Have you seen increased mpg using MMO?

Status
Not open for further replies.
Originally Posted By: kschachn

No, it wouldn't be a definitive test of MMO for fuel economy benefits. There are a myriad of other variables in everyday driving besides the energy content of the gasoline and nearly none of which you control. A controlled test in a laboratory where the only variable is the additive is the only way to test for fuel economy benefits. You just can't do it in real-world driving.


Yes, exactly - - we drive IN THE REAL WORLD!
With all variables firmly in place!
Why do you obsess over the requirement for a lab?

My driving style is very, very consistent.
If I average my numbers over many tanks, THAT IS "REAL" world!

Real world MPG increase on MMO with E-10 -
.2 mpg increase over about 10,000 real world driving miles. (Gas; Diesel engines do better for some reason)
That's honestly within "margin of error",
but it was still there nonetheless.
Whether you like it or not........

However, the even bigger MPG increase goes to E-0 gas.
I can average 17.9 MPG on E-10,
but switching to E-0 jumps it up to 19.2 MPG. That's pretty substantial.
 
Exactly why companies keep high paid R&D teams & costly labs even in tough times. Technology is "applied science" - and when the application is not precise - there is no scientific evidence one can monetize ...
 
Originally Posted By: kschachn
At the risk of angering the one who labels me an ignorant, flat-earth retard here is my opinion. I am not a statistician nor do I claim to be one.

I would guess, and this is only a guess, that the longer the test runs the less statistical correlation there would be. This gives more time for random environmental variables to make an impact, such as barometric pressure, temperature and traffic patterns. The longer you go the more potential there is for your mood and health to change, or perhaps demands on you to get to work faster because there is some project. Another variable that is common here is road construction. The specific roads I take every day to work may be under construction this year but not next. When real-world fuel economy is tested (in the real world) it is always run on the same day and on a closed track (in both directions) using standardized test fuel. Running the test on different days with different drivers will induce errors.

In a previous life when I worked as a Research Technologist for a large company, my job was to come up with tests for materials that would yield statistically valid results which then I had to present to my superiors for review. We had two mathematicians on our staff and their job was to evaluate proposed research for statistical validity. I can tell you this, it was often surprising how large or how small a sample size was required to get valid data. One thing was certain however, it was not something that could be guessed.

Back in college when I ran the test for fuel economy it was an ASTM design single-cylinder engine with a known mass of standard fuel, and everything was compensated for in terms of charge temperature and density. In real-world driving you have none of that, and add into the mixture the attitude and emotional state of the experimenter (you) and it is obvious that there is a lot going on that can't be eliminated out. People like to think that there's a lot of variables cancelling out in real world testing but that is not true. In fact just the opposite is true. Environmental cycles are't actually annualized, what little I've seen is that they cycle on a far longer time scale. In this scenario by then you've moved, took a new job, and are driving on a different road where E10 is now the fuel of choice whereas it wasn't before. Add to it the fact that your car now has 30,000 more miles on it and one of your fuel injectors is leaking just the tiniest bit. All that adds to error and is why I would say that the longer the test runs, the more error will be present.

This whole discussion is just silly on so many levels. It is not rocket science to measure fuel economy and it doesn't take anywhere near two years to do so. If someone (such as the MMO Company) wished to show that their product actually benefits fuel economy then that can be conclusively determined and it doesn't cost a lot of money to do so. Then they can publish those results in their literature and on their website for everyone to see, avoiding long Bitog threads and accusations of folks being ignorant retards. That is how it is really done in the real world.


Firstly I am a freelance statistician with a degree in mathematics. Assuming you keep some variables constant...the longer you run the test out the most valid your results become. This is because...like you said with construction from year to year, this will start to average out in a sense, especially with the longer you run the test. Or weather from year to year, if you only base results from mpg this winter vs last then statistically it will be really hard to discern what is what...but as you increase the number of years the winters will converge to a average value and this is where you can start to notice a difference. Another way to see statistical significance is in fleet operation.
 
Originally Posted By: TheKracken
Firstly I am a freelance statistician with a degree in mathematics. Assuming you keep some variables constant...the longer you run the test out the most valid your results become. This is because...like you said with construction from year to year, this will start to average out in a sense, especially with the longer you run the test. Or weather from year to year, if you only base results from mpg this winter vs last then statistically it will be really hard to discern what is what...but as you increase the number of years the winters will converge to a average value and this is where you can start to notice a difference. Another way to see statistical significance is in fleet operation.

OK so taking into account all the relevant variables, what is a statistically valid sample size in this instance?

Also, I don't know what the statistical term is, but what would be the overall resolution of the measurement? In other words what would be the smallest MPG variance you could reliably measure due to one isolated variable?
 
The sea monster's post was pretty much in line with my understanding of random variables, in that they tend to revert to the mean over a large enough sample size, although I didn't especially enjoy stats when I had to take it.
Road testing to measure the change imparted by the single variable of the use or lack of a fuel additive would be a pretty blunt tool. The reason that I referred to the OP's figure that I inferred to be in the range of 6-7% is that I think that would be a difference that could be reliably attributed to the fuel additive.
A difference of 1% would be no more than noise.
To demonstrate the validity of what you wrote in one of your posts above, I'll note that the Accord got 32 mpg on its last tank, which is about 4% better than average for weather conditions and typical use.
Why? No idea, but that's also a random variable that will revert to the mean.
Had I used a fuel additive, I might have attributed the difference to it, which would have been fallacious reasoning.
 
Originally Posted By: TheKracken
Firstly I am a freelance statistician with a degree in mathematics. Assuming you keep some variables constant...the longer you run the test out the most valid your results become. This is because...like you said with construction from year to year, this will start to average out in a sense, especially with the longer you run the test. Or weather from year to year, if you only base results from mpg this winter vs last then statistically it will be really hard to discern what is what...but as you increase the number of years the winters will converge to a average value and this is where you can start to notice a difference. Another way to see statistical significance is in fleet operation.

What you describe is normal, but what I recall was that it wasn't true when a large number of environmental variables are involved. Seasons are not the same year-to-year, the cycles are on a much longer time scale. Long sampling periods only allowed other often unanticipated variables to present themselves. For us it was completely dependent on the magnitude of the variable under study, if for example you are only looking for an overall temperature effect in a building environment then that is one thing. But if you are looking for a small variable (as would be the case for an additive to fuel) then it was a very different scenario.

But you may be correct and my memory of what we did may be faulty.
 
I agree that if you want to measure an improvement in fuel consumption (especially in regards to something with a small impact such as fuel additives) the best way to do this is in a laboratory. But if you are careful enough in setting up a long term test or a large fleet test you can get statistically significant differences.

If you are tying to measure on a single vehicle without a control on your own by switching between using a fuel additive or not it is practically impossible to accurately measure a difference. There obviously would have to be more calculations preformed but I would assume the higher mileage your vehicle is the worse it will preform which in itself will throw off data. But in fleet usage where you have certain vehicles with the fuel additive and ones without putting lots of miles per year by random placed drivers, all that statistical noise that is an issue with a single vehicle becomes a lot less relevant and the difference doesn't get "lost in the noise".

This is similar to CAFE regulations and thinner motor oils (I don't want to argue about if thin vs thick is better for your vehicle, I am not a mechanical engineer). Thinner motor oils are proven in a laboratory to produce better fuel economy. This in turn can be seen in fleet usage....or in the case of CAFE the fleet is the entire US fleet of cars. As the number of vehicles used to test the results increase the noise from environmental and other factors becomes less of an issue. Theoretically as you let the number of vehicles used (or time in a perfect world) go in infinity you should converge on the value improvement found in the lab. But obviously that doesn't happen.

My point really was that people logging data and trying to see a difference is still important, but certainly not definitive. On our own we will never actually notice a difference that we can calculate, but that doesn't mean it is not there.
 
Originally Posted By: TheKracken
My point really was that people logging data and trying to see a difference is still important, but certainly not definitive. On our own we will never actually notice a difference that we can calculate, but that doesn't mean it is not there.

I agree, with the caveat that correlation is the problem. In my first post about this topic I mentioned that you may see a difference in fuel economy in your tracking, but being able to attribute it to one variable (in this case MMO usage) is not possible. Only in a laboratory where the MMO is the only variable could it be correlated.
 
I support your recommendations however MMO is not a bad lubricant - I even have it as part of a home brew lube for weapons ...
Originally Posted By: Donald
It's a terrible lubricant and an OK cleaner. Any MPG gain is wishful thinking.

Keep your engine maintained using quality synthetic oil and Techron once per oil change.

Forget the snake oils like MMO.
 
The flat-earthers will never stop, even when logic and science are against them. Hilarious, really. And sadly, educational.
 
This forum has changed in the last year or so. Before, users would post their opinions and their reasoning behind them. They would accept other's input and build a constructive thread. Now, if you don't agree with a person's interests or reasoning behind their opinion, the name calling begins almost immediately. It's elementary and quite embarrassing, really. Some people do experience different outcomes with certain products, as seen here. Clearly, not all variables are taken into account, ie. elevation, relative temps, fuel quality in certain areas, humidity, etc. Someone in the NE US will have different outcomes as compared to someone in the Sonoran Desert.

Do I believe MMO adds additional MPG's? I haven't used it long enough to really see a difference. However, I will mention that over a year of use (with logging), I have seen about a 1 MPG difference using Lucas UCL at every fill up. General avg for 2015 was 16.3 as compared to 17.4 in 2016. I used the Drivvo app to log all fill ups. I think that modern engines actually do see some benefit from a UCL. I think that is what is giving people the slight bump in MPG's; Not so much the ingredients for cleaning. Now, you may commence the name calling...
 
Status
Not open for further replies.

Similar threads

C
7-eleven diesel fuel sourced on west coast?
2 3
Replies
43
Views
3K
CleverUserName
victorforge
Extending life of gasoline by storing under argon
2 3
Replies
44
Views
2K
wag123
P
Techron concentrate, better MPG illusion or running lean causing better MPG?
2
Replies
30
Views
3K
carl2591
Graham Piccinini
Marine is better than Automotive?
Replies
11
Views
2K
Graham Piccinini
U
I used 14 fuel system cleaners in 1 OCI and looked at it with an endoscope
2 3
Replies
41
Views
3K
jtay2024
Back
Top