interconnection-impact-mitigation - SKILL.md Agent Skill

name: interconnection-impact-mitigation description: Senior power-engineer playbook for screening and mitigating the grid impact of a new generator, storage, or large-load interconnection. Use whenever the question is whether X MW can connect at a point of interconnection and what upgrades it needs. Triggers on "interconnection study", "system impact study", "POI", "can we connect this plant at bus Y", "short-circuit ratio", or "weak grid". Walks from steady-state screens through weak-grid and reactive checks to the binding constraint and its mitigation set.

Start with a defined injection: MW and MVAr capability, the point of interconnection (POI), and the dispatch scenarios to study.

Pin down the study scenario set: peak and off-peak cases, high-renewable dispatch, and the prior-queued projects that must be modeled as in-service.
Run steady-state screens with the addition: base case and N-1 power flow, thermal loading and voltage screens around the POI. Route violations to thermal-overload-mitigation and voltage-violation-mitigation for the corrective toolkit.
Check weak-grid risk for inverter-based resources: short-circuit ratio at the POI. Low SCR (roughly below 3, critical below 2) flags control-interaction and stability risk — consider grid-forming controls, synchronous condensers, or POI relocation.
Verify reactive capability and voltage control: power-factor range and voltage ride-through at the POI against the applicable interconnection requirements.
Identify the binding constraint and assemble the mitigation set: network upgrades, an output-limitation or curtailment agreement, reactive devices, or phased interconnection — sized to the smallest set that clears the binding screen.

Screen with linear tools (PTDF on the injection) before full N-1 reruns; confirm the short list in AC.
The fault contribution of the new plant changes breaker duty nearby; screen it and route over-duty to short-circuit-mitigation.
Restudy after any material change in plant size, POI, or queue assumptions — interconnection results do not transfer between scenarios.
Separate "needed to connect safely" from "needed to deliver full output" — reliability upgrades and deliverability upgrades have different urgency and cost allocation.