Been auditing a manufacturing client in Chicago and something weird is happening with their new AMI meters ComEd installed last fall. Comparing usage patterns from the old mechanical Schlumberger meters versus the Itron smart meters, we're seeing consistent 12-15% higher readings on similar production days. The client is screaming about a $3,200/month increase on their Schedule D-1 rate. Has anyone else run into calibration issues with these newer AMI installations? The meter accuracy specs claim +/- 0.2% but something doesn't add up here.
ComEd smart meters reading 15% higher than old mechanicals - anyone else seeing this?
Yuri, I've seen similar issues down here with Alabama Power's smart meter rollout. Had a textile mill client where the new Landis+Gyr meters were reading about 8% higher than their old GE I-70S mechanicals. Turns out the old meters had been running slow for years due to worn bearings - the smart meters were actually giving accurate readings for the first time in a decade. Did you check the test reports on those old Schlumbergers? Sometimes what looks like smart meter over-registration is really just correcting years of under-billing.
Val makes a good point about mechanical meter drift. Here in Missouri, Empire Electric had a batch of old Westinghouse disk meters that were running 5-10% slow after 15+ years in service. When they swapped them for Sensus iCon AMI meters, customers thought they were being ripped off but the new meters were just accurate. Yuri, I'd recommend requesting ComEd pull one of those old mechanicals for bench testing. If it's running slow, you can usually get billing adjustments going back 2-3 years depending on Illinois commission rules.
This is exactly why I always document meter accuracy before any changeouts. Georgia Power tried to pull the same thing on one of my clients - claimed the 18% usage increase was due to 'improved accuracy' of their new Aclara meters. Fought them for 6 months and finally got them to admit the old meter was within tolerance. Turns out there was a programming error in the AMI system that was applying a 1.15 multiplier. Got my client a $47,000 refund. Don't let the utility blame it on old meter accuracy without proof.
Derek, that's interesting about the multiplier error. I'll definitely push ComEd to provide the bench test results on the removed mechanicals. The client has detailed production logs going back 5 years, so we can correlate kWh/unit production pretty accurately. If those old meters were truly running slow, the production efficiency ratios should have improved significantly after the smart meter install - but they haven't. This smells like a systematic problem with the AMI deployment.
Yuri, have you checked the pulse output settings on those Itron meters? We had an issue with PG&E where the AMI system was configured for 1 pulse per kWh but the billing system was expecting 10 pulses per kWh. Created a 10x error that took months to catch. Also make sure ComEd isn't double-counting any CT ratios - seen that mistake too many times. The Schedule D-1 has some tricky demand ratchet provisions that can amplify billing errors.
This thread is gold! Taking notes on all these AMI gotchas. Out here in Nevada, NV Energy has been pretty good about their smart meter rollout but I've heard horror stories from other territories. The key is getting that documentation before the old meters disappear. Once they're scrapped, you're at the mercy of the utility's word about accuracy. Always demand those test bench reports!
Update: ComEd finally provided the test results on 3 of the removed mechanical meters. Two were reading within +/-1% of actual, but one was running 4% slow. Still doesn't explain the 15% increase across the board. Digging deeper into the AMI data and found something suspicious - the interval data shows power factor readings that don't match our client's actual PF measurements. Starting to think there's a CT wiring issue or incorrect meter programming. Will report back once we get ComEd out for a field inspection.