Revision history [back]

Hello everyone,

While testing the robust solver for contingency analysis, I observed two behaviors which I cannot explain.

1. Non-Convergence with max Mismatch < 5 MVA

In some instances, the robust solver prints that it is unsuccessful even though the final mismatch is below 5 MVA. My tolerance is set to 1 MVA, but one of the measures that the robust solver usually takes is to increase the tolerance to 5 MVA and hence these models should converge.

2. Convergence with max Mismatch > 5 MVA

Oppositely, in some instances, the robust solver prints that it is successful even though the final mismatch is above 5.

Does someone have an explanation for this?

Thank you in advance!

Hello everyone,

While testing the robust solver for contingency analysis, I observed two behaviors which I cannot explain.

1. Non-Convergence with max Mismatch < 5 MVA

In some instances, the robust solver prints that it is unsuccessful even though the final mismatch is below 5 MVA. My tolerance is set to 1 MVA, but one of the measures that the robust solver usually takes is to increase the tolerance to 5 MVA and hence these models should converge.

2. Convergence with max Mismatch > 5 MVA

Oppositely, in some instances, the robust solver prints that it is successful even though the final mismatch is above 5.

Edit: Case 2 is explained by the creation of small islands and can be solved by turning them off.

Does someone have an explanation for this?

Thank you in advance!