Green-dot, Swedish Steve:
There are so many unknowns. Not disagreeing, just thinking on....
For an arrival in approx freezing temperatures, my calculations suggest a worst case of 500ml of new atmospheric water in the CWT (
from here) which, by the same mechanism we see under the main tanks, will mostly be ice on the coldest inner surfaces of the tank. When the tank is subsequently partially filled with slightly warmer fuel, some of this ice will melt and run into the fuel as suspended droplets. With the aircraft at rest, it is not clear how much of this will collect at the sumping point.
Unless it freezes again, once the aircraft comes alive the scavenge system will deal with whatever water the warm fuel melted. But the ice will remain, presumably to accumulate over successive flights?
I believe there are reports (sorry - can't find references at present) of water-in-fuel warnings at start up which sort themselves out during taxi. The scavenge works. What is interesting is the amount of water needed to trigger such a warning, which as I recall is quite a lot. Such reports suggest that after fuelling, sumping was either not done, not done properly, or the water was somewhere else when the sumping was done, and only collected at the lowest point after start-up.
Given the effectiveness of the scavenge systems, the only time I can see water collecting to be sumped would be when a large amount of accumulated ice has melted after shutdown. The big question is can ice accumulate in this way?