100+ hypermiling / ecodriving tips & tactics

[Christ]

Octane Levels
The most common levels of octane grade are 87 (regular), 89 (mid-grade) and 91-93 (premium). The octane number of the gasolines offered for sale are set by gasoline refiners and marketers based on their perception of the technical and competitive needs in the market. Federal and most state regulations do not regulate minimum octane values, only that the postings on the dispensers accurately reflect the octane number of the gasolines being sold.

Gasoline with a higher heating value (energy content) provides better fuel economy. Traditionally, premium gasoline has had a slightly higher heating value than regular, and, thus, provides slightly better fuel economy, but it is difficult to detect in normal driving. There can be even larger differences in heating value between batches of gasoline from the same refinery, between summer and winter volatility classes, or between brands of gasoline from different refineries because of compositional differences. The differences are small and there is no practical way for the consumer to identify gasoline with a higher-than-average heating value.



FROM:Gasoline -- Is It All The Same? What about Octane?

For all intents and purposes, the heating value (BTU content) of the fuel doesn't change with octane rating.

(The change is so insignificant that it's not obviously traceable.)

[hackish]

Some years ago I had access to a device that measured the MON and the BTU's of heat in a given mass of fuel. We found that the MON and the energy content were strongly linked. Regular 87 octane gasoline had the highest BTU number and Sunoco 94 octane had the lowest. I have no idea what would have happened to the paper on that subject but I do remember the results.

I know fuel in North America is the average of RON and MON but here are a few factoids. First, a lot of premium gasolines use up to 5% ethanol to increase the octane rating. Ethanol definitely has a lower energy content than normal gasoline - look at an E85 engine! Most often the addition of anti-knock agents to the fuel serves to lower it's tendency to ignite and they don't tend to add anything good to the combustion event. Hotrodders will often come up with their own mix injecting water or methanol to increase the effective octane. Again, lowers the number of BTUs of energy in the fuel.

One last thing, a few years ago I was taking a long road trip with my 1992 honda civic. It was properly tuned on both 91 octane and 87 octane. In 2800km of driving I compared mileage on each tune and the 87 octane map won out by nearly 10%. Maybe if I get a new vehicle this year (non-turbo) I will spend some time and re-do the experiment. I'm also curious to see if cylinder balance has a significant impact on emissions and fuel consumption.

On the previous post about the accuracy of these little OBD2 scanners...
The ones that measure injector pulsewidth directly will also need to take into account the battery voltage as the injector opening time is affected by this. This is a tiny difference but over the million or so injections per tank of fuel it can add up to a significant amount. I know this because I did some sub-contracting for a carmaker where they had a fuel economy indicator on the dash - it was approximate at the best of times.

It also takes about 10 seconds to establish an OBD2 connection with an ECU so if you end up with your engine shutoff trick then you may actually be missing the startup and cranking enrichments that the ECU adds while starting the engine. On a cold engine they can be pretty significant - 200% the idle pulsewidth is pretty common. Many cars will also retard the ignition and add extra fuel for the first 15-20 seconds to improve the cat light-off for the cold start emissions some countries mandate.

On one vehicle I know of the reported fuel injector pulsewidth is simply wrong. I verified this with some pretty high tech instruments and I've also spent more than a year disassembling the factory code from inside the ECU.

For me, I usually track kilometres travelled with fuel pumped to fill the tank.

Sorry for making this so long. Didn't have time to shorten it.

[hackish]

Ok, did some further reading on the subject. Seems that my testing may be out of date. The advice to use the lowest approved octane for your vehicle to get the best fuel economy is still good but speaking to a friend who is a chemist for a big gas company he said that the BTU thing is no longer accurate as they've developed far better anti-knock agents in the last decade.

[Christ]

Originally Posted by hackish

On the previous post about the accuracy of these little OBD2 scanners...
The ones that measure injector pulsewidth directly will also need to take into account the battery voltage as the injector opening time is affected by this. This is a tiny difference but over the million or so injections per tank of fuel it can add up to a significant amount. I know this because I did some sub-contracting for a carmaker where they had a fuel economy indicator on the dash - it was approximate at the best of times.

It also takes about 10 seconds to establish an OBD2 connection with an ECU so if you end up with your engine shutoff trick then you may actually be missing the startup and cranking enrichments that the ECU adds while starting the engine. On a cold engine they can be pretty significant - 200% the idle pulsewidth is pretty common. Many cars will also retard the ignition and add extra fuel for the first 15-20 seconds to improve the cat light-off for the cold start emissions some countries mandate.

Most people doing EOC aren't using the key, they're using a kill switch (I hope it's most, I guess I can't really prove it one way or the other, though). That means they're not killing the ECM, thus not killing the ECM connection.

Information from sensors also doesn't need the ECM to actually be active to be obtained. The Sensors work even when the ECM isn't available, which means that things like the MPGuino (which can take inputs directly from sensors) are getting accurate readings from sensors whether the car is on or not.

Regarding whether or not the methods are "tricking" the gauge into displaying false numbers, that's obviously untrue as well. If all we were doing was "tricking" the gauge into showing a higher number than what we're actually able to achieve, it would show up when we (as many of us do) do the math at each fill up. That's why we 'calibrate' the units upon installation, in some cases.

There's also been an argument posted somewhere that all we're doing is adjusting the instant MPG number. This makes me laugh. If the instant MPG number goes up, it counts into the average MPG number. The other gauge we're looking at, ya know?

Tell ya what - Raise your Instant Speed a few MPH and tell the cop that it doesn't count because your average MPH still is lower than the speed limit.

[hackish]

If you are keeping the key on the whole time then you're obviously not losing the connection. This still does not account for the fact that not every auto manufacturer reports the injectorpw correctly. I also know that the estimated fuel economy is approximate because I worked on the math used for several OE's. They gave an acceptable range in their spec and it was 15-20%. We tested it and just scaled the values slightly so if you did the math manually you'd be in the same ballpark. This means if you're trying some special technique and showing a difference it may or may not be real - it may just be tricking the simplistic model. This is why my reliance is more on the old school method of measuring the litres used.

[PaleMelanesian]

Regarding the starter:
- I just had to replace the starter in our van, that my wife drives and does NOT hypermile. 60,000 miles to failure.
- The starter in my civic is just fine, with 177,000 miles on it. I hypermile it heavily.
- Also, the clutch is just fine, at 177,000 miles.
- Basjoos had his clutch last nearly 500,000 miles before he replaced the whole engine.


Whether I'm tricking the gauge or not, the results at the pump show that something seems to be working. In the last year and a half, the total cumulative error between the gauge and pump is only 1/2 gallon.

[Christ]

Originally Posted by hackish

If you are keeping the key on the whole time then you're obviously not losing the connection. This still does not account for the fact that not every auto manufacturer reports the injectorpw correctly. I also know that the estimated fuel economy is approximate because I worked on the math used for several OE's. They gave an acceptable range in their spec and it was 15-20%. We tested it and just scaled the values slightly so if you did the math manually you'd be in the same ballpark. This means if you're trying some special technique and showing a difference it may or may not be real - it may just be tricking the simplistic model. This is why my reliance is more on the old school method of measuring the litres used.

Calibration and averages account for the inaccuracy of the system, like I said earlier.

The gauges are calibrated based on inputs from the vehicle, specifically. After a few tanks are calculated, the output from the gauge can be calibrated to show the same approximate value as the "old math" shows time after time.

Even an abacus is incorrect until you actually do the work to solve the problem.

[MARTINSR]

Originally Posted by MetroMPG

It's probably true to say that jackrabbit starts are wasteful, but some people might take that to mean that glacial acceleration is the most efficient. Which also isn't true for all vehicles/situations.

Sorry if this gets answered somewhere in this thread as I am reading from the start over the last day or so in my spare time and just had to jump in to talk about this particular issue.

I am driving a 95 Ford Taurus SHO with an automatic and have wondered about this, I leave about every stop with a "glacial acceleration" and wonder if it is the best. I take into account when I could p&g a little, I am talking about all the other times.

Is it better or not?

Brian

[MARTINSR]

Originally Posted by cloudjunkie

No, but how about the temptation for someone driving through a safety corridor in the middle of the day to keep their lights off thinking they are somehow saving a bunch of gas? They're not saving much fuel, that's my point.

I certainly have a stake in arguing against keeping your headlights off in such situations: I drive in them regularly and would like vehicles to be reasonably visible.

I may leave my headlamps on, they are automatic and I like it. But I WILL be turning my driving lamps off as I have them on as well during daylight.

Brian

[hackish]

There was a study published a few years ago that was comparing the number of head-on collisions with DRL equipped vehicles versus non-DRL equipped vehicles. As I recall the DRLs reduced your chance of being involved in a daytime head-on collision by about 25%.

HIDs may reduce power consumption as you're looking at about 20-30w versus 100-110w of power.

-Michael