# Hierarchical Reinforcement Learning with Runtime Safety Shielding for Power Grid Operation

arXiv: https://arxiv.org/abs/2604.14032 · HTML: https://mlsanity.com/paper/2604.14032/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in RL for energy, networks and operations, one of 15 directions in Deep reinforcement learning methods.
- Builds on 5 of its 17 references that are in the map.
- First posted on arXiv 15 Apr 2026.

## Placed in

- Deep reinforcement learning methods / [RL for energy, networks and operations](https://mlsanity.com/deeprl/rl-for-energy-networks-and-operations/)

## Builds on (most cited first, 5 of 5)

- [Constrained Policy Optimization](https://mlsanity.com/paper/1705.10528/) (2017)
- [Safe Reinforcement Learning via Shielding](https://mlsanity.com/paper/1708.08611/) (2017)
- [Control Barrier Functions: Theory and Applications](https://mlsanity.com/paper/1903.11199/) (2019)
- [Challenges of Real-World Reinforcement Learning](https://mlsanity.com/paper/1904.12901/) (2019)
- [Learning to run a Power Network Challenge: a Retrospective Analysis](https://mlsanity.com/paper/2103.03104/) (2021)

## Built on by (most cited first, 1 of 1)

- [SiLR: Structure-Preserving Admission and Process Reward for LLM Tool Agents](https://mlsanity.com/paper/2609.04629/) (2026)

## Abstract (arXiv)

Reinforcement learning has shown promise for automating power-grid operation tasks such as topology control and congestion management. However, its deployment in real-world power systems remains limited by strict safety requirements, brittleness under rare disturbances, and poor generalization to unseen grid topologies. In safety-critical infrastructure, catastrophic failures cannot be tolerated, and learning-based controllers must operate within hard physical constraints. This paper proposes a safety-constrained hierarchical control framework for power-grid operation that explicitly decouples long-horizon decision-making from real-time feasibility enforcement. A high-level reinforcement learning policy proposes abstract control actions, while a deterministic runtime safety shield filters unsafe actions using fast forward simulation. Safety is enforced as a runtime invariant, independent of policy quality or training distribution. The proposed framework is evaluated on the Grid2Op benchmark suite under nominal conditions, forced line-outage stress tests, and zero-shot deployment on the ICAPS 2021 large-scale transmission grid without retraining. Results show that flat reinforcement learning policies are brittle under stress, while safety-only methods are overly conservative. In contrast, the proposed hierarchical and safety-aware approach achieves longer episode survival, lower peak line loading, and robust zero-shot generalization to unseen grids. These results indicate that safety and generalization in power-grid control are best achieved through architectural design rather than increasingly complex reward engineering, providing a practical path toward deployable learning-based controllers for real-world energy systems.