Hi,
The head models depend on the geometry or the shape of the head, the cortex (the number and the location of the dipoles), the conductivity of the tissues, and the location and number of the channels.
If you change one of these parameters you need to re-compute the "head model",
So, if you have the same subjects (same MRI), and your raw data contains the same channels (same numbers of and locations), then you can use the previous head models in your new protocols, otherwise, you need to recompute them.