You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently TransformerCircuits with VOLTAGE control can have control_limits that disagree with the regulated_bus's own voltage_limits. We should either take the tighter of each bound or raise an error, open to opinions about either.
And if we aren't already, we should make sure these (and other) limits are finite, non-negative (depending on control), and ordered.
Example downstream issue: on polar networks we have a voltage magnitude variable bounded by voltage_limits that might later conflict with the control constraints.
Both bus voltage limits and voltage control limits default to [0.9-1.1] in PSSE. I think during parsing we should warn if the control voltage limits falls outside of the bus limit that it is controlling and set the control voltage to the bus voltage.
Additional validation: multiple circuits with the same regulated bus. Maybe they have to also have limits that agree, or maybe they shouldn't be allowed at all.
Activity
acostarelli commentedon Aug 14, 2026
Example downstream issue: on polar networks we have a voltage magnitude variable bounded by voltage_limits that might later conflict with the control constraints.
m-bossart commentedon Aug 14, 2026
Both bus voltage limits and voltage control limits default to [0.9-1.1] in PSSE. I think during parsing we should warn if the control voltage limits falls outside of the bus limit that it is controlling and set the control voltage to the bus voltage.
acostarelli commentedon Aug 14, 2026
Additional validation: multiple circuits with the same regulated bus. Maybe they have to also have limits that agree, or maybe they shouldn't be allowed at all.