Answer #23 concerning Noack

[wemay]

From the Pennzoil API SN Plus & 0W-16 Q&A:


Originally Posted By: Pennzoil
• One strength of formulating with PurePlus base stocks is clearly observed in NOACK performance. Though this test is important to specifications, the NOACK is a laboratory bench test known for wider variability in test results relative to other bench tests. This variability in test results makes it more difficult to pinpoint the accuracy of a single result and minimizes the value of comparing results among different products.


Seems like we've been chasing our tails all this time.

 

[JHZR2]

I wrote the question and am a little disappointed by the answer.

Lots of tests have intrinsic variability. It can be due to a lot of things, including the nature of the item under test, the setup of the test, etc.

If NOACK is that bad, it shouldnt be an industry standard test. If it is highly variable, the method should be updated to reduce as much variability as possible. Alternately, if another method ("bench test") performs better, that result could be provided. Beyond that, NOACK should be an objective test, even if with caveats.

I can't imagine Pennzoil does not do the test.

I cant imagine the oil does that bad.

I would imagine that it does as good if not better than others on the market.

I would imagine the other bench tests implied would show similar results.

If they even just put less than the high end of their typical range, assuming its better than other basestocks, it would help inform interested customers.

Makes me question if some of the prior numbers that were passed around and even some of the past tests are relevant anymore. Unwillingness to answer could even imply cheapening of the product. Its simple enough to provide a value.

 

[Nyogtha]

I think this is a result of folks being disappointed with the NOACK test value increase in PUP vs PU. NOACK limits certainly exist to meet some OEM specs so Shell must routinely test for simple batchwise QA/QC just like any other hydrocarbon product. IMO they're just weasling on why they choose not to provide those values on their PI sheets.

I still trust reported random sample test results from PQIA, personally.

 

[demarpaint]

It was quite clear they didn't want to answer the question. Whenever an oil company offers a Q&A they should be better prepared in how to handle questions they don't want to answer. Leaving unanswered questions to the imagination of the people asking them is never a good thing, as seen here.

 

[Rolla07]

I think the variation in NOACK is small between most oils and in being just one single factor isnt all that important. The numerous used oil analysis don't reflect that a specific positive or negative trend can be made for NOACK alone as its simply one factor of the oil itself.can we form a decision of how good an oil is based on an two oils where NOACK is 8 vs 10 lets say? Id say its unlikely.

 

[JHZR2]

Originally Posted By: Nyogtha


I still trust reported random sample test results from PQIA, personally.


Agree. But they don't have everything.

What they need is a base price to do a testing work up, then have a "gofundme" line for different products. People can vote with their dollars on what products get tested. Some standard and randomly sampled products can keep getting done via sponsor money.

So if I want Pennzoil Ultra 5w-40 current stock re-tested, I can donate my $20 along with a bunch of other interested parties, and it will get done...

 

[Linctex]

Originally Posted By: Nyogtha
IMO they're just weasling on why they choose not to provide those values on their PI sheets.


That's exactly how I read his answer.... completely dodging the question with relatively meaningless banter

 

[Nyogtha]

Every such test has statistical limits of repeatability and reproducability, and there are calibration standards, round robins, and all sorts of ways to make sure the testing is being done properly within these limits.

What's interesting is Shell choosing not to include those numerical limits in their response in a technical Q&A. I'll bet TominNJ can tell is those limits for the current NOACK test protocols.

 

[ZeeOSix]

Sounds like a "Rat Test".

I would think it can't be that difficult to accurately measure an oil's volitility/mass loss at X temp for Y time.

 

[Linctex]

Originally Posted By: ZeeOSix
I would think it can't be that difficult to accurately measure an oil's volitility/mass loss at X temp for Y time.


With today's technology of (EXTREMELY precise) precision measuring equipment of all kinds...

Performing this test accurately today - could be done with 50 years old technology.

 

[ZeeOSix]

Pennzoil said:
"Pennzoil doesn't typically report NOACK values due to the inherent variability of the test. That said, Pennzoil products are formulated to have robust volatility characteristics".

How do they know it has "robust volatility characteristics" if they can't accurately measure it?

 

[Shannow]

Originally Posted By: wemay
Seems like we've been chasing our tails all this time.


By chasing our tails, I take it that you mean that we should ignore it so that the koolaid of preference appears as equally as possible.

Here's a discussion from some time back...

https://bobistheoilguy.com/forums/ubbthreads.php?ubb=showflat&Number=3156566

There's a repeatabilty and reproducability with every "standard" test

 

[Nyogtha]

They must test it, but don't have to report it other than internally, except probably checks on meeting OEM specs with that particular OEM or OEM agency, or similar audits.

I found a copy of D5800-08 (2008) revision on the web via a simple Google search but I'm betting it would violate site rules for me to post the link.

There's 3 methods inside the procedure:

A: Manual method
B: Automated method
C: A method that allows for collection of condensed vapors for anysis, helpful in looking for phosphorous and such

Method A is noted for giving slightly lower results for finished lubricants, Method B is noted gor giving slightly lower results for base stocks. However, the r-squared value of all 3 methods together is shown as 0.996

Repeatability for Method A is listed as 5.8% of test result value. So if the test result was 10.0% for example, the limits of repeatability for that result would be + or - 0.058% since the engineering units for the test are % mass loss

Reproducability for Method A is listed as 18.3% of the test result so for a test result of 10.0% the teproducability limits would be + or - .183%.

Since NOACK tesults are specified to be reported as XX.X%, you can see the repeatability for a result of 10.0% would round to + or - 0.1%, and reproducability would round to + or -0.2%

The statistics don't look too wildly variable to my engineer's eyes.

 

[Nyogtha]

Method B shows repeatability = 0.095X^0.5, where X is the average of two test results. So for an average test result of 10.0%, repeatability would be + or - 0.3%. Reproducability limits are shown as 0.26X^0.5, so for an average result of 10.0% reproducability would be + or - 0.8%. So, the automated method definitely has wider limits than the manual method, but is undoubtably more cost-effective in lab tech hours saved over some investment period for the automated equipment.

Still doesn't seem too awful to me except for hair-splitters who chase NOACK differences of 1.0% or less.

 

[Nyogtha]

Method C is a bit different as the statistical limits are not a function of the magnitude of the test results. Repeatability is listed as a fixed 0.81% and reproducability is listed as a fixed 1.62%.

To me as a former custody transfer inspector, a simple thing would be to require the manual method to be the referee method in all cases of dispute.

As far as setting up a home-brewed NOACK test set, one thing stressed is the need for a draft-free environment. The test method lists and describes the necessary equipment for each test method and draft shielding methods and apparatus. That's the part that won't be a poor-boy job, as anything else will undoubtably have wider limits of repeatability and reproducability and be difficult to prove calibration on since the apparatus wouldn't be standardized.

Keep in mind more recent revision(s) to this test method may have different limits, apparatus tweaks, etc. This is a 10 year old revision of this test method.

 

[1JZ_E46]

Their answers regarding NOACK were total cop-outs. Let’s be honest. Same for HTHS.

 

[CleverUserName]

I also asked the same question regarding the Joe Gibbs Racing DP40 and DR40 oils.

Also received the same canned response. 11% sucks for allegedly mPAO based products.

Quote:
The NOACK on both oils is less than 11%. Because the NOACK test doesn’t have great repeatability, we don’t publish the results, but the typical result is under 11%.

Please let me know if you have any questions.

Thanks,

Lake Speed, Jr.
Driven Racing Oil
8649 Hacks Cross Rd
Olive Branch, MS 38654
Cell: 704-239-4401

 

[wemay]

Originally Posted By: Shannow
Originally Posted By: wemay
Seems like we've been chasing our tails all this time.


By chasing our tails, I take it that you mean that we should ignore it so that the koolaid of preference appears as equally as possible.

Here's a discussion from some time back...

https://bobistheoilguy.com/forums/ubbthreads.php?ubb=showflat&Number=3156566

There's a repeatabilty and reproducability with every "standard" test


I was being facetious but also know that quality is difficult to express in a forum.

 

[Shannow]

Ahh, I didn't pick it up.

No

 

[Nyogtha]

I got it wemay!