Archives

I just noticed the FE calculator doesn't seem to make any distinction between city and hwy mileage, even though there's a percentage city input available. I think using that input to calculate the difference in EPA wrt each vehicle's driving conditions, instead of just comparing it to the static combined EPA measurement would be cool.

Hi Lyeinyoureye:

___Tom and I discussed this more then once when we first began building the DBase. The separation is very easy to do in fact but with the varying degrees of city/suburban/highway/stop and crawl jams, where is the cut off for each and every tank? I have 60% of 55 mph limited Interstate, 25% of 65 mph limited Interstate and 15% of a city/suburban condition for a normal commute. Add in a 60 miles of stop and crawl through a tank and where does that leave my own personal stats? Tom and I both travel the Chicago Interstates and arterials … I can turn an average commute to a 60%/15%/25% if I have no stop and crawls plus include my usual 10 + miles per very 200 traveling around town for groceries and stuff but it just is not that easy to rely on or even calculate. Very few of us can ever calculate our commutes this accurately so I based all the 3 hypermiling categories off of EPA combined because it makes sense. Very few can actually separate their City/Highway/suburban with any accuracy over a tank let alone a lifetime so that is why we tried to keep it as simple as possible … Now if we all lived in the country and had all-country road commutes ... the lmpg's would get stupid real quick and the type of commutes would be dead on each and every time for the most part :D

___I hope that helps?

___Good Luck

___Wayne

I figured it'd be a bit simpler since we're only looking at two conditions because the EPA only looks at two conditions, cruising and stop/go. Highway is highway/freeway, even if you get stuck in traffic, and city is everything else. Notes could help sort out any anomalies, like getting all greens during your commute, or going 5mph on the freeway.

It's sort of a moot point anyway, IMO. The EPA numbers of 25 city/32 highway for my Accord don't even begin to reflect what it's actual strengths and weaknesses are when driven in a truly FE manner. Working my butt off for every ounce (though probably leaving something on the table vs. how Wayne would do it) I have seen highway segments in the 45-46mpg range, and think I could have had a 47 or 48mpg segment last week if I hadn't hit the wrong kind of traffic. On the other hand, putting the same amount of effort into my stop-go-stop-light-turn-hill-stop-yougettheidea commute has given me numbers exceeding 50mpg when the stars and planets are in the right alignment. The door-to-door numbers were only brought down because I have to get on the freeway for several miles, bringing down FE like an amateur. So using something closer to the City EPA in that kind of setting would artificially inflate my percentages vs. how good or bad I actually am.

It's a great question to ask, though. For some people and vehicles it might make sense to differentiate.

It's just so odd. We're inconsistently comparing our FE to the EPA's inconsistent combined figure, seems like we're just compounding the inaccuracy. I figure we'd at least compare it to the EPA figure using the same inconsistincies they use to better illustrate how much their numbers can be off, or ditch the comparison all together since we're using techniques that aren't consistent with the average driver.

I wish the EPA would do away with all of their routes and calculations and just give us straight numbers. Honestly, there is probably not one person in the country who drives how the current test is run and there won't be one person who drives like the next version is run. So why not just tell me what a vehicle gets on a level 5 mile straight away once warmed up, at 65mph on a 70 degree day. Then tell me what it gets in stop and go traffic for 5 miles at a top speed of 30mph. That's all I want. I figure if it is that straight forward everyone knows how they came up with the numbers and how their driving compares. they can figure out what their FE should be. If the test is so complex that no one knows what it is or how to compare, how will anyone know what to expect. Give me a real simple "control" and I can figure it out. I know that I drive some hills and that the temps here tend to be hotter, so my mileage will be lower in the summer. I know my city driving involves more stopping and less going, so mileage will be lower. I have no idea whether I make 19 stops per hour with 12 minutes of idle time, 19 left turns, 6 right turns, hit 11 potholes, swerve to avoid the bike messenger, etc. so how would I compare my drive to the EPA test???

Hi Lyeinyoureye:

___You have to remember that the only reason the EPA rating is not exactly the FE you achieve is because you do not drive the exact EPA test cycle, you do not drive it with the same accessories (none), you do not drive it with the same exact number of stops, starts, temps, or anything else. With that being said, your car when new and setup per the manufacturers specs did receive a particular city AND highway FE on that exacting city AND highway test just as an 02 Accord, 04 Prius II, 94 F150, or 03 Jetta TDI did. This is an apples to apples comparison since it was the same cycle in the exact same manner for all vehicles. How your habits, commute, terrain, and temps change the parameters of the test does not change the fact that the 02 Accord, 04 Prius II, 94 F150, and the03 Jetta TDI all ran the exact test and the results were comparable for that test. In fact, it is the most repeatable test at this time. You can take a hundred brand new 05 Accord PZEV w/ an Automatic transmission setup per the manufacturers specs and run it through the FTP75 and HWFET. They will all receive within ~ 2% of one another (24 city/34 highway) for as long as you want to waste you time. The test is repeatable as it is being performed by EVERY manufacturer and every FE test facility in the exact manner in which it is modeled. The variability is practically 0 and that is why I believe it is such a good comparison between automobiles. How that translates to the real world is a completely different matter of course but again, if you run a Prius II on the FTP75 treadmill test with its parameters setup per the factory spec, it will get 60 mpg 98 + % of the time from everything I have read.

___Good Luck

___Wayne

Hi xcel!
I know that there is some disparity in driving conditions between some arbitrary driver over the lifetime of their vehicle, and the EPA tests for the vehicle. Which could probably be expressed with enough data, something like an OBDII/usb terminal, where a laptop could record speed/time and fuel consumption, but anyway... I'm getting off topic.

The EPA cycle is broken up into ~halves, one for city and another for highway. If we compare our mileage and driving habits that may not correspond to the ratio of strict city/hwy driving to the EPA cycle, then we may get a comparison that's even more innaccurate. For example, I thought about this after noticing my mileage comparison was only slightly above the EPA figures. I realized that the comparison was using the specific EPA combined value, and not comparing it to the city driving I was doing. If I were to just look at it, I would assume that the EPA wasn't that far off, but since I know what the city mileage is supposed to be, I know that using DWB and capping my maximum speed at 35-40mph instead of 45-50mph allows a 34% increase city efficiency, instead of the ~17% that the comparison we use gives.

We have the EPA estimates that have specific conditions and proportions, then we have our driving conditions that have different conditions proportions. By using a linear, constant, comparison of proportions that does not adhere to the variation in city/hwy conditions we may face when driving, we are altering the comparison. So when looking at that comparison, I may not realize just how much the EPA conditions were different from my conditions as I move closer to the extremes of highway, or city driving. It's kinda like, X1/Y1 are EPA city/hwy conditions, R1/S1 are the EPA proportions of city/hwy, X2/Y2 are my city/hwy conditions, and R2/S2 are my proportions of city/hwy. Since the combined EPA cycle already has their proportions in it, if I want to compare the EPA and my mpg consistently (X1/Y1~R1/S1)~(X2/Y2~R2/S2), I should include my proportions as well. Otherwise it's not symmetrical and can compound the inaccuracy of the comparison.
:Banane37:

I just added a signature today with my fuel economy. Based upon the percentage of combined EPA that it says I am getting in my van, I realized that there is an error with the combined EPA used to make the signatures. My van has an EPA rating of 20 city/28 highway. This does not correspond to a combined EPA of 24, which the signature uses. According to the EPA's website (http://www.fueleconomy.gov/feg/2001cartablef.jsp?id=21235), the combined EPA of my van is 23. This is in part because driving at lower FE has a more significant effect on overall FE than driving at high FE.

I could not find the exact weightings that the EPA uses to get the combined EPA. For my van, a 50/50 weighting gives 23.3mpg. When the EPA calculates annual fuel cost, they assume 55% city driving and 45% highway driving. Using this weighting gives 23.0mpg. With rounding, it seems that both of these could be right with my van. However, I tried these on another vehicle, (xcel's Honda), and got the wrong answer both times. Xcel's Honda has a combined EPA of 27 (http://www.fueleconomy.gov/feg/2001cartablef.jsp?id=20867). Using the 50/50 weighting, I get 28.1. Using the 55/45 weighting, I get 27.7. Both of these round to 28 and are different than the 27 that the EPA says is the combined.

It is possible that the EPA uses one of these simple weightings, but they use the raw numbers instead of the rounded numbers that we see and I am using. Spreadsheets with all of the various EPA numbers for a given year can be downloaded here (http://www.fueleconomy.gov/feg/download.shtml). You can also just find the combined EPA numbers for your particular car here (http://www.fueleconomy.gov/feg/findacar.htm). I know it is not really important and most people probably do not care, but I noticed it and thought that I would point it out.

It could also be that there are tanks where someone put in specificaly how much highway driving they did so that number is uses vs a 50/50 or 55/45 split.

It could also be that there are tanks where someone put in specificaly how much highway driving they did so that number is uses vs a 50/50 or 55/45 split.

From reading posts in this thread, I was under the impression that the calculated percentage of the EPA was calculated from the EPA combined fuel economy and not from each individuals personal weightings. I am under the impression that the database just averages the number for city and highway. For example, the signature for my van uses 24 (instead of the correct 23) to calculate the percentage, which is just the average of 20 and 28. Because of reading this thread, I did not even indicate a % of city driving. I left it at zero. If this was taken into account, my signature would have calculated versus 28 instead of the 24 that it does. Since the thread says that it is a percentage of combined EPA, it should be using 23.

Also, it is obvious that xcel's signature is calculated based on the average of the city and highway numbers for his car (24/34) which is 29. However, it should be using the EPA combined number of 27. I hope this makes sense.

Thanks.

It does. I guess I need to inspect it closer to see what is going on.