Working on a complex case with Georgia Power GS-TOU-8 rate and need some guidance on interval data analysis. Customer has been on this rate for 3 years and I suspect they're being overcharged due to incorrect on-peak period application. The 15-minute interval data shows usage patterns that don't align with the billed on-peak kWh amounts. Summer on-peak should be weekdays 12:00 PM to 9:00 PM June through September, but the numbers aren't adding up. Has anyone developed a good methodology for validating TOU billing against raw interval data?
Help with Georgia Power GS-TOU-8 Interval Data Analysis
Greg, I've done several GS-TOU-8 analyses here in Atlanta. First thing is to verify the time zone stamps on the interval data match the tariff's defined time periods. Georgia Power sometimes provides data in Eastern Standard Time even during Daylight Saving Time, which throws off the analysis. Create a pivot table that sums kWh by hour of day and day of week, then compare against the tariff windows. Look for discrepancies in the shoulder months - April/May and October when the summer/winter TOU periods transition.
Greg, another thing to watch for is how GP handles partial interval reads. If a meter reading falls mid-month, they sometimes prorate the TOU periods incorrectly. I've seen cases where they apply full summer on-peak rates to April usage when the summer period doesn't start until May 1st. Also check if they're correctly excluding holidays from on-peak periods. The GS-TOU-8 tariff treats New Year's Day, Memorial Day, Independence Day, Labor Day, Thanksgiving, and Christmas as off-peak days.
I use a three-step validation process for TOU interval analysis. Step one: reconstruct the bill from scratch using the raw 15-minute data and published tariff rates. Step two: identify variances between your calculated bill and the utility's actual bill. Step three: document the specific time periods where discrepancies occur. For Georgia Power, pay special attention to the demand billing - they sometimes apply on-peak demand charges to off-peak intervals due to system errors.
Thanks for the guidance everyone. Rachel, you were right about the time zone issue. GP provided the interval data with EST timestamps year-round, but their billing system was applying TOU windows based on local time including DST adjustments. This created a one-hour shift for seven months of the year. I've identified about $4,200 in overcharges so far. Cecilia, your three-step process is exactly what I needed. The demand billing discrepancies are actually larger than the energy overcharges.
Greg, make sure you also check for interval data gaps or estimated reads. PPL here in Pennsylvania had a case where missing 15-minute intervals were being estimated using average hourly usage, which skewed the on-peak calculations. If Georgia Power has data gaps during peak hours, their estimation methodology might not properly account for the customer's actual usage patterns during those periods.
Following this thread with interest. Entergy Arkansas has similar TOU schedules and I've found their interval data analysis tools are pretty limited. Greg, what software are you using to process the 15-minute data? I've been using Excel but it gets unwieldy with large datasets. Thinking about investing in specialized utility billing software for interval analysis.
Helen, I'm using a combination of Excel and custom Python scripts for the heavy lifting. For basic analysis, Excel pivot tables work fine, but for multi-year datasets with thousands of intervals, you need something more robust. There are several utility billing software packages that handle interval data well - UtilityAPI and EnergyCAP are two I've evaluated. The key is having software that can properly handle time zone conversions and holiday calendars.
Greg, one more thing to check - verify that Georgia Power is correctly handling the transition dates between summer and winter TOU periods. I've seen utilities apply the wrong period rates during the transition months, especially for customers with mid-month meter read dates. The GS-TOU-8 schedule has specific effective dates for rate changes, but billing systems sometimes lag behind tariff updates.
Jim raises a good point about transition periods. Georgia Power typically implements TOU period changes on the first of the month, but if your customer has a mid-month read cycle, they might be getting prorated incorrectly. I've seen GP apply full summer rates to partial April bills when the summer period doesn't start until May 1st. Check the meter read dates against the TOU period effective dates in the tariff.
Update on this case - found additional errors in the transition periods that Jim and Rachel mentioned. Total recovery is now up to $6,800 over the three-year period. Georgia Power admitted to the billing system errors and processed the refund. Thanks everyone for the guidance on interval data analysis techniques. This thread has been incredibly helpful.
Greg, excellent work on that Georgia Power case. For others following this thread, interval data analysis is becoming increasingly important as more utilities move to advanced metering. The key is understanding not just what the tariff says, but how the utility's billing system interprets and applies those tariff provisions. Document everything and don't be afraid to challenge inconsistencies between the published rates and actual billing practices.