Why does my car consume fuel even when coasting (09 Accord I4 automatic)
 

My car, the 2009 I4 Accord Automatic, consumes fuel from 0.9L/h to 1.4L/h (seems mainly depend on tempereature) no matter if I am parked idling, or coasting at 35mph or coasting at 65mph.

Even if I am coasting, gaining speed on a down slope in gear, it will still consume whatever it consumes when idling.

The car goes to 5th gear at about 35mph, and I can tell that when it is coasting at 5th gear, the RPM stays with speed, which means the engine IS engaged. But it is still consuming fuel!

I thought modern fuel injected engine does not consume fuel when coasting.

This kinda of upsets me because I coast considerably shorter distance in gear, and I can't even rip the benefit of no fuel injected.

Any time the engine is turning its safe to assume the engine is consuming fuel.
The newest engines may cut fuel if you let your foot totally off the gas petal when rolling.

It's always 'running' when the key is 'on'.

The idea behind 'engine on, neutral coasting' is it should consume less gas.

The 2013 BMW 3 series turn off at stops, then restart w/ the press of the gas pedal.

He said AUTOMATIC.

I would never kill the engine with an automatic.

That now goes to the towing instructions. "Engine off, Automatic in neutral" major no-no.

Just the other day, someone said DFCO uses zero fuel, none. It doesn't seem accurate, but I don't know or use it to make it valuable. I would imagine if the engine is on, it is using fuel.

A friend who just got into ecomodding was putting their car in neutral to coast. We discussed why it might be bad, and when we checked with the UG, his vehicle was using >10% more fuel coasting in neutral than drive, but in drive, was still using approximately .50 G/h.

DFCO is triggered by specific conditions n different cars, and doesn't necessarily show on SG or UG.

My 2002 Nissan Sentra had a a primitive DFCO implementation. It had to be >2500 rpm and decelerating in gear, with the TPS at idle position. With the 6 speed transmission, DFCO didn't kick in under 70 mph in he Sentra in top gear. Worse, when it did kick in, it put the ECU in open loop for the next 4 minutes.

My 2006 Scion is a DFCOing fool by comparison. DFCO kicks in with any engine rpm >1500 when in gear with TPS at idle. That means it DFCOs at any speed >35 mph in top gear, and it goes into closed loop immediately after exiting DFCO.

My'88 CRX did just like the Scion described above. My Dakota, though, keeps pouring on the coal under about 3K, which is 70% to redline in the thing, and even then it takes a few seconds for it to do it. I suspect DFCO in the Dakota is for brakes-failed-bringing-the-boat-down-the-mountain situations.

Jayliu - SentraSE-R is right - your gauge (ScanGauge, UltraGauge?) may not be properly configured to show when your engine is in fuel-cut.

Alternatively, you could also try monitoring the Loop status (Open/Closed). You might see that change (from closed to open) when the engine is in fuel cut mode even if the fuel/hour may not.

And I bet if you went on an Accord forum, someone with the factory service manual would be able to detail the exact conditions under which fuel cut occurs. As was said, it's different for different vehicles, and depends on a number of variables.

So, I went over to SubaruForester.org and searched for "DFCO" and the first result was some guy complaining that his wife coasted in neutral to stops, several users agreed that it was a vile and terrible thing to do, and someone calling himself "TheMathGuy" made some pretty good arguments including:

Unsafe? Please show me the results of an independent study showing that the rate of accidents is higher for people driving in neutral. I want data, not speculation about why it might be unsafe. Does it actually cause accidents?

and:

The law:
Yes, there are laws relating to driving in neutral, including in Oregon. However, it's kind of vague.
Firstly, the Oregon law only applies to downgrades. Coasting on flat ground, like gently coasting up to a light because it's red and you're waiting for it to turn green: no problem.
Secondly, the way it's written, technically you're breaking the law every time you shift gears on a downgrade in an MT. Apparently (according to the text), you have to choose your gear before you crest the hill and stay there until you flatten out again, which is ridiculous.

The rest was people trying to refute what he wrote while largely ignoring it and an eventual consensus that arguing with one's wife inevitably resulted in losing the argument. They did talk about DFCO, but nothing useful.

Well, the law is an ass.

"Ere he shall lose an eye for such a trifle... For doing deeds of nature! I'm ashamed. The law is such an ass." (George Chapman in 1654 - Revenge for Honour)

Now ask yourself who makes such laws, and why and how they are made. Also ask whether they are enforceable or not.

“Laws, like sausages, cease to inspire respect in proportion as we know how they are made”

+1 for different cars - different setups - different variables

My choices are as simple as knowing if I am going to have to stop soon or not.

When I back off the throttle, I can feel the differences between being in 4th on my auto trans, or if I slide into neutral. I go much farther when i am cresting a hill at 60 then slip into neutral, than if i keep it in D and take foot off of gas.

The point I saw on the "Hypermiling an iVTEC Honda" thread was basically yes, one can argue that instant Fuel Consumption is higher in neutral at speed then at DFCO.

But that does not account for the ENERGY lost by having the engine being turned by the wheels in DFCO. The energy lost more than makes up for potential idle consumption for x amount of seconds.

I have no UG or SG, but at least in my car, I have learned the spots on my route, and the situations, to either just stay backed off and in gear (approaching traffic or a light) vs putting in in neutral from time to time (cresting a hill, no traffic behind me to stack up, and a nice long roll to the next flat or hill when i am doing about 45 and put it back into D).

If nothing else, try to 'feel' what your car is doing, and it may answer your question.

YMMV

Cheers,

DFCO and auto tranny's just don't work, don't worry about the fuel useage will idling and coasting, anytime with the cobalt and I'm coasting over 30 mph I'm 99+ mpg. since I have a manual I can DFCO on down hills if I down shift and keep my RPM over 2k, so I get engine breaking, really I should have started my coast sooner if I need the engine to slow me down.

Focus on not needing to use the brakes that's where the best gains are.

only about 10% of all offenders go to prison/jail.
Criminal law is only 15% of all law practiced.

your wiki stat is less than meaningless. it is randomly usless. :rolleyes:

Xist, I found your post at subaruforester.org. You and jayliu1984 need to set up your devices (Ultra-gauges or Scangauges) to recognize DFCO. The UG EM v. 1.2 Injector Cutoff setup procedure is on page 50 of the Owners Manual.

MENU  UltraGauge Setup ..  Injector Cutoff

The SGII v. 6.0 fuel cutoff setup procedure is on page 16 of that Owners Manual.

Once you get your UG or SGII set up to recognize DFCO fuel cutoff correctly, you'll start seeing 0 fuel consumption, or 9999 mpg.

If you rely on a mfr OEM fuel consumption display (as I suspect jazyliu1984 may be doing), you're likely SOL.

Why is this a problem? With the Insight, that's how I recognize that it's actually in DFCO (instead of just using little enough fuel that the realtime mpg gauge pegs out at 150 mpg) on a long downhill.

Or is it that it stays in open loop even after you add throttle again?

FYI, the "instant" display in my dad's '02 Camry 2.4 automatic ( http://ecomodder.com/forum/showthrea...2-a-23818.html ) appears to show fuel cutoff.

Exactly. It stayed in open loop for a full four minutes after I added throttle again. Any fuel savings from DFCO were negated by increased consumption during open loop operation.

The Sentra didn't like to get into DFCO, and after it did, it penalized me for getting there.

I don't follow. Both of my vehicles are automatics and have DFCO. The RAM only uses it for decel with a load--it is hard to get it to kick in with an empty truck. It will upshift a gear, relock the converter, and then cutoff fuel flow. I've watched it several times on the MPGuino.
The Avenger does DFCO in OD lock anytime you slow down from 63+mph...to about 55mph. Engine RPMs still happen, accessories still run for both vehicles.

I agree, learning how to NOT use the brakes so much is the best method.

Hello MetroMPG, thanks for the tip! Now all the puzzle pieces are coming together.

I use a Chinese gauage that display 8 stats (2 are fixed to speed and instanteous MPG). I have configered it to display Fuel Status, and I have noticed very strangely that the car will enter "close loop" very quickly after start but it will almost always switch to "open loop - driving" (direct translation from on screen chinese) which baffled me until now.

I was wondering why would it entre "open loop-driving" no matter how light I step on the gas. And it always swtich to it AFTER I released the padel. I thought it was because lag, now I know better.

Next time I will try to switch to neutrul when it is in "open loop - driving", see if it will switch to close loop.

However, I can't seen to able to configure the device to recognize the mode.

For anyone who is interested, I do strongly recommand this product. Here is the english version on dealextreme which cost $100 (still cheaper than UG)

A301 Vehicle Car OBD Multi-Function Trip Computer - Black (3.0" LCD) - Free Shipping - DealExtreme

And since I am Chinese and have relative come visit me last month, I bought the Chinese version (exact same stuff just language) for only $40 equivalent.

+ another one for different cars, different DFCO criteria.

Some vehicles may not use it very much, if at all. With DFCO there is air, admittedly a small amount of it, being blasted through the catalytic converter. That tends to cool it so there may be conditions where the capability of the cat. to reduce emissions is being protected (DFCO inactive), despite the readily apparent criteria being met.

In addition to the TPS and rpm inputs there's usually a MAF or MAP criterion. High airflow or manifold pressure will kick the ECU out of DFCO.

ECU's will do some interesting things while coasting. I can observe a constant MAP signal while coasting, regardless of differences in speed or gear. I suspect the ECU is moving the idle air control to prevent the MAP from dropping while being motored, avoiding unnecessary pumping loss. (Called sail on.)

I should say with my auto car's DFCO doesn't work where I live.

With the Stratus DFCO will only kick in if I downshift out of OD, I get heavy engine braking (should have started a coast sooner), must N coast as it won't coast very far in gear, and in gear it will downshift while coasting to slow even faster.

Impala and Malibu auto's they unlock the tranny when coasting so only way to get engine braking is to slip them into D, very little difference between Neutral coast and OD coast. Only useful if the hill is so big if I want the engine braking to limit maximum speed.

With the Cobalt engine rpm needs to be over 2,000 rpm (55 mph in 5th) so to get it to work on down hills I need to downshift, but the car goes open loop for a few seconds and consumption goes up to .5 gph, before going to 0. But as soon as rpm goes under 2k, back to 0.3 gph.

Area's with longer steeper hills, and some cars the DFCO to much lower RPM's than mine might be able to take more advantage of it.

Back to the OP's vehicle, an 09 Accord. Most hondas (including my personal 96, 04 and 09) will do DFCO in gear above 1200 rpm. That's all that's required. If your gauge isn't showing it as such, you probably need to calibrate it to match the Open/Closed Loop status.

I asked about DFCO on SubaruForester.org and still have not received any responses, but there were mentions of the RPMs needing to be above 1800\at 1900, which is pretty heavy engine braking. They said that fuel gauges did not recognize DFCO and since they do not recognize lean burn, that makes sense.

As long as the engine is moving, it is drawing in air, right, which is all that is communicated via OBD-II, not fuel consumption?

The guy that said that fuel gauges did not show DFCO said to watch the temperature, it will go down, but except for big hills, I cannot imagine this ever happening, and it seems that when users notice it working, it only does a partial job.

ANALOGY -- Why do you BREATHE when you're asleep, you're not working?

Although ones temperature does drop...

My wife's Toyota Matrix automatic DFCOs all the time - any decel above 1500 in any gear. I can tell on the SGII because:
"LHK" (L/100 km) goes to zero.
"LOD" (percent load) goes from 16-18 down to 8-10.
"LPH" (Liters per hour) goes to zero
"LOOP" goes to "OPEN"
Etc.
If I am in the mood for DFCO, I shift down manually while decelerating to keep the revs above 1500.

I usually do not work when I am awake, either.

Smart gauges like the Scangauge can figure out when DFCO is happening by looking at the throttle position (calibrated with the Cutoff setting) and Open Loop status.

Something else to calibrate? I still have not figured out how to calibrate the fuel. Each time that I fill up [and remember] I go to do that and it says "Full tank!"

No no no no no no no no no!

I need to print out those instructions again.

As I mentioned before, possibly in another thread, when I descended into a canyon in-gear, and then downshifted to third, it said that I was using 0.5 Gph, and if it did not enter DFCO then, I do not have any idea when it would, nor what I could do about it, besides installing a kill switch just for the canyon, to use while in drive, and everyone will tell me that is a bad idea.

Sorry, this is an old post, but somehow I missed it, and then happened to see it again recently. Thanks for pointing this out to me! :)