47% IAM factor Losses using Scandia Model

[linkeshd]

Dear fellow PVsyst users,

I recently was surprized by some very large loss factors when using Boviet 6612P 320 Optimized modules, as shown:

Crazy Big IAM Factor Losses

 

Investigating the IAM curve, the Scandia curve looks really strange:

Scandia Lab's IAM factor versus others

 

In general, I thought that the Scandia model was experimentally confirmed. Is this the case? Has anyone seen this before? What would you consider a reasonable approach to this?

Thank you,

Link

[André Mermoud]

I really don't understand how you got this result.

The Sandia model is hard-coded in PVsyst, the values are unmodifiable and give a result very close to the normal Fresnel's model.

This curve is probably a user-defined IAM profile, specified in your PV module, with completely erroneous parameters.

 

Sandia model and Fresnel, normal glass and with AR-coating