Strange situation with a ComEd customer here in Springfield. Their interval data is showing negative demand readings during certain 15-minute periods, particularly during overnight hours. We're talking about -50 to -150 kW readings when the facility should never export power. The customer has no generation equipment, just typical commercial loads. ComEd is billing these as zero demand, but I'm concerned there's either a meter wiring issue or a billing system problem that could be affecting accuracy. Anyone seen negative demand readings like this before?
ComEd interval data showing negative demand readings - billing system error or meter issue?
Darlene, negative demand readings usually indicate CT polarity issues or phase rotation problems. If the current transformers are installed backward or if there's a phase reversal in the meter connections, you can get negative readings. This is definitely a meter installation issue, not a billing system problem. ComEd needs to send a meter technician to check the CT installation and phase connections. Don't let them dismiss this as a minor issue.
Derek's right about the CT polarity issues. I've seen this with FirstEnergy in Ohio. The negative readings are a red flag that the meter isn't measuring accurately, which means all the positive demand readings could be wrong too. If they're billing the negatives as zero, they might be undercharging for actual positive demand periods. This could go both ways - you might find underbilling or overbilling once it's fixed.
Had a similar case with National Grid in Syracuse. Turned out to be a loose connection in the CT secondary wiring that was causing intermittent negative readings. The key is to correlate the negative readings with weather patterns, operational schedules, or any maintenance work. If the negatives happen randomly, it's likely a wiring or connection issue. If they follow a pattern, it might be related to specific equipment cycling.
Thanks for the insights everyone. I looked at the pattern more carefully and the negative readings always occur between 2 AM and 5 AM, which is when their HVAC system cycles off. That timing suggests it's not random wiring issues but possibly related to how the meter handles very low load conditions. Could this be a meter sensitivity or resolution issue when loads drop to near-zero levels?
Darlene, that timing pattern is interesting. In Alaska, we've seen some newer smart meters have issues with very light loads, especially if there's any power factor correction equipment that creates leading power factor conditions. Capacitor banks or newer LED lighting with power factor correction can sometimes create apparent "reverse power flow" in meters that aren't properly calibrated for leading power factor loads.
Georgia Power has had similar issues with some Landis+Gyr meters when customers have significant power factor correction equipment. The meter might be measuring reactive power flow and incorrectly interpreting it as reverse real power. I'd check if your customer has any capacitor banks or power factor correction equipment that operates on timer controls during those overnight hours.
This is getting into some pretty technical meter behavior territory. Idaho Power has protocols for testing meter accuracy under light load conditions, and many older meters lose accuracy when loads drop below 1% of full scale. If this facility's overnight load is very light, the meter might be hitting its accuracy limits. ComEd should perform a full meter test including light load accuracy verification.
Update - ComEd sent a tech out and found exactly what Greg suspected. The customer has automatic capacitor banks that switch on during peak hours and off during overnight hours. When they switch off, the meter briefly sees reverse reactive power flow and interprets it as negative demand. ComEd is upgrading to a newer meter that handles power factor transitions better. Great troubleshooting help from everyone!
Excellent detective work on this case! The capacitor switching explanation makes perfect sense. This is a good reminder that interval data analysis often reveals equipment operational patterns that aren't obvious from just looking at monthly bills. The negative readings were actually diagnostic information about the customer's power factor correction system. Smart analysis by the whole group.
This thread shows why interval data analysis requires both billing expertise and electrical engineering knowledge. The interaction between capacitor switching and meter behavior isn't something most billing departments understand. Darlene's systematic approach to analyzing the timing patterns was key to solving this mystery. Great collaborative troubleshooting!
Cases like this are why our profession requires continuous learning. Ten years ago, most of us wouldn't have encountered smart meters with power factor interaction issues. Now it's becoming common as more customers install LED lighting and power factor correction equipment. The technical complexity keeps increasing, but so does the value we provide to our clients.
The power factor correction angle is something I hadn't considered for negative demand readings. In Alabama, most of our commercial customers have pretty basic electrical systems, but as more facilities upgrade to efficient equipment, we're going to see more of these complex meter interactions. This case study should be required reading for anyone doing interval data analysis.
This has been an outstanding technical discussion that demonstrates the expertise of our AAUBA community. Darlene's systematic approach to analyzing the negative readings, combined with everyone's technical knowledge about meter behavior and power factor interactions, solved a complex problem that most utility billing departments would have missed. This type of collaborative problem-solving is exactly what makes this forum so valuable for our professional development.