Need some advice on a challenging case with NorthWestern Energy here in South Dakota. Large commercial customer on Schedule 15 TOU rate has interval metering, but when I add up the 15-minute interval data for on-peak periods, it doesn't match the on-peak kWh shown on their monthly bills. The difference is significant - about 8-12% variance each month. NorthWestern claims their billing system is correct, but the math doesn't work. Has anyone encountered similar discrepancies with interval data vs. billed amounts? Could this be a meter reading issue or billing system error?
NorthWestern Energy TOU Billing - Interval Data Doesn't Match
Diane, I've seen this type of discrepancy with Avista where the interval data export was missing certain time periods due to communication errors between the meter and their head-end system. Check if there are any gaps in the 15-minute interval timestamps. Even short communication outages can result in missing data that gets estimated differently for billing purposes. Also verify that the interval data you're analyzing covers the exact same billing period dates as the monthly bill.
Diane, another possibility is multiplier errors. Some commercial meters have current transformers or potential transformers that require multipliers to calculate actual kWh usage. If the interval data export is showing raw meter readings without multipliers applied, but the billing system is correctly applying multipliers, that could explain the 8-12% variance. Check the meter nameplate and compare the CT/PT ratios against what's being used in your analysis.
Nancy and Kim, thanks for the suggestions. I checked for data gaps and found several instances where 2-3 consecutive 15-minute intervals were missing, usually during overnight hours. But that wouldn't explain the systematic variance across all months. Kim, the multiplier theory is interesting. The meter does have CTs with a 400:5 ratio (80x multiplier), but I've been working with the interval data as provided by NorthWestern. Let me verify if they're providing raw or multiplied values.
Diane, I'd also check the time synchronization between the interval data timestamps and the TOU billing windows. Even a small clock drift in the meter can shift usage from on-peak to off-peak periods or vice versa. Duquesne Light here in Pittsburgh had an issue where their AMI meters were losing time sync, causing TOU billing discrepancies. The 8-12% variance you're seeing could be explained by usage shifting between rate periods due to timing errors.
Following up on Walt's timing comment - I've seen cases where the utility's interval data export uses a different time reference than their billing system. For example, the interval data might be timestamped in UTC while the billing system applies TOU windows based on local time. This creates discrepancies especially during daylight saving time transitions. Ask NorthWestern to clarify the time zone and daylight saving treatment for both their interval data exports and TOU billing calculations.
Update - found the root cause! Kim was right about the multiplier issue. NorthWestern was providing interval data in raw meter units (not multiplied by the CT ratio), but I was comparing against billed kWh which had multipliers applied. Once I applied the 80x multiplier to the interval data, the numbers matched perfectly. Lesson learned: always verify whether interval data exports include meter multipliers or if you need to apply them manually during analysis.
Diane, glad that solved it! The multiplier issue is one of the most common mistakes in interval data analysis. For others reading this thread, always check the meter configuration and verify how multipliers are handled in both the interval data export and your analysis spreadsheet. A simple 80x error can make a perfectly accurate billing look completely wrong.
This is a great reminder about the technical details that can trip you up in interval data analysis. I keep a checklist now that includes: verify time zones, check for data gaps, confirm multipliers are applied correctly, validate holiday calendars, and cross-reference TOU window definitions. These technical details matter a lot when you're trying to validate billing accuracy against raw meter data.
Helen's checklist approach is smart. I'd add one more item: verify the billing determinant methodology. Some utilities use the highest 15-minute interval for demand billing, others average multiple intervals, and some use the highest hour. If your interval analysis assumes one method but the utility uses another, you'll get discrepancies even with perfect data. Always confirm how they calculate billing determinants from the raw interval readings.
Gil makes an excellent point about billing determinant methodology. Idaho Power uses a 15-minute window for demand billing, but they calculate it as the average of the four 15-minute intervals within each hour, then bill the highest hourly average. If you're just looking for the single highest 15-minute peak, you'll get different results than their billing system. These calculation nuances can be buried deep in the tariff language.
This thread highlights why interval data analysis is both powerful and complex. The data gives you unprecedented visibility into billing accuracy, but you need to understand all the technical details of how the utility processes that same data. Small errors in time zones, multipliers, or calculation methods can completely invalidate your analysis. Take time to understand the utility's methodology before questioning their results.
Sylvia, you've perfectly summarized the key lesson from this case. The interval data is incredibly valuable for auditing TOU billing, but you have to make sure you're analyzing it the same way the utility does. Thanks to everyone for the guidance - this thread probably saved me from making similar mistakes on future cases.
Excellent discussion everyone. This thread demonstrates the importance of understanding the technical infrastructure behind utility billing. As more utilities deploy advanced metering and TOU rates, these interval data analysis skills become essential for effective auditing. The devil is truly in the details - time zones, multipliers, calculation methods, and billing determinants all matter. Keep sharing these technical insights as they help the entire community improve our audit capabilities.