Join Date: Aug 2020
So I have been chasing a "vapor lock" issue with my 1990 E350 since about 2016.
Long story short, carringb is correct - it is the fuel pump, not the engine compartment.
In 2013 my OEM fuel pump burned up in my 1990 E-350 RV. It has a 460 CID engine with fuel injection.
The first symptom was the Fuel Pump Relay burned up. Had this replaced, but about a month later the fuel pump (in-tank) itself burned up. It actually melted metal on the pump itself. When this happened, it also burned up the inertial cutoff switch and the relay.
Had these replaced. They replaced with an Airtex pump, but put the wrong one in and so my fuel gauge ran backwards. Got Airtex to send a replacement (E2060S), and had the wrong pump replaced with the right one.
In 2016 I had an overheat problem on an 11-hour trip. Limped home with the heat going full-blast in the cabin and in the house. I replaced the radiator and thermostat, and installed a digital temperature sensor and gauge for both the thermostat housing and the E4OD transmission test port, just for good measure.
Ever since then, when there were high ambient temperatures, if I was running down the road everything was usually fine, but if I slowed down due to traffic or getting off the interstate onto a surface street the RV would die as if it was out of gas. Trying to start it would result in it sputtering like it was trying to catch, but it would not. Sometimes it would start, barely, but die on idle very quickly.
I assumed that there was some "cooked" component on the engine that had been damaged during the overheat in 2016. So, based on lots of "internet" advice on what could be wrong, I spent a lot of time and money, and eventually replaced every electronic thing on the engine:
Distributor (and thus PIP sensor)
Ignition Control Module
Engine Temperature Sensor (this actually was bad, but unrelated to shutoff problem)
Air Charge Temperature Sensor
Idle Air Control Valve
Fuel Pressure Regulator
EGR Control Solenoid
Plugs and Plug Wires
High Pressure Fuel Pump (mounted on frame rail)
Catalytic Converter (this actually was bad, but unrelated to shutoff problem)
I even found one of the old, vintage EEC-IV computer diagnostic/recorder systems that you would hook in between the computer and the man harness to intercept, read, and record signals to and from the computer, but never got that far into using it.
Ultimately, I began to suspect vapor lock. Vapor lock is not supposed to be possible on fuel injected systems, due to the return-circuit the fuel follows. Basically fuel is pumped to the engine at high pressure, the engine drinks what it needs, and the excess fuel is routed back into the fuel tank. So fuel does not spend much time sitting in a hot fuel rail to boil.
Still, due to the temperature-related component, and the feeling of being out of gas, I suspected vapor lock. My initial assumption was that the fuel rail, which runs a circuit around the top of the intake manifold, was getting hot enough to boil the gasoline. Especially since now ethanol gasoline is commonplace, and it has a lower boiling temperature than the gasoline that was commonplace when this engine and vehicle was designed.
To test this theory, I installed a GlowShift digital pressure gauge on the Schrader valve on the fuel rail. Since it was a duel-pressure gauge I also installed a sending unit on the oil pressure pick-up, to boot.
Unfortunately, you have to wait until summertime to catch this in the act. However, as luck would have it, I came back from another 11-hour long haul (same place I had the overheat with back in 2016), and when I got home, I just pulled into the storage facility parking lot and it died, and I could see on the fuel pressure gauge I was only getting 12-20 PSI on the rail. Normally it is 30-40 PSI. So, I had finally caught it in the act. This was winter of 2019.
At this point, I still did not suspect the in-tank fuel pump, since, after all, I had replaced it in 2013. So, my suspicion fell on the fuel filter and the high-pressure, frame-mounted fuel pump.
You see, Ford vehicles of this era have TWO fuel pumps. One is in the tank, which is low pressure, and one is on the frame under the driver's seat. That is the high pressure pump that pumps it up to the pressures needed for fuel injection.
So, I replaced the fuel filter (I cut it open and it was pretty gunky inside), and I replaced the high-pressure pump.
But, the problem happened again!
This time, I also installed digital volt/ammeters. These are "slip on" type meters that slip over the wire being measured. I installed one at the power lead going to the high pressure pump, and another at the harness for the inertial cut-off switch. I could not get to the fuel pump on top of the tank to attach the meter for just the in-tank pump, or I would have (and I am still going to see if I can hunt down its harness so I can get the power just for the in-tank pump). But by looking at the high-pressure pump I could see how much power it was pulling (and NOT just whether it had power or not), and by looking at the inertial cut-off harness I could see the combined power pulled by both the in-tank and high-pressure pumps combined. This should always read more than the high-pressure pump alone.
So now I could see, via digital gauges, my fuel pressure on the rail, and the operational status of the two fuel pumps. In addition to this, I tightly taped (by winding electrical tape around it) a thermocouple to the return side of the fuel rail, right by the transmission where it leaves the engine compartment. This ran to a digital temperature gauge I just left sitting on top of the dog house.
What I discovered is that when you have been running with high ambient temperatures (high-90Fs), for some time (1 hour plus at interstate speeds), the fuel in the rail would reach about 125F. If you stopped the vehicle and let it sit for about 10 minutes, the temperature on the rail would go to about 145F. I have used an infrared thermal gun to measure the temperature of the fuel tank itself from outside/under the van, when stopped (obviously), and it measures about 110F.
Bear in mind that ethanol gasoline begins to boil at around 175F - at normal atmospheric pressure. This is important.
Finally, with my RV now wired up like the Space Shuttle, I finally caught it in the act. And sure enough, I was getting 12 PSI on the fuel rail, and no current at the in-tank pump (the current at the inertial cut-off harness was the same as the current at the high-pressure pump, telling me the in-tank pump was not running).
I then confirmed this by using a mechanic's stethoscope to listen to the tank while my son turned the key to on/off. Sure enough, no sound in the tank. I then disconnected the fuel inlet to the high pressure pump and stuck it in a container while my son hit the key again - sure enough, no fuel coming from the fuel tank.
My in-tank pump was not functioning.
Here is what I think is happening. When the in-tank pump fails, the high-pressure, frame rail-mounted pump is strong enough to actually suck fuel all the way from the pump, through the dead in-tank pump, and pump it up to the engine, no problem.
But remember, the high-pressure pump is pumping a pressure head of about 30-40 PSI. This means it must be sucking from downstream 30-40 PSI if it's doing all the work itself with no assistance from the in-tank pump. Not to mention that the in-tank pump is now acting like an obstruction.
The boiling temperature of liquids is directly proportional to pressure. This is why you can't boil potatoes to cook them on Mt. Everest. The lower atmospheric pressure there (only about 4.8 PSI as opposed to 16 PSI at sea level) means that water boils there at only 160F, instead of 212F at sea level. So about a 10PSI difference in pressure lowers the boiling temperature of water by about 50 degrees!
So if ethanol starts to boil at 175F, and the fuel temperature is 125F, that is a 50 degree difference. If the high pressure pump is pulling a 30+ PSI negative pressure upstream of itself, my theory is that this is enough of a pressure delta at 125F to boil ethanol gasoline. I have not had any luck finding pressure/boiling temperature data for ethanol gasoline (gasoline doesn't have a singular boiling temperature since it is made up of many different volatiles), but that' my bet.
My bet is that the high-pressure pump is sucking enough of a negative pressure on the hot fuel to cause it to flash to vapor (boil). When this happens, the high pressure pump cavitates and can no longer push any liquid fuel to the engine.
If you stop and wait 20 minutes or so, the engine fires right back up and everything seems normal - until it does it again.
The internet is full of posts like this where people have shutdown problems with high ambient temperatures. No one ever comes back with a fix. I suspect this is for a couple of reasons. First, these vehicles are now well beyond their normal service life. 1990 is 30 years ago as of this writing. So many of these vehicles simply aren't around anymore. A lot of the internet threads on this problem date to 2006 or later. The second problem is that the problem is highly intermittent. It only happens with high ambient temperatures. And when things cool off, everything runs fine. So a mechanic is not likely to ever see it when it's in the shop nice and cool. They fire it up and everything runs like a Swiss watch.
So, having finally, after 3 years of chasing the shutdown-while-hot problem, I nailed it down. The in-tank pump fails, and the high-pressure pump is able to carry on under most conditions, thereby masking the failure of the in-tank pump.
So, about a month ago I had my shop replace the E2060S Airtex pump that was replaced back in 2013. They replaced it with another Airtex E2060S pump. You can't get Motorcraft replacements anymore.
I had them give me back the Airtex pump that was "dead". I hooked up 12V power to the terminals on the top of the pump assembly, and sure enough, the pump did not work. I then wiggled the spade terminal connections at the pump motor itself, and suddenly the pump spun to life! There was nothing wrong with the pump. There was simply a bad electrical connection at the connectors at the motor! Boy was I pissed. It cost $250 for the replacement pump, and about $400 in labor to replace it when there was nothing wrong with it! But I figured, "Hey, the pump was 7 years old, I guess now I won't have to worry about it for a while).
Well, I got it back, and the second weekend I ran it down the interstate, it did it again! Got it home, it died in the driveway. Listened to the tank, and yup, dead pump again.
So I finally searched the internet high and low and think I found one of the last remaining New Old Stock (NOS) Motorcraft fuel pumps. It is being installed Monday. I hope this solves the problem for another 20 years.