Billings folks - running into weird issue with NorthWestern Energy interval data accuracy. Hospital client has their own revenue-grade metering system that records 15-minute intervals, and we're seeing consistent 3-5% discrepancies vs what NWE reports for TOU billing. Client's Schneider PowerLogic system shows 2,847 kWh for January 8th peak period (1pm-7pm) but NWE billed 2,976 kWh for same window. That's $67 difference just for one day at Schedule E-3 peak rates. Meter accuracy testing shows both systems within +/- 0.2% specification. Anyone dealt with interval data validation disputes like this?
Interval data validation - utility vs customer meter readings don't match
Wendell that's a significant variance for revenue-grade systems. Up here with PSE we've seen similar issues caused by CT ratio differences or timestamp synchronization problems. First check - are both meters using identical CT ratios and PT ratios? Even small multiplier differences compound over daily intervals. Second issue could be interval boundary definitions - some utilities read "ending at" vs "beginning at" for 15-minute periods. Hospital loads change rapidly so interval timing matters. Also verify both systems using same timezone and DST settings.
David's right about CT/PT ratios being critical for accuracy. Also seen interval boundary timing issues cause systematic discrepancies like what you're describing Wendell. Hospital environments are tricky because loads fluctuate rapidly with medical equipment cycling. The 3-5% variance suggests systematic error rather than random measurement differences. Request detailed interval exports from NWE showing actual meter register values before any scaling factors. Compare raw pulse counts if possible, not just calculated kWh values.
Randy and David covered the technical basics well. From billing dispute perspective, you need to establish which meter reading has legal precedence for rate calculations. Usually utility meter wins unless customer can prove systematic error. The $67 daily difference adds up fast - probably $2000+ annually on TOU peak charges alone. Document everything and consider requesting independent meter accuracy verification. Duke Energy Ohio has process for third-party meter testing when discrepancies exceed 2% consistently.
Cecilia's suggestion about independent testing is good but expensive for most clients. Wendell what's the hospital's monthly TOU billing impact from these discrepancies? If it's substantial, worth pursuing. Alabama Power required independent verification when we challenged their AMI accuracy on large medical campus. Key was proving consistent variance pattern across multiple months, not just isolated incidents. Hospital client probably has detailed energy management records that support your case - use their historical data to establish credibility of their metering system.
Albert makes good point about using historical data patterns. Down here in Kentucky with KU and LG&E we see interval discrepancies usually trace back to communication errors between AMI network and billing systems. The consistent 3-5% variance suggests data processing issue rather than meter hardware problem. Wendell have you checked if NWE applies any load profiling or estimation during communication gaps? Some utilities fill missing intervals with estimated data that can skew TOU calculations. Request communication logs showing successful vs failed meter reads for disputed periods.