Idle second equivalent of starting an engine
 

This stems from a recent thread (https://www.gassavers.org/showthread.php?t=4700) about how much fuel it takes to start an engine compared to idling. I was surprised by the number of opinions expressed there when no one has done what appears to me to be be a simple experiment for anyone with a scangauge:

* Warm engine to normal operating temp and turn engine off.
* Reset scangauge.
* At t=0 seconds start the engine and idle
* At t=10 seconds kill the engine
* At t=20 seconds start the engine and idle
* At t=30 seconds kill the engine
* Repeat for say 10 minutes (5 minutes of actual starting+idling and 5 minutes of engine off, with 30 starts in total).
* Record total fuel consumption from scangauge.
* Start engine and at t=0, reset scangauge.
* Record total fuel consumption of simply idling for 5 minutes.
* The difference between the two measurements, divided by the number of starts, is the additional fuel consumption per start.
* Divide this by the idling fuel consumption per second to determine the number of seconds of idling that is equivalent to starting the engine.

I don't have a scangauge (my car is pre '96). Does a scangauge have fine enough resolution to accurately detect the small difference in the amount of fuel consumed in this test? If so, can someone try this? I think everyone would be curious to see some hard results for different vehicles.

An alternative method would be to probe the injector during a single start with a program such as skewbe wrote to record the injector voltage through the audio input of a laptop. A small modification to the program could simply add up all the pulse widths during starting. This could be compared to the sum of pulse widths for the same duration of pure idling.

Some vehicles may have a separate "cold start injector" in the intake manifold. To ensure that this isn't supplying unmetered fuel during starting perhaps it should be disconnected (shouldn't be operating on a warm engine anyway).

Please don't let this thread turn into an exchange of opinions on whether killing the engine is a good or bad practice. Good or bad, I'm interested to see if anyone has determined the "idle second equivalent of starting" through actual experimentation on their particular vehicle. Thanks.

Great idea.

When I did my cold weather idling warm-up experiment

I saw 0.5 L/hour (0.13 US gal/hour) once the engine had warmed up & temp had stabilized (17 minutes @ ambient: -7 C / 19 F).

The highest resolution of the SG is on the metric setting.

@ 0.5L/hour, the increase in consumption on my car would have to be greater than 20% to be measurable.

Another consideration: depending on the current drawn by the starter, it may take longer than 10 seconds for the alternator to replenish the energy used in the start.

If the repeated 10 second restart interval puts the battery into a deficit situation, the fuel consumption will be artificially raised as the alternator works longer & harder to try to maintain battery charge levels. In other words, the fuel consumption may get steadily worse through the repeated start/stop cycling. Enough to show up on the SG? Don't know.

Interesting. I hadn't considered the battery state of charge and it's effect on fuel consumption. In the cranking test, this would result in an artifically low fuel consumption since effectively, some energy is being supplied by the battery, rather than the fuel. In the idle test to follow, the results would show an artificially high fuel consumption as the engine recharged the battery. I don't know how significant this effect would be but it could easily be accounted for by doing a 5 minute idle test before and after the 10 minute crank test and comparing all results. Any additional fuel consumed in the final idle test compared to the initial idle test can be assumed to be a result of cranking and should be divided by the number of starts in the crank test and added to the fuel consumption per start. Makes sense? This assumes the final 5 minutes of idling is enough to bring the battery back to full charge.

Makes sense to me!

Maybe I'll give it a shot next week. It'll give me a reason to start the car and move it around to stave off the rusty-brake-rotors-from-dis-use problem.

I know this isn't exactly what you are asking for, but I'd like to offer my thoughts:

There are basically three seperate fuel mappings that affect fuel on startup. The engine/coolant temperature map however is the one with the largest swing, and once the engine is up to temperature, the effect of it is effectively nil and the others are statistically irrelevant.

Expect SGII results to be inconclusive. ;)

My SG stops recording as soon as the engine shuts down... Even in hybrid mode... I think it's something to do with the signals the ECU sends to the OBDII port when the engine is killed :/

Does it matter if the SG stops recording with the engine off? As long as it records during cranking it should work. Will it do that?

There's a lag from ignition to recording - at least I notice one (1, maybe 2 seconds)... Given we're focusing on startup, 1-2 seconds during the most critical measurement period is significant (30-60 seconds of missed data over a 10 minute cycle).

At least, that's just me.... From what I've read - other members don't seem to have their SG shut down like mine does when they kill their engine :p

-----
Hrmm.... this has me thinking, and questioning myself... I wonder if going to ignition on before starting would fix that issue. I know if the engine is off and I move the key from off to run - the SG will start recording data and maybe continue after startup without shutting down first? BUT here's the rub, even with the engine completely off - I get a gph reading of .3 (same as my a/c off idle). I'm not 100% sure why it does that (it also throws off trip/tank FE readings), but it makes sense that there would be something there - to prevent division by zero :p

I need the SG Chuck Norris edition - Chuck Norris can divide by zero...

-----
Next time I'm in my car, I'll double check.. Even so, I don't think my car is suitable because of that .3gph - it's not even offset, it's just .3 minimum :thumbdown:

Here's another method:

Gauges?
Both of my cars have FE gauges, the truck is on fleet fuel (which computes mpg for me).

There's the cost of a starter, and the starter's ratings (i..e.: on average, how many times does your engine's starter crank before failing?). Then, divide the figures and there's your cost for starting it, per crank. Yes, because you're not getting 20 years out of that starter anyhow, and it will need to be replaced somewhat more often, so we need the cost of this.

Then if anyone plans on coasting / starting on a regular basis, I wouldn't even question it:
Run synthetic oil.
Without this one assumes extra engine wear, it might not matter to the new car buyer who trades it in every 40,000 miles, but it will matter around the 200,000 mile marker or beyond.
And, synthetic oil is still cheaper than rebuilding the engine.
So now add the cost of this oil on top of regular oil changes, this gives us the cost / mile.

Still need to figure out the cost / start vs. cost / mile correlation, this will vary between city and hwy driving, and per car and per driver, but roughly speaking those are the costs, on top of fuel.

These costs are as important, the starter and either an engine replacement or synthetic oil (I am telling you synth is cheaper lol), we can not assume fuel savings alone will do it, fuel savings have to off set these costs before true savings are realized.

The battery?
Hmmm, I'd just get a 1,000 cca and be done with it, no questions there.

Beyond that all I know is below 20 seconds off time I see no increase in mpg.
Once I start to approach 30 seconds, I do.
In addition this has to be productive off time, such as I'm not just waiting out the extra few seconds to get to 30 even thou the light is green :p

Don't assume ;) Get oil analysis ;) I run dino oil with 10,000+ mile service intervals - lots of starter use etc... No abnormal wear, dilution, etc. ;) Unless Blackstone is consistently telling me untruths :p I do have a engine off criteria though - minimum coolant temperature of 166F.

-------

Has anyone on GS had starter failure that can be remotely attributed to the additional crank cycles from hypermiling? I remember there was a thread about component failures as a result of hypermiling, but I also remember everyone was scratching their heads.