> Returning fuel from the rail to the sump is
> what got Paul Connor in trouble.
> Gravity feed main(s) to sump, return to main(s) ... Jim S.
>
> How do we KNOW that the return to the sump is what caused the prob? Are
> we talking his redesign or the original install?
>
> The reason I am SO interested in this particular issue is that the
> Velocity has a sump tank as well, but it holds nearly an hour of fuel at
> lower power settings.
Dave,
An hour of fuel "consumption" is obviously NOT an hour of fuel draw. I think
it would help to discuss the pros and cons of sump based systems depending
on where you send the return. The way I see it there are cons in both
directions.

Return to the sump
Risk pressurizing the sump. Needs overflow vent to the mains.

Return to the mains
Risk negative pressure in the sump and failure to gravity feed fast enough.

A sump is nice in theory because it gives you an extra reserve, but in
practice I don't see a "perfect" way to install one. Perhaps I'm missing
something. Could someone explain the details of a good sump based system and
show how the above issues are dealt with?

John