What you get on the highway is your "highway mileage" though. For that period of time you are getting that 25.6 mpg and if that's disclosed in the discussion, there shouldn't be a problem with that. Someone that expects to match that over a full tank of mixed driving deserves the hurt feelings they get when they can't do it, IMO. The person giving the info isn't being any less truthful just because the reader doesn't understand basic statistics.

A full tank that covers 500 miles or so is going to cover a lot of different types of driving (city, highway, towing, idling, etc), rendering that information completely useless to anyone else that won't be doing the same exact type of driving. You're also bound to get different numbers from tank to tank making that info useless to yourself. Mileage of one type (like highway) with fewer cold starts is much more comparable between different people and run to run for the same person. When the EPA tests new vehicles and comes up with the city/highway mileage estimates seen on the window stickers and elsewhere, they aren't running full tanks. Additionally, if I were to do a full tank of fuel on the highway (almost never happens) and posted the same mileage I get in 30 miles of highway driving, would it be any more "real world? Of course not. It's not even more useful as a basic statistic... It's the same thing, try it.

If I reset the meter at the beginning of my work week, and I get 21 mpg over the 3 days, I can expect to get exactly that if I did the same thing for a full tank. The hand calc will be within a couple tenths of the meter. So when I say I get 21 mpg on my mixed driving work commute both ways including cold starts, that's exactly what I get. Averages are very simple math. Whether I do that over 3 days or a full tank, the result is the same and just as credible either way. If I give single trip averages or one way averages I always call them that.

Anyone that is only calculating full tank averages is really not interested in getting good mileage anyway. You can not accurately test different driving styles, methods, or routes if you wait to use a full tank and muddy the results with other irrelevant driving. There are sites and forums all about hypermiling for the rare individual interested in this sort of thing.

I realize that there are people out there that will inflate the numbers just to be "cool," but that doesn't mean that everyone that gets good mileage should be lumped in with them by the people that get poor mileage due to their technique and circumstances. Some of us have very solid testing methodologies, and might even do this sort of thing for a living, and can usually be spotted by the quality of the info (or disclaimers) given along with their mileage numbers.

Sorry for the length, driving for mileage is a long time hobby of mine.

 

Well, I reset my mpg readout the other day while going downhill and it said 99.9 mpg for about 300 ft, so I suppose I'll start telling people I get 99.9 mpg. After all it must be true if the computer said so.

 

Not sure how to break this to you, but you were then getting 99.9 mpg for that particular stretch of road in that example (or more if it was maxed out. Coasting fuel cut gives an MPG of infinity). If you tell people that it gets 99.9 mpg downhill for 300 feet, it's an accurate statement. That's obviously a useless test though, and is only good for posting facetiously on the internet to make yourself feel better about getting crappy mileage.

That example is a far cry from someone posting 30 miles worth of highway driving and calling it highway mileage. I get the point you guys are trying to make, but it's a useless one IMO. Maybe I don't come here often enough, but I haven't seen any really outlandish claims or obviously bad testing methodologies. From my experience anything up to ~23 mpg is perfectly achievable in certain conditions. Rather than focus on how these people must be inflating the numbers, focus instead on how to duplicate those results, or get as close as you can. Trust me, it's a good time.

I recently got 23.x mpg on a ~50 mile drive on back roads to drop my motor off at the machine shop, and 22.x on the return trip, door to door. Last weekend I got 22.x mpg going to NED and back for the NHRA top fuel event door to door, ~75 miles each way. I get 20.x on average on my mixed driving work commute both ways door to door. If you think I can't keep this up for an entire tank, you don't understand how this works, go find your nearest 7th grader to help you with basic statistics. As I said earlier, it doesn't make it any more "real world" than doing it for 50 miles. Would I still get that towing? Or in the dead of winter? Of course not, and I'd be a real jackass to expect it to, or to use that as evidence that the ~22mpg numbers must be bogus. They were accurate in the conditions in which they measured. Next time I'm going to take one of those trips I'll fill up first, get the mpg meter reading, then fill up again, and get the hand calculated tank average that will be within a couple tenths anyway. Then it will be acceptable to everyone.

As far as "the computer" being a derogatory term, keep in mind that the alleged "lie-o-meter" is an actual measurement of miles driven divided my an actual measurement of fuel consumed. It's not an estimate of any kind. It's a measurement. Anyone that has worked with factory EFI systems, and some aftermarket, understands how this works. As with any instrument, there is room for some variability and calibration. It is still quite likely to be more accurate than 99% of hand calculations done using the pump's auto shutoff feature, which is far more variable than the highly evolved EFI systems in use on modern vehicles. Hope that helps.

 

Not really, because you're still burning some fuel. The only way it can go to infinity is if you were to switch to neutral and kill the engine, where you're using zero fuel to cover the distance while coasting. It's still going to be up pretty high, though, sometimes over 99 mpg where even the instant number will give up reporting, as in if you reset it as you were coasting.

I don't reset the mileage every tank, normally, partly because I rarely wait till the 50 gallon Titan is running low before filling it up. I keep a pretty close eye on the instant readout, and lately ever since the Mythbusters episodes where they tested hypermiling concepts I've been focused on trying to be smooth and efficient. Long-ish term MPG right now is at 16.0 (lots of city driving), while the overall lifetime average after nearly 30k miles including towing goosenecks and 5ers is at 14.5. Just For a "real world" number to add to the mix.

 

I agree 100% with you. I keep trying to tell people the same thing.

 

I agree too

 

Kjewer,

When I went to school and we learned what "averaging" meant it was taking the total sum of the combined numbers and dividing them by another number, getting the "total average". See if I wasn't so stupid, I'd know to ask if people were giving the whole tank average or their best 2 gallon stretch ever mileage? The real world number is what I care about since one does have to indeed start & stop their vechicles at some point. Yeah baby, I get up to 99.9 mpg's in my 1 ton!

 

I understand that you consider yourself the authority on 6.7 MPG, but you apparently missed the part where I said that I manually downshifted to reduce speed. Comprehension is your friend.

Please do yourself a favor and read the posts.

 

You would normally be correct! But this is where modern EFI systems throw you for a loop. When you are coasting (0% throttle position, or idle switch true) above a certain speed (usually about 5mph), the ECU completely cuts fuel, so you are burning no fuel at all. This is done for emissions reasons, but there is a fuel economy benefit. The type of cars that I race were doing this as early as 1990. Taking advantage of this requires that the TC is locked up for auto trans vehicles, which is why I mentioned that previously. If the TC unlocks and the motor falls to idle, then you are correct, you are still burning some fuel. This is where the TC control strategy of the 6.7 liter trucks helps out, it's far better for fuel economy than the 5 speed auto in my 2004 that would unlock any time you let off the gas, effectively defeating coasting fuel cut if you weren't going fast enough to keep the engine above idle.

Whether or not to employ coasting fuel cut or put the truck in neutral and coast at idle depends on whether or not you can maintain your speed with the engine braking. It's always most beneficial to stay in coasting fuel cut of course, but if you need to coast in neutral to maintain speed it's the next best thing and still beats running at maintenance throttle by a long shot.

I'm not disagreeing with you, people need to disclose the exact conditions of the test that resulted in the numbers they are posting. That goes for many things, not just fuel economy. My point is simply that if someone states that they got 29 mpgs on their best 2 gallon average ever, that's fine, because they disclosed it. I can now go out and do the same 2 gallon test and see if I can match their results. That's obviously an absurd example, but you get the point. If someone posts 100 miles of "highway" driving even when already up to speed, I should be reasonably able to duplicate that performance in similar conditions or I need to work on my driving or setup.

The real question is what distance gives a statistically accurate or repeatable result. Purely mathematically, it doesn't matter. But there is some variability that can be reduced by going longer distances or using more fuel. For example, the variability in when the pump shuts you off can be made statistically less significant by dividing it over more fuel. A variability of .1 gallons means much less when it's out of 30 gallons vs 1 gallon. So I'm still all for longer running averages. The problem is that for most people in their daily lives, they can't maintain the same set of conditions for very long. If I want to see exactly what my fuel economy average is only going back and forth to work for example, pulling a trailer every Friday makes it impossible to do a full tank average.

That being said, the nature of averages is that averaging the same number over and over again results in the same average. If I get 21 mpg on a rount trip to work, it doesn't matter if I do it 1 time, or 50 times. The average is still 21. Which is also to say that it doesn't matter how many gallons you do it over. 3, or 30. Aside from the reduction in variability mentioned above, which is why when I measure my round trip to work average, I do it over a full 3 or 4 day work week.

While we're talking about testing methodologies, there is the subject of whether or not to include cold starts, getting up to speed, and all of the other concerns mentioned in this thread. Typically that should all be included if you want to know what your "cost" to go to work and back is, to use the above example. I reset the MPG meter on my first work day before I even start the engine. That way the cold startup, warmup, and getting up to speed and fully warmed up 10 miles later is all included. Purposely leaving that out and then not giving that disclaimer with your numbers is being intentionally misleading, I agree with that point.

There are times however when you don't want to include cold startup etc. Typically when working on your driving technique. You can't have variability in cold startup and warm up fuel consumption messing with your results. Part of doing good testing, in any field, is knowing how to reduce variability, and how to partion, or segment, exactly what it is you are trying to test. If I want to work on how I handle a particular section of highway, and try different ideas and see what works best, I need to reset the meter strategically to only show me what I want to see. The truck has to be fully up to temp, similar weather conditions, and so on. You can be sure I'm not posting those numbers as my "highway" mileage though...

Sorry for the length, again. As I said, I enjoy this topic. There is always disagreement in this area, but I appreciate the discussion just the same. It's a vast topic that could go on forever. As previously mentioned there are entire forums, enthusiast groups, competitions, etc, dedicated to hypermiling. My tow vehicle is a strange choice to apply this all to, but hey, it's a fun hobby, and gives me something to do while I'm stuck behind the wheel anyway.

 

Apparently you missed the part where I was agreeing with you 100%. That's why I said your theory (as you put it) was good.

I'm not the authority on 6.7 anything, believe me, but I am highly experienced in EFI and internal combustion engines, and hypermiling, and make a living from statistical process control in the semiconductor industry. Skills that readily transfer to the topic of discussion here. I'm not trying to argue just for the sake of arguing. I don't like to see misinformation being spread as fact, and if 1 or 2 people learn one thing from my incessant rambling it's worth it IMO. I'm bound to learn a thing or two myself, which is why we're all here.

 

I apologize if I misunderstood the reply -"That's good theory. Turning fuel into brake dust doesn't get you anywhere."

I was wrong when I assumed that the second sentence was a direct reference to what I wrote.

 

No problem here! That's what I get for trying to make the same point and be funny at the same time. I'm off my soap box now, I promise.

I'm doing a small experiment this week. This morning I stopped on the way home from work and filled up. I have two work days to do a test of low mileage mpg meter vs hand calc measurements. No special tricks to raise MPG, but I need to get the number high enough that the non-believers could have called BS on it. Almost 1 day into I'm at 22.x mpg, I'll try to keep it there. My only fear is that pumping 20 gallons this morning will result in more foam shutting the pump off earlier than it will on the 3 or 4 gallons I'll end up pumping Wednesday morning. We'll see how it turns out. I may fill up tomorrow morning as well to normalize the foaming issue, I can still add the two fillups together on Wednesday. Any thoughts on that?

 

I haven't noticed a correlation between the number of gallons pumped and the foam. I can actually continue filling slowly until diesel spills out of the filler tube.

That being said, I do not know if there is trapped foam in the tank that I have been unaware of, basically haven't thought about it because I "assumed" that the fuel flow through the injectors was measured/metered to calculate MPG.

I'm not sure that you will change any minds, but think that you could prove your point. My take is that some trucks will get different MPG just because there are so many variables - elevation, wind resistance, driving habits, quality of fuel, modifications (tonneau cover, lifts, tires, rims, etc), ambient temperature, A/C on or off, etc..................

 

Yea - the variables are probably the thing - In my '11 6.2 I never got what the other folks driving them would get. My spreadsheet proved to me the diefrence between the original bumper/ air dam and the Ranch hand bumper that I put on it - dropped about .5 mpg...

On my '13 6.7 I reset the milage on Monday morning & drove in to town 50 miles. Some speed limits 55 some 65 some 75mph. No lights, no traffic... I got to office with 20.4 mpg. Ran a few errands - stayed in town last night - ran around a little this morning - all stop & go... maybe 30 more miles on the truck and mileage reads 14.7 right now...

City/ranch kills mpg, Idle kills mpg, Texas hill country - always pedal pressure hurts -

On that note - best mileage I ever got with the 6.2 was coming from Corpus to Tilden, Texas after leaving the coast - about 300' rise per 100 miles... I'd call that flat!

 

Ideally you would be able to fill up to a point that you can see actual fuel and be able to repeat that over and over again, so that would be good news. On my 2004 you could "stuff the tank" with another 4-5 gallons if you pumped it super slow (we're talking ~5-10 minutes here), way too much of a pain in the ***. Word on the street (forums) was that the diesel truck shared the fuel tank with the gas trucks, which need room for expansion at the top of the tank (above the vent fitting). This led to "harpooning" of the tank and so on, which I never really looked into and have no personal experience with. I'll try filling this truck up to the filler neck and see if it's more feasible. That would eliminate the variability due to foam buildup. Thanks for sharing that.

Curiousity got the best of me this morning and I filled up again. I'll do so again tomorrow morning. I'm glad I did, because the result was pretty surprising, in that it did what I predicited it might do yesterday, but even more than I thought it would. I'll post the numbers tomorrow when I get another fill up from only a few gallons down. I really need a third one to fully normalize it properly, but I'm out of work days. I work 4 days next week and may try this then. The problem this week is that even if the shutoff is more accurate tomorrow, it is still compounded by the inaccuracy of this morning, so it's a "bad" test for the two days. The individual result for the second day will be more accurate, but still less so than the third day would have been. If I didn't fill up this morning I would probably have missed this. It highlights how easy it is to get a "bad" test even though it seems like it should be so simple.

I agree completely. I might have rambled on somewhat about that earlier. In addition to mods and driving habits, people with different terrain, routes, traffic patterns, etc may struggle to get the same mileage as someone with different circumstances. My only request for those people is that rather than call BS on the people doing better for whatever reasons, they see the differences for what they are and acknowledge the possibility that others could be getting much better mpg. They can work on the things within their control (technique, mods, etc) as I do and try to optimize it. The results can still be surprising even in less than ideal circumstances.

I'm not really looking to change any minds, but I would like to see people at least apply the laws of mathematics and physics properly so that we can reach comparable conclusions and learn new things. With the race cars I deal with, I don't like to try to get other people to think the same way that I do. Instead I help them to properly instrument their own cars, review the data, and come to their own conclusions. In this way we learn a lot more (and more quickly) collectively than I do on my own.

Byram, I'd kill for a flat road like that. You can't go more than 100 feet around here on flat ground, it's all never ending hills up here. I could do all kinds of testing. MPH sweeps in 5mph increments, cruise control vs load control, etc. I find it mildly disturbing that for a moment there I got excited about that prospect and even considered driving a good distance to find such a place...