AWR-fix: e53 Fuel pump / siphon pump details

[xbimma]

@andrew: I agree with your theory. I have always noticed how the needle does not move for a while after topoff. I bet Rental car companies save lots of money because of that.

[andrewwynn]

The rental car company loses out also: I have always filled until the needle matches eg 1/1 = 92% full. Poor "next guy" who actually tops off :-)

No longer a "theory" scientifically tested and proven. A "92L tank" that shows 82L when full that maintains 82.0L until 10L is consumed there is one possibility. The top 10L is off-scale and not measured by the gauge.

Several people have already corroborated my testing it's now a known fact.

The precision is the only question. I haver "rounded" to 10L. The tank may be 93-94L and the unlisted amount could be 11-12L. I've seen 93 or 94L but that probably includes the 3.6L or so below the gauge as well so I like to use 92L for "full" 10L for the unmeasured amount above 82L 1.4L for the "topped off" surge tank that holds about 5L. That leaves almost exactly 1 gallon (3.6L) that is the true reserve amount that is remaining when the gauge reads 0.0L of gas.

That 1 gallon is more than double what is needed to keep the pump submerged and cooled as well as enough to keep the pump submerged at any angle the car can drive without tipping over. It's a genius solution I've not seen in any other car.

Again "in the 0.001% club" of you've read this thread, not one person in 10,000 groks the fuel supply system in the x5.

I spent over 20-25 hours researching this as I mentioned in the first post: 100% of the dozens of pages I found


Sent from my iPhone using Tapatalk Pro

[Fifty150hs]

Originally Posted by andrewwynn

It actually makes sense. Watch the dial as you drive. The first 10L of consumption doesn't move the gauge. In other words the top 10L is "off scale". I think the car assumes you fill it up whenever the gauge doesn't bobble from 82L.

It would take some alien tech to avoid the problem.

I'll draw a diagram to help explain:



Laws of physics dictate when you have a float some portion will be above and some below the level of the fluid.

When the float is at the bottom it is possible to more or less be able to measure the bottom as there is zero liquid and the float no longer does.

However web at the top, the float will hit the top of the container and stop rising for the last cm or so. That works out to about exactly 10L of fuel.

It's crazy that this has only been discovered in 2017! Silly "everybody on planet Earth" for assuming that the gauge on our cars went from 0-100% when they really go from about 5 to 92%.

"Everybody" knows that the first 50-80 miles the fuel gauge doesn't move. Now pretty much exclusively the few people to read this thread and the actual designers of the fuel gauges know why.

The float is pinned to the top of the tank from about 92 to 82L in the case of the E53. The fuel gauge will be pinned at 1/1 until you consume about 10L at that point the computer will read 82 L and dropping and the needle will start to drop.

I'm not aware of any other sensor to confirm the tank is actually full so if you only filled to 84L I'm pretty sure the computer will over estimate your range and make an abrupt adjustment when the floats are no longer pinned to the top.

My test showed that my computer saw 82.0L for the first 60 miles or so then finally started to drop absolutely confirming what I've described above.

The tank could be made with a bump in the top to make room for the float but that would reduce the capacity of the tank by lowering every place besides the float bump. It's a reasonable trade off and every car on Earth does the same thing just the 0.001% club now knows why.

Would this have anything to do with why my OBC consistently over reports my MPG by about .4 MPG?

[andrewwynn]

How are you comparing the mpg calculation by the obc vs manual computation?

It could exactly be why though if the computer assumes you have 10L more than the gauge at fill up but you actually have 12L more for example.

Do you "top off" when filling? (Bad practice for a few reasons especially in states where the pumps vaccum up the fumes as you pump). You could be throwing off the math. Your gauges also could be reading off: you should see 82L full and 1.4L or so right tank when left is empty and you should get to 0.0L right tank when you get to 0.0 DTE.


Sent from my iPhone using Tapatalk Pro

[Fifty150hs]

Originally Posted by andrewwynn

How are you comparing the mpg calculation by the obc vs manual computation?

It could exactly be why though if the computer assumes you have 10L more than the gauge at fill up but you actually have 12L more for example.

Do you "top off" when filling? (Bad practice for a few reasons especially in states where the pumps vaccum up the fumes as you pump). You could be throwing off the math. Your gauges also could be reading off: you should see 82L full and 1.4L or so right tank when left is empty and you should get to 0.0L right tank when you get to 0.0 DTE.


Sent from my iPhone using Tapatalk Pro

I compare the reported mpg by the OBC to actual at every fill up. then I reset one of the mpg readings. I leave the other one go for long time average. My manual calculation is consistently .4 mpg under reported mpg.

I do top off as I have to. When filling the pump generally cuts off after I've only put anywhere from 12 to 16 gallons in. I know I need more than that as the tank is nearly empty. At least by the gauge. Comparing what goes into the tank seems to be comparable to what I would expect based on the tank capacity.

[80stech]

It would be interesting to see if/how far and where the vent extends into the tank.

[Spunbearing]

A special thank you to andrewwynn for taking the time to post all of this info. I wound up dead on the side of the road myself, replaced the fuel pump only to find my sucking jet pump (That is some name....) was actually leaking all the gas back into the left side. I would not have known what to look for if it was not for this post. I live outside Chicago and do not have a heated garage so I can sympathize with you working in the freezing temps. Last weekend's warm up was well received!

My X5 has 227k miles on it now with the original parts so I am just going to replace the level sender and jet pump with new at this point. That combined with a fuel filter/pressure regulator and the new pump should last me to my 300k+ goal without having to worry about it breaking. I just hope I can get the parts while the weather is still "warm".

Sorry to hear that other people are having failures sooner then I did. I think this design was a little over the top and the parts do feel cheap to me but they appear to work ok.

[andrewwynn]

@ spun: thanks for the feedback.

After evaluating the second sucking jet pump with the exact same failure it's clearly a design flaw but it tends to fail about 5-7,000 hours so should get to 130-230,000 miles (big range of average mph).

When replacing, strengthen the o-ring connection so it is a bit stronger.

I may disassemle my left side today and if so, I may install a screw. If you drill a 7/64 hole into plastic it may as well be tapped for use with a 3mm metal screw. Since it won't be a problem if a small amount of gas drips back into the tank it's not critical if it leaked but it should be sealed as well.

[80stech]

If you put in a backup-ring the o-ring will not get pushed out like that. It almost looks like it was designed to have a backup ring installed and they forgot to put it in at the factory.

[andrewwynn]

Originally Posted by 80stech;

It would be interesting to see if/how far and where the vent extends into the tank.

I think basically exactly the top of the tank (not the filler neck).

Gas in the filler neck may end up in the charcoal filter (why I don't recommend filling to the top of the neck).

I run the pump full speed until auto off then ONE time at 1/3 to 1/2 speed.

Modern pumps are very sensitive to sucking liquid fuel up the vapor collection vacuum that's what shuts off the flow. When the vacuum feels added restriction of the liquid getting sucked in it shuts off (which is often why they dribble after shut off: topping off, fuel gets sucked up the vaccuum then dribbles out after it stops, typically on the fender)

There is a drain for overfill at the filler door but I believe that is outside the seal and only drains while (over) filling. If you top off slow enough and with a non vaccum pump you will just notice the car peeing behind the wheel.

Back to the charcoal filter:

The line to the atmospheric vent I think is attached to the top of the tank and the filter is actually lower.

During operation a vacuum develops and will pull liquid gas from the filter as evaporated vapor but too much liquid will cause problems.

I think the system may actually pressurise the tank to check if the gas cap seal is working.

The car has an automatic valve that closes in case of roll-over to prevent gas leaking from the vent. That valve can fail closed and cause a too much vacuum problem.