Power system state estimators: designed for reliability or accuracy?

It is the experience of this author that state estimation programs suffer mainly from the effort to preserve maximum accuracy in the presence of changing measurement availability and topology of the power system. The paper deals with the issue of how the loss of a single measurement or the loss of a single line can render a state estimator inoperable and asks why state estimators are built that way. A solution for this problem is offered through the use of load forecast measurements.