Additive For Sticking Fuel Pump Anti-Drainback Check Valve.

[crainholio]

Anybody had luck with fuel additives to free up a sticking in-tank fuel pump anti-drainback check valve?

Vehicle info:
I've been troubleshooting a long cranking time issue on our '05 Mercury Mariner (Ford Escape) 3.0L V6, and have it narrowed down to the fuel pump drainback check valve. The ECM reports several useful PIDs including fuel rail pressure, desired fuel rail pressure, and fuel pump duty cycle. It's a returnless system with no pressure regulator, using a pressure sensor on the fuel rail and pulsewidth modulated fuel pump in the tank. The fuel pump is a Delphi replacement assembly installed February 2016 when the factory pump failed.

Symptoms:
I noticed around June timeframe a long crank time of ~5 seconds, which only happens after the car has been sitting for ~1hr or so. ECM has been reflashed with updated firmware per an old TSB for this issue, no effect.

Troubleshooting:
Scan data shows the ECM runs the pump at 75% duty cycle continuously when the key is in the RUN position, engine not running. Desired fuel rail pressure is 80 inHg, and it reports 80 inHg actual. There is also an audible hiss from the pump for ~1sec on initial key turn to RUN, but I can't scan during that initial turn (no data acquisition)...the additional hiss suggests a 100% duty cycle initial priming run.

However, with the fuel pump fuse removed, I see rail pressure drop at a rate of ~20 inHg per 10 minutes. Very slow dropoff, and by 20 minutes it's reporting ~38 inHg. Haven't been able to let it sit for a full hour as yet.

Yesterday, I pinched off the rubber fuel line running from the tank to the filter canister with a plastic clamp. With the fuel pump fuse removed and key in RUN, the ECM reported the pressure held steady at 76 inHg for >30 mins testing time. This rules out the injectors as the leakage source, it's likely the fuel pump drainback valve.

So...before I open up the fuel tank and take Delphi up on their lifetime warranty, I figured I'd experiment with additives. TC-W3 at 640:1 was loaded on Monday and it'll be burned by the weekend. Any other recommendations appreciated.

 

[StevieC]

Unfortunately I don't think any bottle of magic is going to fix this one.

 

[SOHCman]

1oz per 5 gallon TCW-3?

 

[crainholio]

Originally Posted by SOHCman
1oz per 5 gallon TCW-3?


Yes, the 640:1 ratio.

 

[crainholio]

Originally Posted by StevieC
Unfortunately I don't think any bottle of magic is going to fix this one.

I'm also skeptical. The problem is a low-level nuisance, which lends itself to experimentation. Replacing the pump isn't a horrible chore, Ford put an access plate under the driver side rear seat base. If I can't cure it with TC-W3, MMO, Techron, Redline SI1, or whatever I'll swap the pump assembly under warranty before the weather gets cold.

 

[StevieC]

In that case I vote for Redline Si-1 for the trial. If it's gunk holding it back then that would be my choice because of it's cleaning ability for gunk. Not necessarily carbon but greasy/oily gunk

 

[rubberchicken]

TC-W3 ? As a long time boater and 2 stroke Waverunner owner, I was confused why you would try this instead of something like Techron, although if you had it on the shelf I guess it would not hurt. I was not aware that TC-W3 oil has any special cleaning abilities. Could you enlighten me ?

 

[crainholio]

Originally Posted by StevieC
In that case I vote for Redline Si-1 for the trial. If it's gunk holding it back then that would be my choice because of it's cleaning ability for gunk. Not necessarily carbon but greasy/oily gunk


Fair point, but the fuel pump discharge port submerged in the gas tank is not likely to have greasy/oily gunk inside. I've seen these check valves on Walbro gerotor and Delphi turbine pumps, usually just a metal ball in a plastic cage with a spring. The fuel pump pressure blows the ball off the seat against spring pressure, and when the pump shuts off the spring pushes the ball back onto the seat to stop reverse flow.

 

[crainholio]

Originally Posted by rubberchicken
TC-W3 ? As a long time boater and 2 stroke Waverunner owner, I was confused why you would try this instead of something like Techron, although if you had it on the shelf I guess it would not hurt. I was not aware that TC-W3 oil has any special cleaning abilities. Could you enlighten me ?


This seems to be the discussion that started it all: https://www.ls1.com/forums/f48/been-testing-oil-91206/

There's also a few threads here on the subject available by search function.

 

[StevieC]

Originally Posted by crainholio
Originally Posted by StevieC
In that case I vote for Redline Si-1 for the trial. If it's gunk holding it back then that would be my choice because of it's cleaning ability for gunk. Not necessarily carbon but greasy/oily gunk


Fair point, but the fuel pump discharge port submerged in the gas tank is not likely to have greasy/oily gunk inside. I've seen these check valves on Walbro gerotor and Delphi turbine pumps, usually just a metal ball in a plastic cage with a spring. The fuel pump pressure blows the ball off the seat against spring pressure, and when the pump shuts off the spring pushes the ball back onto the seat to stop reverse flow.


Good point.

 

[Trav]

I would try Berrymans Chemtool, a full bottle in a full tank. It is far better with these types of deposits than PEA based cleaners.
If you get it free then use Redline SI-1 2-3oz per tank to keep it lubricated.

New cans..

 

[crainholio]

Originally Posted by Trav
I would try Berrymans Chemtool, a full bottle in a full tank. It is far better with these types of deposits than PEA based cleaners.
If you get it free then use Redline SI-1 2-3oz per tank to keep it lubricated.


Berrymans aerosol has always done well for me, never tried the gas additive. I'll grab a can and load it on the next refill, thanks.

 

[SOHCman]

Originally Posted by rubberchicken
TC-W3 ? As a long time boater and 2 stroke Waverunner owner, I was confused why you would try this instead of something like Techron, although if you had it on the shelf I guess it would not hurt. I was not aware that TC-W3 oil has any special cleaning abilities. Could you enlighten me ?

TCW-3 is an excellent lube, I thought a lube might help the valve not stick. Cleans the combustion chamber as well. win win.

 

[crainholio]

Originally Posted by SOHCman

TCW-3 is an excellent lube, I thought a lube might help the valve not stick. Cleans the combustion chamber as well. win win.


I figure it's most likely that the check valve seat surface is imperfect either due to corrosion or buildup. I ran a tank of MMO mix at the label rate of 4oz/10gal, can't say it had any measurable effect on the drainback. TC-W3 is up at bat now, to be followed by Berryman's B-12.

It's nice having scanner data showing the actual fuel rail pressure vs. ECM desired pressure, much easier than a pressure gauge on the fuel rail.

The bad news is that even if this experiment does prove successful...it will be impossible to know which additive did it. Refilling at 1/8 tank (16 gal spec capacity) still leaves 2-3 gallons of the prior additive on board. And there's the possibility of lag between the introduction of the additive and it's effect on the problem.

The check valve is directly in the flow path of the fuel, as well as submerged in a sloshing pool of gas while the vehicle is moving. So it's getting maximum exposure to whatever I put in the gas.

 

[crainholio]

Update: TC-W3 seems to be correcting the problem a bit. I haven't added the can of Berryman's as yet. Wife reports the car is back to nice quick starts, and I've also been unable to reproduce the long crank time. Tried the usual first thing in the morning drama, fired right up after ~1-2 sec of cranking. Tried it after sitting for ~1 hour, again fired right up.

However, it's still losing pressure as shown in my data below from last night's test session using both a scanner and mechanical gauge. I'll continue TC-W3 for another week and re-test. Seems like the fuel pressure dropoff may not have been the cause of the long cranks.

Results: (note: all times are EDT):
***************************************************
8:08pm: 78 inHg actual, 80 inHg desired (Slightly below 40 PSI on mechanical gauge)
8:09pm: 71 inHg actual... (35 PSI on mechanical gauge)
8:10pm: 62 inHg actual...
8:11pm: 54 inHg actual...
8:12pm: 48 inHg actual...
8:13pm: 44 inHg actual...
8:14pm: 40 inHg actual…
8:15pm: 35 inHg actual…
8:16pm: 32 inHg actual…
8:17pm: 29 inHg actual… (Low end of 15 PSI mark on mechanical gauge)
8:18pm: 25 inHg actual…
8:19pm: 23 inHg actual…
8:20pm: 21 inHg actual…
8:21pm: 19 inHg actual…
8:22pm: 18 inHg actual…
8:23pm: 17 inHg actual…
8:24pm: 17 inHg actual… (Slightly below 10 PSI mark on mechanical gauge)
8:25pm: 16 inHg actual…
8:26pm: 15 inHg actual…
***************************************************

 

[crainholio]

Update: after trying Berryman's B-12, Amsoil PI, Red Line SI-1, MMO, and TC-W3 with no measurable effect on the fuel pressure leakdown...I pulled the fuel pump assembly and brought it back to Auto Zone for a free replacement under Delphi's lifetime warranty. The new unit holds pressure quite well, with pressure rising from the 80 inHg at initial hot shutoff to 97 inHg in the first 2 minutes, then 86 inHg at shutdown + 10 minutes, then at 54 inHg at shutdown + 20 minutes. I presume that the initial rise in pressure (fuel pump fuse removed) is heat soak at the fuel rail.