You will have both in your spectrum if both P and B are present in the sample. However, i am not sure but i think the shape and features of these two edges should be different; hence, if you know what is supposed to be the chemical environment around both of them, you can use pure spectra of P and B (collected on reference compounds) and try to make a linear combination fit with Athena software, in order to determine the ratio of the two components.
Is there a way to distinguish between overlapping energies in XAS graphs from the P L1-edge from the B K-edge? is a very interesting question that has to be understood and answered properly. However, as it is completely outside the field of my specialization I will not be among the first researchers to answer your question. On the first place I will follow the discussion to gain knowledge about the issue that you have raised. Thereafte, i.e. among the second stage I will be able to make a contribution. Thank you very much for taking us this very interesting question.