Suddenly won't start after recent tandem pump change

TAABVW said:
To be clear, the one way valve on the fuel feed (starting as close to the pump as reasonable possible) is to help diagnose where air is leaking in. It's not intended as a permanent fix. It will give you one of three results:

1. If the leak is between the tandem pump and the one way valve then the car will start and very little air will be seen in the clear fuel return pipe. This is because the one way valve will prevent the head difference between the leak site and the level of fuel in the tank from sucking air in at the leak site.

2. If the leak is between the one way valve and the tank then the car will start and lots of air will be seen in the clear fuel return pipe (or the engine will splutter to a stop). There will be no air between the valve and the pump hence it starts, but then air will be drawn in through the leak site further upstream. This will be bled out by the pump and seen in the return, or it will stop the engine.

3. If there is still a very poor or no start then the leak is in the pump.

Given the history of the problem you have described, I would suggest that the original pump was probably not at fault and it was an air leak in the feed pipe all along. You have disturbed this air leak site in the fitting of the new pump making it worse. This gives a clue as to which fuel feed pipe to replace first. philward suggested the feed pipe an unlikely source of the leak (while acknowledging it's an easy fix) but I have found the more they are pulled off and replaced, the more likely they are to leak. I once spent about a month (not full time) chasing around a slow starting issue. I clamped up that feed pipe nice and tight every time I refitted it but that's what it turned out to be. If the pipe "clunks" into position when you push it on you might think that's a nice fit, when it's actually indicating the rubber has gone hard and is no longer malleable enough to seal.

Answering some of your questions directly:

Your plastic tube clearly shows air and fuel coming out of the pump, therefore the feed pipe is the other one.
The lengths of the pipes are irrelevant provided they are air free.
You don't need a clear pipe on the fuel feed necessarily, just the return to see what air the pump is dealing with and when.
Any 8mm diameter diesel fuel pipe will do for testing and indeed replacement if necessary.
If the leak were at the filter housing it would likely leak diesel out given the filter housing is low down in relation to the pump level. This kind of (no diesel out) leak tends to be high up in elevation hence starting near the pump.
philward suggestion of brimming the tank and parking on a (steep) downhill, or car service ramps might well yield results in the form of a physical diesel leak. You would have to go round and carefully check every joint in the feed line (perhaps using kitchen paper) for signs of diesel damp.

I have had about 12 audi fuel delivery systems (not all A2's) in many pieces and adapted them to run vegetable oil over a period of about 20 years. The oil hardens the rubber hoses quicker than diesel and the extra viscosity of the oil creates a bigger vaccuum in the feed pipe than diesel (which is why i fit a lift pump at tank level). Thus I probably have more experience in this area than many. I apologise if this experience might be clouding my assistance.
Click to expand...
TAABVW said:
To be clear, the one way valve on the fuel feed (starting as close to the pump as reasonable possible) is to help diagnose where air is leaking in. It's not intended as a permanent fix. It will give you one of three results:

1. If the leak is between the tandem pump and the one way valve then the car will start and very little air will be seen in the clear fuel return pipe. This is because the one way valve will prevent the head difference between the leak site and the level of fuel in the tank from sucking air in at the leak site.

2. If the leak is between the one way valve and the tank then the car will start and lots of air will be seen in the clear fuel return pipe (or the engine will splutter to a stop). There will be no air between the valve and the pump hence it starts, but then air will be drawn in through the leak site further upstream. This will be bled out by the pump and seen in the return, or it will stop the engine.

3. If there is still a very poor or no start then the leak is in the pump.

Given the history of the problem you have described, I would suggest that the original pump was probably not at fault and it was an air leak in the feed pipe all along. You have disturbed this air leak site in the fitting of the new pump making it worse. This gives a clue as to which fuel feed pipe to replace first. philward suggested the feed pipe an unlikely source of the leak (while acknowledging it's an easy fix) but I have found the more they are pulled off and replaced, the more likely they are to leak. I once spent about a month (not full time) chasing around a slow starting issue. I clamped up that feed pipe nice and tight every time I refitted it but that's what it turned out to be. If the pipe "clunks" into position when you push it on you might think that's a nice fit, when it's actually indicating the rubber has gone hard and is no longer malleable enough to seal.

Answering some of your questions directly:

Your plastic tube clearly shows air and fuel coming out of the pump, therefore the feed pipe is the other one.
The lengths of the pipes are irrelevant provided they are air free.
You don't need a clear pipe on the fuel feed necessarily, just the return to see what air the pump is dealing with and when.
Any 8mm diameter diesel fuel pipe will do for testing and indeed replacement if necessary.
If the leak were at the filter housing it would likely leak diesel out given the filter housing is low down in relation to the pump level. This kind of (no diesel out) leak tends to be high up in elevation hence starting near the pump.
philward suggestion of brimming the tank and parking on a (steep) downhill, or car service ramps might well yield results in the form of a physical diesel leak. You would have to go round and carefully check every joint in the feed line (perhaps using kitchen paper) for signs of diesel damp.

I have had about 12 audi fuel delivery systems (not all A2's) in many pieces and adapted them to run vegetable oil over a period of about 20 years. The oil hardens the rubber hoses quicker than diesel and the extra viscosity of the oil creates a bigger vaccuum in the feed pipe than diesel (which is why i fit a lift pump at tank level). Thus I probably have more experience in this area than many. I apologise if this experience might be clouding my assistance.
Click to expand...
Just to add my recent experiences to Trevors extensive knowledge on this. I've been SLOWLY converting my 90 to run on veg oil. Coincidentally I finally got it all running last weekend, but still have a couple of issues to sort. the veg oil conversion involves installing a clear inline prefilter, so its easy to see any air in the system - i have ALOT. where is it getting in? well, i have disturbed both the return and feed fuel pipes where they meet the metal pipes in the engine bay. The feed pipe i initially reconnected with the spring clip rather than a jubilee type. when i started the engine there was a fountain of diesel/veg mix shooting out of the feed pipe union - sealed that with a jubilee clip instead, no fountain, but can still see tiny bubbles and slight moisture coming through top of pipes (both feed and return) - i'll be replacing both. cheers
 
TAABVW said:
To be clear, the one way valve on the fuel feed (starting as close to the pump as reasonable possible) is to help diagnose where air is leaking in. It's not intended as a permanent fix. It will give you one of three results:

1. If the leak is between the tandem pump and the one way valve then the car will start and very little air will be seen in the clear fuel return pipe. This is because the one way valve will prevent the head difference between the leak site and the level of fuel in the tank from sucking air in at the leak site.

2. If the leak is between the one way valve and the tank then the car will start and lots of air will be seen in the clear fuel return pipe (or the engine will splutter to a stop). There will be no air between the valve and the pump hence it starts, but then air will be drawn in through the leak site further upstream. This will be bled out by the pump and seen in the return, or it will stop the engine.

3. If there is still a very poor or no start then the leak is in the pump.

Given the history of the problem you have described, I would suggest that the original pump was probably not at fault and it was an air leak in the feed pipe all along. You have disturbed this air leak site in the fitting of the new pump making it worse. This gives a clue as to which fuel feed pipe to replace first. philward suggested the feed pipe an unlikely source of the leak (while acknowledging it's an easy fix) but I have found the more they are pulled off and replaced, the more likely they are to leak. I once spent about a month (not full time) chasing around a slow starting issue. I clamped up that feed pipe nice and tight every time I refitted it but that's what it turned out to be. If the pipe "clunks" into position when you push it on you might think that's a nice fit, when it's actually indicating the rubber has gone hard and is no longer malleable enough to seal.

Answering some of your questions directly:

Your plastic tube clearly shows air and fuel coming out of the pump, therefore the feed pipe is the other one.
The lengths of the pipes are irrelevant provided they are air free.
You don't need a clear pipe on the fuel feed necessarily, just the return to see what air the pump is dealing with and when.
Any 8mm diameter diesel fuel pipe will do for testing and indeed replacement if necessary.
If the leak were at the filter housing it would likely leak diesel out given the filter housing is low down in relation to the pump level. This kind of (no diesel out) leak tends to be high up in elevation hence starting near the pump.
philward suggestion of brimming the tank and parking on a (steep) downhill, or car service ramps might well yield results in the form of a physical diesel leak. You would have to go round and carefully check every joint in the feed line (perhaps using kitchen paper) for signs of diesel damp.

I have had about 12 audi fuel delivery systems (not all A2's) in many pieces and adapted them to run vegetable oil over a period of about 20 years. The oil hardens the rubber hoses quicker than diesel and the extra viscosity of the oil creates a bigger vaccuum in the feed pipe than diesel (which is why i fit a lift pump at tank level). Thus I probably have more experience in this area than many. I apologise if this experience might be clouding my assistance.
Click to expand...
One or two points I would seek clarification on.

1. A NRV is positioned before the leak and the tandem pump then the engine will start better because fuel does not return to the tank. However, the bubbles will increase because the NRV increases the pipe losses between the tank and the pump increasing the negative pressure at the leak.? A way to simulate and test would be to but a brake clamp on the fuel line, tighten up slowly and see if the bubbles increase.

2. If the leak is between the NRV and tank, the fuel in the line has +ve pressure until the pressure becomes negative due to pipe losses and filter losses. On balance I would say the bubbling would remain roughly the same? These are the leaks that I indicated could result leakage onto the drive. Most likely to occur around the fuel filter or at the tank. But such a leak would not cause starting problems until the tank was empty.

3. A NRV will not diagnose the pump? A known good pump or a know good fuel supply is required to diagnose the pump.

The point regarding hard pipes clicking is a good one. They also don't clamp up. This is the reason why I thought the author would have noticed bad pipes. It is possible and an easy fix.

I note your diesel experience and my experience is primarily petrol engines. However this is fluid mechanics that apply to both, what are your thoughts on the points I raise?
 
Nod,

In order not to confuse Johny I would point out that the reason you got a "fountain of fuel" from a leak site is because you have fitted a rather effective lift pump which Johny doesn't have. Also the prefilter you mention must be near the tank, upstream of the lift pump.

Phil,

1. The NRV's I use are for viscous fuel. They take minimal pressure to open and even less to keep open. Your point is valid but I have never found this to be so. The brake clamp idea would work.

2. When running all the fuel feed pipework will be under negative pressure. When stopped you would get a small positive pressure in the fuel feed pipework where the pipework is below the level in the tank. However this might not be enough to cause an external fuel leak. Put another way, it sucks much harder than it blows!

3. Fair comment but if the NRV rules out all the fuel feed system except the pump and the pipe connection to the pump then it leaves little else other than the pump. Granted though, not proof.

Trevor
 
An NRV will simply prevent fuel flowing back to the tank; it should have no effect on pressure. The pressure between the tank and the pump will be negative (except for early TDis, which had an electric pump at the tank). All Lupo diesels had an NRV between the tank and the pump,

RAB
 
As others, not a diesel man, but could the return line be blocked, so the air cannot be separated from the flow through the pump?
Mac.
 
PlasticMac said:
As others, not a diesel man, but could the return line be blocked, so the air cannot be separated from the flow through the pump?
Mac.
Click to expand...
The pump has two pumps, a fuel pump and an air (vac) pump. Air and fuel should never mix. Either the pump is pumping air into the fuel (seal failure) or a pipe connection is leaking air in. Once air is in it only separates at the fuel tank.
If the leak is in the pipe before the pump drawing fuel from the tank causes the pressure inside the pipe to be less than atmospheric. Atmospheric pressure forces air in and the pump draws air and fuel. Thats where the air stays, in the fuel, until the fuel is returned to the tank.
I think, I'm going to take one apparat and find out.
 
Last edited:
philward said:
If the leak is in the pipe before the pump drawing fuel from the tank causes the pressure inside the pipe to be less than atmospheric.
Click to expand...
The pressure between the tank and the pump inlet will always be less than atmospheric - the fuel is being sucked by the pump. As such, any loose connections will allow air to be sucked into the fuel.

RAB
 
philward said:
The pump has two pumps, a fuel pump and an air (vac) pump. Air and fuel should never mix. Either the pump is pumping air into the fuel (seal failure) or a pipe connection is leaking air in. Once air is in it only separates at the fuel tank.
If the leak is in the pipe before the pump drawing fuel from the tank causes the pressure inside the pipe to be less than atmospheric. Atmospheric pressure forces air in and the pump draws air and fuel. Thats where the air stays, in the fuel, until the fuel is returned to the tank.
I think, I'm going to take one apparat and find out.
Click to expand...
Interesting you mention the seal, do you mean the gasket?
Could it be a simple case that I've not torqued it up enough using my brand new wrench?
It's not a cheapo 14.99 thing but then again it's not a mega expensive brand.
Also the quality of the gasket might be suspect.
I'm willing to remove the pump again after trying the new fuel pipe and NRV method.
Thanks again for input guys.
 
johnyfartbox said:
Interesting you mention the seal, do you mean the gasket?
Could it be a simple case that I've not torqued it up enough using my brand new wrench?
It's not a cheapo 14.99 thing but then again it's not a mega expensive brand.
Also the quality of the gasket might be suspect.
I'm willing to remove the pump again after trying the new fuel pipe and NRV method.
Thanks again for input guys.
Click to expand...
No I'm going to take a pump apart and see what is in there. An internal seal or gasket failure is just speculation on my part. The inlet pipe is a good option but I have always assumed that you would have been aware if it was damaged or not fitted correctly so that has always been my second option.
Have you considered fitting clear pipe for a test before the pump? If it is drawing bubbles through from the fuel filter you will see them. If they disappear at the return you have your culprit; the bit of pipe before the pump.
 
I think Phil means the seal between the vacuum and fuel sides of the pump - there must be one. On a new pump that should be unlikely to fail. Much more likely to be a poor joint/failing rubber tube between the tank and the pump.

RAB
 
philward said:
No I'm going to take a pump apart and see what is in there. An internal seal or gasket failure is just speculation on my part. The inlet pipe is a good option but I have always assumed that you would have been aware if it was damaged or not fitted correctly so that has always been my second option.
Have you considered fitting clear pipe for a test before the pump? If it is drawing bubbles through from the fuel filter you will see them. If they disappear at the return you have your culprit; the bit of pipe before the pump.
Click to expand...
In a video I shot it shows both fuel lines using clear tubing, I thought I had uploaded it to the YouTube channel I have created.
I'll check now
 
Just rewatched your video in post #17. There is clearly air in the fuel feel pipe before the pump. I'm sure the pump is fine. It's not at all surprising it won't start with that much air going in the pump.

Trevor
 
TAABVW said:
Just rewatched your video in post #17. There is clearly air in the fuel feel pipe before the pump. I'm sure the pump is fine. It's not at all surprising it won't start with that much air going in the pump.

Trevor
Click to expand...
Second vote for this conclusion, I have a pump apart on the bench and have watched the video again.
 
Just for clarity, the video where there a lot of bubbles is when I fitted the pump the first time.
Since refitting there was way fewer bubbles but some bubbles none the less.
 
Johnny have you tried the arrangement as suggested above as used for diesel purge? It’s really easy to do and determines if the issue is in your pump or fuel lines etc.
 
An update.
still FUBAR.
Strong crank, half a tank of fuel, cant get hold of new fuel pipes so having to use old ones.
Think its about time I threw a bag over this, ive not got the money to take it somewhere even if I could.
Ill take off the new BOSCH pump and put the old one back on, then I can post it for sale on here in the near future with all the details.
Thanks lads for all your input and suggestions, its much appreciated.
 
Hi Johnny, it’s worth disconnecting the two fuel pipe at the cambelt side of the engine and sticking them in a container of diesel. That rules out the fuel lines, filter tank etc. it’s literally a 10 min job. I did it at the the weekend doing a diesel purge
 
George Hogg said:
Hi Johnny, it’s worth disconnecting the two fuel pipe at the cambelt side of the engine and sticking them in a container of diesel. That rules out the fuel lines, filter tank etc. it’s literally a 10 min job. I did it at the the weekend doing a diesel purge
Click to expand...
Wish you had taken photos George.
 
You must log in or register to reply here.
Back
Top