← Back to Portfolio

01 — Data Generation & Setup

Generates a synthetic 8-week chiller plant dataset with three embedded operational faults for diagnostic training.

chiller-plant synthetic-data fault-injection
Pythonnumpypandasmatplotlib

Key Findings

  • 1,344 hourly data points simulating a central chiller plant (June–July 2025)
  • Three hidden faults injected: low-delta-T syndrome, cooling tower degradation, night-time fan bias
  • 23 columns covering temperatures, flows, power, and derived KPIs

Synthetic chiller plant dataset creation with embedded operational faults for energy management diagnostic training. The dataset simulates real-world HVAC conditions including weather drivers, occupancy patterns, and cooling load dynamics.

Notebook