Raleigh office building client switched to Duke Energy smart meters in January. Now seeing 15-minute interval demand readings of 850 kW when building load never exceeds 400 kW. Mechanical meter never recorded anything over 420 kW. Anyone seen phantom demand readings?
Duke Energy AMI interval data shows demand spikes that don't exist
JEA in Jacksonville had similar issues with Landis+Gyr smart meters. Voltage transients during switching operations were being recorded as demand spikes. Firmware update fixed it.
Arizona Public Service had phantom readings too. Problem was interference from nearby radio transmitters. Had to install RF filters on the meter sockets.
Avista in Spokane saw this with Elster meters. Current transformer polarity was reversed causing negative power factor readings to flip positive and double the demand calculation.
Update: Duke sent a tech out with an oscilloscope. Found loose connection in the CT cabinet causing voltage spikes during load switching. Fixed the connection and demand readings dropped back to normal.
Had the same issue with Eversource in Hartford. Loose CT connection was causing erratic readings. Always check physical connections before blaming the smart meter.
Phantom demand readings are usually wiring issues not meter problems. Smart meters are actually more accurate than mechanical meters for detecting electrical system faults.