Skip to content

Z139-Lab/bounded-criticality-power-grids

Repository files navigation

<<<<<<< HEAD DOI: https://doi.org/10.5281/zenodo.19686209

# Bounded Criticality in Power Systems

A reproducible research repository for finite-width critical bands, suppression of cascading failure, and scale-dependent transition structure across IEEE-39, IEEE-118, IEEE-300, and IEEE-2383 benchmark systems.

Live research portal:

---

## Overview

This repository studies **bounded criticality** in realistic power-grid systems under stochastic loading.

Unlike classical sharp critical-point pictures, the results here support a **finite-width critical regime**:

a bounded transition region in which collapse probability changes continuously and control effectiveness becomes maximal.

The repository integrates evidence across four benchmark systems:

- **IEEE-39** — proof-of-concept suppression regime

- **IEEE-118** — emergence of a finite-width critical band

- **IEEE-300** — structured critical regime with stronger peak suppression

- **IEEE-2383** — near-discontinuous critical edge

---

## Key Findings

- **Finite-width critical band** is observed in realistic power grids rather than a single sharp threshold.

- **IEEE-118** shows a critical band around **30–50 MW**, with memory-dependent suppression near **40 MW**.

- **IEEE-300** shows stronger suppression, with peak **Δ ≈ 0.332** near **126 MW**.

- **IEEE-2383** exhibits a near-discontinuous edge, with a narrow transition around **29.05–29.10 MW** and extreme suppression.

- **Peak susceptibility does not diverge**, supporting a bounded rather than classically divergent critical regime.

- The empirical width of the critical band decreases with system size, but **no scaling law is claimed** from the limited width points.

---

## Research Object

- **Domain:** Power systems / cascading failure

- **Phenomenon:** Bounded criticality

- **Systems:** IEEE-39, IEEE-118, IEEE-300, IEEE-2383

- **Mechanism:** Logistic softening with memory feedback

- **Observables:** Collapse probability, suppression Δ, empirical band width, peak susceptibility

- **Goal:** Identify and compare bounded transition structure across system size

---

## Reproducibility

Clone the repository:

```bash

cd bounded-criticality-power-grids

Install dependencies:

pip install -r requirements.txt

Generate the main paper figures:

python run_all.py

Main compact dataset:

data/real_data.csv

Figure-to-data mapping:

docs/figure_mapping.md

Figures

Figure 1 — IEEE-118 critical band + memory ablation

Generated from IEEE-118 rows in data/real_data.csv.

Figure 3 — IEEE-300 finite-width band

Shows baseline, controlled collapse probability, and suppression Δ across the critical regime.

Figure 4 — Empirical critical band width

Uses the compact width points for IEEE-118, IEEE-300, and IEEE-2383.

No power-law fit is used.

Figure 5 — Peak susceptibility

Shows non-divergent peak susceptibility across system size.

Repository Structure

data/ compact paper dataset

docs/ figure mapping and notes

paper/ manuscript and generated figures

scripts/ figure generation scripts

case_studies/ieee39/ proof-of-concept case

Related Repositories

Keywords

bounded criticality, finite-width critical band, cascading failure, power grid stability, memory feedback, IEEE benchmark systems, complex systems, critical phenomena, reproducible research

Citation

Please use the repository citation metadata in CITATION.cff.

Author

Juan Adam

Independent Research