# Doubt about energy convergence with different k points

During the convergence test of k points, I observed my energy start increasing instead of decreasing. I am doubtful is this correct or I am doing something wrong. can I take the appropriate K point in this type of converged graph in which energy is inceasing. I am doing this convergence test for bulk material (spin polarised calculation) and K points I took like 2X2X2 then 3X3X3 and so on and observed the corresponding energy.

Here are my input file settings:

SYSTEM = GRPH Electronic Relaxation 1 ISTART = 0 ICHARG = 2 ISPIN = 2 LWAVE = .F. LELF = .F. MAGMOM = 128*1.0 PREC = Accurate ENCUT = 500 LREAL = A NELM = 100 NELMIN = 4 EDIFF = 1E-5 EDIFFG = -1E-2 NSW = 500 IBRION = 2 ISIF = 3 LORBIT = 10 #DOS related values ISMEAR = 0; SIGMA = 0.05

• Welcome to our community!
– Camps
Sep 27, 2021 at 22:45
• Kindly tell your system, input setting as well for clearity Sep 28, 2021 at 6:05
• +1 energy should decrease as you change the energy cutoff because it is a variational parameter (you increase the number of plane waves in your basis), but there is no reason why the energy should decrease as a function of the number of k-points, so this looks ok to me. Sep 28, 2021 at 6:22
• It is a bulk system with 68 atoms in the unit cell. Lattice parameter I took from the literature. Here is my input file setting. SYSTEM = GRPH Electronic Relaxation 1 ISTART = 0 ICHARG = 2 ISPIN = 2 LWAVE = .F. LELF = .F. MAGMOM = 128*1.0 PREC = Accurate ENCUT = 500 LREAL = A NELM = 100 NELMIN = 4 EDIFF = 1E-5 EDIFFG = -1E-2 NSW = 500 IBRION = 2 ISIF = 3 LORBIT = 10 #DOS related values ISMEAR = 0; SIGMA = 0.05 Sep 28, 2021 at 8:27
• +1. Welcome to our new community and thank you for contributing your question here! We hope to see much more of you in the future !!! Sep 30, 2021 at 21:06

Let's just remind ourselves what k point sampling is. It is used to approximate the integral of a function over the Brillouin zone by replacing the integral with a sum over a finite number of points: $$\int_{BZ} F({\bf k})d{\bf k} \approx \sum_j w_j F({\bf k_j})$$