Kiln_Red

Regretfully, I have to agree. Porsche is, and always has been, a pioneer brand in the auto world. They make guinea pig cars. They really pushed the envelope with the 928 and took it to unchartered territory in so many of its design aspects. The 928 certainly isn't a perfect car and many of its well-intended design efforts need a careful re-tracing. This is why guys like Porken and Constantine exist. In fairness, most of what makes up a 928 is very robust - much more so than an ordinary automobile of its era. More so than most cars being made now even.

AO

Well a warm tank certainly adds to diagnosis. Still surprised, but makes sense.

So, how would an in-tank pump have resolved this? It's not like the in-tank pump adds to the flow or pressure, per se. I believe the in-tank pump's primary purpose is to prime the external pump. I understand that any fuel pump will warm up the fuel to an extent, but I'm not so sure that the in-tank pump will minimize this. And if you already have warm fuel... you see where I'm going?

Pretty hard to replicate identical conditions.

I will say, however, after many years on this forum and being in the 928 community, this is the first time I've every heard of this issue on a stock 928. Tuomo had an issue some time back, but I think due to different pumps, flows, pressures, etc.

FredR

Andrew,

The in tank pump does add pressure that is why it works. We are not talking huge amounts of presssure here- just enough head to suppress cavitation- a concept that is difficult for most folks to understand but something that is vital to pumped liquid systems that are at or near to their boiling point.

Once the liquid is at its boiling point the in tank pump has to add sufficient head to overcome the suction losses into the main pump and then satisfy the NPSHR [nett positive suction head required] for the pump in question. The head due to liquid level inside the tank is next to nothing thus why this priming pump is immersed inside the tank.

Do not let the test that Jim did confuse the issue- I suspect that what Jim did was what we call an "end of curve test". This means that the pump is allowed to flow with no resistance so it delivers quite an amount of fuel at little to no head. When it is connected to the main pump the delivery pressure will increase but the flow will drop to suit the demand of the main pump. Under such conditions I would think that the in tank pump will deliver at least 3 feet of head and maybe more. The only way to check this is to fit a pressure gauge in the delivery line or make a test rig with a valve in the discharge and monitor the flow/pressure as the discharge is throttled to the point where the flow is the rated capacity of the main fuel pump.

Rgds

Fred

dr bob

Interesting.

So for future reference and for those to whom this might happen in the future, the fuel in the tank will cool itself given the chance, by opening the fuel filler cap after you put out your cigar. There's pressure built up in the tank (the fuel is 'boiling'...), so do this slowly and hang on to the cap. I had one launch across the desert the first time I did it, and had to go find it. The fuel in the tank will cool as some boils off at the reduced pressure. You'll know that it's boiling in there by the burbling noises from the tank. Wait until all that's all done and for a little while longer, then put the cap back on and restart the car. The fuel will have cooled to somewhat less than ambient thanks to the rapid evaporation, to somewhere near its "dew point" if you will. From there the pumps will see liquid again, and you can proceed on until the fuel heats up again. If the tank level is pretty low, adding fresh cool fuel will help, not just from the cooling but from the added liquid head available to the pump.

Reminder that the vapor from boiling fuel is a HUGE fire hazard as it ejects from the tank and mixes with air. No sparks or other ignition sources please. Of course you shouldn't inhale mr president. Nice if there's a breeze to carry the vapors away.

BC

I wonder if this is what happened on my 88 the other day. 80-100 From northern LA, down through good traffic, finally hit Southern OC, traffic stopped, slogged through that until Oceanside.

Stopped, turned off the car at pump, filled up about half way (I don't fill up 100%) and the car would not start.

Rolled it into a spot, jumpered fuel pump, still no start, but pump was running. Waited about 3 minutes, and car started.

No AC.

FredR

I don't think you have to wonder too much [see my earlier post]. This type of problem will be exacerbated if the fuel pressure regulator does not hold tight [or you have a leaky injector] as a pressure drop will induce vapour formation. I am pretty sure that at some point of heat saturation the fuel in the rail starts to boil.

Rgds

Fred

dr bob

Possible but unlikely.

As Fred mentions, there's local heat-soaking in the fuel rails with the engine off. At the same time though, you were adding relatively cold fuel to the tank, which should pretty much guarantee at least initial liquid flow into the pump. The pump delivers a constant volume based on the back pressure presented by the fuel pressure regulator, with very slight differences from pipe friction. Fred and Jim C described this previously. Anyway, a constant-flow pump pushing liquid through the rails would pretty quickly clear any vapor present in the loop upstream of the pressure regulator.

There's always the chance that the pump itself was already hot, and so was vaporizing fuel just from the heat in the suction piping. But that's not much volume really, and the fuel is in a hose on it's way to the pump. So interesting but unlikely combinations of things could cause the symptoms you saw.

If it's any comfort, the in-tank pump solves at least the initial issue of boiling in the lines from the tank to the main pump suction.

jcorenman

Good advice, thanks! And besides reducing the temperature, "boiling off" the more volatile components of the fuel will also reduce the vapor pressure of what's left.

the flyin' scotsman

another reason to give up those cigars

great thread all!

hacker-pschorr

I don't think anyone has ever argued against the benefits of the in-take pump when it's working correctly in an otherwise stock 928.

The debate is choosing between the potential of the hose breaking starving the engine of fuel, or vapor lock.

Is the internal hose failure potential overblown?

dr bob

There's enough evidence of failure from list members to make looking at that hose a Good Idea. I'm not sure that there's a firm formula that can guide owners on when to replace that hose. Even though I preach about needing the pump working, my only proactive steps so far have been to stockpile the parts and pieces needed to replace the pumps. So far everything has been working OK, so the project moves to "when the tank is almost empty" .and. "there's lift time available in the garage." That will be sometime this coming winter I suspect. There's 3/4 of a tank in the car that I hope to be able to use up before the car gets put up for the winter again.

John Veninger

Wonder if the one in my GT is still in tack after all theses years

FredR

The internal pipe most certainly fails and then it is pot luck as to what happens. On one occasion I did not really notice it [cool season?] and on the other first I knew was when my car stopped and i had to be towed in- a piece of the rubber jammed the main pump.

Rgds

Fred

dr bob

Easy enough to check. Remove the hose from the main fuel pump inlet and connect to it with a piece long enough to reach back into the fuel filler. Put 12V on the inner pump wire and see if fuel is pumped around.

John Veninger

Good idea, but that won't tell me if the hose is just cracked. Would tell if completely split or pump is dead.

I know I have to eventually drop the tank since I smell fuel back there when I first fill it. I'm pretty sure the tank has the famous top crack