Hi,

I've been testing different brands of gasoline available in Huntsville AL and found Shell 87 appears to give 11% higher energy. The intermediate report and thread in the Prius forum that led to this conclusion is available here:

https://www.greenhybrid.com/discuss/...6&postcount=76

Bob Wilson

 

I did my only little informal test, I went and found a Shell station and have been filling my hybrid up for the past month on Shell 87. To make the test unbiased I did not clue in my spouse on the idea being floated on this board. She does 99% of the driving on the hybrid so I was curious what would happen with the switch. So far she has not noticed any difference in MPG compared to the slightly less expensive Haffners (non-brand) fuel and Mobil 1 tests.

I don't dismiss the data presented by Bob thorough analysis, and I don't think Bob is suggesting that a 11% higher energy (Shell 87) increase translate directly to a 11% mpg increase (or are you), but I have not seen any measurable increase between the 3 fuel brands that we use.

Anyway, I don't know what to conclude from this. The data shows one thing but I can't match the results.

 

It has always been know that 87 Octane of any brand has more energy than 93 Octane for a given low compression….and vise versa for higher compression. There are so many variables missing that it would render this test absolutely useless. What of waste of time.

 

This is good and I agree that more testing is merited. During my testing I was informally trying to track mileage on the samples but I couldn't see any of merit. But we're also going through the spring weather cycles, cold- nice- hot- nice- hot- cold and I've been doing some other short term tests.

I just filled my tank with Shell 87 and plan to run it dry. We'll see what the mileage comes in at in about two or three weeks.

Bob Wilson

 

bwilson4web — I'm a bit puzzled. In your interesting tests (in the Prius thread "Fuel Matters," post #76) you give the measured energy produced (in watt.sec) per gram of fuel, but say that you have still to measure the densities of the fuels used. You say that the fuel consumption was calculated from the volume of gasoline squirted into the cylinders by the injectors. So, how did you convert this volume into mass (grams) without knowing the density?

Stan

 

Nice work Bob. I'm looking forward to reading your final assessments .

Cheers;

MSantos

 

In my original calibration runs, I had used gallons, a measure of volume, to calculate based upon published report of gasoline density the grams of fuel. Obviously, this was an error.

For other reasons, I needed a more accurate gasoline density measurement and have ordered a gasoline scale, hydrometer. Once it arrives, I'll be able to go back and calibrate everything. But for now, I've only got access to 'published' values for gasoline density. The hydrometer will solve that problem.

Still, the data has value as a relative measurement of specific energy available per unit of different brands. I felt it was better to publish what is known.

Bob Wilson

 

bwilson4web — I understand now. So, your current efficiency graphs are plotted "per gram," but are actually calculated using the same density for all the gasolines tested, as you don't yet have their actual densities. Thus your current interim graphs are really a "per volume" not a "per mass" comparison.

Some suggestions:
(a) Why not use 'joule' instead of 'watt.second'? They are the same, of course, but joules is a more universal unit of energy. Then you can give your results in kilojoules or megajoules per gram (kilogram?) or per liter (gallon?).
(b) I suggest that you graph your results in both "joules per volume" and "joules per mass." The former is useful in comparing gasolines, since they are sold by volume and not by mass. The latter is good since the consensus seems to be that it's the mass of the fuel consumed that relates directly to its energy content, and not the volume. Once you have the densities, and factor them into your calculations, you'll be able to do this. It will be interesting to see whether this reduces the spread of the data, or whether it remains! That's the leading question, of course.

You're doing very valuable and good work here, Bob, and I would like to thank you for your dedication and careful experimentation.

Stan

 

It is a little more complex. To come up with an initial constant to translate "injector time" into a quantity of gasoline, I used a 534 mile tank recorded in October of 2006. Later, I ran some tests using air "mass flow" to further refine the constant. However, the "mass flow" is reported in grams/second, which is where I made my initial mistake. In my haste to convert gallons per injector second to a metric constant, I mixed up my units (grumble, grumble.)

I agree. The Watt.second comes from the way the value is calculated and I hadn't done the trivial task of calling it a Joule. What I'll probably do is include a note pointing out the relationship between "Watt.second" and Joule.

I will try it and see how well it works out. But first, I need those density measurements.

BTW, I did a little 'back of the envelope' calculations the other day:

• Assume the ICE is at 33% efficiency at 2,400 rpm
• 16,000 Joules/gram -> 48 MJ/kilogram
• Wiki reports gasoline specific energy as 46.9 MJ/kilogram
• "Gasoline FAQ" reports a range, 42-44 MJ/kilogram

This suggests my data is fairly close to the range of gasoline specific energy values from published sources. But the final report will 'tighten up' the units and numbers. I'll also include a statement about the loss of water vapor heat.

Bob Wilson

 

There are some smart cookies in here! Thanks for all your hard work on this project. When you come to a conclusion make sure to provide a dumbed-down version for me (in other words, "buy brand x gas instead of brand y.")