2004 Honda Civic Hybrid, AT-PZEV, CVT
(Tank on AT-PZEVs only holds 11.9 gallons)
90% 53 MPH on freeway on cruise control.
Low fuel light just came on at 602.9 miles, so I could still go 60 to 100 miles more I think.
Somone here asked for a photo:

Attachment 3780

Hi Kenny:

___Great job on the tank! I looked at your last tank doc and did you fill her up with just 9.7 gallons? If so, you have almost 2.2 gallons left which is good for 125 + miles, not 60 – 100!

___Now, drive w/ load and use 53 mph as your target and see what happens. Buzz70 will be eagerly awaiting the results ;)

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

Yes 9.7 gal filled it up.
I don't top it off though.
When the automatic gas nozzle thingy stops, that's it.
I don’t want to top it off since gas evaporates.

Also, I fill up when I have only one bar on the gauge.
I am afraid to go too dry.
Doesn't crap get sucked up into the fuel injection if you run too low?

I'm not sure I can really drive with the load.
My route to work does have several slight up and down inclines, but I really do not feel safe allowing the car to fall below about 53 MPH.
Which means I would be letting fall to 53 as the lowest at the top of the inclines.
That means I’d be going OVER 53 downhill, and that seems wasteful to me.

Also, my mileage is about 2 MPG worse since I changed my oil a week ago.
I measure my morning commute every day so I have a lot of data.
The change is not noise; the reduction in mileage is real.

I put in exactly 3.4 qts of Mobile a 0W-20.
When I check the oil in the morning (completely cold) it is hard to tell the oil level because the oil is so thin and clean, but it looks like it almost reaches the upper hole in the dipstick.
Maybe if I let a little oil out my mileage will improve.

62.4!?! Great job!

I just have to say again how much I love the look of the HCH gauges. The integration of the analog speedometer with the other digital readouts is very cool.

You're darn tootin' I'll be watchin'. I'm actually more interested in what mpg's you get for the whole tank, not how far you can go.

It's nice to see that you're at these mpgs early in your hybrid ownership. I've had mine 2 years, and just now I'm getting these figures. Does this mean the gas is different, I've just learned to drive it that efficiently, or dumb luck? I do have to wonder about the gas, because my mpgs jumped about 4-5 mpg over the last 2 months, and I have no idea why....

I think the only difference between mine and yours is 1) I have the manual transmission, and 2) I don't drive at 53 mph with cruise. As a matter of fact, I hardly ever use the cruise control, and when I am on the highway, I'm doing 65. Most of my high mpg tanks are going back and forth to work, which is in high traffic! Figure that one out.

By the way, Kenny, have you been putting the dash mpg's in your tank data? I think you're supposed to use the calculated value.

I don't top off my tank, either. Like you, on the first click, I stop.

Great job Kenny that's simply fantastic.
Good to know I'm not the only one who carries a camera in their car for dashboard photos! :lol:

Keep up the great work!

I fixed my entries in the database to reflect calculated at the pump and not on the dashboard MPGs.

FYI it made it an average of 3.4% lower.

That's still fantastic figures, Kenny. Congratulations, and keep up the good driving!!

I think part of the problem w/ the display versus calculated is the difference in the displays gallon and the gas stations gallon. The gas station's gallon can be 10% (or something close) off as far as acuracy. The display should be calibrated to a proper gallon.
Cheers,
Steven

Steve
That would mean we should be posting the dash board numbers.


Maybe I should go back and fix it again. :(

kenny,
It WOULD, but then those w/o dashes would be skewed in the results...i'm just making a point as to why its different. I think the calcualated should be what is put in.
Cheers,
Steven

But this is a hybrid board
It would be nice to compare apples to apples.
Every hybrid has a dash MPG read out.

Pump numbers have a lot of variability because of pump inaccuracies, temperature during fill up, gas expands when hot, gas tank also expands at hot.
I read here that the federal government only requires the pumps to be accurate within 10%.
That means if the pump says 10 gallons, you may be getting anywhere from 9 to 11 gallons of gas.
I thnk we all know which side they error on.
That would explain why we all get worse MPG at the pump.
We are getting less gas than we think, and it is legal.

If every HCH owner put in dash reading it would be more fair and repeatable.
External variables are eliminated.

People who drive Corollas or other cars can enter pump data.

Next time I'm filling my 5 gal container for the tractor I'll pay more attention to how much the gas pump says at the 5 gal fill mark.

......Assuming my can is marked correctly!

Good idea.
But that will only tell us about one pump and one gas can, at one temperature.
More variables.

Hi Kenny:

___I have been filling 5 gallon cans and various cars for years and never had a pump display anything out of the ordinary other then the first time I placed almost 14 gallons into the Insight. Even that was explained with the 3 gallons heading into the vapor recovery HW. The displays are what are all over the place, not the tanks you fill them from. Over time whether a car has a bladder or not, cold or warm, filling to 1 click or right up to the filler neck, whatever … odometer reading/gallons filled over longer period (like that of the database) is more accurate then the FCD’s.

___Although my LMPG from the FCD is within .1 of my actual calculated LMPG, individual tank over tanks can sometimes be off as much as 2 mpg calculated vs. the FCD. Posting FCD numbers is the wrong thing to do given these inaccuracies. The Toyota Prius is a habitual over achiever in terms of the FCD w/ a usual + 2 to +3 over actual over an entire tank from what I have read. The Buick Le Sabre I rented last month was ~ 2 over actual. The HCH sounds a bit closer sometimes but it is + or – depending on whatever.

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

I don't think you can generalize things to the Prius display being 2-3mpg over manual calculation. I have four tanks on record so far, two of them are over and two of them are under. As for lifetime, my number calculated from the consumption screen is .7mpg above gas volume calculation. Not 2-3mpg.

Given the way the the Prius calculates mpg number on the display, it can't possibly be all over the place. It's perfectly possible that there is a car dependent calibration error, but I think the numbers are stable enough for tank by tank comparison on the same car.

I don't know if you've actually filled a Prius before, but you can't see the gas level down the neck at all as there is a flap over the neck. If you start to force things, you can easily overfill and you won't know it until you pull the nuzzle out then the car will throw up all over you (don't ask how I know... :-). There are supposedly poeple who rock the car side by side to burp it (just like you burp a baby so he/she won't throw up as easily :-) Never tried that myself so I don't know how well it works.

Now if you keep things on the cautious side, you won't know how close you are really. These days, I squeeze in half a liter at a time until a little bit of gas bubbles up the flap and then go back down. But that is not precise as I have been able to do that twice in a row without the car throwing up. So there is a non-trivial slack in the "fill to the neck" method. As for first click, it's highly dependent on pump pressure. It won't work unless you go to the same station and same pump all the time. Most of us don't do that.

So manual calculation is good for lifetime because it doesn't have a calibration error, but the individual tank number is seriously in doubt, on the Prius at least. Thus the display number with correction factor method I propose eariler. But in the end, which ever method is fine I think. This stuff is not serious like the Olympics eh :)

Andy

My '96 Vette had a gauge that was about dead on and I trusted it more than the gas station pump readouts. The 2000 is always high, anywhere from 1 to 3.5mpg, so I don't trust it. Manual calculation is the only way to go one the 2000. I agree that fill up inconsistancies make single tank numbers meaningless but they are accurate in the long run.

BTW gas stations have their tanks undergroud and the temp of the gas is almost alway around 55deg. Do a search on ground coupled heat pumps to learn more.

Hi Accwai:

___I do not know if you have visited Priusonline or not but the stories of the FCD’s being off and/or varying gallons during tank fills are way to numerous to ignore. There are too many posts to dismiss so you are back to odometer miles/tank over tank gallons to get an exacting number. It isn’t the Olympics but you should strive for closer numbers so you know how much temperatures, traffic, techniques, or even setup is affecting mileage. We aren’t speaking of 1/2 the fuel economy unless you drive with your foot to the floor at 90 + one time vs. 50 + and being gentle but tenths of a gallon difference to see what helps and or hinders. If it were the Olympics, we would have our glass beakers out at each and every fill ;)

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

Actually, I did say both the consumption screen is off and the fill is inconsistent. So I'm not disagreeing. My last message was more about the nature of the consumption screen error and whether or not the fill can be made consistent, as you suggested.

For the former, the error on the display is such that it's more or less the same tank over tank, on the same car. That's what I meant by not all over the place. It might not be dead center, but it's at more or less the same place all the time. And apparantly it varies from car to car too. So you can't say all Prius are off by 2-3mpg.

For the latter, I've explained why you can't consistently fill a Prius either to the neck or to the first click. So that doesn't work either.

Now what we know is this: The number on the display is a good relative indicator for tank to tank comparison. The total miles/total gallons number is good if averaged over a lifetime but it can be quite misleading on individual tank. Therefore, my proposal was to computer a correction factor based on the ratio between lifetime mpg calculated from the display number vs. lifetime miles/lifetime gallon filled. Then you apply the correction to each tank mpg from the display. With this correction, you get accurate lifetime mpg as well as meaningful tank mpg. Best of both worlds right?

The only way this would not work is if the screen measurement doesn't remain stable from tank to tank. In the case, you have to explain to me where the error would come from. My understanding is that the Prius uses the number of times the fuel injectors are fired for fuel volume measurement. If that doesn't remain stable, I think the ratio of the combustion mixture delivered to the cylinder would be unstable as well. In that case, I think we have a much bigger problem then inaccurate mpg guage...

Andy

Hi Accwai:

___Here is a question for you. Does your FCD mpg for 4 tanks = what your total miles/total gallons purchased over 4 tanks? The second is accurate. The first is off for many reasons. IIRC, all of the FCD’s use injector pulses for their calculations. Do the calculations have density compensation based on temperature? How about fouling of the injectors themselves? Any pressure compensation as the fuel pump ages? Miles traveled/gallons purchased doesn’t have these kind of issues other then loss of ~ 7/32” radius as the tires wear toward 2/32’s tread depth. If the fueling station pumps you use are within that 1% number previously discussed, the manual calculation is the only true number you can use because it is in fact a real number and not calculated from a number of variables that are themselves varying as temperature or age change them.

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

Like I said earlier, my mpg from the display over 4 tanks is .7mpg over the total miles/total gallons.

As for injector drift, I don't have any solid statistics on that so there might well be long term effects, I don't know. (Do you have any stats?) But what I do know is that with the bladder on a Prius, there is no such thing as a mark that you can try to fill to. My estimate is that the slack is at least 2 liters. 2 liter out of 45 is 4.5%. Personally, that's a much more immediate problem than the injector drifting. Yes, the error will even out over time and you have accurate lifetime number, but the inidividual tank numbers are indeed "all over the place".

I think using this type of tank data to benchmark technique and condition changes is misleading. For example, the calculated mpg for my first tank is actually higher than my second tank. Which is clearly not the case, as I was much more consumption conscious on the second tank. I think the reason is the pump I used for the first fillup was really weak. It probably cut off about 4 liters early and I didn't know how to force it at the time. And of course the second tank took the hit for that.

Anyway, I don't know what more I can say on this topic. It's an imperfect situation all around. I just want to make the best of it...

Andy

Great discussion here, a credit to the knowledge of the posters!
It is about which is a more reliable data, the dash or the pump calcualtion.

Steve.
I quoted you when I thought pumps could be 10% off.
Do you know what your source was?
Thanks

One thing seems clear, the dash *almost* always reads higher.
The gas stations have a financial incentive to adjust their pumps to give us as little gas as legally possible.
Call me a skeptic or cynical but I think that assuming gas stations are ethical, and pumps are accurate, is naive.
99.999% of consumers don't verify pump accuracy, and they know it.
Then again, car makers have an incentive for getting the dash to read a little high.
They look better, and customers are happier.

Very good points brought up about car variables like tire diameter wear and diameter of engine parts.
However I think we are looking for the largest variable.
Considering the consistent direction of the error, and the financial incentive I think pump accuracy deserves our main attention.

We need a source for pump accuracy.
Also information on enforcement, or lack thereof.
Who does it, state, federal, county?
How often?
Are they slacking off from budget cuts?
What do they find?
Is there corruption?
Follow the money! I say.

I'm sure some are thinking, "Man! why take it so seriously?"
Answer: "I don't know."
It's all for fun, actually.

Hi Accwai:

___Let me put it another way. My Insight’s lmpg is dead on. My segment can vary +/- 6 mpg depending on how deep I can fill her vapor recovery HW. If it’s a full fill (~ 13.9 – 14 gallons) the next tank with a short 11.0 gallon fill causes extreme anomalies of course which I have to average so the “Real Hybrid Mileage Database” isn’t skewed nor is my own knowledge of what the Insight is worth form the FCD or not. You know what a 10.8 gallon fill over 1,400 - 1,500 mile tank looks like? Believe me, the mpg over a tank achievement record would look like it was created by Bozo the clown if I placed them in there without averaging the two ;)

___Back to my point, If your Prius is within .7 mpg to the + side at each tank fill over the last 4, you are not having a bladder induced error as it is dead on (over the last 4 tanks anyway). With that, if your FCD is .7 mpg over actual miles traveled/gallons filled, it is .7 mpg over reality period. Your data is to close imho not to believe actual calculated vs. the FCD. Real miles traveled over real gallons filled is the most accurate way to measure because FCD’s are not that close. I still don’t see why you would use the FCD knowing full well it’s already consistently > .7 mpg above actual today? When you keep your spreadsheet over 10’s of tanks, you will more then likely see the FCD not meeting reality more then what you have posted so far. The threads in regards to this in the other forums about the Prius II in particular are to overwhelming to ignore as stated above. In other words, the FCD will induce errors over the short and longer terms that cannot be corrected accurately. The Miles traveled/gallons purchased will not need any correction because they are actuals.

___Kenny, in regards to pump accuracy, there is one item overhanging each and every station owners head. If the scales inspector shows up and you are pumping that 1% off, you will be paying a nice little fine and having someone come to re-rig the pumps at your expense. Do you think BP is going to deal with a law suit from your truly for an over/underage of .3 gallons/20 gallon fill? Not a chance and I know what my various tanks hold and what they don’t. Anyone showing up with a 5 gallon can will know instantly that there is a problem and if they perceive the problem and it’s reproducible, in comes the local news team with an investigation. $2.00 +/gallon gas brings this kind of fraud to the forefront might quick if you know what I mean ;) There is to much attention to this for a station owner to play games with and I suspect most don’t have the expertise to jerry rig the later pumps anyway given the all digital nature.

___Sorry all if I am coming across less then understandable. I had a nice 16 and 8 hour shift in the last 36 hours with a hell of a drive on both sides and in between.

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

Sigh... Could you show me where I said to go with the screen number and ignore the lifetime manual calculation please.

While there is a .7mpg difference between display and manual over 4 tanks, the difference is as high as 4 mpg on individual tank. The latter fluctuation is far too high to provide insight into tank over tank changes. And we have a perfectly good explanation as to why is so high too. To me, claiming it doesn't need correction and not doing anything about it is unacceptable.

I never said to use the display number straight. Instead, the display number should be corrected using a calibration factor. The factor is the ratio between the lifetime numbers from manual vs display. This factor is recalcuated and reapplied to display tank data on every new tank, so things are always up to date. This way, the .7mpg difference is fixed, and the 4mpg fluctuation is fixed as well. It might not be perfect, but I believe this is closer to reality than either straight manual calculation or straight display number.

I've repeated this quite a few times already. If this doesn't sink in, I don't have anything more to say...

Andy

___I never said ignore the lifetime but you are trying to ignore the individual tanks and go with a much less precise FCD X a correction factor which in itself can be off as much as 7% by your own reckoning. 4 mpg off vs. your actual average of 57 mpg as an example.

___I asked you how your last 4 tanks actuals were in comparison to your FCD and you said they were all off by .7 mpg. In further detail you said they are all over the place by 4 mpg.

___This appears similar to mine and everyone else with an FCD from my understanding. I am off at worst 6% of actual in the winter/spring temps and I do know my tank holds ~ 11.0 gallons without a vapor recovery fill? Placing correction factors over many tanks does not rid the actual mpg shown of a single tank outlier. A correction factor on a variable that isn’t consistent leaves you with all kinds of inaccuracies and they are not always correcting to the right side of actual. An example. Let us say you have a + 4 mpg on your next tank which would apply a negative correction factor to the next tank(s) following even though the last tank fill after the first + 4 may be negative. If it is negative, then what do you have? This will skew this last tank even more! A 7% FCD display discrepancy is almost as good as the Insight’s 6% max that I have seen in mid-winter and again, I can tell within a few tenths of a gallon on a non-vapor recovery fill if actual makes sense before I actually do the calculation in my head or by the spreadsheet when I get home.

___I do need sleep but I also know FCD’s are not that accurate over a single tank. It is one thing if it always + 2 to + 3 but it isn’t always + 2 nor is it always even a + on a tank over tank? Well not on my Insight anyway so maybe with the Prius it is different? You will have to tell me if you have received a + .7, + .7, + .7, + .7, and a + 4 vs. actual. With that, it sounds like you need a CF of ( – 1.36 ). With the negative CF, what happens if your next tank is – 2 per the FCD vs. actual? In other words, the FCD shows 55 vs. an actual of 57 but the RMDB will add the negative CF to give an actual 53.64? Eventually the CF will turn out to equal a true number to match the LMPG but it still doesn’t give a single tank outlier any more chance of being correct.

___Another example … Let us imagine it is tank # 100 and you have a CF of ( - 1.3 mpg). If you ever get a lesser FCD vs. actual (it is bound to happen) even after 100 tanks, then that individual tank is as bogus as no CF applied at all?

___I am also not a statistician so don’t take my thoughts on this as proper for the discussion. I just see the Insight receiving great summer time mileage with the FCD being very close to actual. In the Winter, the FCD has a bit more variance from actual but not enough to concern me knowing after another tank, they will average out correctly or darn close anyway.

___Good Luck

___Wayne R. Gerdes
___Hunt Club Farms Landscaping Ltd.
___Waynegerdes@earthlink.net

Sorry I misread. I did say .7mpg was lifetime display vs. lifetime calculation and the four tanks by themselves are all over the place. By the way, the 4mpg is an absolute value. I didn't imply they're all positive. In fact, I did mention earlier that two are positive and two are negative.

<!--QuoteBegin-xcel@Aug 22nd 2004 @ 10:44 PM
[...] I can tell within a few tenths of a gallon on a non-vapor recovery fill if actual makes sense before I actually do the calculation in my head or by the spreadsheet when I get home.
[/quote]

You see that's the difference between us. There is no such thing as consistent fill on a Prius. I would be really lucky if I can get close to half a gallon consistency tank over tank. When you see a difference between the display and the tank, you automatically assume it's the display number that's the problem. My assumption is that the tank can't be filled to a high degree of accuracy and therefore the display is a more stable base for tank to tank comparison. So I apply one correction factor, based on lifetime differerce, to all tanks. The factor is recalculated and all tanks recomputed on every fill. So the corrected display tank consumption and tank distance yeilds a lifetime mpg that is exactly the same as total volume/total distance calculation.

Since we cannot agree on what causes the difference between the display and the tank, I don't see we'll ever come to a closure on this one. I see no point continuing like this.

In any case, I've been using display numbers in the database so far because the display vs tank fill difference is quite small. I will start applying the correction factor on my next fillup. So by that time, lifetime number on JHX will be exactly the same as manual calcuation.

Andy

This is going nowhere.
How about two columns?
One for dash
The other for calculated at the pump.


Then all the information is there and each reader can draw their own conclusions.