Flash point, base oil quality, performance

Status
Not open for further replies.
Joined
Oct 20, 2005
Messages
11,956
Location
PA
I always thought there was generally a strong correlation between an engine oil's flash point and the quality of its base stocks, such that an oil with a higher flash point could generally (though not always) be expected to be more resilient at high temps: less shear, fewer deposits, etc. I fully expected there to be exceptions, but I thought that was more-or-less the general trend.

Now I feel like I might have been wrong about that. On the one hand, Castrol Edge Supercar HP -- their Indy Car oil -- has a flash point of 220º C; Mobil 1 FS 0W-40 is at 226º C; and most (all?) of the Pennzoil Ultra line is below 230º C, with some oils only just above 200º C. On the other hand, the dino 5W-20 I'm running now has a flash point of 230º C, and Idemitsu makes an $8/qt 10W-30 with a flash point of 260º C.

Is there no correlation between flash point and base oil quality? Can anything about an engine oil's performance be reasonably inferred from its flash point? If so, what?
 
The problem with flash point is the test reproducibility by ASTM D 92 is +- 9°C (+- 16°F). To this test reproducibility you need to add batch-to-batch variability, which can give quite a range to the formulation. In addition, the figure published by the manufacturer for the formulation may be the average of their batch data, or the highest, or even the lowest.

In my day I ran thousands of flash points, and monitored lab data on thousands more so I could report an accurate average on our ester data sheets. I knew exactly what the flash point should be for a given pure ester, yet the flash points listed on competitor data sheets for the same ester were all over the place.

If you had a true flash point on the motor oil base oils, I would expect some correlation to base oil quality within a given chemistry, but when you start blending base oil chemistry and adding additives, all bets are off - too many variables. For example, a 4 cSt Group I stock has a flash point of around 400°F while a 4 cSt PAO is about 425°F, a diester about 440°F, and a POE 480°F. These differences make it difficult to compare unknown blends.

Noack volatility on the finished motor oil is probably a better measure of base oil quality since it isolates volatility from flammability, but even here there are many variables. For all the above reasons I pretty much ignore flash points.
 
How did Noack came up with 250°C?
Or who came up with 250C?I

Do we want the Noack test temperature (250°C = 482°F) be above most (all?) oil flash points to be an effective test or does it matter at all?
 
Originally Posted By: Tom NJ
The problem with flash point is the test reproducibility by ASTM D 92 is +- 9°C (+- 16°F). To this test reproducibility you need to add batch-to-batch variability, which can give quite a range to the formulation. In addition, the figure published by the manufacturer for the formulation may be the average of their batch data, or the highest, or even the lowest.

In my day I ran thousands of flash points, and monitored lab data on thousands more so I could report an accurate average on our ester data sheets. I knew exactly what the flash point should be for a given pure ester, yet the flash points listed on competitor data sheets for the same ester were all over the place.

If you had a true flash point on the motor oil base oils, I would expect some correlation to base oil quality within a given chemistry, but when you start blending base oil chemistry and adding additives, all bets are off - too many variables. For example, a 4 cSt Group I stock has a flash point of around 400°F while a 4 cSt PAO is about 425°F, a diester about 440°F, and a POE 480°F. These differences make it difficult to compare unknown blends.

Noack volatility on the finished motor oil is probably a better measure of base oil quality since it isolates volatility from flammability, but even here there are many variables. For all the above reasons I pretty much ignore flash points.


Thanks, Tom!

As a largely-academic aside: is there ever a case where a tougher base oil is less volatile but not less flammable?
 
Originally Posted By: d00df00d

Thanks, Tom!

As a largely-academic aside: is there ever a case where a tougher base oil is less volatile but not less flammable?


Yes, different chemistries may have different flammabilities. For example phosphates would be less flammable at a given viscosity that say a hydrocarbon. There may also be flammability differences between linear hydrocarbons and aromatic hydrocarbons or esters or metal containing additives, but I do not have data on this. Molecular weight distribution is also a factor. The flash point test gathers oil vapors in the space above the oil surface, mixes the vapors with air and attempts to ignite the vapor/air mixture with a flame. The Noack test sweeps the vapors away making it a more pure volatility test. Since both tests are substantially affected by volatility there is a correlation but I wouldn't expect it to be perfect since there is no flame involved in the Noack (not to mention different temperature profiles).
 
Originally Posted By: OilUzer
How did Noack came up with 250°C?
Or who came up with 250C?I

Do we want the Noack test temperature (250°C = 482°F) be above most (all?) oil flash points to be an effective test or does it matter at all?


I don't know who decided on 250°C - hopefully it was correlated to some real world performance parameter as all tests should be, but I have also seen what I call "management by round numbers".

Noack and flash point are quite different. Noack evaporates the light ends of an oil by blowing air across the surface at a constant 250°C and measures the weight loss of the sample. Flash point slowly heats the oil from ambient to the point at which the vapors give a flash when a flame is passed over. The flame is not applied until the oil is within 27°C of the expected flash point, so much of the light ends are already evaporated off. If you added say 5% of a solvent to an oil, the Noack loss would go up by 5% since all of the solvent would be evaporated off under the conditions of the test. The flash point, however, would not be affected since the solvent would be evaporated off before the flame is passed over the oil surface.
 
Just wanted to mention that I find Tom NJ's contributions here to be very valuable and I always keep an eye out for his posts.
Thanks so much for sharing from your wealth of oil industry knowledge, Tom NJ!
 
"Just wanted to mention that I find Tom NJ's contributions here to be very valuable and I always keep an eye out for his posts.
Thanks so much for sharing from your wealth of oil industry knowledge, Tom NJ!"

+1
 
Thanks! Since I am retired for 11 years now I have to do something with all this useless knowledge floating around in my head before it all fades away!
grin2.gif


Now out to the garden to harvest some garlic, onions, peppers and beans.
 
Originally Posted By: Virtus_Probi
Just wanted to mention that I find Tom NJ's contributions here to be very valuable and I always keep an eye out for his posts.
Thanks so much for sharing from your wealth of oil industry knowledge, Tom NJ!

Seconded!
cheers3.gif
 
Originally Posted By: Tom NJ
Thanks! Since I am retired for 11 years now I have to do something with all this useless knowledge floating around in my head before it all fades away!
grin2.gif

Now out to the garden to harvest some garlic, onions, peppers and beans.

What time is lunch? I'm all for some fresh produce...
;^)
Sounds like you have found new ways to be productive in your retirement!
 
Originally Posted By: bluesubie
I wish there was an applause emoticon but Tom will have to settle for
thumbsup2.gif
.

cheers3.gif

11.gif
 
Originally Posted By: Virtus_Probi

What time is lunch? I'm all for some fresh produce...
;^)
Sounds like you have found new ways to be productive in your retirement!


Retirement is so busy that I don't know where I ever fund the time to work! Been growing vegetables for over 40 years, just on a larger scale now. Lunch is beet salad with onions, blue cheese, and walnuts. Better hurry over!

Oh, and I brew beer!
cheers3.gif
 
Originally Posted By: Tom NJ
Originally Posted By: d00df00d

Thanks, Tom!

As a largely-academic aside: is there ever a case where a tougher base oil is less volatile but not less flammable?


Yes, different chemistries may have different flammabilities. For example phosphates would be less flammable at a given viscosity that say a hydrocarbon. There may also be flammability differences between linear hydrocarbons and aromatic hydrocarbons or esters or metal containing additives, but I do not have data on this. Molecular weight distribution is also a factor. The flash point test gathers oil vapors in the space above the oil surface, mixes the vapors with air and attempts to ignite the vapor/air mixture with a flame. The Noack test sweeps the vapors away making it a more pure volatility test. Since both tests are substantially affected by volatility there is a correlation but I wouldn't expect it to be perfect since there is no flame involved in the Noack (not to mention different temperature profiles).


So, to make up another example:

- Engine Oil A has high-ish volatility but low flammability
- Engine Oil B has low volatility but is more flammable

If I'm understanding you correctly, I'd think Engine Oil B might show a lower flash point even if it were made of stouter stuff overall. Does that sound right?

On another point: my amateur brain is having a hard time letting go of the idea that flash point numbers can't be relied on for useful information, if only because a lower flash point should mean faster burn-off and more deposits. But I think what you're saying suggests that it'd depend on exactly WHAT is volatizing and burning off, and what the rest of the chemistry is like. Is that correct?
 
I for one can't understand why anyone thinks flash point is generally relevant. Why would we care what temperature it ignites in the presence of an ignition source? Overall volatility yes, but why flash point?

Formulators would want to know for material characterization but how is it relevant to me the end user?
 
Would flash point in a used oil analysis be a more accurate snapshot, since it is one sample of oil, complete with all additives, (albeit used)? In other words, when I look at a UOA from my engine, is the flash point measured from that sample accurate? I know it won't relate to the flash point in the virgin oil, obviously because of use and contaminates, but maybe it is a useful number to tell how that specific oil in the engine held up?
 
Originally Posted By: kschachn
I for one can't understand why anyone thinks flash point is generally relevant. Why would we care what temperature it ignites in the presence of an ignition source? Overall volatility yes, but why flash point?

Formulators would want to know for material characterization but how is it relevant to me the end user?

So this isn't a defense of the idea; just an explanation of the reasoning.

When you look at a list of base oils' flash points like what Tom posted, it's easy to imagine that a higher flash point in a finished lube suggests a greater proportion of high-end base oils.

It's also easy to imagine that an oil with a lower flash point will burn off more easily, and possibly leave more deposits.

Then you might notice that well regarded brands like Red Line tend to produce oils with flash points on the high side. And that oils with higher and/or "tighter" viscosity grades tend to have higher flash points (e.g. 15W-40 vs. 0W-20).

Without better info, all of the above taken together can seem like a pretty good rationale.
 
Originally Posted By: d00df00d

So, to make up another example:

- Engine Oil A has high-ish volatility but low flammability
- Engine Oil B has low volatility but is more flammable

If I'm understanding you correctly, I'd think Engine Oil B might show a lower flash point even if it were made of stouter stuff overall. Does that sound right?


Possibly, depending on the differences in volatility and flammability. Flash point measures the amount of vapor produced under certain heating conditions (volatility), and the flammability of those vapors when a flame is passed into them. Noack measures just volatility. In general, an oil with a low volatility will produce less vapors and within a chemistry produce a higher flash point, but that doesn't make it a better base oil. For example a straight cut 5 cSt Group I stock may have an equal or higher flash point than a dumbbell blend of a 4 and 6 cSt Group II, but is not a "better" base stock.

Flammability comes more into play with blends of different chemistries making it more difficult to compare. For example in a Group I/POE blend vs a Group III, the POE may raise the flash point of the blend up to the Group III level but the poorer oxidative stability of the majority Group I would be a weak link making it inferior to the Group III.

While flammability differences are a relatively small factor in motor oils, if you just want to look at volatility, remove the flame. Plus the fact that flash point may completely miss low volatiles that boil off before the flame is passed, and has such a poor reproducibility, makes it an unreliable test.

Originally Posted By: d00df00d
On another point: my amateur brain is having a hard time letting go of the idea that flash point numbers can't be relied on for useful information, if only because a lower flash point should mean faster burn-off and more deposits. But I think what you're saying suggests that it'd depend on exactly WHAT is volatizing and burning off, and what the rest of the chemistry is like. Is that correct?


Yes. While there is a general correlation between flash point and Noack, I would look to Noack for a better understanding of the overall volatility of the oil, and in most cases a better assessment of base oil quality within a chemistry. One can still be misled however. Vegetable oils have a very low Noack, but the unsaturation makes them much less oxidatively stable and more prone to deposits. Volatility is just one factor in assessing motor oils.
 
Status
Not open for further replies.
Back
Top