For the record, I rarely gas up before the warning light trips. Once it goes off, I typically go another 50 miles or so. I once figured out that (on MY GD3) I have about 2.7 gallons left past the warning light tripping. I know that I can get at least 32 mpg, so I don't usually worry until I go 60+ miles past the light. I once went 65.4 miles past the alert, but I still had at least .6 gallons left.
I use 32 mpg as a conservative guide. I am recently averaging 39.27 mpg, so I could theoretically go 39.27 mpg X 2.7 gallons = 106.029 mpg past the warning light.
In regards to the fuel pump failing when running at low fuel levels, I can state that I am still on my original pump. I have been using this fueling practice for 235,000+ miles now. So if you are lucky, running the fuel low like this (i.e. below 1/4 tank) could last at least 236,056 miles!
Just yesterday, I went 47.7 miles past the warning. I always reset trip odometer "B" as soon as the light goes off. That way I know I have at least 60 miles or so left to find a gas station.

 

I have replaced the fuel filter that is part of the reservoir, and the way it all goes together with the fuel pump, it sucks up fuel and seems to pump it round the reservoir, the over pressure flows back into the reservoir. from having it out I can see how the level in the reservoir will be higher than the tank level (only when the car is running) as the pump, pumps more than the car uses, the pump intake is via a small grit filter below the reservoir, this with the over pressure (not sure if the phrase is correct, but you know what I mean) puts extra fuel in the reservoir.
Not sure if I explained properly, but anyone that has had it out will know what I am trying to say.

 

You could probably go 2000 miles then on a full tank of gas. Have you ever noticed the gauge going the other way as you drive?
Your calculations sound radically inaccurate.

 

What are you saying with this picture? Your power steering light went off by how full the tank is?

 

Thanks for trying dude, but sorry, your description does not make sense to me.
I get your point, you are supporting steve but I don't see how you got to it.

 

I replaced the fuel filter, and had the whole unit in my hands, you have to dismantle everything, and put it all together again, and I left out a second O ring on the fuel pump, so ended up taking it apart 3 times before I found what I did wrong, and in doing that I worked out exactly how it works, as I said it is difficult to explain.

 

You have been shown that a reservoir exists inside the fuel tank around the pump. Your assumption that 1/4 tank or more is needed to adequately cool the fuel-pump is a self-made unsupported assumption. There is no source supporting your assertion other than a mechanic in LA (state, not city) showing a primitive tank-embedded electric fuel pump with no reservoir. Got anything else?

Brian is more savvy than me: he's had the thing apart. If you don't want to take his word the reservoir acts as a reservoir to keep the pump immersed and from sucking air, I can't convince you. If you choose to fill at 1/4 tank, that's fine. You're not hurting anything but the frequency of your trips to the gas-station. How's that for personal?

Edit: I see where Ford uses a pump with a strainer that appears to have no reservoir, at least on this 1999 truck. This is similar to the diagram on your Louisiana mechanic's site.

GM trucks use a reservoir with a "jet-pump" capable of keeping the reservoir filled to "help prolong pump life" and "prevent fuel pump starvation."

Even in the case of Ford, the fuel is pumped through the body of the pump, keeping it cool. It shouldn't matter whether the body is immersed for temperature control. Sucking air is another matter.

I don't see any videos for Hondas. Draw your own conclusions.

 

I'll take Brian's word for it and I understand what he's saying. The pump diagrams i've seen support this though I don't remember where I saw them unfortunately.

 

The complete unit shown in the video is very similar to the Honda unit, right down to the springs. When you put it back in the tank you have to squash it down into the tank.

 

A little further information in terms of pump wear, if you're interested:


Most pumps use the fluid travelling through it to keep cool and lubricated. Where the fluid travels is where the most friction occurs, and therefore the most effective location for cooling. Because of this, pumps generally do not build up much heat. It is the motor powering the pump which heats the most, and at that, it is very minimal with a small electric motor. The only situation where a motor would be creating a lot more heat would be where it is very inefficient. With high efficiency motors, there is less loss of energy due to heat.


Where the precaution arises towards pulling in air, here is how it works. If you have a fluid pump that is running strictly air (no fluid at all) then what you have is a pump that is not required to do much work. As a result, the motor also is not required to do much work. Since minimal work is being done, the power consumption of the motor and pump is very low, as it is essentially running at idle, and little heat would be generated. This is the easiest operation situation, and you will not see a motor burning out as a result. The only real concern is if the pump has fairly tight tolerances that require fluid to keep running smoothly. Here excess wear may occur, but the situation would have to be a bit extreme to cause any serious issues over a short period of time.


The largest concern with these pumps would be a mixture of air and liquid. What happens here is called cavitation. The air/liquid mixture enters the pump at a low pressure. It is compressed as it passes through the pump, and due to the high pressure resulting, these air bubbles can implode as they are forced to dissolve into the liquid. This results in many constant microshockwaves, which are surprisingly powerful, and the most common cause of pump failure, as the shockwaves erode the interiors of the pump and latter system.


I hope this helps to understand pump wear and cooling concerns.