Ever since the cold weather arrived, I've started leaking fuel from the fuel line connections at the float bowl nipple connections. Dual SU's totally rebuilt, all new fuel lines with stainless braiding and new clamps.

It first happened after a cold night. When I arrived at work, I smelled fuel and saw that gas was dripping out of all 3 nipple connections (the main line connection, and the connector line between float bowls. I tightened all three clamps and it was fine for a day or so. Then, after the next slightly colder night the same thing happened. This time I replaced the lines, and put on new clamps. Now, this morning, after an even colder night, I get to work again with the same issue, fuel out of all three clamp points. I know it isn't the bowl gaskets, because the gas is on top of the float bowls. Gaskets are also brand new. I did remove the caps to check that nothing was amiss with the floats.

Is this a common problem when things get cold? This morning it was below freezing. I am using standard screw tightening hose clamps, are these the wrong types? I've never had these issues when warm, but now have the lines squeezed so tight, I'm amazed they still leak.

Any ideas?

"I am using standard screw tightening hose clamps"

There are hose clamps and then there are hose clamps. The "Jubilee" screw type that screws around a "thread" in the clamp aren't suitable for small diameters like this. The type used for fuel lines use a small bolt threading into a trapped nut. These put an even pressure on the pipe, giving a better seal. Clamps sold as suitable for fuel injection piping should do the job. Clamp on the rubber, not on the braiding.

I went out and purchased some of the trapped nut hose clamps. The package said it was for fuel injection lines. I replaced the leaking clamps and test ran the engine without issue. The leaking usually started after the engine was fully warm and driving for a while, so I'll report back tomorrow after my sub zero commute!

It was 21 degrees this morning, and no visible leakage after the commute. Looks like the worm drive clamps where the culprit. I also have noticed that rubber fuel lines get really hard when it gets this cold. Are there silicon or other material options out there?

Fingers crossed for the new clips. "Hard" is normal when cold. It's not real rubber nowdays. Old pipes age harden too but carry on working. Surprisingly the original rubber pipes were just push fit on with no clips. Correctly sized they had enough squeeze to make a good seal. Metal stubs are long and there should only be about 2-3lbs psi anywhere in the system.

OK, so there's been a success and a failure. The failure is that it turns out I was never leaking from the fuel lines at all, it was a red harring. The success is that I've discovered the real source, which is the overflow holes on the top of the float bowl, the tiny ones just beneath the fuel nipple.

It appears that each bowl is overflowing at random times. That is, I can pop the hood and one bowl will be dry, while the other one is trickling fuel out of the hole. I can drive for 5 minutes, and open the hood again, except this time, the other bowl is leaking, and the first one is now dry. Drive another 5 minutes, and either both can be leaking, or neither. The only correlation is that is appears they are more prone to leaking after I get off the freeway where I've been at high RPM's for a while.

My only guess is that each of the floats are getting stuck in the down position, so that fuel is allowed to continue pumping into the bowl, causing the overflow. Why both needles would fail at the same time, but not 100% of the time is a mystery to me. The needle check valves in the floats are 6 months old, same as the rest of the carb rebuild. Can these needle valves fail in this manner, or is there another issue? I opened the bowls and everything looks clean and free moving, no binding that I could see in the needles.

One thought would be that water in the fuel would freeze inside float needle and stick it open, causing overflow. However, why it would only get stuck down while at high flow for a long time when the engine is warm doesn't make much sense to me.

The other idea that people have mentioned is a failing fuel pump. Mine is 6 months old, like the rest of my rebuild, and I have a hard time believing the failure mode of a mechanical pump like this is to produce a higher pressure. Is there something I'm missing there?

Any insight would be helpful.

How cold is it? If you are in central Alaska I could see the rubber shrinking that much...

The first two nights it got to 35. Last night it got to 30. Not "cold" compared to what others drive these cars in. That is the only recent change that could be responsible for this leaking.

I'm really lost with this one. The fuel lines are the right size, the rubber is soft and supple, and the clamps have been fine until the cold weather. The nipples themselves are in perfect condition, no folds or ripples that could prevent a good seal.

Are we talking HS6 or HIF6?
Is it possible the fuel is leaking between the brass tube and the float lid/carb body?

These are HS6's, and I've taken the float bowl lid off and inspected where the brass tube joins with the lid. I have seen splits in that joint before, but mine look ok.

Sorry, no ideas here.
I've got the same setup and no leaks year round.

Sure you don't have too much fuel pressure? For SUs, you'll only need about 2 psi or so.

I've been running a newly installed stock mechanical pump all summer without issues. My inline fuel pressure gauge reads 2-3 PSI while running.