Before my fuel pump problems the car would start within 1 or 2 seconds of cranking. After I replaced the fuel pump it took maybe 10 seconds of cranking to fire up. I expected that because there was probably a lot of air in the fuel lines. The next time, maybe an hour later, it started normally within 1 or 2 seconds. Since then the pattern is: if the engine has been started within the last 48 hours it fires up within 2 seconds of cranking. If it has been more than 48 hours since the engine ran it will crank and crank for 8 or 10 seconds and then 2 or 3 of the cylinders will "catch" and it will cough and sputter for 1 second, then runs fine. I convinced myself I stirred up something in the gas tank during the pump change so I replaced the almost new fuel filter. I thought it got better but now I suspect I had confirmation bias because it had to crank 10 seconds this morning.

Fuel pressure, per the code reader and Torque app, is solid at 53-54 psi as soon as the key is in the accessory position. Could there be air in the system that the fuel pump has to compress on start up? Is there a burp procedure for the pressurized fuel system?

 

Hi dh53,

You are definitely having some issues there with your car.....sorry you are having such problems.
Sounds a bit like it is drawing some air back into the system from somewhere.
There is often a valve on the fuel rail that you normally would attach a test gauge to, but you could operate the valve quickly manually to see if you can purge any air out of the system.
The question becomes, where is the air getting in?
Is it being gradually drawn into the line with tank vacuum, if so is it entering in at the top end and starving the injectors right away or is it entering in lower and then working its way up the line to the highest point.
I recall you mentioning you have changed the fuel pressure sensor on the fuel rail, is that new unit letting some air gradually seep in?
Did you remove the rail to get to that sensor and now got a slight seal problem to one of your injectors?

 

dh53, like h2o2 said, it sounds like you are getting a bit of air into the system. Now, this could be from a slight leak in the shrader valve (may or may not be in your car as I know they got rid of this valve around the time of building your car) or you could have a fuel injector that is now quite sealing like it should. Could be a number of reasons. What I want you to try is when you have let the car sit for over 48 hours, roll the key to the RUN position and let it sit for 3-5 seconds. You may be able to hear the pump start and then stop during this time. The pump is priming the system, getting an initial pressure in the line. Once the pump stops, now, roll the key to the RUN position. Does the car start within the 2 seconds? If yes, then we have confirmed a loss of fuel line pressure. The trick is figuring out where the leak is coming from. This is where it can be so many things and the leak is so minor, is it worth hunting down. In the past, when this condition occurs, it is the fuel pump that is the culprit. the check valve in the fuel pump starts to leak and result in a no start condition.

 

Thanks Mark, Thanks Chris. I do not have the schrader valve on the fuel rail, must have been deleted before VIN E59xxx. I did take the fuel pressure sensor off when I was trying to diagnose the fuel pump problem but I don't think I replaced the old one... I'll ask the folks at SNG Barratt, surely I will get invited to their company parties after all the parts I have ordered. I have had the fuel rail and the lower intake manifold out just before replacing the fuel pump (thinking it was a misfire problem so swapping coils and injectors) so it could be any of those gaskets and o-rings. The car will be sitting again this week waiting for a rear axle so I can try Chris' test.

Interesting that you mention the check valve in the fuel pump...the tank has been near empty the whole time I have had this issue. If the tank was full and the fuel pump was fully submerged a bad check valve couldn't suck air into the fuel line right? That is a cheap and easy test to do. Also if it is leaking out under pressure maybe I should add a fuel sniffer to the toolbox...

 

I tried Chris's test yesterday. The car had been sitting 4 days without starting. I put the key to Run, waited 5 or 6 seconds, then rolled the key to Start. The engine fired up strong almost immediately.

Question: Is the system designed to hold pressure in the fuel rail? Meaning, I expect about 50 psi at the fuel rail during normal operation. If the car sits for a few days should I expect a normally operating car to still have about 50 psi in the fuel rail? Or should a car that has been sitting for days without starting bleed off that pressure? I will try and get a voltmeter on the fuel rail pressure sensor to see if I get any insight. Unfortunately the OBD-II reader with Torque takes longer to connect to the car than this problem exists so I can't see what is happening via a code reader.

 

FPS?

 

dh53, the car should be able to hold the fuel rail pressure for a bit. But, keep in mind, your car is on the older side. So, component are getting old and they may not seal quite as well. So, having a fuel rail bleed down over a few days, I would consider that "normal" for an older car. From seeing what other people have dealt with, this is simply your fuel pump leaking internally, resulting in the lost pressure. You may have to remember to pause for just a second in RUN before starting the car. Granted, if it gets much worse, then I would be looking at the fuel pump. That is unless you can spot a fuel leak somewhere else on the car. It could always be one of the injectors leaking. But that would be a hit and miss type of adventure to find that unless it got bad enough that it was causing a cylinder to run rich.

 

I hate to think it is the fuel pump itself because I replaced that a few weeks ago with only 100 miles on the car since. This problem started as soon as the new fuel pump went in. It was an eBay pump so could very well have been a factory reject for a bad check valve. More likely is I messed up some seal or I ripped an o-ring or gasket somewhere.

I'm trying now to get the output of the fuel pressure sensor with a voltmeter so I can see what happens before the Torque App boots up. I think there will be knowledge gained from knowing the fuel pressure instantly, not averaged by the ECU then serialized and sent to OBD-II and read in by more software that does who knows what with averaging over data bits and time. Also thinking about taking it to a garage to use a sniffer to find leaks, but finding it myself is more satisfying.

 

Hey dh53,
Just a left field thought on a Friday.......could you possibly install in a one way check valve before or after the fuel filter?
It won't stop the pump pressure being lost at the pump, but it could technically hold the fuel pressure within the main length of fuel line.
Line pressure might still immediately drop as soon as injectors fire as there is no compression factor in the retained liquid, so might prove not to be a real world fix unless there were a small reservoir that has a trapped bell of air that could maintain the line compression briefly.

If you had an old fuel filter that you could cut the spigot off, then flexi-hose a check valve in without having to do any factory pipe mods.
You might just have some better access there to make a non intrusive and reversible modification.......

 

Good thinking Mark. This probably doesn't translate well outside the US but I can have a check valve designed to work with gasoline for about $4 plus $10 shipping and have it Saturday since it is only Thursday night here. If it were Friday I wouldn't get the part until Monday which is Tuesday for you... So I could potentially get the part delivered 3 days earlier in the US by the Theory of Relativity? That must be why I didn't do well in Physics.

 

If you find the pressure collapses too quickly when the injectors fire before that pump can backfill, you could also add a T joint in your rubber hose mod, then add a short vertical riser tube that is blanked off at the top end to create the small air pressure reservoir.
(Like you used to do over there with your plumbing risers to stop the pipes getting hydraulic shock with fast action flick mixers and other pipe hammering sounds).
You might need to play with final length to find the sweet spot to provide enough coverage for pump lag versus the possible reduced responsiveness of real time fuel delivery demand changes in the system for acceleration.
You never know, we might be on a winner here.

Thumbs up to your theory of relativity.....cracked me up!
Everyone says coming to New Zealand is like stepping back 20 years.
To my mind that puts my 2005 X-type 4 years still ahead of it's time....loving it!

 

dh53, if you want to install that check valve to prove it is the fuel pump, then I would just remove the fuel filter for a few days, install the check valve in its location, start up the car to get things up in pressure, then let the car sit for 48+ hours. See what you have then. If the car fires right off, then you somewhat have proven there is a bad seal somewhere in the fuel pump. If the problem persists, then either the pump can't maintain the pressure fast enough with the extra check valve or the problem is up near the motor. This is where you could possibly set up the rig to have a pressure port to see what the fuel rail pressure is before you start the car. That would help answer the question about the engine maintaining the pressure.

 

Connected the code reader and started the Torque app and logged fuel pressure, it is obvious how bad the leak is. Turn the key to on (engine not started), fuel pressure jumps to about 52 PSI then decays immediately. I can't find much info about how long the pressure is maintained in a fuel rail but I found something saying if fuel pressure decreases more than 10 PSI per hour you have a problem. I'm losing 10 PSI in 20 seconds.

I should have a chance to get the inline check valve installed in the next day or two and see if that points to a bad check valve in the new pump but I am leaning that way. I do not want to do that fuel pump job again.

It is curious that with such a rapid loss in pressure the car will still start fine 24 hours after it was last running, it takes longer for any problem in starting.

 

I cut the fuel line between the tank and the fuel filter and added an inline check valve. Before I put the check valve in I tried to run gravity fed gas through it. Flowed fine in one direction and not at all in the other so the valve is good. Repeated my pressure measurements and... it loses pressure slightly less quickly. Fitting a curve to the data says I will bleed off to zero pressure in an hour instead of 20 minutes. Doesn't seem like much help.

Stuck injector? This is a return-less system so if pressure isn't bleeding back into the tank it is leaking into a cylinder or leaking to atmosphere. If that much fuel was leaking that quickly to the outside air I think I would smell it.

 

dh53, you are not looking at that much fluid to raise the pressure. So, a drop or 2 could be the difference between 60 psig and 30 psig. So, really not that much. I would not expect it to be noticeable. Most fluids are not really compressible. Gas has that same property. What allows you to see a gradual pressure drop is the small sections of rubberized fuel line.

 

I just had a CEL for P0193 today, fuel rail pressure sensor circuit high, while I was futzing with the fuel pressure sensor and trying to power it separately from the car and trying to isolate the pressure drop with clamping the fuel line in different places. For today I will assume the P0193 was related to my tests and not what is wrong with the car.

Today after not cranking over the car in 24 hours I turned the key and it fired up immediately. Will try again this week after 2 or 3 days of not running the car.