I selected the Harwood–Susquehanna 230kV line (PPL Electric Utilities) as it's one of the most famous day-ahead DLR projects. The goal was to see if I could replicate the official PJM values using only public sources before the ratings are published.
The setup was quite simple:
1. I extracted the line geometry directly from OpenStreetMap.
2. I assumed the conductor to be widespread'1590 Lapwing/ACSS'
3. I selected the maximum operating temperature of 105°C/221°F.
4. I ran the thermal analysis using Gridraven Claw with hyperlocal weather forecasts.
Results are in the chart below:
· Black: PJM Day-Ahead market data (normal rating).
· Blue: Ambient Adjusted Rating (AAR).
· Orange: My P50 DLR prediction (Gridraven Claw).
· Green and red: My P95 and P98 predictions (Gridraven Claw).
Ideally, the Orange line (Prediction) and Black line (Market) should be identical. Currently, there is a visible offset because I had to estimate the specific conductor parameters. However, the correlation is strong (0.87). The model successfully captures the volatility and the specific constraint patterns.
Note: I am treating the Day-Ahead value as "perfect" ground truth here. In reality, real-time telemetry would be the superior validation.
