Thin cleans better, allows longer OCI than thick!

[Shannow]

Originally Posted By: Gokhan
Absolutely! BOQI does predict the base-oil quality (PAO, GTL) and you pay for the PAO and GTL that goes into the 0W-xx's.


It does nothing of the sort...

I just showed you 6 oils with 100% PAO basestocks, and your BOQI ranked then from 71 to 46.

 

[Garak]

Originally Posted By: Gokhan
Yes, most 5W-40's (PCMO and HDEO) I've seen are made of the cheapest base stocks they can get away with and are loaded with a very high dose of the cheapest VII, which makes the oil even worse, potentially leading to deposits and sludge. There are probably some good 5W-40's though. Most 0W-40's are good bu there are exceptions there, too.

Which ones?

 

[Virtus_Probi]

Originally Posted By: Gokhan
[
It's very simple: BOQI ~ 1/CCS/NOACK ...

I would read this as BOQI ~ NOACK/CCS, which would suggest a lower BOQI for a lower NOACK loss.
Do you mean BOQI ~ 1/(CCS*NOACK)?

 

[Gokhan]

Originally Posted By: Virtus_Probi
Originally Posted By: Gokhan
[It's very simple: BOQI ~ 1/CCS/NOACK ...

I would read this as BOQI ~ NOACK/CCS, which would suggest a lower BOQI for a lower NOACK loss.
Do you mean BOQI ~ 1/(CCS*NOACK)?

Yes, mathematically, in standard notation, 1/x/y means (1/x)/y, not 1/(x/y), as division is not associative and you carry out the operations from left to right in standard mathematical or computing notation. However, I understand the confusion. That's why it's a good idea to put parentheses when writing a paper so that it's not misread.

Yes, BOQI ~ 1/(CCS*NOACK).

 

[4WD]

Arithmetic heiarchy aside ... back to the total formulation discussion at 10 pm ...

 

[Gokhan]

Originally Posted By: Gokhan
John Rosenbaum (not me) actually patented it:

Chevron US patent on predicting the base-oil quality through CCS and NOACK

Last night I gave quite a bit thought on John Rosenbaum/Chevron's patented "base-oil-quality prediction" vs. my (unpatented) "base-oil-quality index (BOQI)."

The two approaches are similar but he makes base oils by mixing two base stocks to generate an entire curve of NOACK vs. CCS with the varying ratios of the two, whereas I look at a single base oil with a single NOACK and CCS value.

The main difficulty in both his patented approach and my approach is that CCS strongly varies with the temperature. In fact, CCS asymptotically reaches infinity at a certain low temperature, depending on the type of the base oil. This is hard to account for, especially because the variation of CCS with the temperature strongly depends on the type of the base oil -- like a "cold" viscosity index.

My conclusion was that both his patented approach and my approach have their applications. However, I concluded that I actually like my approach better than his patented approach. There are two reasons. Firstly, mine is more general, as in his case the temperature is fixed, while in my case, I have a heuristic constant that allows comparing across different cold-viscosity grades (0W-xx, 5W-xx. 10W-xx, etc.). Secondly, my calculation is simpler, and simplicity often deserves merit, especially when you are trying to come up with benchmarks.

Both his patented approach and my approach are far from perfect. Nevertheless, you can use my BOQI values to get a good feeling about the base-oil quality and/or composition of a given oil. While it's a benchmark, it shouldn't be taken as an absolute indicator of the base-oil quality but more like a correlation with the base-oil quality and composition.

Before I conclude, I include again the reference table for BOQI of various oils tested by PQIA:

Reference table of BOQI for 0W-20 and 5W-30 oils

Should I have my BOQI patented as well? I'll pass.

 

[Shannow]

And it gives six fully synthetic, PAO based oils six wildly different answers regarding the "basestock quality".

 

[Gokhan]

Originally Posted By: Shannow
And it gives six fully synthetic, PAO based oils six wildly different answers regarding the "basestock quality".

Wildly?

(1) The NOACK values you used are expressed in a single significant figure, which means you can express the BOQI only to a single significant figure (50, 60, 70, etc.).

(3) There is also no guarantee of accuracy on either CCS or NOACK for these internal research oils.

(2) They are all above 50, which is only attainable by PAO and GTL -- so, it checks.

(3) The VII content seems to decrease BOQI, as it increases the CSS and probably the NOACK as well -- not a bad thing because we don't like higher VII content either.

(4) There is no such thing as PAO base stocks with different viscosities in the same base-stock slate should have the same BOQI. In fact, if a base stock has a lower CCS without the NOACK being too high, this should indicate a better base stock.

(5) BOQI is a benchmark that should be considered on its own -- tells you how good your CCS and NOACK are -- and the quality and type of the base oil is correlated with it -- not necessarily precisely given by it.

 

[PimTac]

All this math makes my head spin. As a member of the math deprived portion of the boomers, the only division I know is the hangman style. Within a couple of years the schools switched from old math to new math then back to old. I am one of those Sputnik kids.

 

[Shannow]

Originally Posted By: Gokhan
Originally Posted By: Shannow
And it gives six fully synthetic, PAO based oils six wildly different answers regarding the "basestock quality".

Wildly?

(1) The NOACK values you used are expressed in a single significant figure, which means you can express the BOQI only to a single significant figure (50, 60, 70, etc.).


OK, I'll cede that...not that it's important here 50 to 70 for the six formulations...all of which have 100% PAO basestocks...160, 140, 100 for the individual basestocks...I stand by "wildly".

And thus looking at your ranking using the BOQI...it changes things to. Is the "zero" at the end of each NOACK a significant figure or not ?

They ALL happen to end in a zero, which is pretty amazing if it's a measured zero.


Originally Posted By: Gokhan
(3) There is also no guarantee of accuracy on either CCS or NOACK for these internal research oils.


Surely the data sheet for the base oils themselves can be trusted ??

Originally Posted By: Gokhan
(2) They are all above 50, which is only attainable by PAO and GTL -- so, it checks.


Originally Posted By: Gokhan
(3) The VII content seems to decrease BOQI, as it increases the CSS and probably the NOACK as well -- not a bad thing because we don't like higher VII content either.


Again, I'll use "wildly", as there's no correlation between them other than the numbers are different.

I'll posit that my CCS/MRV ratio (no I'm not going to patent it) gives a better understanding of the impact of VII on the oil...



Originally Posted By: Gokhan
(4) There is no such thing as PAO base stocks with different viscosities in the same base-stock slate should have the same BOQI. In fact, if a base stock has a lower CCS without the NOACK being too high, this should indicate a better base stock.


If it's an index of "Base Oil Quality", then SURELY spectrasyn same manufacturer, same product line would at least be "close" to each other.

Originally Posted By: Gokhan
(5) BOQI is a benchmark that should be considered on its own -- tells you how good your CCS and NOACK are -- and the quality and type of the base oil is correlated with it -- not necessarily precisely given by it.


You can see the CCS, MRV, and NOACK...that tells you what and "how good" they are...and I disagree with the last point...the WILD variance does not allow anything to be drawn from the number.

Originally Posted By: Taleb
Wen you develop your opinions on the basis of weak evidence, you wll have difficulty interpreting subsequent information that contradicts these opinions, even if this new information is obviously more accurate

 

[aquariuscsm]

Hey Shannow,which oils are the 6 pao oils? Some of these graphs and technical papers go way over my head

 

[Shannow]

Hope that the link works...This is a source that both Gokhan and I have used variously.

Hope link works

 

[Gokhan]

As I said the thinner base stocks tend to have a higher BOQI and the amount of VII content decreases the BOQI in finished oil, which explains the gradual BOQI fall toward the right in your table.

If you're uncomfortable with that, then use the BOQI to compare only within the same SAE xW-y viscosity grade.

Now, the question remains whether the thinner base stocks in the same base-stock slate, before you add any VII, actually have more paraffinic (linear) content and therefore less oxidation than thicker base stocks or not. I would think the answer is probably yes, as when the base stock gets thicker, the molecules would have to become not simply longer yet still linear but more branched and therefore less paraffinic (less linear), which would increase the tendency toward oxidation. The paper I posted in my OP is the only thing we have so far and feel free to do research on it and let us know if you find anything.

 

[Shannow]

Originally Posted By: Gokhan
Now, the question remains whether the thinner base stocks in the same base-stock slate, before you add any VII, actually have more paraffinic (linear) content and therefore less oxidation than thicker base stocks or not. I would think the answer is probably yes, as when the base stock gets thicker, the molecules would have to become not simply longer yet still linear but more branched and therefore less paraffinic (less linear), which would increase the tendency toward oxidation.


PAO aren't "linear"

Linear is not the definition of Paraffinic, it's the degree of saturation...and they can be branched, and still paraffinic.

 

[Gokhan]

Originally Posted By: Shannow
Originally Posted By: Gokhan
Now, the question remains whether the thinner base stocks in the same base-stock slate, before you add any VII, actually have more paraffinic (linear) content and therefore less oxidation than thicker base stocks or not. I would think the answer is probably yes, as when the base stock gets thicker, the molecules would have to become not simply longer yet still linear but more branched and therefore less paraffinic (less linear), which would increase the tendency toward oxidation.

PAO aren't "linear"

Linear is not the definition of Paraffinic, it's the degree of saturation...and they can be branched, and still paraffinic.

Shell US patent US7129197B2: Synthesis of poly-alpha olefin and use thereof

EXAMPLE 8

A 500 ml pressure reaction bottle with a magnetic stir bar was thoroughly vacuumed followed by purging with argon. Then the bottle was charged with 50 ml of dried toluene (distilled over potassium) and 100 ml of 1-decene (dried over 5A molecular sieves). At 50° C., 20 ml of 3.3 M methylaluminoxane in toluene solution was added to the reaction bottle and stirred for 20 minutes. Then 4 ml of 0.049 M bis(isopropylcyclopentadienyl)zirconium dichloride in toluene solution was injected into the bottle to start the reaction. Temperature of the reaction system was controlled by a constant temperature bath within ±1° C. range. After 1 hour, 100 ml of 10% HCl aqueous solution was added to the bottle to quench the reaction and the resulted mixture was stirred for two hours. The organic layer was then isolated and further washed twice with 150 ml of deionized water. Toluene solvent was subsequently removed from the organic layer on a rota-evaporator. Analysis of the product mixture by high temperature simulated distillation (ASTM D2887, Modified) indicated oligomers yield was 96%. The oligomer mixture contained about 43% dimer, 24% trimer, 11% tetramer, 5% pentamer and 17% higher oligomers (FIG. 1).

The oligomer mixture was hydrogenated and then fractionated by vacuum distillation. A fraction (Fraction E8-I) containing 89% dimer, 7% trimer, and 4% higher oligomers was collected and further characterized (FIG. 2). The viscosity properties of this fraction were determined to be kinematic viscosity at 100° C. (KV100) of 1.8 cSt, kinematic viscosity at 40° C. (KV40) of 5.3 cSt and viscosity index (VI) of 138.9. The PDSC induction time was recorded to be 44.2 minutes. In comparison, a commercial poly-alpha olefin obtained from British Petroleum (BP) under the trade name of Durasyn 162 with KV100 of 1.8 cSt, KV40 of 5.7 cSt and VI of 94.1 was measured to have PDSC induction time of 27.3 minutes. Durasyn 162 was determined to contain 85% dimer and 15% trimer (FIG. 3). Both the higher VI and longer PDSC induction time (better oxidation stability) of the newly synthesized poly-alpha olefin indicates it has substantially higher linearity than the existing commercial poly-alpha olefin product of equivalent viscosity.

 

[Shannow]

This is what they mean...not that they are making molecules that more closely resemble broomsticks.



Edit...per you patent, here's one of the trimer sub-blocks as they build the PAO structure.



Note the lack of broomstick like structure.

 

[Gokhan]

Yes, PAO molecules are oligomers (short-chain polymers) -- dimers (most commonly), trimers, tetramers, etc. -- and never monomers alone that make up these oligomers. "Branches" refer to the branches in the monomers and they are not desirable, as partially pointed out in your picture above, and they lead to more oxidation and poorer volatility, CCS, etc. These are the different possible types of monomers (source):

The Shell patent says: "However, it has been observed that boron trifluoride catalysis causes excess skeletal branching during the oligomerization process. An increase in the amount of skeletal branching directly correlates with an increase in the number of tertiary hydrogens in the molecule, which are prone to oxidation, and therefore exhibit poor stability when used in lubricants."

Coming back to BOQI ~ 1/(CCS*NOACK), this is how ExxonMobil promotes their higher-BOQI (lower-CCS-and-NOACK) PAO -- called PAO plus or SpectraSyn Plus™ (link):

As you can see, their better PAO, or PAO plus, is directly promoted as having a higher BOQI than their standard PAO -- plus simply meaning higher BOQI -- which immediately translates into longer oil-change intervals (OCI's) or less oxidation as their first selling point. It even claims less wear, probably because of a lower coefficient of friction.

ExxonMobil PAO plus = higher-BOQI PAO

Higher BOQI obviously means less branching and a better PAO.

 

[JAG]

I ran formulation #4 in the following link through the calculation as:
1,160,000/(6500*7.14) = 25
I used a numerator of 1,160,00 because it’s what you suggested for 10W-30 oils, as this oil is.
The link: https://patents.google.com/patent/US20170029734A1/en
The oil’s base oils are entirely PAO, ester, and alkylated napthalene.
25 is a relatively poor score but the overall quality of the base oils is very high, considering all relevant properties.

This BOQI metric is wasting your time, in my opinion. It’s not a good indicator of base oil “quality”. It is a mathematical function of CCS (big is bad), volatility (big is bad), and a fudge factor to correct for viscosity grade. True quality is not determined by just CCS and volatility with equal weighting; the relative importance of those two factors depends on the engine itself, the manner in which it’s used, and the ambient temperatures. Some base oils have poor CCS and/or volatility but they are overall desirable to include at some treat rate to provide those desirable properties. If you insist on continuing exploring it, I suggest you take your large dataset of CCS and volatility numbers and do a scatter plot of one vs. the other. It would be interesting to see that plot.

 

[Tom NJ]

4 cSt PAO is about 90% trimer of 1-decene and about 10% tetramer. 6 cSt PAO is about 50/50 trimer/tetramer.

 

[Gokhan]

Originally Posted By: JAG
I suggest you take your large dataset of CCS and volatility numbers and do a scatter plot of one vs. the other. It would be interesting to see that plot.

Chevron has done these plots and I put them a few posts back there.

The idea is not to study NOACK and CCS on their own. You don't need an index for that. Therefore, weighing them separately is irrelevant for any physical purpose. Besides, the units wouldn't match and the resultant physical quantity would make no sense. When you multiply them, the unit works out to be that for pressure (stress).

BOQI is not really intended for Group V oils, which could be anything. The ester oil you pointed out is not even for regular runs but for maintenance. Nevertheless, ester base oil has poor hydrolytic stability and that's probably what the lower BOQI is telling you.