I hope someone with more experience using that specific PDK might offer a better answer but in the meantime I can offer some ideas that may help. I suspect that it is a problem of "scaling" with the source netlist or maybe just that one property in the source netlist. I don't see any other devices in the netlist so I don't yet know how the other devices are scaled. My thought is that the other devices may already be scaled and this one particular property might need to be listed as $EA=1u instead of $EA=1. If the problem was more widespread then it may help to add an option to the netlist such as:
...but I'm concerned that may result in proper checking of just that $EA property while possibly all the pther properties then become wrong by a similar magnitude. So experiment with that as you see fit.
Hope it helps,