We've found that locally (or regionally), many sedges are identifiable at an early stage before flowering, but that it requires a combination of knowledge of the morphology of the local species as well, often or at least sometimes, a knowledge of habitats. Many local species can be identified vegetatively even without habitat clues, but somewhat fewer. I think a person could develop a key to local sedges by vegetative traits.
In fact, I remember now that there is such a key for Newfoundland sedges, and I've just found that it's available online: http://cfs.nrcan.gc.ca/bookstore_pdfs/24645.pdf. This is focused only on Carex. It might give you some ideas about how to proceed in writing a key to your local species.
Surely, if you have a basic knowledge of Cyperaceae morphology, they could be identified many times even in vegetative stage ( The three angled culms, the ternate arrangement of leaves, The channeled mid vein on upper side etc are some of the distinguishing characters. Cyperaceae do not creep on the ground like grasses and here creeping branches are formed from root stolons) What I mentioned are only some general characters.
You should begin with knowing which are the local sedges in the place; as Andrew said, some knowledge of habitats is essential. Then sometimes morphology is enough to identify them, but a plus in anatomy of leaves and/culms could be fine. Most of the possible doubts that you have when you only rely on morphology can be solved with a tissue section under the microscope (see Anatomy of the Cyperaceae by Metcalfe)