Sorry, yes, it provides the count of the cells >0 rather than the count of the rows with a cell value >0. My mistake. I was looking for a total in which a row with two values >0 is only counted once.
Tony Valko's formula above works for my purpose.