5w-20; Ford 4.6L engines; UOA testing

dnewton3

Staff member
Joined
May 14, 2007
Messages
11,370
Location
Indianapolis, IN
Here are two UOAs from our cars at home. These are both 4.6L MGM cars; one an '05 and the other an '07, both identically equipped. My wife and I both drive nearly identical routes to work; about 32 miles each way and we only diverge in the last mile (she goes east and I go west). Because she occasionally works from home, I even drive her car to balance out the miles. One was run on dino oil; the other car was run on synthetic.

Tell me, if you can, which sample is the syn and which the dino oil? Which is a brand-name lube and which a house brand? Tell me, if you think you can discern from the data, how many miles did I run each lube; did I run both the same amount, or one of them half of the other? How many miles are on the cars; are they similar or is there a large disparity in accumulated miles? What filters did I run? Did I "over-run" the OEM recommended O/FCI? Did I run the filters for a longer or shorter duration than the lube? Which used more make-up oil, or were they they same? It's OK to guess, but that's not very definitive. If you believe you know, then state WHY you feel your answer is right; what data proves your assertion correct?

Obviously there is a method to my madness here. Most of you know how I like to test the boundaries and shatter myths. I'm trying to get some of you oil bigots to quit staring at the bottle label, and rather look at the results. Quit focusing on VOAs and brand names, and pay attention to the data you pay for in UOAs. What goes into a bottle is not nearly as important as what comes out of the crankcase.

I'll be on vacation the next few days, away from the computer. I'll let you all discuss it and I will return this weekend to tell you which is which, and edit the fields to give the detailed info.



Html:


UOA sample # A B

Brand RK Peak

type dino syn

grade 5w-20 5w-20

filter TG TG

Oil miles 10k 10k

Veh miles 90k 230k

make up oil 1.2 2.4





Blackstone Data

w/ macro analysis

Univ std

Avg dev

@ 5.2k



Al 3 3 4 1.1

Cr 1 0 1 .5

Fe 10 6 15 9.5

Cu 1 5 5 2.9

Pb 0 0 2 .3

Tn 0 0

Moly 37 65

Ni 0 0

Mang 0 0

Silver 0 0

Ti 0 0

Potas 2 1

Boron 2 43

Si 14 10

Sodium 4 14

Calcium 2013 1942

Magn 15 20

Phos 612 615

Zinc 707 686

Barium 0 0







Sus V @ 210 51.8 54.9

cSt V @ 100 C 7.82 8.73

FP 415 435

Fuel
Antifreeze 0 0

Water 0 0

Insol .2 .3

TBN

TAN
 
Last edited:
Originally Posted By: mooferz
Yup, VI of 117 vs 136.

How do you know this? I don't see any lower temp viscosity numbers, only 100C/212F numbers.
 
Yes indeed this is difficult to tell as all numbers are too close. And having worked in a lab doing testing myself, these numbers could change slightly(though not necessarily switch) if run again.

A: Name brand syn/Hav
B: House brand dino/AAP

Very hard to tell even when using PQIA site as reference. It's a guessing game and has been fun. Thanks for the test!
wink.gif
 
Last edited:
Originally Posted By: dnewton3




Html:


UOA sample # A B

Brand * *

type * *

grade 5w-20 5w-20

filter * *

Oil miles * *

Veh miles * *

make up oil * *





Blackstone Data

w/ macro analysis

Univ std

Avg dev

@ 5.2k



Al 3 3 4 1.1

Cr 1 0 1 .5

Fe 10 6 15 9.5

Cu 1 5 5 2.9

Pb 0 0 2 .3

Tn 0 0

Moly 37 65

Ni 0 0

Mang 0 0

Silver 0 0

Ti 0 0

Potas 2 1

Boron 2 43

Si 14 10

Sodium 4 14

Calcium 2013 1942

Magn 15 20

Phos 612 615

Zinc 707 686

Barium 0 0







Sus V @ 210 51.8 54.9

cSt V @ 100 C 7.82 8.73

FP 415 435

Fuel
Antifreeze 0 0

Water 0 0

Insol .2 .3

TBN

TAN


a )A is dino. B is synthetic cos of better add packs in higher Mo and Boron ppm as well as higher FP.

b )Dino A has lower OCI of (guessing) 8-10K miles, cos of
-lower residual KV@100*C of 7.82 cSt indicating viscosity shear thinning (from typically virgin 8.5 cSt)
-despite no fuel dilution issue, and
-insolubles of 0.2 ;
-10 ppm Fe

c )Synthetic B has higher OCI of (guessing)14-16 K miles, cos of
-oxidative thickening as shown by increased residual KV@100*C of 8.73 cSt from (guessing virgin 8.5 cSt) and
-strictly no fuel dilution issues , and
-higher insolubles of 0.3 indicating prolonged OCI, and
-6 ppm Fe.
 
A basic UOA is not going to reveal comprehensive insight about an oils performance. Additionally, although the driving route is the same, that doesn't make this a controlled experiment.

And the bottle label is worth staring at because it tells you the specs the oil meets and more and more specs require synthetic oil to meet increasing performance requirements.
 
At first, I may have thought that "B" would have been the synthetic oil when only comparing this posts "A" & "B" oils. However after reviewing thr PQIA site on VOA 5W20 oils listed all in a row, it is difficult to tell VOAs & UOAs. I'm sticking with my:

"A" is the name brand syn
"B" is house brand dino
 
info updated 4-28-178 at 11:45 EST; check the lead post for the data reveal



The Rural King dino is $1.59/qrt every day walk-in price. The Peak is $2.99/qrt locally. Considering that both oils are within the 1st stdev of "normal" wear, they are essentially doing the same job. The OEM OCI for the 2005 is 5k miles, but in 2007 Ford upped the OCI to 7.5k miles, yet there were no changes in the design or equipment manufacture. Both oils here were run to 10k miles. And yet there is a ton of life left in the lubes; neither is anywhere near a condemnation point.

Two points to understand from this data:
1) the OEM OCIs are grossly conservative
2) for these conditions there is no benefit to running a syn; it cannot pay for itself because it did nothing statistically significant in terms of wear control to justify the cost differential
 
Last edited:
Will you do another UOA using the same oils but swapping cars? It would be interesting to see if the results are the same or if the dino has more wear on the higher mileage car. And also how much make up oil the higher mileage car uses on dino as it did use more than its lower mileage comparison.

Edit:

No offense but this is hardly shattering any myths at all unless you run a mirror UOA swapping the oil brand of each car and repeat. For all we know, the results may be that the high mileage vehicle will drink the dino and or have even higher wear as it's probably more worn out. The results may be similar once again and you may be 100% right. But the test conditions are not identical, they're very very similar and that's great. But one has more than double the miles on it so that's a large varying factor right there.
 
Last edited:
Testing is continuing; yes.

The points I made are valid.

1) OEM OCIs are ultra conservative, typically. Neither lube was anywhere near compromised. This is certainly not true of all conditions for all equipment; one must test to understand where they are at. This clearly shows I have no reason to OCI at 5k or even 7.5k miles. 15k miles would be easily done.

2) Macro data shows both are both below "normal" wear, and the variation is well within standard deviation. The typical up/down nuances of any UOA series are greater than the numbers both units and both lubes exhibited. I don't have nearly enough data to do micro analysis on either; macro data is both convenient and poignant.

The use of each is certainly similar enough to disregard it's variation in macro data; that's the beauty of large sample groups as it normalizes environmental and use factors.
 
Last edited:
Good to hear that you will continue tests. So each car will be doing 64 miles per day. Approx 1200 per month. 10k before xmas maybe? Will be interesting to see the next results.

I'm assuming mostly highway.

I agree with your points. Just still undecided on whether the dino is producing the same results until it is swapped to the other engine. Even if within a small margin, if the syn continues to show slightly lower wear numbers on each car, then it can be concluded that it is protecting better, even if only by a small margin. For some people, all the 1% here and there add up and are worth it for them.

As nickdfresh did state though, I'm not sure 5w20 can even be 100% dino. If not, then it is semi syn as a minimum, which explains the relatively close results. Your point I'd very true in that the cost does not justify the results. I wish I could find $1.59 per quart oil!

Also was there a TBN or am I going more blind?
 
I would what the TBN and TAN is for both if these oils?? You know exceptionally well that the TAN can easily exceed the TBN and still be serviceable. I just wonder what those results would be here. I would bet good money that the TBN and TAN are either equal or a slight amount of TBN above the total acid number. Which means keep on trucking.
The viscosity of both oils are very good as are the flash points.
 
Last edited:
This is about macro data and variation. There is always going to be some amount of give/take in the UOA process and wear numbers. For any product to be "better", it has to distinguish itself in terms of being able to produce a result outside of normal variation. Neither of these lubes did so; they both performed well within "normal" variation. We cannot say that either did better than the other, but we CAN say that neither did better than the other! So for about 1/2 the cost, the dino did the same job, under these conditions.

If you note, the syn had lower Fe wear (6 vs. 10ppm), but higher Cu (5 vs 1ppm) contrasted to the dino oil. I see that about everyone has overlooked that fact ... the syn had 40% less Fe wear, but 500% more Cu wear. That seemed to evade most folks ... Perfect example of how macro data is important to understand in terms of variation. Knowing the mean is only 1/2 the issue; one also has to know the stdev to understand true "normal" reactions.

I ran a dino oil (one of the cheapest around anywhere, but API certified) for 2x the OEM OCI, and yet it did an outstanding job, just as did the syn, with wear rates lower than average. For the syn to pay for itself, it would have to do one of two things:
1) provide wear rates statistically significantly lower equal to a magnitude of the cost delta (it did not)
2) last in use for 2x the lifespan of the dino lube (I currently have no idea just where that dino limit is, so I have no means to be able to calculate an ROI for the more expensive choice)

I chose not to pay for TBN/TAN. I've run many experiments on my Dmax and other vehicles, I've watched TAN cross over, and never seen any detriment in terms of wear metals. The theory is that a highly acidic lube will cause pitting, etc as the acid acts on the materials. That theory has never materialized in anything I've ever tested. The reason I believe this to be true (and also does Ryan Stark at Blackstone) is that because the sulfur components are now so much lower in fuels and lubes, and because the crankcases are now sealed very well (which limits moisture intrusion), it's become a moot point. Most assuredly, a grotesque example may be discovered as a result of neglect, but that would NEVER apply to a BITOGer. Essentially, OCIs happen frequently enough that the TAN crossover just does not materialize into a manifestation of real wear concerns; it's an unrealized theory. The theory is valid, it's just not common at all. I've run OCIs to 3x the OEM recommendation limit, and never seen TAN affect wear metals. After all, if the theory is that acids cause pitting, then the metals would show up in the UOA. It's not like the acid will carve out a huge chunk of Fe or Cu, so large that a UOA cannot see it. No - that cannot happen. Rather, the acid would lead to small pitting and micro-galling; stuff that would be sure to be seen in a UOA. And yet, I've not seen any TAN so great as to affect wear metals. Could it happen? Sure - with total neglect. But not in normal use. And so, after my own experiences, and seeing countless others in terms of TBN/TAN, I've just decided not to pay the extra cost; there's no benefit to me paying for something that does not manifest into a wear issue.

These two UOAs prove that a very cheap API lube did every bit the same job as a more expensive brand name syn. Macro data shows these two units were completely in line with desirable wear rates. The small differences are completely expected and WELL within typical variation. So why pay for more when you don't get more? The syn here did absolutely nothing to distinguish itself, other than cost about twice as much. This bears my repeating:
We cannot say that either did better than the other, but we CAN say that neither did better than the other!


What I find ironic is that some folks want to nit-pick this experiment, but they don't understand what it's supposed to prove. If you don't want to trust macro data analysis, that's OK with me. But then I ask that you never, ever post a UOA and try to compare/contrast it to any other here on BITOG, because if you don't trust my data, why would you trust your own? Until you have an absolute bare minimum of 30 samples, you cannot understand the typical variation of a process. Even if you ran 5k mile UOAs, it would take you 150k miles just to understand your one car's variation, holding all inputs the same. As far as I know, I'm the only one that has that kind of micro data. I've never seen a single BITOGer come even close to that kind of commitment in micro data collection. Therefore we are left with macro data analysis. And that, my fellow BITOGers, clearly allows us to conclude that the dino did everything the syn did here, for half the price.

I am in the process of continuing the experiment, but I am withholding the details, so that you all can take another stab at it. I am including a twist or two, so that the obvious reversal of lubes isn't going to be the easy guess. I'll update when the data is in.
 
Last edited:
Originally Posted By: dnewton3
This is about macro data and variation. There is always going to be some amount of give/take in the UOA process and wear numbers. For any product to be "better", it has to distinguish itself in terms of being able to produce a result outside of normal variation. Neither of these lubes did so; they both performed well within "normal" variation. We cannot say that either did better than the other, but we CAN say that neither did better than the other! So for about 1/2 the cost, the dino did the same job, under these conditions.

If you note, the syn had lower Fe wear (6 vs. 10ppm), but higher Cu (5 vs 1ppm) contrasted to the dino oil. I see that about everyone has overlooked that fact ... the syn had 40% less Fe wear, but 500% more Cu wear. That seemed to evade most folks ... Perfect example of how macro data is important to understand in terms of variation. Knowing the mean is only 1/2 the issue; one also has to know the stdev to understand true "normal" reactions.

Statistically can one say Cu wear in this dino of 1 ppm is better than the synthetic's 5 ppm ?
Whilst Fe wear is normal for both samples.

Originally Posted By: dnewton3
I chose not to pay for TBN/TAN. I've run many experiments on my Dmax and other vehicles, I've watched TAN cross over, and never seen any detriment in terms of wear metals. The theory is that a highly acidic lube will cause pitting, etc as the acid acts on the materials. That theory has never materialized in anything I've ever tested. The reason I believe this to be true (and also does Ryan Stark at Blackstone) is that because the sulfur components are now so much lower in fuels and lubes, and because the crankcases are now sealed very well (which limits moisture intrusion), it's become a moot point. Most assuredly, a grotesque example may be discovered as a result of neglect, but that would NEVER apply to a BITOGer. Essentially, OCIs happen frequently enough that the TAN crossover just does not materialize into a manifestation of real wear concerns; it's an unrealized theory. The theory is valid, it's just not common at all. I've run OCIs to 3x the OEM recommendation limit, and never seen TAN affect wear metals. After all, if the theory is that acids cause pitting, then the metals would show up in the UOA. It's not like the acid will carve out a huge chunk of Fe or Cu, so large that a UOA cannot see it. No - that cannot happen. Rather, the acid would lead to small pitting and micro-galling; stuff that would be sure to be seen in a UOA. And yet, I've not seen any TAN so great as to affect wear metals. Could it happen? Sure - with total neglect. But not in normal use. And so, after my own experiences, and seeing countless others in terms of TBN/TAN, I've just decided not to pay the extra cost; there's no benefit to me paying for something that does not manifest into a wear issue.

What OCI beyond which TBN/TAN crossover maybe of significance for dino as well as fullsyn ?
15K miles for dino?
20K miles for fullsyn?
 
Last edited:
Originally Posted By: zeng
Statistically can one say Cu wear in this dino of 1 ppm is better than the synthetic's 5 ppm ?
Whilst Fe wear is normal for both samples.


What OCI beyond which TBN/TAN crossover maybe of significance for dino as well as fullsyn ?
15K miles for dino?
20K miles for fullsyn?



Ehhhhh - sort of yes, sort of not ... Mathematically, you could make a reasonable argument for it. But I look at it more from a pragmatic viewpoint. The dino Cu reading is just exceptionally good. And it's likely to bounce back up a bit on the next OCI. When you have really low values, it's hard to call something "better" because, while lower is desirable, it gets into residual effects and it's hard to manage data in tenths if you look at wear rates. If you want to define "better" as simply lower, then that's up to you. I accept that wear metals are always in flux in an OCI, and when we take a snap-shot, we don't understand the fluidity of the numbers. Can we say the dino Cu in my UOAs here is lower? Yeah, but I don't really think that wear this low is anything to define as truly monumentally "better". Both lubes did a great job; the Cu wear was below average for both dino and syn. This is why I am very hesitant to claim something is "better", but very willing to define things as "normal". The Cu in both these are likely to go up/down a bit, as will all the metals. To call one lube "better" here, I'd have to see at least three of the five main wear metals (Fe, Cu, Pb, Cr, Al) all be statistically lower; outside of normal. That is just rare; only seen it a few times. It's very typical to see UOAs where a few metals may be lower, but others higher. But to call something statistically better, I need to se a propensity of data showing an overall reduction in a majority of metals, not be lower in one and higher in another. Here in mine, the Cu was lower with dino, but higher with Fe. But BOTH did "better" than average values. Hence my point I keep driving home; we cannot say one was overall better than another, but we can say that neither was better overall than the other!


As for the TBN/TAN issue, there is no set OCI factor. Obviously it will depend upon the specific lube and application. The crossover is going to depend upon the add packs and other factors; for example, there are times when a dino may have more Ca and Mg over a syn in fact. I've seen no evidence to show that there will be a predictable manner for generalization. The way to know it to track the wear metals, and look for a statistically significant shift in wear rates. And this is done with the greatest accuracy when both micro and macro data is practiced together. Since most of us don't do micro data, we have to rely on macro data. But it's generally moot; no one here runs an OCI anywhere long enough to ever see that escalation of wear rates.
 
I feel like you are trying too hard to make some black and white conclusion here. The vehicles have drastically different miles on the engines, driven by 2 different people, potentially different previous service history, different driving routes (even if they are similar). This 1 UOA hasn't proved anything beyond the shadow of a doubt.
 
Last edited:
Back
Top