View Single Post
Old September 21st, 2008, 16:02  
Veteran Member
Join Date: Aug 2007
Location: Chapin, South Carolina, USA
Default 2005 BEW PD lift pump operation and internals

If you end up needing new IN TANK fuel hose, look on this page, then scroll down a fair bit. You can find the corrugated hoses (my favorite) or you can get the solid elastomer (rubber looking) hoses.

I have personally seen hoses turn to mush when put into fuel tanks - just fuel rated hose will not work.

Edit: The "no-suffix pump" with the white canister, external hose to the eductor and blue/grey top plate is now the only pump available from any source, vendor or dealer. See post #26, #57 and #70 for comparison photos.

I got to examine and test an aftermarket lift pump (1J0 919 050 with no suffix letter) as sold by the vendors on the site today (12/19/2008), and I now DO now recommend it for PD's, I no longer recommend doing Left Coast resident's suction modification to these pumps. The modified suction lines were failing much more often than the pumps themselves. Though the idea of the mod is sound, the execution causes problems.

Edit: Skip to post #7 for some simple tests for your PD fuel system.

Edit: November 2011 replacement pump and motor is now available for the original pumps, this means you can now just replace the bad part, for less money, and do not need to buy the new model pump and modify it.

I took my PD lift pump out and apart today, I was intending to install LCR's alternate suction path
, to ensure my tandem pump could get a suction if the lift pump failed.
I was also searching for any failure modes of this assembly that could cause elevated fuel temperatures, when you have a known good fuel thermal tee and fuel temperature sensor.

I found that I did not need to do the mod; my pump assembly is not even close to the one that LCR pictured, and I can pull fuel from the assembly even with the pump completely failed. Enough fuel to limp home, for sure, if I do not have any air leaks occur when I start pulling a vacuum on the usually pressurized lines.

My pump is part number
1J0 919 050 B, Built out of POM in FRANCE, with a Pierburg electric pump inside it.
and as far as I can tell it is the original pump for the car, which was built in 4/2005.

MY lift pump flows .85 gallons per minute, or 3.2 liters per minute. This is flow to the fuel filter; the electric pump actually flows more, but some is used in the eductor.

This flow rate is enough to equal the cam driven pump's potential flow (which is 1.3 lpm per 903 rpm) at 2200 rpm, but not higher. Above 2200 rpm flow through the fuel rail will be essentially constant, as the positive displacement tandem pump will be limited in flow to that supplied - it will have voids inside the pump. (As the tandem pump draws a vacuum on the lift pump assembly's discharge it will rob a slight bit from the eductor, not enough to matter to the fuel rail, but enough to impact the eductor performance.) The fuel rail will be held at about 15 psi by the outlet pressure regulator - so your injectors will never see any voids.

This may help explain why fuel temperatures tend to run over 70 C at speeds of over 65 mph or so. Below 65 my car will hold 70 C unless the fuel tank is near empty.

The shut-off head of my lift pump was about 15 psi - using a 150 psi gauge, so the reading was not too accurate. It did solidly peg my 10 psi gauge.

Pump assembly overview:

The pump sits in a canister with a fairly open top structure, and only one small check valve on the very bottom to allow fuel to enter the canister. There are no other openings below the top of the canister.

This photo shows the top of the pump (black round circle with red and black wires to it), you can see the discharge hose going back into the canister. You can also see the pump discharge hose coming back up out of the canister.

The normal fuel flow path is fairly complex. The only way to get fuel out of the tank (when level is below about half) is for it to pass through a rough screen, then through a check valve, through an eductor, and into the canister main body where the pump sits. The pump has a fine metal screen (grounded) on its suction. The pump discharge is sent out the top of the pump, through a discharge line, back into the canister body to drive the eductor. There is also a pump pressure control valve (spring loaded valve) on this line, that will relieve into the canister if a low flow/high discharge pressure condition occurs. The remaining flow (normal flow path) is directed to the exit of the pump assembly, where it passes through another spring loaded check valve, to prevent return flow to the tank in this line.


This photo shows the pump discharge hose coming in from the left and entering the tee where it feeds the pump overpressure control valve (white cylindrical object just below the pump discharge hose), and the eductor (not much to see of the eductor from this view, it just looks like the base clipped to the canister.)

Under the suction screen on the outside of the canister, you can see the inlet check valve. All fuel that leaves the tank passes through this check valve. You can see the rounded casting for the eductor to the right of the check valve.

The return line (from the tandem pump, thermal tee, and fuel to air heat exchanger) goes into a path that is dumped right at the top of the canister. This return flow is 1.3 liters per minute at idle. During normal operation the eductor pulls in more fuel than is discharged to the fuel filter. This causes a continuous overflow out the top of the canister, carrying the warm fuel from the return line out the top, and ensuring a steady supply of cooler fuel from the fuel tank.

When the tank level gets so low that the eductor is pulling air instead of fuel, the return flow from the tandem pump will fall into the canister, ensuring a pump suction until it runs completely dry.

If the pump quits running, the tandemp pump (or a mityvac) can pull fuel up through the eductor jet (reverse of normal flow). This hole is small, but would provide enough fuel to get home at low power levels. I doubt it could support full power operation. The tiny opening would be subject to plugging if you had debris or gel in your tank.

Last edited by DanG144; January 25th, 2017 at 06:39.
DanG144 is offline   Reply With Quote
Page generated in 0.26680 seconds with 7 queries
[Output: 21.41 Kb. compressed to 19.79 Kb. by saving 1.61 Kb. (7.54%)]