Gen4 Subaru Outback, High Mileage Synthetic OCI?

[Random_0101]

Originally Posted By: bluesubie
IMO, unless are exceeding 7,500 mile intervlals there is no advantage of running synthetic in an naturally aspirated EJ. You'll be fine with any conventional 5W30 from Wal Mart.



Thanks for your reply. I did pick up some M1HM from Walmart on Rollback with a pretty good promotion (ends up costing me about $12 per 5qt bottle) and I'll probably consider switching to a High Mileage 5w30 synthetic blend unless the price difference is minimal (within a few dollars)

 

[oil_film_movies]

Originally Posted By: BufordTJustice
I've reviewed all of them. Nobody is recommending jumping two grades, and the percentage increases in FE you've stated are virtually all based on going from a non friction modified Xw-20 to an Xw-40 oil, or from a dino 10/15w-40 to a 5w-30 syn oil.

Wrong. The one chart you picked proved I was right about the generally 1.5% improvement going up one grade.
Take that picture, showing going from a 40 to a 20 weight, 2.8% total, cut that one in half, and you get 1.4% to get a 20 vs. 30 comparison.
Other statements and tests all over those tech articles and dozens of others show the average is around 1.5%.

 

[BufordTJustice]

Originally Posted By: oil_film_movies
Originally Posted By: BufordTJustice
I've reviewed all of them. Nobody is recommending jumping two grades, and the percentage increases in FE you've stated are virtually all based on going from a non friction modified Xw-20 to an Xw-40 oil, or from a dino 10/15w-40 to a 5w-30 syn oil.

Wrong. The one chart you picked proved I was right about the generally 1.5% improvement going up one grade.
Take that picture, showing going from a 40 to a 20 weight, 2.8% total, cut that one in half, and you get 1.4% to get a 20 vs. 30 comparison.
Other statements and tests all over those tech articles and dozens of others show the average is around 1.5%.


Still no.

Maybe you should talk to Garak about the actual statistical analysis involved in calculating gas mileage in the real world. You obviously didn't read anything I wrote, nor did you correctly contextualize the pictures I cut and pasted from your own examples.

Further, outside of the lab-based studies from which you've cherry picked your data, less than 3% is literally within the statistical noise of one's ability to accurately calculate gas mileage in the real world. 1.4% could be accounted for by the difference between filling up early in the morning versus late in the afternoon....alone. You're misleading the OP with unsupportable claims about his gas mileage being negatively affected when he doesn't even have enough empirical resolution to discern the effects you state will be there. So, even if the negative effects you assert are actually there in the real world (I do not cede this point to you, nor do any of those studies support it in earnest) are there, you're STILL wrong because the OP won't be able to reliably detect them (that's called statistical insignificance).

You really should read about Garak's professional work done in the field of statistical analysis in the area of accurate MPG calculation. Really.

 

[TheKracken]

Originally Posted By: BufordTJustice
Originally Posted By: oil_film_movies
Originally Posted By: BufordTJustice
I've reviewed all of them. Nobody is recommending jumping two grades, and the percentage increases in FE you've stated are virtually all based on going from a non friction modified Xw-20 to an Xw-40 oil, or from a dino 10/15w-40 to a 5w-30 syn oil.

Wrong. The one chart you picked proved I was right about the generally 1.5% improvement going up one grade.
Take that picture, showing going from a 40 to a 20 weight, 2.8% total, cut that one in half, and you get 1.4% to get a 20 vs. 30 comparison.
Other statements and tests all over those tech articles and dozens of others show the average is around 1.5%.


Still no.

Maybe you should talk to Garak about the actual statistical analysis involved in calculating gas mileage in the real world. You obviously didn't read anything I wrote, nor did you correctly contextualize the pictures I cut and pasted from your own examples.

Further, outside of the lab-based studies from which you've cherry picked your data, less than 3% is literally within the statistical noise of one's ability to accurately calculate gas mileage in the real world. 1.4% could be accounted for by the difference between filling up early in the morning versus late in the afternoon....alone. You're misleading the OP with unsupportable claims about his gas mileage being negatively affected when he doesn't even have enough empirical resolution to discern the effects you state will be there. So, even if the negative effects you assert are actually there in the real world (I do not cede this point to you, nor do any of those studies support it in earnest) are there, you're STILL wrong because the OP won't be able to reliably detect them (that's called statistical insignificance).

You really should read about Garak's professional work done in the field of statistical analysis in the area of accurate MPG calculation. Really.


I have a degree in statistics and from what you just wrote you obviously do not have a working knowledge of statistical analysis. Calculate over hundreds of gas tanks you certainly could see the variance between a heavier oil and a thin oil. ceteris paribus

 

[Garak]

That certainly applies over hundreds of tanks, and fleet usage. That gives one enough of a sample size to (hopefully) weed out certain variables. Me hopping around grades in my own personal vehicle isn't going to tell me anything, except demonstrate the danger of a large error bar.

Of course, it has to be said that going to a lighter grade of oil will save me money, even as a single user example. I may not notice and I won't be able to pick the difference out of the weeds, but it's still there. The problem we get on this site is that someone changes oil viscosity, and on their first tank, they're lamenting that they flushed half their paycheque down the toilet on gas, or they've somehow saved enough gas to earn CAFE credits for trade on the open market.

 

[TheKracken]

Originally Posted By: Garak
That certainly applies over hundreds of tanks, and fleet usage. That gives one enough of a sample size to (hopefully) weed out certain variables. Me hopping around grades in my own personal vehicle isn't going to tell me anything, except demonstrate the danger of a large error bar.

Of course, it has to be said that going to a lighter grade of oil will save me money, even as a single user example. I may not notice and I won't be able to pick the difference out of the weeds, but it's still there. The problem we get on this site is that someone changes oil viscosity, and on their first tank, they're lamenting that they flushed half their paycheque down the toilet on gas, or they've somehow saved enough gas to earn CAFE credits for trade on the open market.


I completely agree with just about everything you have said in the posts I have read of yours. My issue was mostly with the other poster throwing around terminology that either didnt apply or was almost the opposite of what they were trying to imply.

I do not think anyone should worry about running a thin oil for the purpose of fuel economy, they should weigh the benefits of thin vs thick oil on their own and the fuel economy should just be an added bonus. Of course in fleet usage it is a very simple way to see a small but certainly a statistically valid amount of fuel savings. This is where the argument about if thin oils are now being implemented due to cafe standards or not. I am sure CAFE has a huge influence in that, but I also believe thin oil has its benefits in certain application (I am no engineer and this is out of my expertise).

Also if anyone is worried about the fuel saving portion of the oil there are much easier ways to save money in regard to oil...find a cheaper oil? Use the oil to its maximum potential...ect.

 

[Garak]

As an aside, with the taxis, back in the day, my dad was a very early adopter of the ILSAC type oils. Even when I suggested a 5w-30 HDEO back in the day for personal use (long before the modern iterations of the E6 type 5w-30s), he thought I was nuts, asked me why I'd consider thick "garbage" like that, despite how much oil my car burned. His verbiage was considerably stronger. He must be spinning in his grave with me running a 5w-40 now, or a 15w-40 when I had the Audi.

A lot of the older mechanics around here, who are around from his times actually were very appreciative of the 10w-30 and similar options, given what they put up with in their youth in this province.

Of course, CAFE matters. When you get to a fleet size that of the U.S. general public fleet, it certainly does. For personal fuel economy, how hard one uses one's feet matter most. And, to expand on what you mentioned, even in fleet use, spending wisely on a lubricant and using it to its maximum potential can bring about significant savings. Of course, there is nothing wrong with using the specified grade, or the lightest of a few specified grades, to save fuel. After all, every little bit helps, but we do have to be realistic about what one saves.

CATERHAM once suggested to me that running an ordinary ILSAC synthetic would save me money over the Delvac 1 I've been using (which I had, historically, gotten a wonderful price on), enough to pay for the difference in oils. He's probably right, but that difference in price is so tiny. Heck, buying a stash rather than going to the store three times pays off, too, I suppose.

 

[oil_film_movies]

Originally Posted By: Garak
That certainly applies over hundreds of tanks, and fleet usage. That gives one enough of a sample size to (hopefully) weed out certain variables. Me hopping around grades in my own personal vehicle isn't going to tell me anything, except demonstrate the danger of a large error bar.

Statistics is certainly one valid way to gather the clear evidence that thinner oils save gasoline. As an engineer, I look at the very definition of viscosity, and understand its a drag on the engine (viscous friction). Certainly in the hydrodynamic portion of the Stribeck curve, and also in oil pump losses.

Maybe the real proof to some is you can run the FTP EPA drive cycle for MPG, and run one with 0w20, and the next with 0w30, and you will see the difference easily.

Again, as I've said, it does depend somewhat on the specific engine design & other variables. In by far the vast majority of cases, you will save somewhere in the range of 0.3% to 3% in gasoline by dropping 1 grade when its safe to do so. It is another matter what the minimum viscosity you should run for wear should be. The main point is to use the smallest viscosity HTHS you can get away with to get an MPG benefit without sacrificing wear. Run too low a viscosity for your specific engine type and you do risk increasing wear. For example, if GM says to not go any lower than 0w20 in their pickup trucks, as they do now, then don't put in 0w16, but also you don't need 5w30 necessarily either. Its "safe" to run 5w30 in a 0w20 engine like that, yet you will probably only lose MPG without decreasing wear in that application.

 

[Garak]

From first principles, it's, as you mentioned, obvious that lower viscosity will save gas. Quantifying it in the real world on an individual basis is always the problem.