Anyone else seeing weird demand spikes in LG&E interval data that don't match actual usage patterns? I've got a client with Schedule RTS-2 showing 15-minute peaks of 847 kW when their actual load never exceeds 650 kW. The spikes always happen at exactly :00, :15, :30, :45 intervals which seems suspicious. Their demand billing jumped from $8,200 to $12,400 last month because of these phantom readings. Has anyone successfully challenged LG&E on interval data accuracy issues?
LG&E interval data showing phantom demand spikes?
Jack, I've seen similar issues with Alabama Power's AMI system. Usually it's a meter communication problem or voltage transformer issues. Request the raw pulse data from LG&E - sometimes the interval aggregation software double-counts pulses. Also check if there were any power quality events during those spike times. I caught APCo billing a client an extra $15K because their meter was recording phantom demand during momentary outages.
This happened to one of my Gulf Power clients in 2014. The interval data showed impossible load jumps that violated basic electrical physics. Turned out to be a firmware bug in the Landis+Gyr meters. Gulf Power eventually credited back $23,000 in erroneous demand charges after I provided load research data proving the spikes were impossible given their connected load. Document everything and demand a meter accuracy test.
Thanks for the feedback. I'm pulling together load research data from their building automation system to compare against the LG&E interval data. The phantom spikes always occur when the building is unoccupied (weekends, holidays) but somehow show higher demand than peak operational hours. That's physically impossible for this facility. Going to request a formal meter test under Kentucky PSC regulations.
Jack, make sure you get the CT accuracy tested too, not just the meter. Duke Energy Ohio had a batch of current transformers that were reading 15% high during certain load conditions. Also request the meter installation and maintenance records - sometimes loose connections cause erratic readings that spike the interval data. I've recovered over $40K from similar cases by proving meter/CT system errors.
Had a similar case with Entergy Louisiana last year. Customer was getting billed for demand spikes that occurred during scheduled maintenance shutdowns when all equipment was physically disconnected. Turned out the meter was storing ghost readings in memory and dumping them randomly into interval data. Entergy credited back $18,500 after we proved the facility was completely de-energized during the phantom demand periods.
Update: LG&E agreed to a full meter test after I showed them the interval data anomalies. Turns out the demand register was malfunctioning and double-counting certain 15-minute intervals. They're replacing the meter and crediting back $14,200 in erroneous charges. Thanks everyone for the advice on documenting the load research data - that's what convinced them to take action.
Great outcome Jack! Always satisfying when the utility admits their metering error. Make sure they provide written confirmation of the credit amount and adjust any capacity charges that were based on the erroneous demand readings. Sometimes they forget to fix the downstream billing impacts.