Fuel pump high current draw on hot restart

[Porsche Barry]

Hello all,
This is actually on a 2008 535xi but I figured I would post in the E90 section because it might get a little bit more feedback and is also an N54 car.

So this one has me a bit stumped so far.
96k miles, original fuel system to start with. The car will drive for about 20 minutes on the road and then throw a fuel pump malfunction and cut off. Typical story where you wait 15 or so minutes, clear the codes and then you’re able to drive again.
After spending some time diagnosing, I found that the car would say the fuel pump was getting 18.5A at failure when it shut off, and it would read that amperage even with the engine stopped until you waited some time and then it would let you clear the codes and drive again. This part is strange because it shouldn’t measure any draw at all with the engine off—it’s like it was frozen on that last bit of data and as long as it thought it had an 18.5 amp draw, you couldn’t start the car for more than a second. Maybe this has to do with something else and it’s irrelevant, but I thought I should note it.

I saw that when the car would restart, it would usually say the fuel pump was pulling (and I’m rounding to the nearest whole number) 5a and 7v while driving, with fluctuation for throttle inputs. As long as it was pulling those numbers, you could drive the car for an hour if you wanted. No issues. I checked both low pressure and high pressure rail sensor outputs and I was getting a good target of 72psi and high pressure side was somewhere in the range of 700 unless throttle was applied.

Here’s where it gets tricky: if I left the car for 10 minutes or even an hour and came back and started it, I would usually see the pump now running at 10a and 10-10.5v. Every time it showed these numbers, that’s when it would go into shut down after 15 or 20 minutes of driving. Something is making this pump run way too hard and I believed the fuel module was shutting it down.

So the first thing I did was throw in a brand new EKP and code it to the car with an Autel Elite. No change at all. So I moved to low pressure fuel pump, put a new fuel pump in as well as a new filter and sender assembly. I did notice that the right side of the tank was completely full and the left side was basically bone dry. I also noticed there was a ton of crap in the left side of the tank that seemed to have clogged the filter, so I thought I had it nailed and maybe the pump was actually sometimes working too hard from crap in it and overheating. Wrong again! Damn!

I don’t believe it’s the low pressure sensor because it does show relative fuel pressure values based on the pump working at nominal current draws or high current draws. When it’s running at its high levels where the car always shuts down, I am seeing something like 89 psi at the sensor then of course 72 when it’s not in suicide mode.

I don’t believe this car has a tune or anything on it as it certainly doesn’t feel like it has any extra power at all. So now with a brand new low pressure pump assembly and the original EKP installed again, I’m coming to you guys for some expertise.

Any ideas?

[Porsche Barry]

I have read through probably a dozen threads on this issue and none of them actually pertain to this type of specific failure. I also have the car running the latest version of software on all modules.

[marvinstockman]

Just for fun, try running with the gas cap off, or when it stops, open gas cap, and if you hear a hissing sound, tank is under a vacuum. This can cause stalling and high loads on the pump.

[BC535is]

I am curious as to what the "crap" in the fuel tank was. Having one side empty is not normal I suspect.
The fuel pump module was on my mind until you mentioned the stuff in the tank.

Sent from my XT1635-01 using Tapatalk

[Porsche Barry]

So here’s an update, I just started the car after letting it sit overnight, and I’ve got 9a and 9V at idle with 88/89psi at the low pressure sensor with the increased LPFP duty. HP sensor numbers are normal. I’ll attach a pic of the crud in the tank here if it’ll let me. (Will edit this from a computer when I’m home, image is too big).

interesting theory about the vacuum with the tank. Vent valve at the engine? I’ll check that.
I’m getting around 15v on the screen from the OBD readout at all times, kind of doubt that has much to do with anything but it is a thing...

[Porsche Barry]

Here is the crud (had to encode the pic in here because no matter what I do, it says the pic is above the quote size even at 77kb)

[BC535is]

It is kind of difficult to say from the photo as to how much of that dirt is in the tank, but it does make it plausible that the pump screen is getting clogged after 20 minutes if there is enough of it.

Sent from my XT1635-01 using Tapatalk

[Porsche Barry]

Yeah, that’s the theory I had as well. The filter was fairly well clogged but I have replaced all of those parts already and the behavior is exactly the same.

I just don’t know why sometimes it wants to run the fuel pump so hard and other times not at all. Is there something aside from the EKP that could be “moody” in its fuel requests? I can’t say it’s a short because it will pick either a high load or a normal load and not fluctuate whatsoever the entire time the car is running.
Is it maybe possible that the high voltage output of the alternator is playing tricks? I’m seeing 14.8 to 15.1v consistently.

[Porsche Barry]

Going to try registering a new battery and see what it does.

[Porsche Barry]

I realize nobody has an answer for this and I’m not getting anywhere, but I’m going to keep updating this until there is a solution in the event that somebody in the future has this exact same strange problem. This has been incredibly frustrating and the car is going to the dealer as this is no longer worth my time. Final update before kicking it out the door to the dealer is below:

New battery registered, group 49 AGM, and voltage at idle is 14.9 to 15.6 with an average of 15.1-3. Thinking alternator is not behaving and that is the last thing I can come up to check.
I programmed and recoded every module in the car with the latest software and no change. Still starts and runs at normal pump load or it runs at high load, no rhyme or reason. If it’s at high load, I can restart the car and half the time it will back down to normal, usually within 2 restarts it will run normally.

[pimpmyride101]

You have an Autel Elite? That's a nice scanner for these cars. Do a full scan and show us the codes after the fault happens.

[Porsche Barry]

It’s been pretty handy so far, but disappointed in other areas.

codes are 6292 and 6293. Voltage too low and current too high.

[Porsche Barry]

Edit: alternator was not the issue. The owner went ahead and replaced the HPFP, pressure sensor and alternator and it still does it.
WTF

[marvinstockman]

Barry, have you ever run your car without the gas cap, as I suggested earlier?

[BC535is]

I am still going with the theory that crud is clogging the pump screen and then after sitting it falls back into the tank. I know you replaced the pump and screen, but if the tank still has a bunch of particulates in it that would clog the new pump screen. Treating symptoms vs. cause?

Sent from my SM-T580 using Tapatalk

[Porsche Barry]

I’ve been told that the car is fixed.
After the latest round of parts were installed and made no difference, a BMW tech visited the car with some bootleg BMW software and stated that my Autel Elite didn’t register the new battery and that when you replace the low pressure pump, you need to also replace the fuel pump module as well. I’ve never heard that before and am skeptical but hey he’s a bmw tech at a dealership...
I had run a programming and coding sequence on the old module to “refresh” it in a way and flashed to the newest software, but he says that the battery being registered by his bootleg software and [another attempt at] a new fuel module cured it.

so if the new fuel module that I had previously tried made no difference and the battery registration really didn’t work, that means that the Autel Elite is basically an overpriced piece of crap and I need to try something else. Maybe a Texa unit or something... I’m becoming less and less impressed with this scanner if this guys $200 clone software did more than my $3000 unit & J box. Bummer!

on the bright side, the car is running and has a nice list of new parts and my friend can enjoy it once again.