Skip to Main Content
SFT is successful in building stable thermal branch constraints using an ac power flow method in conjunction with an optimal generation dispatch algorithm used in the SPD. While solutions may not converge with a maximum SPD-SFT iteration of 1, a mechanism of passing constraints forward to subsequent schedules allows the constraints to be continuously refined as the dispatch time frame is approached. The SPD-SFT convergence logic was found to be working as designed but needed to address two issues: removal of reserve variables from the convergence logic and extension of the logic for handling multiple iterations in order to make the algorithm more robust. This latter issue is of reduced significance given the pseudo-iterative approach adopted operationally. Automation of the constraint generation process is a significant departure from previous manual methods. A particular operational challenge exists in terms of maintaining a suitable set of voltage and reactive load profiles that provide consistent and adequate modeling of the reactive behavior of the network while maximizing the likelihood of a full ac solution. Initial results from parallel operations, however, have shown the algorithm to be producing both sensible results and valid constraints with little to no operator intervention. It is also able to adapt to situations not anticipated by the outage planners in preparing manual constraints.