Just wrapped up a brutal audit for a manufacturing client on Georgia Power TOU-7. The utility had them classified under the wrong time windows for 18 months. Peak hours were being applied from 1-7 PM instead of 2-8 PM during summer months. Client was paying peak rates during what should have been shoulder periods. Total overcharge came to $47,300 before we caught it. Anyone else seeing TOU window errors with Georgia Power lately? Their billing system seems to have some serious bugs.
Georgia Power TOU-7 schedule error cost client $47K
Derek, that's a significant error. We've seen similar issues with Duke Energy here in the Carolinas, but usually it's incorrect seasonal transitions rather than wrong peak windows. Did Georgia Power acknowledge the error when you presented your findings? How cooperative were they with the refund process?
Duke was actually pretty good about it once we showed them the tariff language and our interval data analysis. Took about 6 weeks to get the credit issued. The frustrating part was their explanation - apparently a system upgrade in early 2011 corrupted some TOU parameters for industrial customers. Makes you wonder how many other accounts are affected that don't have auditors watching.
This is exactly why we always verify TOU parameters against the published tariff before starting any analysis. FirstEnergy here in Ohio had similar issues after their merger integration. We found three clients being billed under completely wrong rate schedules. Always check the basics first - rate code, TOU windows, seasonal definitions. The devil is in the details with these complex tariffs.
Jim makes a good point about verifying basics. ComEd has been pretty reliable with TOU applications, but we caught them applying weekend rates to a Friday once. Only a $200 error, but still shows these systems aren't foolproof. Derek, did you use any specific software for the interval data analysis or just Excel?
We used a combination of Excel and some custom scripts I wrote in Python. Excel for the initial analysis and visualization, Python for processing the massive interval datasets. When you're looking at 15-minute intervals over 18 months, that's a lot of data points. The key was creating hourly summaries and then mapping those against the correct TOU periods from the tariff.
Python scripts for utility analysis - that's impressive Derek. Most of us are still stuck in Excel hell. Xcel Energy up here has been pushing more complex TOU structures, and manual analysis is becoming nearly impossible. Would you be willing to share any of those scripts? Obviously sanitized of client data. The forum could really benefit from better analytical tools.
Let me clean up the code and I'll post something in the tools section. Nothing too fancy, but it does automate the tedious stuff like interval aggregation and TOU mapping. Glad to help the community - we've all been there staring at spreadsheets until our eyes bleed.