As a first step perform a bulk calculation, use the tetrahedron method and increase the number of k-points till your calculation is converged to the required accuracy. What is the required accuracy: Sorry, no general answer to this question, if you want to calculate surface energies within 10 meV you should probably increase the k-mesh till your energy is converged to 1 meV. Mind that a slab used to model the bulk usually contains around 20-100 atoms. This means that you need a very accurate bulk energy to get reliable surface energies. Once again, be as careful as possible. If you generate the param.inc file automatically chose "PREC=High" for the bulk calculation.
As a second step switch from the tetrahedron method to a finite temperature method. There are two possible choices at this moment:
Choose a reasonable set of k-points and the energy cut-off you want to use for the surface calculation and calculate the equilibrium lattice constant. Avoid wrap around errors ("PREC=High" for the makeparam utility). The lattice constant you obtain must be used as the lattice constant in the surface calculations. The free energy is the references value for all further calculations. Also calculate the entropy (i.e. the term
entropy T*Sin the OUTCAR file) or write down the total energy and the physical energy ( ).