Thank you two, zh and Blom.
In fact, I have calculated that system under varied biases, only 0.2V-result has two "0 eV" levels, and the others have no "0" output. I know there may be degenerated levels and they are existent in all my results. Therefore, I think the little problem shouldn't attribute to the script or the configuration. I just can't know the meaning of the two "0" values.