Indeed, at higher injection levels the carrier concentration inside QWs increases, and electrons and holes are separated due to
the built-in field induced by the spontaneous polarization and piezoeffect. Separation of electrons and holes induces electric
field opposite to the built-n field. In that way, carriers compensate ("screen") the built-in field.
This effect is considered in SiLENSe, as the electric potential is calculated self-consistently with the carrier concentration. No
additional fitting parameter is required.
I know APSYS users often decrease the built-in field multiplying it by a factor of about 0.5 (and many people vary this factor to
account for the partial strain relaxation). It can be done with strain relaxation option in SiLENSe. Also, I guess that default
material properties used in APSYS are taken from a popular paper of Fiorentini et al. (V.Fiorentini, F.Bernardini, and O.Ambacher,
“Evidence for nonlinear polarization in III-V nitride alloy heterostructures”, Appl.Phys.Lett. 80 (2002) 1204), while we think
these results overestimate the built-in fields about two 1.5 - 2 times.