Hi @evozovoy,
Have a read through this thread which I believe covers your situation nicely. In short, the easiest approach would be to import/demultiplex and process each run separately then merge after. Once demultiplexed, the barcodes won't matter and only the sample IDs are relevant. The merging options depend on whether there are shared sample IDs between your runs, or they are simply different samples being analyzed together. The link above covers this nicely as well!
Hope that helps!