There are 3 scans that I apply while solving problems at Chess Tempo:
The only reason that traps are number one in the list is due to the biased problemset of CT, which contains an abundance of mates. During a game it will be the last scan to perform, convergence being the first.
The following problem is rather simple, but it provides a clearcut example of how to scan for invasion squares where two or more of your pieces converge.
White to move.
The white queen and rook converge at e6. The scan cuts down dramatically on the candidate moves. You first look for the invasion squares. If no pattern is triggered and you cannot make use of the square, you can always look at other candidate moves. But that is what a good scan does, it prioritizes your candidate moves without you consciously being aware of it.
1.Re2+ Kd6 2.Qxe6+ Kc7
The queen and bishop converge at c6.
The queen and rook converge at e8.
There is something very logical in looking at the squares where your pieces converge. From all the squares in hostile territory where you can put a piece, these are the most logical candidates. Since if you put a piece there, it is automatically protected. Convergence means cooperation of your pieces. How likely is it that there is a working combination without your pieces cooperating?
I find this the most useful scan by far I and I make use of it almost every move.