If your simulated choice scenarios essentially match the same number of concepts + None for your experimental tasks (that captured the data for estimating utilities), then you really shouldn't have to change that value different from 1.0 to get appropriate estimates of what new respondents would pick in those new choice scenarios.
So, to turn None "off" you set it to 0. To turn it "on" with its default weight you set it to 1.
Another issue is that respondents tend to understate their None likelihood in market research surveys (they typically tend to exaggerate purchase likelihood). If you have external data indicating some direction regarding what the None rate should be, then you might think about adjusting the None weight to something like 1.5, 2, or 3, etc.