This is an area of active research at NIST.
http://www.nist.gov/bfrl/fire_protection/fireservice/hose_stream_charact_effect_modeling.cfmThe key would be Validating that the nozzle in your model has some basis in reality.
Also, there are some physics of water suppression that are not accounted for by FDS, it is fairly complex from a physics and chemistry point of view.
The best approach is to put together a simple model of your nozzle spray and test various parameters related to how it operates.
Best Regards,
-Bryan Klein