This project demonstrates a Cloud-Native approach to Automated Grid Compliance. Instead of running manual load flows on desktop software, this pipeline automatically ingests grid models, performs physics-based simulations in the cloud, and validates results against German Grid Codes (VDE-AR-N 4110).
It addresses the industry need for "Cloud-Bursting", leveraging AWS Lambda to handle peak simulation loads during Redispatch 3.0 calculations without maintaining expensive on-premise hardware.
Data Flow:
- Ingest: Grid models (JSON/Pandapower format) are uploaded to an AWS S3 Bucket.
- Trigger: The upload event triggers an AWS Lambda function.
-
Simulation: The Lambda function:
- Loads the grid topology using
pandapower. - Runs a Newton-Raphson Power Flow calculation.
- Checks node voltages against the
$\pm 10%$ tolerance band defined in VDE-AR-N 4110.
- Loads the grid topology using
- Storage: Simulation results and violation flags are stored in Amazon DynamoDB for auditing.
- Grid Logic: Python, Pandapower, NumPy
- Cloud Infrastructure: AWS Lambda, S3, DynamoDB
- Infrastructure as Code: Boto3 (Python SDK)
- Automated Physics: Runs full AC Power Flow (not just DC approximation).
-
Regulatory Compliance: Hard-coded logic for VDE 4110 voltage stability checks (
$0.90 < V_{pu} < 1.10$ ). - Scalability: Capable of processing hundreds of grid scenarios in parallel using Serverless concurrency.
- Integration with AWS Kinesis for streaming smart meter data.
- Expansion of compliance checks to include line loading (thermal limits).
- Automated generation of "Traffic Light" (Ampelphasen) signals for DSO control rooms.
Created by Clifford Ondieki - Power Systems Engineer & Grid Data Architect