Died on Highway: Fuel Issue

2003 Jetta Wagon ALH Manual
Stock other than Franko6 Cam and 520 Injectors
290,000 Miles
Southeast Michigan
Original Injection Pump
Half tank of fuel (Historically have used high quality fuel)

Wife and son called last Sunday telling me car lost power on the highway, kept running as they got onto the exit then eventually stalled as they coasted to the shoulder. They tried to restart and got nothing. I went out, looked for obvious issues then towed it home. I began checking the easiest possible suspect areas first. Chain of diagnostic events follows:

1. Checked for relay 109 symptoms...none
2. Opened IP connector below pump to look at terminals...nothing irregular
3. Tested, based on info here, the crank speed sensor...all appears good
4. Tested fuel cutoff solenoid...good
5. Got worried and checked the static timing of the pump, crank and cam...good
6. Tried to run pump off a jar of Power Service...it would not run
7. Realized I probably needed to bleed the pump all the way to the injectors to get it to run off the Power Service if it had indeed run dry.
8. Primed pump with mityvac
9. Loosened the four injector nuts, cranked over to prime each of the injectors
11. Tightened four nuts, cranked and it came to life
12. Idled for a bit, shut it down
13. Decided it was probably a clogged fuel filter. Ashamed to admit I somehow let the Cat filter go for 74,000 miles. I changed it out, filled new filter with Power Service, started up and took a ride around the block
14. Drove the car the few miles to work then later that evening to an event, wife drove it a couple miles to work then put it in the garage. Total miles driven were about 45.
15. After sitting for about 4 hours she went out to pull it out, it started fine, she got about 6 feet then it stalled, tried to restart and it stumbled, stalled and would not restart.
16. I am heading out there now to have a look.

I suspect this is an air/priming issue with the fuel system. I am not convinced the original highway issue is the same cause of the latest no start. Based on my research the air leaking into the fuel system would need to be significant to take down a running engine. I think the original issue was indeed the clogged filter and the new issue is a loss of prime issue. I plan to inspect the fuel sight for presence of fuel first then begin looking at the thermo-tee. Any guidance would be appreciated. Some say I don't even need the thermo-tee with the Cat system. I've been running it since about 20,000 miles on the original thermo-tee and o-rings I think.

tacolifestyle said:

....... [/B] Some say I don't even need the thermo-tee with the Cat system. I've been running it since about 20,000 miles on the original thermo-tee and o-rings I think.

Click to expand...

seems like you've got many things covered and a plausible theory. The last thing you wrote stands out - you're running really old o-rings? That would be the first thing I would change - and would you consider going back to the stock filter setup?

I don't know anything about the "Cat system". I've always used the stock filters and the formal alternating drain/replace service interval. My car also has 290k miles, original pump and I do have new nozzles since 100k ago.

last thing that went on my car was the regulator/brush set - battery charge light came on. If you have not done a new alternator or brush and regulator set - I highly recommend that as well! very easy to do from the top of the car and can stop you getting stranded - for about $50.

My car has an aftermarket lift pump installation that might help prevent me having problems bringing the fuel from the tank to the pump.

The CAT system is just a head on which a 2-micron CAT filter screws.

The head, however, has a hole for the original Thermal-T. You can put an O-ringed plug in there and maybe you should, because doing so prevents the lift pump in the IP from pulling vacuum on the filter in a way that draws air into it through that thing. They DO crack, and the O-rings DO get old and stop working.

Will that stop a running engine? You bet it will, if enough air gets in there. If the car ran on bottle of fuel directly connected, then the pump and engine are fine. You either have a severe restriction in the fuel system (possibly some gunk in the tank pickup) or you're getting air into the system, probably through the T (they're known to crack) or its O-rings.

Not in Southeast Michigan...... but, my brother never changed the OE fuel filter on his 01 Jetta until just shy of 171k miles (lives northeast of Atlanta)... no issues, but no gelling weather either.


Quick story worth sharing which may shed some light on the overall big picture..

In about 2000, my brother was moving from Nashville, TN to Columbus, OH. In the process of relaying his vehicles (helping him move), I was driving his 1984 Jetta diesel from here in southeast Kentucky. When we got to northern Kentucky just south of Cincinnati, the fuel gauge had barely moved off full (about 175 miles)..... Well, a fill up and it only held about two gallons.

Long story short, the vent line hand collapsed shut (long story in of itself). Without air taking the place of fuel as it was burned, the Injection Pump collapsed the fuel tank.

So, yes, the IP in an ALH engine has the ability to suck air in around the little O-rings when the filter is dirty, gelling, etc.

Thus, I suspect the OP's issue was/is a combo of the dirty filter and air getting in the system.

Sounds like your problem may be....

Your sending unit/fuel pick-up, that is in your fuel tank, has a very small orifice at a check valve that can and often will bridge/plug with debris, ice, and/or gelled fuel.
When this happens the vacuum created by the inj pump trying to suck fuel will cause air to be sucked into the fuel system at the weakest link, usualy the thermo T at the filter. This all happens more commonly in cold temps.
See....
https://www.youtube.com/watch?v=yRMTEAKsScE
If this is your problem search "ALH TDI blocked check valve in sending unit".
The check valve can be drilled out and removed.

More info..

17. I went to the garage and the fuel sight line was dry.
18. I pulled the thermo-tee to inspect. Having never changed the o-rings I looked closely and really didn't see anything. It did slide out of the filter head without much fuss. I went ahead and changed the o-rings and lubed them to reinstall. Seems like a much better fit now but overall inconclusive.
19. I then went to pull vac on the return line coming out of the IP to prime the pump. Initially I pulled some fuel from the filter then nothing. I had forgotten to plug the thermo-tee barb so suspect I was just sucking air after the vacuum of the fuel broke.
20. I then thought I should pull vacuum across pump directly at the return with the inlet plugged. This is all directly at the fittings on the pump. I notice that it will not hold vacuum across the pump. I suspect that isn't normal. I don't see any leaking of fuel from anywhere on the pump. Is this a case of IP seals needing to be replaced?

Thinking through this the only points that could be venting to atmosphere when pulling vacuum in this manner are the injector hard lines & connections, injector return lines, injectors through the engine or the pump seals. It is my understanding that the injectors themselves are unlikely to leak to atmosphere in this manner. The lines show no signs of leaking fuel. Considering they are under pressure during operation and show no sign of fuel spilling out I don't think they are a concern. Especially at the rate vacuum is diminished. I can pull 15lbs if I pump the vac rapidly then it dissipates within 1-2 seconds.

Anyone have thoughts?

As I illustrated in my other post, the vane pump inside the IP has the ability to collapse the fuel tank (metal tank in the 84 Jetta) if the venting is obstructed. To keep my post as short as possible, I did not elaborate on the restriction. Well, the alignment shop jack smashed the vent pipe when raising the vehicle. Until we found the problem, we left the fuel tank cap loose.

So, a restriction will induce a strong vacuum pull from the IP. The weak point in the system would be the O-rings at the Thermo T on the fuel filter.

maxmoo said:

Sounds like your problem may be....
Your sending unit/fuel pick-up, that is in your fuel tank, has a very small orifice at a check valve that can and often will bridge/plug with debris, ice, and/or gelled fuel.
When this happens the vacuum created by the inj pump trying to suck fuel will cause air to be sucked into the fuel system at the weakest link, usualy the thermo T at the filter. This all happens more commonly in cold temps.
See....
https://www.youtube.com/watch?v=yRMTEAKsScE
If this is your problem search "ALH TDI blocked check valve in sending unit".
The check valve can be drilled out and removed.

Click to expand...

Check ^. I just had the same issue an that was the culprit. Be careful with the gauge sending unit, I killed mine in the process.

Thanks. I am going to inspect the fuel pickup today. I've never looked in there and at this mileage even if not the culprit wouldn't hurt to see what's going on in there. I am still wondering if I should be able to hold vacuum across the pump return and a plugged inlet. Maybe it's not definitive.

I have never checked if a dry pump will hold vaccum but I’d assume it should, gurus chime in here. An easy way to check for tank restrictions is to put a vacuum on the line from the tank to the filter, the fuel should flow easily with little vacuum. If it does try the line from the filter to the pump as the filter could also be plugged. By any chance were there air bubbles in the clear line when running normally?

Did not look for bubbles in the sight line while running. Filter is new and that is when we got it back running pretty easily then drove 125 event free miles now dead again. One thought is if the filter was clogged that was the restriction that caused the IP to pull air through somewhere that is now perhaps continuing to allow air into the system and losing prime. I have changed the o-rings at the thermo tee as well.

Heading out in a bit to pull vac on those fuel lines to see what flows and what doesn't....will report back.

Maybe some progress...

I went out just now and pulled vacuum from as suggested at the line between the filter and the tank. Fuel flowed in what my opinion was freely with 10 lbs of vacuum into the catch can on the mity-vac. The fuel looked good and not thickened at all. I then put everything back together, cracked the nuts on the injectors, cranked, saw fuel at each, and tightened the nuts. It took a bit of cranking but it eventually fired up. It was in the single digits last night so this all seemed reasonable. I observed no bubbles of any size flowing through the site line. I let it idle for a bit in the alley then put it back in the garage with the hood open so I could see the sight line. I turned off the ignition while watching the sight line and only saw an inch long bubble rise to the highest point in the sight line. I am going to let it sit for a while and check every 30 minutes to see if that line goes dry slowly.

The IP will not be totally dry inside. There will be enough fuel to lube everything and allow the Vanes to pull fuel.

There is a restrictive orifice (overflow restriction) in the return fitting on top of the pump. Also, there is a check valve (pressure control valve) in the inlet fitting of the pump. So, based on the purposes of those two fittings and their components, I don't think the IP body will hold vacuum. However, during operation, there is "pump body" pressure.

Glad you got it going..

Yeah...I suspected it would run as the symptoms when it died both times were similar. I will monitor and report back if I experience another episode or have anything further to report. This is a great forum and I always try to bring any of my issues presented here to closure in hopes of helping the next guy when he's searching out his problem.

In the meantime I will be keeping some tools in the car to field bleed the injectors in case this crops back up.

With any luck, the bubble won’t get bigger and your saga is over. Keep an eye on the bottom of the pump for leaking fuel. It could be something as easy as the qa seal and you can probably fix it easily. I would run some Anti-Gel additive, I use Howes. Don’t drill the check valve if your pump is loosing prime, it will make it worse.

Update: I went out this morning and the bubble had grown to almost consume the sight tube. I tried to start it and it took a bit of cranking but it eventually fired over. I pulled it out of the garage and let idle for about ten minutes. I then put it away and watched as I turned off the ignition. A one inch long bubble returned to the sight tube from the pump. My recollection is that there is always some air in that line when the engine is off. About 4 hours later that bubble has doubled to two inches mostly growing towards the pump. I believe that indicates I have a leak allowing the pump to lose prime. I sprayed it down with brake cleaner to try to see where I might be leaking. I am ordering the reseal kit in anticipation of the QA or cover seal being the culprit. I also sprayed down all the injectors, returns and lines to clean it all up and be able to see what it what after all the recent bleeding. Any other advice to more easily identify where the leak is occurring?

Very fine filters like the 2 micron cat filter will plug up much easier with wax crystals than the stock filter. I'd try cutting the fuel with some kerosene in addition to the powerservice.

tacolifestyle said:

Update: I went out this morning and the bubble had grown to almost consume the sight tube. I tried to start it and it took a bit of cranking but it eventually fired over. I pulled it out of the garage and let idle for about ten minutes. I then put it away and watched as I turned off the ignition. A one inch long bubble returned to the sight tube from the pump. My recollection is that there is always some air in that line when the engine is off. About 4 hours later that bubble has doubled to two inches mostly growing towards the pump. I believe that indicates I have a leak allowing the pump to lose prime. I sprayed it down with brake cleaner to try to see where I might be leaking. I am ordering the reseal kit in anticipation of the QA or cover seal being the culprit. I also sprayed down all the injectors, returns and lines to clean it all up and be able to see what it what after all the recent bleeding. Any other advice to more easily identify where the leak is occurring?

Click to expand...

Yeah, you're leaking air somewhere. You need to find it; the top culprits are in and around the Thermo-T (either a crack in it or the O-rings around it) although the pump can be the cause too. If the pump seals are involved you will usually (but not always) have fuel leaking out where the air is leaking in, although the fuel leak can be very small.

If there IS a fuel leak you need to fix it immediately because beyond the no-start problem you'll find all sorts of rubber things get eaten by the fuel, including hoses and the hard-pipe coolant O-ring that is under the IP -- and while that O-ring is cheap it is also a five-alarm pain in the butt to replace.

Does the pump seem like it’s getting wet anywhere? When the Qa seal died on my pump I found fuel wetness on the bottom of the pump. The only reason I’m saying it’s a pump issue is because you said it wouldn’t hold vaccum, by any chance was anyone able to confirm it should? Usually I start with the thermo t, try putting grease on it to improve the seal.

Andybees says the pump body will not hold vacuum in the way we think. I didn't quite follow his explanation but he seems to know his stuff so I am going to trust he is right.

I sprayed down the pump and related areas with parts cleaner last night to clean-up any spilled fuel from all this testing. I will take a look tonight to see how much air is in the tube and if there are any oily spots where fuel might be leaking. Might need to run it to return pump pressure to the body of the pump. I did change the o-rings on the thermo-tee and added some grease there as well. It fit nice and snug. I could not see any obvious cracks in the tee. I think I might run it tonight then at shut off pinch the inlet hose off between the filter and the pump to isolate the pump and see if the bubble increases over time. That should mean it's from the pump right?

One other thing I just thought of, I have had an issue once with the seal on the water drain on the filter being bad. On another note, is it a good filter or an autozone special as I have had issues with cheap chinesium filters leaking at the thermo t. They don’t seem to be manufactured correctly and the size of the hole is incorrect causing air to leak in. I definitely trust Andybees and missed that post, as he has been much further into that pump than I have. In my experience it’s usually a beat fuel line, bad clamp or thermo t not the pump.

KyleMillione said:

One other thing I just thought of, I have had an issue once with the seal on the water drain on the filter being bad. On another note, is it a good filter or an autozone special as I have had issues with cheap chinesium filters leaking at the thermo t. They don’t seem to be manufactured correctly and the size of the hole is incorrect causing air to leak in. I definitely trust Andybees and missed that post, as he has been much further into that pump than I have. In my experience it’s usually a beat fuel line, bad clamp or thermo t not the pump.

Click to expand...

Sorry just realized that you have a cat filter mod. Filter is probably not the issue. Check the elbow that attaches to the hard fuel line and make sure it’s not loose. But definitely go over all the lines and clamps. Also I would run the car to get pressure up in the pump to check for leaks. The best way I can think of to check if it’s the pump leaking would be to prime it to run off fuel in a jar with a send and return line, run it for alittle bit, then leave it and see if it forms the air bubble. If the air bubble forms you have a pump leak. That’s how I determined that the pump was leaking in my wife’s old 97.

If it was mine I would still check the pickup in the tank......certainly before replacing the inj pump.
I assume you mean 10 inches of vaccum not 10 lbs.....still even 10 inches may be more than there should be.....I've never measured it.
It dosn't take a lot of resistance to make tired inj pump seals leak.

I went out tonight and the air in the sight line had grown to take over most of it.

More info based on the most recent replies above.

Went out to see what pulling fuel from the tank up to the filter looked like with the mity-vac. I used the catch can. I started by pulling the lever once then twice slowly building up vacuum. After three pumps I saw no fuel entering the can. This made sense as fuel certainly drained back to the tank when I pulled the line from the filter. Or should it? I assume there is no back flow prevention in the sender. If I pumped to 10 in Hg I could get fuel flowing into the catch can but vacuum was quickly dissipated. I found it would hang at 5 in Hg until I released the vacuum. Maybe that was equilibrium between the back suction of the fuel trying to flow back to the tank and my pump pulling the other way. Any thoughts on all that?

I then pulled both lines from the pump to run it off of a quart bottle of fuel. I set up two clear lines clamped to the in and out of the pump. It's about 18F here and the car has been sitting for a few days. I let the glow plugs do their thing then cranked it over. I did not prime the pump because the in line still had fuel to the top and the fact the sight line was not fully dry. It started slowly but not too bad. Maybe cranked for 4 seconds? I ran it on that bottle of fuel for a bit observing the clear lines. The return line had some real aeration with tiny bubbles at first. Then the inlet started to pick them up as well. I adjusted the lines to minimize this. I shut it down leaving two loops of the clear line up high to trap any air bubbles. After about three minutes the small bubbles were no longer trickling to this high point. A bubble about 1/2" long settled there.

My plan is to monitor that to see if that bubble grows overnight. I also am confused why so many little bubbles were being generated in the outlet line. Then I thought maybe I had the inlet sucking some air or the original air gap in the line had just gotten distributed and recycled into the bottle. I may need to revisit this with a better set-up in a more IV like set-up.

I need to fully understand how air getting into the system results in a large air bubble in the site line if different scenarios.

1. Engine running with some restriction in the fuel flow anywhere between the tank and the pump: Air enters through the path of least resistance into the fuel flow. This could be anywhere in the line from tank to filter plus the thermo-tee interface (and maybe the IP seals?). This small amount of air gets aerated into the fuel but not enough to stall the engine.

2. Engine off with condition above: Upon shut off this air mixed finely with the fuel found in the relatively high volume of fuel in the pump and gathers slowly at the sight line.

3. Engine off without condition #1: As fuel leaks from compromised seals in the IP air enters to replace it and gathers in the sight line.

Is that all right? Am I missing something?

Yes, there is a check valve in the fuel sending unit in the tank. Read the first post in this Thread... about check valve in the tank sending unit.
http://forums.tdiclub.com/showthread.php?t=399479

The tiny bubbles were coming out of the IP as some larger ones were sucked in when you first set-up the IV fueling arrangemnt. Notice that the return fuel fitting is at the highest point on the IP. That position helps to remove air.

The air is either getting in from the Thermo T or a seal(s) on the IP.

I recently (Thanksgiving weekend) purchased my brother's 03 Jetta. The IP leaks like crazy. He was lucky to get to Kentucky from Georgia. The next day after they got here, it would not start.......... air in the hose from the filter to the IP. When this cold weather breaks (0 F here tonight), I plan to install a used pump.

EDIT: To clear up my previous post... The IP has a check valve on the "in" side and a restrictive valve (not sure of the term) on the exit fitting(return). Thus, this arrangement will allow the IP body to build some pressure during operation, obviously not a lot of pressure.

tacolifestyle said:

Maybe that was equilibrium between the back suction of the fuel trying to flow back to the tank and my pump pulling the other way. Any thoughts on all that?

I need to fully understand how air getting into the system results in a large air bubble in the site line if different scenarios.

Click to expand...

I think it's like siphoning fuel out of a car.
The tank is lower then the filter and injector pump.
Once you get it going to a lower point it keeps flowing.
An air leak up front would let it flow, slowly until air is back to the tank.
Also the fuel cap opens the tank vent when its on so there wouldn't be pressure against the flow.

I wonder if putting a can of fuel under the hood with 2 clear lines going to the fuel filter would eliminate the tank & lines from it as a problem.
If it still gets air in then run the clear lines to the injector pump to see if it still leaks back. This might confirm the pump sucking air.

Last thought, the injectors have a 3MM hose between them going back to the pump.
I think I recall someone posting one of them was leaking slightly. Maybe letting air in through there?
I had the last injector rubber cap split on one of my Rabbit diesels. I spotted it as a leak. Could have been a point for air to get in.

Here is a link to an exploded view of the IP.

Scroll down the list and look for parts 135 and 136 on the list. (control valve and overflow valve).

http://pics.tdiclub.com/data/500/Manual_Pump_Parts.pdf

wonneber said:

I think it's like siphoning fuel out of a car.
The tank is lower then the filter and injector pump.
Once you get it going to a lower point it keeps flowing.
An air leak up front would let it flow, slowly until air is back to the tank.
Also the fuel cap opens the tank vent when its on so there wouldn't be pressure against the flow.

Click to expand...

^ Thats the idea. It’s not the bubbles collecting- air is being sucked in as the natural vaccum is siphoning the gas back to the tank on the return side. Put the fuel canister below the pump, I think I may have forgot to mention that, so that gravity is pulling the fuel back to the can. I believe that the check valve is also your problem, I get fuel with 1 pump, with literally no restrictions. Something is actively letting air in, which has to get resealed, but the restriction is causing the pump to work harder more pressure on the tired seals.

tacolifestyle said:

"I need to fully understand how air getting into the system results in a large air bubble in the site line if different scenarios.

1. Engine running with some restriction in the fuel flow anywhere between the tank and the pump: Air enters through the path of least resistance into the fuel flow. This could be anywhere in the line from tank to filter plus the thermo-tee interface (and maybe the IP seals?). This small amount of air gets aerated into the fuel but not enough to stall the engine.

2. Engine off with condition above: Upon shut off this air mixed finely with the fuel found in the relatively high volume of fuel in the pump and gathers slowly at the sight line.

3. Engine off without condition #1: As fuel leaks from compromised seals in the IP air enters to replace it and gathers in the sight line.

Is that all right? Am I missing something?"

Click to expand...

Yes, based on my experience that all makes sense.