DIY: E53 X5 Siphon Pump with Improved O-ring and Teflon Wrap

[andrewwynn]

Originally Posted by cn90

Thanks andrewwynn,



The photo you posted above should NOT be used to show your point.

That photo I took is when I pried the Coupler out.

It was halfway out so it is NOT an in-situ photo of factory setup!



I am very convinced the root cause is either:

A. O-ring shrinking with time.

B. The one-way valve (Plastic+Spring) momentarily seized,

pressure builds up and blows the O-ring outward...



Just some theories...





BTW, I think one can remove the Plastic+Spring combo (I did not) w/o any issues. This is an over-engineered part.

Thanks for the feedback I thought that photo looked partly disassembled but it works perfectly to showcase the issue because it's exaggerated somewhat. I've updated my post to make it clear.

I didn't take a picture from the correct angle to show the problem mine is already fixed by the time I snapped the shot.



This photo shows the o-ring pushed out of the seat. In both mine and wife's case the photo marked up with the blue and red lines perfectly showcases the issue though it was a little tighter.

Neither A nor B is correct. Both my wife's and my car had the same exact problem where the o-ring writhed out after the part twisted like shown (by chance) as shown in the photo above.

B is not accurate because there is a pressure relief valve to deal with back pressure and dump over pressure to the right side tank

The problem could be exacerbated by the o-ring being smaller than it could have been but without movement of the plastics the o-ring would be lifetime.

[cn90]

Even though I applied the zip tie, I don't think it is the solution.

When I removed the siphon pump, mine was intact with intact O-ring, zero issues. Everything aligned perfectly from factory.

I still think the O-ring shrinks with time, allowing it to bow under pressure.
When I disassembled the Coupler, the part slides out easily, telling me that O-ring has reduced in size. It should have been tighter when sliding it out.

Anyway, the new O-ring is much better, a tight fit.

The one-way valve is over engineered, if you remove the one-way valve can be removed without issues.

I understand that you did a lot of pioneering work in the fuel siphon pump issue, kudos to you. But I don't think anyone here knows for sure the EXACT cause of failure...

Everything here is conjecture...

[andrewwynn]

What I'm saying is that yours just didn't fail yet. It's great if one can catch this issue before it fails because it's so simple to prevent it from happening. The o-ring will never come out of its seat until there is enough room for it to do so. You didn't see It because yours didn't fail. Sadly I didn't take a pic from the best angle to show why I'm absolutely certain that regardless of other valid email examples of what can help the problem along, that without the tilting of the seat of the o-ring this o-ring failure would be non-existent.

It's best if somebody can catch it before it failed because then fish a zip tie though without even disassembly and you are done. For a non-pro I wouldn't even recommend messing with the other "seal" that will only drip when it fails.

Simple engineering and geometry plus laws of physics dictate why I'm making my claims. There is a 1" cross section with 18psi pushing down about 1" from the fulcrum. 18 in · lb of torque applied to a plastic part for 4000 hours is apparently enough to bend it enough to get the o-ring close to where it will squeeze out of the seat.

Very possibly a toward the thinner side of spec on the o-ring cs or fatter opening on the plastic part or a spike in pressure will be a contributor to the failure it's rarely just one thing. My contention is that in spite of any of the other contributor factors it will simply not ever fail if the o-ring seat is always square.

You could overcome the problem with a fatter o-ring but it's unnecessary since the primary issue is the not square seat of the o-ring.

The fatter o-ring fit makes a lot of sense and you are more correct than it sounded like I was suggesting only it was not the o-ring that shunk it was the plastic that grew: case in point the other non-seal "seal". I measured that the outer part was a full 0.2mm larger than the inner part and that is without the intense pressure of an o-ring pushing on it for 17 years in my case. I am willing to bet the outer part has grown at least half a mm making a chubbier oring make more sense than the OE spec that will work just fine in my opinion, the fatter cs option will work fine-er.

[andrewwynn]

Originally Posted by cn90

BTW, I think one can remove the Plastic+Spring combo (I did not) w/o any issues. This is an over-engineered part.

Actually I think that part is critical. I think it's job is to make sure the siphon jets go only from left to right.

Her is what I mean: when the car is pitched up or down one siphon jet will be sucking air and may not produce any voiume maybe it's job is to make sure there isn't reverse flow going in a circle from back to front of the left tank.

This is absolutely a guess. I can't even remember the direction of flow on the thing I just remember it's a very soft spring that would "suggest" the direction of flow. Maybe more to keep the siphon jet lines primed when the car is off? The seal is plastic to plastic it's not going to hold forever more of a suggestion, perhaps to prevent damage to the siphon head in the case of overpressure from the FPR?

[semcoinc]

This is a very timely review of this system for me as I have a replacement pump ready to install into my 113K mile E53 4.4i

I ran Test 6 on a 20 mile highway drive today and say right tank levels in cruise rang8ng about 2.0 liters to 3.8 liters. During some braking and turns I saw as low as 0.5 liters and was pretty freaked out but all kept running fine and the right tank numbers recovered.

I currently have about 26 liters total in test 6 and started my trip with about 30 liters.

I’ll be opening up the left side of the tank to evaluate my siphon jet pump but wanted to get an opinion on whether my current test 6 values show anything amiss in the Jet pump.

Thanks in advance to you guys who have dug deep into this system and provided excellent photos of the areas involved.

Mike

[andrewwynn]

The right side should hover close to 1.4L once the left gets below about 28L. If on level ground the right side starts to drop it means the siphon jet or not keeping up with the gas being consumed from the right side "well" and the siphon jet needs addressing.

The value will vary wildly when cornering or braking/acceleration.

Example: a hard right turn will slosh a bunch of fuel out of the well to the left side. You can make a hard right and once you go to level constant speed, the right side will be lower like 0.9 and slowly build back up to "1.4"

You should be able to drive down to year six of 000 for the left and 014 on the right with no problem. At that point you have 5L of gas left. At 000 000L there is about 3.6L in the tank about 3L you could actually consume.

When my siphon jet was leaking I noticed that the right side would hover at 0.9 for a good while then start to drop while the left side had 18-20L. That's the exact symptom of leaking siphon jet. The good news is that the right side has a good 10-15 miles once the right side drops below 1L on test 6.

[semcoinc]

Thanks Andrew, that's very helpful.

Mike

[cn90]

semcoinc,

- The above tests are EXACTLY what I had in my 2006 X5 when fuel is below 32L. And that is normal as my siphon pump worked perfectly before I did the preventive maintenance.

- I just did the experiment the other day, the YELLOW warning light comes on when there is about 10L in the LEFT tank.

- As andrewwwynn pointed out in other threads, I am redrawing the fuel tanks so you can see. It is just descriptive drawing for the ideas (not to scale).

- First photo shows the real tank shapes. NOTE that the 5L well, where the fuel pump sits, is EXCLUDED from the level sensor calculation, so this 5L well is somewhat "sacred".

- The 2nd photo shows my understanding of the cross-section, as if you look directly from the FRONT of the car.

- Now you can see why when you turn R or L, accelerate or decelerate, the level sensor moves up and down, either by the swaying motion or the turns/braking etc. (without liquid moving it) or by liquid sloshing over the right side moving the R level sensor...

- The important lessons are:

1. The 5L pump well is "sacred".

2. Calculation is done via 2 level sensors and the sum is reported to the instrument cluster.

3. Fuel Level Sensor issue: sometimes the gauge shows 40% but the car dies from lack of gas. This happens either because:
a. The level sensor is stuck in the 40% position.
b. The resistor (the level sensor works by varying the resistor Ohm values) is kaput and reports an Ohm values consistent with "stuck at 40%", no matter where the mechanical float is...

4. If your car dies and you suspect failed siphon pump, try to jack the left side up (drive it up on a curb so the left side is up on the curb), this should slosh about 5-6L to the right side, allowing you to get to the nearby gas station. Or do a hard left turn lol...

Hope this helps...


---

[andrewwynn]

That's a great depiction of the tank and how it works. 10,000% better than BMWs own since they copied the page from the sedan which does not have the pit and they drew the left and right lobes the same size and symmetrical.

A minor difference is that the right gauge will actually read about 1.4L of the 5L pit so the unlisted "sacred" amount meant to keep the pump always submerged is about 3.6L

I would have flipped the drawing to have right on the right side but see you drew to match the photos taken from the front.

If you catch yourself as you are about to run out of gas from a failed siphon jet you can make a hard left turn (with any luck a U-turn to a gas station you just passed) I bumped into several threads where people mentioned success with this trick in sedans while doing my fuel system research.

This trick or the one to Jack the left side will likely not work on the X5 however, because as the o-ring seal fails, it still will siphon some gas it just relies on the depth of the fuel to help that's why some people run out of gas at 30 miles to empty some people run out at 75 like when happened to me. The wear level of the electric pump comes into play as well: stronger electric pump will push more fuel volume and pressure back to the siphon and will work longer with the leak.

The other problem is you don't get to use all the emergency gas you put in: anything over 5L will spill info the left lobe of the tank and the pump will suck out more than you consume and send it over to the left side so you can't drive very far with the siphon jet failed and a boost of 2gal from a plastic can of gas. (eg though 2 gal should get you 30-40 miles it might only get you 5-10. )

One last thing: test six works when the car is on not running so if you do stall while driving you can perform test six to confirm a siphon jet vs electric pump failure. If both sides have plenty of gas eg 402153 expect fuel pump failure. If you see 254000 it's a sure sign of siphon jet failure. If indicating siphon jet failure you don't need a tow just 5L of gas to drive to a near station where you can add more than 30L to flood to the right lobe. Then do the fix above and it'll be good until the elec pump fails.

[cn90]

Originally Posted by 270_BMW

..Did you use PTFE or Gasoline Rated Teflon Tape? Will keep it from dissolving.

- Used regular household plumbing White Teflon Tape.

- I did an experiment: White Teflon (household plumbing) tape inside a jar containing gasoline. Two (2) weeks later, zero issues, tape is still intact.
So I am not worried. I will report this experiment result in another 3-4 months.

- See photo of experiment...

- The Improved O-ring is a tight fit, very good O-ring, better than factory O-ring.


---