Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Subjects: | |
| Online Access: | https://doi.org/10.5281/zenodo.19129740 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p>Extreme price spikes in electricity markets are not global events — they are local structural failures.</p> <p> </p> <p>This study investigates the structural origin of extreme price spikes in electricity markets using one year of real-time data from PJM Interconnection.</p> <p> </p> <p>Conventional interpretations attribute system-wide price spikes to global supply-demand imbalance. However, this study demonstrates that such signals are instead generated by localized structural constraints in the transmission network.</p> <p> </p> <p>Using zonal price dispersion (p95–p05) and transmission congestion metrics (flow / limit), we identify a consistent pattern in which extreme price events coincide with simultaneous saturation of multiple transmission constraints.</p> <p> </p> <p>The results show that:</p> <p> </p> <p>- Extreme price spreads cluster near congestion ratios of 1 </p> <p>- Multiple constraints approach saturation simultaneously </p> <p>- The same constraint clusters repeatedly appear during spike events </p> <p> </p> <p>These conditions create localized network separations that prevent flow equalization across regions. As a result, price signals diverge spatially while aggregate metrics incorrectly suggest system-wide instability.</p> <p> </p> <p>The findings indicate that instability in electricity markets is fundamentally structural and spatial, rather than global.</p> <p> </p> <p>More broadly, this suggests that in networked systems:</p> <p> </p> <p>- Large-scale signals may originate from localized structural constraints </p> <p>- Aggregate indicators can obscure underlying topology </p> <p>- Simultaneous constraint saturation is a key mechanism of system-wide phenomena </p> <p> </p> <p>All analyses are based on publicly available PJM data and are fully reproducible.</p>