choosing to run the oil longer than you currently are. Note that ppm calcium and magnesium stay roughly the same – it’s their ability to neutralize acids that decreases.

chart showing the TBN and TAN of Mobil 1 5W/30 at various mileage intervals

Figure 2: This oil has an average starting TBN of 7.5. Note the roughly inverse relationship of the TBN and TAN readings; as the TBN decreases, the TAN increases, as less “active additive” is available to neutralize acids.

chart showing the TBN and TAN of Mobil 1 Annual Protection 0W/20 at various mileage points

Figure 3

Figure 3: This oil has an average starting TBN of 7.9. The chart shows the same fairly predictable drop in TBN as miles increase. Interestingly, the TAN is less predictable, probably due to factors outside the scope of this newsletter.

Is more better? A look at two novel blends

It’s easy to see how you might feel like you want an oil with a starting TBN that’s as high as possible. But Figure 1 makes it clear that oils with all sorts of starting TBNs are available. Did the manufacturers at the low end of the scale just cheap out on additive? Not at all. Oil manufacturers have to cater to an array of unique engine designs, operating conditions, etc. As technology evolves, so does oil.

Joe Gibbs

See, for example, Figure 4, which lists a few different samples of Joe Gibbs Driven D140 oil. It had the lowest average starting TBN (4.4) thanks to fairly low levels of calcium and magnesium. This left the TBN between 2.0 and 1.0 after just 5,000-6,000 miles. But the engine that produced those numbers was a Porsche 911 that had excellent wear trends (see Figure 5). This oil is specifically formulated with lower calcium and higher moly to combat low speed pre-ignition and reduce abnormal combustion and wear. While we can’t say whether this oil really does reduce LSPI, it seems to work as well as others do and we see no problems with the novel additive blend.

Figure 4

Figure 5

Figure 5: Wear trends for the five samples in Figure 4 (and one additional sample, not included there because TBN and TAN were not requested). Wear is consistent over time and compares favorably to averages, despite the low TBNs.

Chevron Delo

Chevron Delo 600 ADF is another oil that breaks the traditional additive mold, and it’s fairly new to the market. The 15W/40 and 10W/30 formulations hold the 2^nd and 3^rd place spots for lowest starting TBN in Figure 1, which is surprising, since they’re formulated for diesel engines – diesel oil tends to have more dispersant additive than oil designed for gasoline engines (most of the oils in Figure 1 are gas engine oil). Figure 6 shows the Delo 600’s TBN reaching our “1.0 limit” starting around 9,090 miles. Chevron also had particular goals in mind for this oil – it uses “ultra-low ash additive technology” and is meant for engines with SCR and EGR emissions systems that need to meet state emissions standards.

Figure 6

Figure 6: This oil had an average starting TBN of 4.7. The chart shows the low levels of calcium and magnesium that resulted in fairly low TBNs after typical oil runs for diesel engines.

Since additive packages tend to be proprietary and Chevron never did respond to my email, we can only speculate as to how the elements we find in our testing relate these constraints. Maybe such low calcium and magnesium reflect a reduction in calcium sulfonate and magnesium sulfonate (the compounds that register as calcium and magnesium). While these compounds work well as detergent/dispersants and their alkalinity helps buffer acids, their presence would also boost the sulfur content – a potential problem for emissions goals. But the additive package is unique in other ways too.

Oil report for a virgin sample of Chevron Delo 600 ADF 15W/40

Figure 7

Figure 7 is a virgin sample of Chevron Delo 600 ADF 15W/40. Note the high levels of molybdenum, potassium, and boron, and low levels of phosphorus and zinc, in contrast to the more typical additive package shown in the universal averages column. Moly seems to be providing most of the anti-wear properties that phosphorus and zinc ordinarily would. Potassium is noteworthy and caught our attention right away, since it’s one of two potential markers for anti-freeze.

We’re not certain what additive compound registers as potassium in this oil, but because potassium is alkaline, perhaps it performs some of the same functions calcium sulfonate and magnesium sulfonate do in more traditional additive packages.

Interestingly, even when potassium (and sodium, which is also alkaline) is truly from coolant contamination, it can skew the TBN. Figure 8 is an example of an engine suffering from coolant contamination, which is taking a heavy toll on the bearings and physical properties of the oil. An oil change (and probably major repairs) are needed, yet out of context, the 10.0 TBN looks great. But that doesn’t mean the oil is ready for more use; rather, coolant is skewing the reading. That’s why we never judge a used oil sample by a single data point!

Oil report for a used sample of Valvoline 10W/30

Figure 8

Figure 8 shows a sample of Valvoline 10W/30, which has an average TBN of 7.2 out of the bottle. The oil was used 3,000 miles in an engine with a major coolant problem, seen in very high levels of potassium and sodium, a thick viscosity, high insolubles, and high wear levels. The TBN is very high at 10.0, but that doesn’t mean the oil is ready for more use; rather, coolant is skewing the reading.

As for Chevron 600 ADF? The jury is still out on what kind of results this oil will produce over time, since most of the samples we’ve tested so far are from young engines going through wear-in. It will be interesting to see how these engines mature, but we suspect in the end, this unique oil will perform as well as any other in the most crucial ways: lubricating, cleaning, and cooling engine parts. We’ll just have to give it some special treatment on our end, to avoid false positives for anti-freeze, and avoid putting too much stock in “low” TBNs.

We hope you’re walking away armed with knowledge and a pretty good idea whether adding a TBN or TAN is going to serve your particular aims. If you’re wanting to extend your oil changes, go for it! If you just want a basic assessment of how your engine and oil are holding up, not to worry! We can provide that with the core tests in the standard analysis. Stay tuned for Part 2 next newsletter, where we venture into the lab, and learn about the effects of heat on TBNs and TANs.

About Suzanne Herman