Category: Physics

Summary: Finding how much fast-support misallocation across stressed islands an islanded microgrid can tolerate before small-signal rescue fails.

Islanded microgrids rely on fast support and reserve-sharing mechanisms to remain stable after separation from the main grid. This experiment asks when that stabilizing layer stops working because support is distributed too unevenly across stressed islands.

The model builds dense non-symmetric swing Jacobians for modular islanded microgrids and uses GPU eigensolves, iterative deepening, and bisection on a support-imbalance parameter. The target is the point where a previously rescued system loses small-signal stability again, turning imbalance itself into the main control variable.

This matters because fast support is often treated mainly through average gain or total reserve. The experiment instead isolates allocation imbalance and asks how unfairly support can be routed before the rescue mechanism breaks down.

Method: GPU dense non-symmetric Jacobian eigensolve with iterative deepening and bisection on fast-support imbalance in modular islanded microgrid models.

What is measured: Critical support-imbalance threshold, leading-eigenvalue stability boundary, rescue failure status, and threshold bracket width.