Answers are grounded in this paper and cite their sources.
The loadability margin measures how far loads can grow along a direction while some operating point, with full redispatch, satisfies the power-flow equations and every limit. Local solvers estimate it without proof; convex relaxations bound it with unknown looseness. We compute an interval that provably contains it. Its upper end comes from the dual of the semidefinite relaxation's maximum-loadability problem, in one conic solve on all but the 2000-bus grid. Where the relaxation is loose, certified bound tightening proves single loads infeasible. Interval Newton tests prove its lower end along a path of interpolated optimal dispatches or along the optimal point continued with binding limits held. Exact rational arithmetic re-verifies every upper end. We bracket the margin along 130 directions of 17 systems with up to 2000 buses. Over all 130 directions the bracket is 1.00% wide in the median, and 0.01% along the 35 where the relaxation is within 0.1% of the local maximum. Bound tightening, run along 54 of the 82 directions where the relaxation is over 1% loose, proves the margin at most 0.1% above the local maximum along 32, on systems with up to 300 buses.
No takes yet. Share an insight, caveat, or question.
Yuan Shen (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: