Effect of fuel octane number on FE
 

I was using 93 octane fuel for quite some time, and I was getting about 45 mpg from my 1.5L 5MT City. (This was before jackrabbit pulses adoption).

Then I moved to 87 octane and adopted jackrabbit pulses. FE improved. But now I am experiencing that after the first jackpot tank netting 56.66 mpg, my FE has come down from the high tank, and is ranging ~52 for the next two tanks (the second is only half way through, but I can sense the trend from my fuel consumption trend).

Question is, was the jump in FE because my ECM was used to burning fuel in a certain way, and I gave it a shock by moving over to lower octane fuel? The ECM then adopted to the new fuel, and therefore the transient effect of the first tank is slowly settling to the new steady state parameters for the new fuel?

Alternately, was the cocktail of 93 octane fuel (~2 gal remaining in the tank before the 87 fill-up) and the 87 octane fuel resulted in better FE than any one of them individually?

Are temperatures a factor? the best tank came in Feb-Mar 09, where temps were kind of moderate ~60 to 70 deg F?

I am following the same routine through all these tanks, and am pretty consistent in the way I use the car, so that should not be a factor.

Yes, temps are a factor, because engine intake temps will increase in warmer months. This means higher chance for detonation.

Assuming you have a 12 gallon tank, your fuel's octane rating on your change-over tank would have been approx:

2 x 93 = 186
10 x 87 = 870

870+186 = 1056
1056/12 = 88

So you'd have had 12 gallons of 88 octane fuel, approximately. Simple averaging is all you need to determine a mix of octane ratings.

Yes, it is possible that the ECU had to experience a re-learn with the new fuel. Being OBD-II compliant kind of makes it difficult to speculate without actually knowing how the ECU compensates for potential knock (detonation).

Thanks for the reply, Christ! I have two points here -

1. With increase in temp I saw a decrease in FE. Should it not be the other way round?

2. The ECU mistook my ~88 octane fuel to be 93, and gave me waay higher FE. What did it do differently with that assumption that it adjusted for the later tanks? (If I know, I can mimic it, so I get back the FE) :D

1. Again, this deals with more things than just intake air temperature. Specific density of the air, dynamic compression ratio in the engine, etc - all affect detonation characteristics. Without more information, I can't say for sure, or even in thought, that higher intake temp would actually increase your efficiency.

2. The ECU may have kept telling the engine to run at higher timing specs, in a more efficient range than the ~88 octane fuel could normally be run at according to the ECU's programming.

This is about as much as I can tell you specifically... hopefully, someone who knows more about your specific vehicle can chime in.

Everything else being equal, on normally expects better FE in warmer summer temperatures. Since you switched from 93 octane to 87 octane (the first tank being a mix of ~88 octane??? the second tank being almost exactly 87 octane), everything else isn't equal. Looking at you fuel logs, I also see that you had it in for an oil change on the last tank. Any chance the guys servicing the vehicle bumped the tire pressure down from 50 PSI??? (This the very first thing to look at.) I also see that you've used the AC more often.

It's a misunderstanding to think the ECU mistook ~88 octane for 93 octane. Octane is basically a measure of how subject the air/fuel mixture is to pre-ingnition. As the piston compresses the air/fuel mix, it's temperature rises. If the temperature and pressure are high enough, the mixture spontaneously ignites instead of being ignited by the spark plug. The lower the octane number the lower the temperature and pressure at which the mixture pre-ignites. When the combustion of the mixture is induced by a carefully time spark, a wave of flame travels though the mixture from the spark plug tip to the top of the piston and cylinder walls. When the mixture pre-ignites, it happens more or less everywhere in the mixture at once. The burn is a lot quicker and resembles an explosion. Pre-ignition can quickly damage an engine. If the ECU detects that pre-ignition is (occasionally) occurring, it adjusts the spark timing to prevent pre-ignition occurring.

Anyway, if the tire pressures are still up and you really didn't use AC enough to account for the FE hit, then octane may be the culprit. To test this theory, at your next fill start with ~3 gallons of 93 octane then top off the tank with 87 octane.

octane/mpg
 

I just read your post,and input from Christ and I believe the timing may be the issue.All grades of fuel have the same energy content,and since the higher octane allows more ignition advance without detonation,and mpg is best at the highest timing advance,then the lower octane fuel may have triggered the vibration-sensor ( knock-sensor ) to retard the timing a bit to thwart detonation,and in so doing,move the engine into a less efficient timing regime.

The warm air is causing your engine to be more knock prevalent.

The colder air was allowing your engine to run on lower octane at the same spark location(roughly) without incurring knock(detonation in the cylinder as the gases expand and heat after the spark begins traveling towards the opposite side of the cylinder). In that sense the 88 octane would have helped as well as its slightly more capable of being at the higher compression or timing than 87.

Trying to replicate those results on 87 fuel is inadviseable. What your engine is doing is relieving stress in the engine by doing any number of things to reduce pressure in the cylinder at the cost of FE.

If you trick your engine into not doing that it will knock/detonate/ping more often and then you are probably shortening your engine's lifetime while also putting it within range of death on a particularly bad day. For instance lets say you are in a hurry for once and are accelerating kind of hard and your engine is experience knock because you tricked the sensors into not opening valves slightly early or retarding ignition. In this situation its possible for a component in your cylinder to overheat and create a glow plug. Then you are just a bad combination from pre-ignition and likewise a new cylinder and piston.

So, should I go back to higher octane fuel now that the air will stay warm for ~6 months...?

I mean I can then generalize this to a recommendation that one should use higher octane fuel in summer and lower octane cheaper fuel can work equally well in winter, maybe sacrificing a tank's FE in transition each way.

Wot u say?

It depends. It might be worth a test run on both. I gather you started hypermiling when you switched? if thats true you ought to get a baseline on both and then tabulate the extra cost of higher octane against the lower octane FE and cost.

I'm going to gamble FE losses are as someone else pointed out that it could be the tires or the AC as well. I highly doubt the higher octane is going to yield a greater savings for you. then. . .if you really want to be ecologically friendly lately premium gas is not selling as high in demand as regular and therefore as long as there is excess premium being created and not consumed thats not ecologically friendly.

That said I'm not going to pay for premium just because exotics and high compressioners stopped buying gas. Stick with low octane get a diagnostic unit for your car, check tires and cut down on AC as much as possible(lowering windows slightly is better than AC if it still keeps you comfortable(all the way down is still more efficient(on most cars) than the AC on full blast)). If that diagnostic unit tells you its encountering knock all over the place I'd switch though.

Is the car supposed to run premium? I kinda of doubt it. I am pretty sure its a not too distant cousin of our FiT, and they don't require premmy fuel. Yeah stateside Fits require unleaded as the only requirement(lowest grade of fuel available here)

Answering a few open questions -
1. Yes Oil changed with the service. As synthetic was 2.5 times as costly as unbranded mineral, I chose mineral.
2. Yes they got the tyre pressure down to 32 (habit slaves!) I insisted and made them take it back to the compressor and filled it back up to 50. Lots of weird glances endured... I check and top up regularly at each weekend using a digital gauge and battery operated pump.
3. AC use was not significant. It was only two / three instances, each not lasting 5 minutes before I got nervous about the FE hit and switched to window cracks. :D
4. I P&G regularly since Dec 08, (when I got to know about it on joining ecomodder). The change I had with the octane change was switching from granny mode to jackrabbit mode for my acceleration, on reading about the ABA testing done by SVOboy.
5. My car is almost identical to FIT (same platform), has 1.5L 5MT IDSI engine (not the popular Vtec variety), and does not require premium fuel. the premium fuel was touted to make the vehicle run smoother, so was using it.

Unchosen: What is the diagnostic unit you are talking about?

aerohead: Is the knocking thing equivalent to lugging?

Diagnostic units are small calculator sized computers that read the signal the ECU receives from the various sensors in your car and then puts it in a code for you.

Basically its a box that you plug into a receptacle in your car(mine is on the passenger side behind/below the glove box) that tells you what a check engine light means. They usually come with a book that tells you what specific codes mean. Its what the dealer will do everytime you bring a car in because its easy.

Some vehicles have a memory. So it records important events. not like "I turned 12 weeks old today!" but "detonation in cylinder 4" and then also records all the temperature readings(intake, exhaust cylinder rpm all kidns of other stuff too).

It basically alerts you if something is wrong before it needs fixing. Typically I am going to say its not necessary. I have one because there are 3 Del Sols in my nuclear family along with 2 other cars that can use the same diagnostic and codes so its totally worth it for us.

Knock/detonation/ping occurs when the cylinder gets a little too warm or pressure gets a little too high.

The air gets compressed and the spark plug ignites the fuel air. As the spark front travels across the cylinder(from the crown/top) to the piston(bottom) the air at the bottom becomes extremely hot and suffers extremely high pressure. At this point that small pocket of air at the bottom detonates(normal engine cycle just involves the fuel burning not exploding) and causes abnormal pressure and stress against the cylinder walls and piston head/crown/top. Its not really dangerous in non-high performance cars(Ferraris Lamborghinis Honda NSX. . .) as long as its doesn't happen every stroke or every other stroke. If it happens every stroke the spark plugs, valve heads or the walls of the cylinder themselves can cause the fuel to ignite at the bottom of the compression cycle. At that point the engine is toast. Its going to have to compress super-heated high pressure air for 180 degrees of movement(opposed to just a few normally) during which time the heat and pressure are increasing rapidly. In effect you will know when that happens because the car stops immediately. it melts a hole in the piston, maybe the walls, maybe the crown and definitely cooks the valves.

That said your car is smart enough to avoid knocking becoming too intense as long as you don't short-circuit the knock wiring in hopes of higher FE. So don't worry about it unless you are really trying to push the envelope and get lean burn and high compression with the sensors unplugged.

I am very much interested in pushing the envelope on my FE achievements - my current short-to-medium term goal is 60 MPG and eventually, in long term, 70 MPG.

BUT I sincerely don't want to be stupid enough to allow all sorts of damage to happen to the engine in the name of FE. That's why this enquiry.

Can I get some links for the diagnostic unit you are talking about? Any brands/models you can point me to?

AMPERIT OBD-II Pocket Scanner : JC Whitney: Auto Parts & Accessories

I believe is the unit I have. I'm not sure if the fit has a memory unit attached to its ECU, but all you have to do is connect it before shutting it down and it will tell you what has been going on while you were driving.

As I said its mostly to diagnose problems. I sincerely doubt you are causing enough knock to worry about and also if you are going to be exploring the lean burn regions you will need to re-map your Fuel-air ratios. If and when you do that you will probably have an on-board computer that runs a program to change the fuel maps, but then you can also get programs that read the diagnostic codes. If you straight-up hardwire the map(set it on a computer connect an ECU tuner to the computer and then unplug and connect the tuner to the engine) then you will want one of the above tools to make sure you aren't damaging something.

Honestly I would wait until you have decided you are going to go lean-burn settings and are getting the equipment to do that. Thats where that scary dangerous pre-ignition lair is. As long as you aren't tuning the ECU to inject alot less fuel than normal its not likely to pre-ignite.

Will a scangauge work as well? The JC Whitney device looks like a younger sibling of the scangauge, just giving out codes and not MPG feedback. :(

However, you are right about what I might be up to, I don't intend to go all that nerdy about messing with the engine.

Any pointers then to improve from where I am to where I want to be?

you could try some aero improvements although the fit/Jazz/city is pretty good initially. I know the US version did not come with a belly pan on the earlier models, you might take a look at covering the under side of the car with some inexpensive semi-heat resistant material. Wheel skirts for the back wheels using a similar material (like coroplast, linoleum something relatively light and easy to manipulate(and cheap)) improve aerodynamics.

If your version(I am less familiar with the overseas models) has roof racks or a wing/spoiler you could cinsider deleting those(although be wary because open bolt holes for either could cause rust so be sure and cover those). You could also delete interior components you don't use.

Another thing you could do, but depends on your driving hours, is to delete the alternator and replace it with maybe a good 70-80 pounds of batteries and a few "marine application" solar panels. The panels can go on the interior of your car on the dash(Honda likes to make very dark large dashes that generate heat mine is the same(del sol)). The panel would reduce the interior temperature somewhat heavily if sun is striking the panel instead of the dash. Also the panel would allow your batteries to charge when the car is sitting at work or school so you wouldn't have to pay legacy costs on energy.

knocking/lugging
 

hummingbird,knock is synonymous with detonation,a condition which occurs in the cylinder of the engine associated with piston slap caused by uncontrolled combustion which can lead to rapid engine damage.------------Lugging is described as a condition which exists when operating a vehicle in a gear too "high" for the rpm of the engine,which causes the vehicle to "buck" and shudder,immediately before the engine stalls.Lugging causes "jerk",which is an instantaneous acceleration in the power-train,and if maintained,can lead to bearing failure or gear failure under these instantaneous loads.

I would argue with that. Its not really uncontrolled any more than normal. It is suspected that it happens if/when the flame front strikes a pocket of stoich mix while the surrounding area is lean. the stoich burns faster and causes an uneven pressure spike on the bottom of the cylinder(against the piston) which ignites that fuel and creates very small pockets that detonate under the intense pressure and heat.

All said and done the combustion event is no less controlled(as there are no means for controlling the flame front).

Under ideal conditions, the flame front advances from the spark plug. What is important, though it that it IS less controlled in that the timing of the ignition is before the intended ignition. This causes a pressure wave that has to be absorbed by the piston, cylinder walls, head, valves, and plug. That affects power and durability. If unchecked, it can raise the heat level to a point where you can get pre-ignition and that's when you blow holes in pistons, bend rods, or throw them through blocks (which can be quite dramatic and entertaining if it's not your car).

When the knock sensor detects knock (basically, the knock sensor is a microphone that picks up the little pings through the metal block), it tells the ECU, which switches to a more conservative engine map. This would be retarded ignition timing (cools things down inside the cylinder, but is lousy for power or fuel efficiency) and possibly increased fuel, which cools the engine even further. I think some really paranoid ECUs might even lower the rpm limiter, but that's just what I heard and I don't know for sure. Not that I doubt it, but it could have just been the retarded timing and rich fueling naturally limiting the maximum rpms. Anyway, if the knock sensor stopped working in my old Subaru, it was supposed to go into a 'limp home' mode which did the ignition cut and the fuel enrichment (I heard of people getting 12 mpg from 2.2 liters like this!). Unfortunately for me, the early Legacys had knock sensors not up to the rest of the car's durability standards and my knock sensor went, but the ECU never noticed and I ended up yanking a real ringer of an engine (the N/A in that car was as quick as some stock turbo versions up to around 70) with a relatively young 250k miles because of it when I lost a cylinder (still maintained 85 mph). Even more unfortunate was that I had a turbo engine I was rebuilding in my kitchen and so I just swapped it out the next day (running the lower compression block normally aspirated. Went down to 23 mpg from about 26 and lost any breathing above 5k rpm) and all the ended up being wrong with the 'dead' motor was a melted $12 exhaust valve! I was pretty ticked at that point (especially when I saw how beautiful, even almost new, the rest of the internals looked), but committed to the other engine. So anyway, the point is, don't try to save a couple dollars by risking the engine. I'd play it safe and go with whatever the minimum octane is that works until you can monitor what's going on. Really, you shouldn't notice a difference, but if it feels slower or is getting worse mileage, then you're too low. If the recommended octane seems too low, you might have a problem. Could just be something other than knock is setting off the knock sensor (bad sensor, odd vibrations from something) or the plugs might be the wrong heat range or bad.

Not correct. The spark timing and ignition event take place as normal. literally the only detectable difference between a knock combustion event and a normal one is the pressure wave is non-homogenous and in no way whatsoever could be considered quasi-equilibrium(properties are different for different parts of the volume by substantial measures).

As I originally said what happens is the fuel and air for some reason don't mix properly. Theorized reasons are the intake air may have pockets that are warmer and disperse the fuel more rapidly increasing the air's saturation for that pocket. Valve opens it all gets sucked in. Now you have a pocket of rich fuel particles undispersed and a bubble of ignition ready 14.7 AFR. Go spark. Flame front travels normally across the sliver or dz. The flame front then encounters several things at the same time. It encounters the pocket of 14.7 that is hot and almost to ignition due to pressure, it encounters a pocket with too little fuel and is not quite ready to ignite(it needs a touch more pressure and heat) and a pocket that has fuel droplets(not in the vapor state). Flame front ignites the stoich mix, pressure increases on the lean section causing it to ignite and the flame front expands rapidly through both. Meanwhile the pressure climbs dramatically along with the temperature, but the increased pressure "holds" the droplets(or at least some small part of the droplets) from becoming a vapor. At that point the pressure is akin to that of diesels and the droplets literally explode, not ignition, but a pulse of pressure that rattles the block. This is how Isobaric bombs work. Fuel-air bombs are more dangerous against open air unarmored targets but in a closed system(caves or engine cylinders) causing the same fuel to detonate causes a much more devastating effect.

Any time ignition happens before your engine normally wants it to is pre-ignition. period. Even if it happens 1-2 degrees in advance of normal your engine will not survive more than a few strokes of it. If it occurs any more than about halfway through compression its toast and is likely to put your other cylinders in jeopardy from the heat soak of the molten slag.

I was only incorrect in my horribly unclear wording. While the ignition event during detonation may take place at the correct time of the piston cycle, the direction of the flame front(s) is still uncontrolled in that it does not ignite in the intended order. The ignition event is supposed to be a controlled burn from the spark plug out. Any part of the mix that ignites prior to the intended flame front reaching it would be sooner than anticipated, but not the same thing as pre-ignition, where an ignition source other than the spark plug initiates the event before intended ignition is to take place. I didn't mean to imply that they were the same. Hopefully, that's a little more clear.

In that case we totally agree

argue
 

theunchosen,I'm relying on memory going back to the 1970s,with my I.C.Engine Lab class.While "ping" was acceptable under load,"knock" was considered dangerous to the engine and is the reason for anti-knock additives in gasoline,as tetra-ethyl lead,Methyl Tertiary Butyl Ether,and Ethanol.Engine manufacturers will recommend an octane rating for a particular engine to protect it from the knock caused by detonation which is created by a non-uniform flame front propagating from the spark source.The octane helps guarantee "controlled" combustion at maximum spark advance,allowing the engine to achieve peak efficiency for its compression ratio.

Ping and knock are the same thing. Ping is one of its forms because it can be defined partially as a certain frequency that occurs in the engine. In that term it refers to a very short area with a very high output, the term also fits things like sonic-resonance where you ping the ground with a high frequency and listen for the ripple waves to come back. Lately the term more than most anything means a networking event in which a computer sends a packet that immediately returns after arriving to determine how much disconnect is between the two computers or however many machines or whatever array of machines.

Knock denotes what who was it . . . you also referred to as piston-slap for its tendencies to occur traveling out from the spark plug at the fringes of the system but almost always against the piston head itself.

knock/ping/piston-slap whatever is detonation. The knocking or the pinging don't cause the detonation. The jerk to the piston-driveshaft-trans-wheels-diff-whatever else comes because of that pulse emitted from the "instantaneous" conversion of fuel-air to CO2+heat+water+small amounts of other stuff.

All knock is detonation, but logic does not mandate that all detonation is knock. Such is the case here.

Fuel manufacturers are avoiding the fuel detonating during compression. Those compounds increase the amount of energy it takes to ignite the fuel, normally in open air this is bad. This is not open air though and the fuel-air is almost achieving its activation energy during the compression stroke so it just takes a little zap to make the ignition flame kernel start and expand. The activation energy would be a waste to burn it in open air because the fuel is then having to overcome that additional AE for no reason, but since its going to be compressed and we are going to add that enthalpy anyway, it doesn't hurt the cycle.

I believe the op is driving a Honda Jazz/Fit. Since the fit has anti-knocking ability its moot, but its not a very high power-weight engine so its not going to hurt his engine very much at all for ping to occur unless he runs it for say an hour under knocking. Then whats very likely to happen is something inside that engine is likely to outpace the cooling systems and pre-ignite destroying his engine. If that happens he will know because the engine's check light will come on and it won't accelerate properly at all and it will rev low at idle.

Piston slap is not always detonation. There are several cases of it happening on production vehicles which did not involve detonation at all. The culprit, in most cases, was either worn piston rings or worn wrist pins/rods. Less often, it was an out of round condition in the cylinder's bore.

duly-noted and thanks for the. . .edification. I was pretty unfamiliar with the term and associated it pretty strictly with knock.

Google "Ford Piston slap" and "Chrysler piston slap" and "Chevy Piston slap" etc... you'll see it numerous times in off-the-lot vehicles.

same thing
 

theunchosen,I dug out my old text and it basically concurs with everything you've posted.There were over a dozen pages which dealt with "autoignition" which the text suggests is present in all "real" engines,from sub-audible high-frequencies,to the destructive "severe" detonation at the extreme end of the spectrum,where impulses from supersonic pressure pulses create stresses on the engine structure,enough to cause failure.---------- My thought,with respect to hummingbird's situation,was that if all other variables were held constant,that the lower octane fuel would prevent maximum timing advance,forcing the engine to operate at a lower efficiency,leading to the lower mpg.

I can agree with that. I don't think I ever disagreed with you on that. I think I just got over excited and wanted to tweak something just a little.

I'm not certain about the italics part. actually I agree. But with respect to overall savings I think the engine is definitely smart enough to avoid damaging itself and just run at lower hp output and get slightly less mpg, but in the long run cost less(premium here is I think 30 odd cents more than regular). Its hard to estimate this though without several tanks using premmy and several tanks using regular with maybe 3 tanks of one and then 3 of the other in succession to make sure the premmy fluid clears the 2-3 tank to get good results for the test.

My inclination is always to go with the cheapest possible fuel if it doesn't hurt the engine(or radically reduce FE), but I don't have all the data or maybe even enough data to get a feel for that(seasonal changes in op case).

Phew! Guys, that was some colorful analysis!

I am somewhat OK using premium fuel for the sake of FE as the difference in prices is actually closer to ~40c. The questions are, will full premium boost FE or simply increase chances of engine foul up? Will a blend be better? What ratio?

My ambient temp is on the rise gradually and would be anything between 90 to 120 deg. now for about 4 months. Having to haul kids and stuff in the heat of the summer for leisure activities is not helping either, and probably is going to pull me below 50 mpg for this tank. But one does not lose faith or enthusiasm so soon, after getting such spirited support! Keep it up guys!

Humming, I'm going to have to defer to someone else. My jury is out because I don't own a Jazz/Fit so it could be more advantageous for FE to upgrade or it might not. Its more a car specific thing than a general rule. Sorry it will increase FE of the engine but it might be more costly. You might get more MPG but it might cost more.

The only thing I can recommend is try a test of 2-3 tanks of each and compare keeping in mind any changes that you make either to the car or the driving.

boost or foul
 

hummingbird,my thought now,is that there are so many variables in your driving environment,that even if the fuel could alter your cars performance,it would be lost in the scatter of all these other variables.Seems like it would almost require a laboratory environment to sort it all out.I don't think premium can foul anything but your purse,if the car will operate satisfactorily on a less expensive grade of fuel.

Aero, thanks for the assessment. my earlier two tanks had completely identical driving environment and the only two factors differing were the temperature and the fuel grade. That prompted me to ask.

However, I will build some data up by switching between 87 and 93 using 3 identical driving environment tanks each, so I get an idea.

Thanks guys for all the help!

carbon deposits