I've just studied the Least Herbrand Model, that is the intersection of every possible Herbrand Base.
So, i guess that it's possible to have an Herbrand Base with more elements than needed . Is it correct ? If no, i dont understand the need to introduce the Least Herbrand model...
See Minimal models:
Note that minimal does not imply unique, in general. A set of formulas might have several minimal Herbrand models.
For example, $\{ a ∨ b \}$ has the following Herbrand models: $\{ a, b \}, \{ a \}, \{ b \}$.
The models $\{ a \}$ and $\{ b \}$ are both minimal.