n1-contingency-dispatch

N-1 security-constrained DC-OPF RL environment. Fourth rung of a power-systems vertical (economic-dispatch → dcopf-grid-verifiers → multiperiod-dispatch → this). The skill axis is security: the dispatch must survive the outage of any single non-islanding transmission line, with post-contingency flows within emergency ratings. The naive answer this environment defeats is precisely the previous congestion environment's correct answer — the ordinary DC-OPF optimum, which leans on fully-loaded corridors that have no margin left when a parallel line trips. Measured: 68% of instances make the plain OPF optimum insecure.

Task

Given a meshed network (lines with reactance, normal rating, emergency rating), generators, and loads: output {"dispatch_mw": [...]} minimizing cost such that (1) the base case is feasible at normal ratings, and (2) for every single-line outage that leaves the network connected, the same dispatch remains feasible after DC power flow reroutes — every surviving line within its emergency rating (preventive SCOPF).

Ground truth validation

Two independently formulated solvers must agree — different variable spaces and different algorithms:

• Primary: monolithic security-constrained LP in angle space — variables [P, θ⁰, θ¹…θᴷ], one angle block per contingency state (scipy/HiGHS).
• Cross-check: injection-space solver using per-contingency PTDF matrices and iterative constraint generation (solve, find violated contingency constraints, add as cuts, repeat), written from scratch with no shared assembly code.

Rewards (weighted rubric)

Anti-reward-hacking, factory standard (6 regression-tested attack gates): the insecure-cheap attack (submitting the plain OPF optimum — cheaper and base-feasible — earns optimality 0.0), NaN/Infinity literals, >4300-digit integers and bracket bombs (rejected, never crash), and tolerance-rent (violations inside the ±0.5 MW grading tolerance settled at 2× the highest offer, below-optimum cost penalized symmetrically).

Usage

pip install -e .
N_INSTANCES=200 python tests/test_validation.py   # dual-formulation cross-check + attack gates
python calibration/measure.py --n 300             # difficulty distribution sweep

# give frontier models room to reason; -r 1 matches published baselines
vf-eval n1-contingency-dispatch -p anthropic -m <model> -n 50 -r 1 --max-tokens 16000

import verifiers as vf
env = vf.load_environment("n1-contingency-dispatch", num_examples=300)
# disjoint train/eval: load_environment(1000) + load_environment(200, seed_offset=1000)

Instance generation

Meshed random networks (spanning tree + loops, 4–8 buses, extra-edge probability 0.35 — N-1 only bites where parallel paths exist), tiered baseload/mid/peaker costs, ~9 non-islanding contingencies per instance. Line ratings are anchored to duty: normal ratings to each line's flow under a nominal proportional dispatch (tight draws at 1.05–1.35×, loose at 1.8–3×), emergency ratings to the line's worst N-1 duty (tight 1.02–1.20×) — so ratings correlate with usage the way real thermal ratings track expected duty, and the nominal dispatch is N-1 secure by construction (draw rejection 7.7%, no survivor bias; see CALIBRATION.md for the two rating constructions the metrics rejected). Non-islanding screening only, as in real N-1 practice for radial taps. Fully deterministic per seed.

The vertical

merit order → congestion (space) → ramps (time) → N-1 security (uncertainty). Next: LMP/nodal pricing verified against LP duals.