Unusually long engine cranking time during cold weather- 2015

Hey guys!

I've had my 2015 Legacy Touring (45 000 miles/70 000KM) for the last 2 months or so.

Temperatures have been dropping in Canada and I have been noticing an intermittent issue which I'm looking to diagnose before taking it to the dealer.

Whenever temperatures are low (50f or less/10c or less), the engine takes more time to start than is normal. Enough that it sometimes fails to start.

If the engine does fail starting, I have to crank it while giving gas for it to start running.

Of course, it is normal for the engine to be (a bit) harder to start during cold weather. That being said, I'm worried as to what will happen this winter, since temps drop to -22F/-30c around here.

Doesn't seem to be a battery issue since voltages are normal (I have a small OBDII reader) and the start motor sounds fine.

I'm thinking fuel pump. Any other ideas?

Edit: Still under Major components warranty, but I'd rather go to the dealership with a few suggestions, given that it's an intermittent issue

If the engine does fail starting, I have to crank it while giving gas for it to start running.

That's not the correct starting procedure for modern fuel-injected engines in general, or for our Subarus in particular. Normal start is with the throttle closed (i.e. foot off the pedal). Only the case of a flooded start calls for throttle input, in which case full throttle is recommended until the engine fires. "Pumping the pedal" has no effect on a fuel-injected engine.

I'm thinking fuel pump. Any other ideas?

That's not the correct starting procedure for modern fuel-injected engines in general, or for our Subarus in particular. Normal start is with the throttle closed (i.e. foot off the pedal). Only the case of a flooded start calls for throttle input, in which case full throttle is recommended until the engine fires. "Pumping the pedal" has no effect on a fuel-injected engine.

 

A bad fuel pump is unlikely, but low fuel pressure is possible. Other causes could be low battery voltage, poor spark, or a bad engine temperature sensor. Poor-quality fuel or summer-blend fuel could also be a factor. If the problem persists, see your dealer and at least get your complaint on record while you're still covered by the drive train warranty.

It's the original battery. However, I haven't had any other issues with it and voltage readings on it seem normal, at around 12V when the car is off.

Regarding using gas to start the car, I'm fully aware that it isn't normal. I've only done it when the car wouldn't start at all. And I was not pumping the gas, but rather fully pressing the pedal.

I will call my dealer and let them know. My main gripe is that the issue is intermittent. Thus the dealer probably will just end up saying "I can't reproduce the issue, come back again".

I'll call my dealer and get back to you guys!

It's the original battery ... voltage readings on it seem normal, at around 12V when the car is off.

No-load battery voltage (100% state of charge) should be ~12.6 volts. A voltage of ~12.0 volts indicates only a ~25% state of charge.

You might put the battery on a charger overnight to see if it bumps the amperage some. There was a TSB to improve the charge cycle (TSB:11-174-17R), and if your car didn't get it, it may undercharge perpetually.

I would still replace that anemic original battery. IF that doesn't fix it, get other help.

Mine is still OK, never failed, but at the first whiff of trouble, I'm upgrading to over 500 amps.

No-load battery voltage (100% state of charge) should be ~12.6 volts. A voltage of ~12.0 volts indicates only a ~25% state of charge.

Hmmm. I get around 12.6 when the engine is running.

I guess I'll ask the dealer to check the battery specifically.

I would still replace that anemic original battery. IF that doesn't fix it, get other help.

Unfortunately, I wouldn't be able to install a charger, although I do have access to electricity near my car, I live in a shady neighborhood and have had plugged in items stolen before.

If I do end up replacing the battery though, I won't buy one from the dealership. Does anyone have experience with AGM's ? I've read before that they're better suited for modern vehicles.

Especially considering the cold climate I live in.......

It's the original battery. However, I haven't had any other issues with it and voltage readings on it seem normal, at around 12V when the car is off.

Its the CCAs that the battery is putting out that matter than the actual voltage read. Don't know if that can be read via the ODB II; but down here any auto parts store is able to hook up some gadget to the battery terminals and read that value. I'm sure you have similar access at auto part stores up north without having to go to the dealer.

My guess is that the battery is not putting out enough juice to turn the engine.

I'm no expert, but I'm pretty sure you should be at around 14 volts with the engine running, no less than 13.8 volts or so. So in addition to having the battery tested, I'd have the alternator tested too, they'd probably do it anyway while it's there.

I know you say the battery seems fine, but I'm pretty sure that as time goes on, the total CCA capacity drops. So being an OEM battery, it probably had a pretty low CCA to begin with, and and after a few years it has probably dropped even further.

A fully charged battery at rest, should be 12.6 or 12.7V

With the car running, you should be 13.5-14.2V

If you have 12.0V at rest, and only 12.6V when running, you most likely have a charging problem.

1. check, clean & tighten all connections on the battery and alternator

2. with a 4 year old battery, you should replace it at this point.

3. re-check voltage with the car running. If still not charging, you may need to check the alternator.

I get around 12.6 when the engine is running.

Under those conditions the system voltage should be ~14.3 volts, varying somewhat with ambient temperature. There's probably nothing wrong with your alternator ... the "battery management system" as originally designed into the ECU will sometimes command a much lower voltage than normal, which will result in a low battery state of charge under certain driving scenarios (e.g. frequent short trips). TSB 11-176-17 (16 Nov 2017), "Reprogramming to Optimize ECM for Improved Battery Life," was issued to address this exact problem. Your dealer can update the ECU firmware under warranty per the TSB. I'd replace the battery, too, even though it's no longer under warranty.

I found this thread interesting and wanted to test this out for myself. Before I started my car I checked the voltage using the head unit reading and it was at 12.0v. When I started the car the voltage dropped to 11.9v or lower and then the headunit restarted as usual(which is why I say the voltage was 11.9v or lower since the screen turns off during starting). The starting feeling/sound of the car felt as strong as it always has, and I would describe it as typical.

When I brought the voltage page back up it showed the battery maintained between 14.2v-14.3v during idling in park. Once put into drive, the battery voltage dropped back down to 12.0v. I noticed for my whole trip home that the only time the battery is in a charging state is when my instant miles per gallon gauge reads 99.9mpg and occasionally at stops. I say occasionally because it charged at probably 2 of the 5 stop lights I encountered, one at the beginning of my trip and one at the end. The car never was charging the battery during acceleration and never went above 12.3v during a non-charging state. When I pulled into my driveway after the ~15 minute trip my voltage was at 12.1v and when I turned just the radio back on after turning the car off it was back to 12.0v like it had started.

My battery is about 28 months old and I have had the TBS performed to update the ECU for charging.

I guess my question to everyone who says the voltage "is supposed to be x during y" is - are we 100% of this, or does Subaru charge their batteries differently than previous experience may tell us? I could vary well have a dying battery and am not ruling this out, but my experience from actually looking at the data has my car behaving as @plm42 described originally. Or maybe a 15 minute trip is too short to fully charge a battery?

This is something everyone can very easily monitor while driving using the headunit so it would be neat if a few other people repeated this test to compare and see what "normal" is. I will try to check this again tomorrow as the temperatures in the morning are supposed to be in the high 20's (F) where I am.

Once put into drive, the battery voltage dropped back down to 12.0v. I noticed for my whole trip home that the only time the battery is in a charging state is when my instant miles per gallon gauge reads 99.9mpg and occasionally at stops. I say occasionally because it charged at probably 2 of the 5 stop lights I encountered, one at the beginning of my trip and one at the end. The car never was charging the battery during acceleration and never went above 12.3v during a non-charging state. When I pulled into my driveway after the ~15 minute trip my voltage was at 12.1v and when I turned just the radio back on after turning the car off it was back to 12.0v like it had started.

That pretty well matches the extensive testing I did on my 2015 Legacy before installing the ECU firmware update per TSB 11-176-17. The "battery management" system in the ECU seemed to be implementing a simple kind of energy recovery (a.k.a. regenerative braking) during deceleration; system voltage would often drop to near 12.0 volts during acceleration and constant-speed cruising, except when the headlights were in use, and then abruptly jump above 14 volts during deceleration.

After the update the quiescent battery voltage is now ~12.5 volts (typical) and my system voltage with the engine running stays pretty close to 14.3 volts under all driving conditions.

Hey guys!

Just a quick update regarding the issue.

It hasn't been possible for me to go to the dealership, due to long wait times during the winter tire changing season (fairly usual around here for all dealers).

That being said, the issue leaves me quite baffled. I've seen the car have trouble starting up at around 33F/0C, yet start almost instantly in the following days during a -4F/-20C cold snap.

Baffling to say the least. Anyhow, I've an appointment set for the first week of January.

I'll also have the dealer check for any loose suspension components, since I've been having handling issues on the highway (Just like many around seem to have; car wanders on the highway, a lot...)

try this:

turn car on but do not start

turn off

turn car back on wait 3 second and start

report back results

try this:

 

turn car on but do not start

turn off

turn car back on wait 3 second and start

 

report back results

I don't see any difference for now, since the issue is decidedly intermittent... Will try and follow up!

OK, the reason I asked you to do that is to check and see if the fuel pressure is draining off. It should not drain off in a day but if the car sits for an extended period of time it will drain off somewhat. if the fuel pressure is draining off it means one of the following is true

1. bad check valve

2. bad fuel pressure regulator (if so equipped)

3. leaky injector (this would not likely be intermittent)