I went out to the storage yard and climbed under my school bus to be sure that the fuel flow path was as I remembered. It almost was. I've attached a modified diagram to show the differences. Essentially, the two fuel filters are between each tank and the selector valve, rather than in series after the selector. That will alleviate some of the problem - the fuel on the intake side of the pump only goes through one filter instead of two. I also found a filter on the output side of the pump, and there is a fuel strainer attached to the intake of the pump.

So, while the fuel only goes through one filter on the way to the pump, there is another aspect to the fuel path that will make things worse. Maybe. The fuel pickup is (obviously) in the bottom of the tank. But the fuel exits from the top of the tank, so there is about a two foot elevation gain. From there the fuel line loops back down to about a foot below the bottom of the tank to the fuel filter, then goes back up about three feet to the selector valve before coming back down to the fuel pump. Hey - don't blame me, I didn't design this! In the normal course of events, these elevation changes wouldn't matter, there is a net elevation loss between the fuel pickup and the pump so fuel would flow to the pump through siphon action even if the pump were not developing suction. But in our case, where the vacuum in the lines may drop the line pressure to the point that the gasoline boils, the siphon may break. I'm not sure about this. Maybe the ups and downs are irrelevant. I've attached a drawing more or less to scale showing the vertical travel of the fuel on the way to the engine.

There are a few other data points that support the idea that the fuel is boiling on the intake side, not from temperature but from vacuum.

(1) The previous owner swears he never ever, not even once, had a vapor lock problem. After I purchased the bus the filters were added on the intake side of the fuel pump because the fuel tanks were filthy inside with the remnants of 20-year-old gasoline. (That's how long the bus had sat parked, and why it cost me so much to get it going). The shop that did the work no doubt thought they were doing the right thing.

(2) There is virtually no engine or exhaust heat anywhere near the fuel tanks and fuel lines until after the fuel pump, yet the problem is absolutely hot-weather related. I'm not talking about weather hot enough to boil gasoline on its own, but warm enough to help it boil in a reduced pressure situation.

(3) Every time I had vapor lock occur, it was apparently on the input side of the pump because it was going "thocka-thocka-thocka" just as it does when one of the fuel tanks runs out (there is a fuel gauge on only one tank, so I run the other tank to empty before switching to the "gauged" tank).

The shop that did the work on the bus did a good job of cleaning the tanks. The fuel filters have clear plastic shells, and after 4000 miles one filter is still completely clear and the other one has just a trace of sediment on the bottom. Both filters are simple in-line filters and all I need is a screwdriver and a couple of feet of 3/8" fuel line hose to take them out of the system.

When I do that, I think I should replace re-route and insulate the fuel line between the pump and the carburetor to get it further away from the exhaust headers.

This has been an interesting and informative thread for me. I consider myself quite knowledgeable about internal combustion engines, but this was new territory for me.

I would, however, like a definitive answer as to whether vapor lock can occur on the output side of the pump, that is, between the pump and the carburetor.

tanstaafl.