PG&E's SmartMeter program has been running for years now and I'm getting 15-minute interval data on all my commercial accounts. On one hand, it's amazing for load profiling and identifying demand anomalies. Caught a faulty HVAC controller that was cycling every 20 minutes, costing my client $340/month in unnecessary demand charges. On the other hand, I'm drowning in data. Anyone developed good tools for parsing this stuff efficiently? The raw Green Button XML exports are unwieldy as hell.
Smart meter interval data - blessing or curse for auditors?
Pete, I feel your pain. Alabama Power's Smart Meter data comes in 96 data points per day per meter. For large clients with multiple accounts, that's thousands of records monthly. I built a Python script that imports the CSV files and flags unusual patterns automatically. Looks for demand spikes >150% of baseline and interval-to-interval jumps >500 kW. Saved me probably 20 hours per audit. Happy to share the code if you're interested.
The interval data is gold for proving utility billing errors. Just caught Indianapolis Power & Light double-charging demand on a GS-3 account. Their billing system was reading both the monthly peak from the AMI and adding interval peaks during the same period. $2,100 error going back 8 months. Without 15-minute interval records, I never would have caught it. Old mechanical meters would have hidden this completely.
Greg makes a good point about billing system errors. Black Hills Energy in South Dakota had a software bug that was calculating TOU energy charges using peak period rates for all intervals during summer months. Affected about 200 commercial accounts for three billing cycles. Total overcharges around $180,000. The detailed AMI data made the pattern obvious once you knew what to look for. Mechanical meters would have just shown monthly totals - no way to catch the TOU miscalculation.
ComEd's AMI rollout has been a mixed bag. Love having real-time usage data for my clients, but their web portal is trash. Takes 3-4 clicks just to export interval data, and the CSV format keeps changing. Plus they're now charging $25/month for "detailed usage reports" that used to be free. Feels like they're monetizing data that customers should own. Anyone else getting hit with new AMI data fees?
Yuri, Xcel Energy Colorado started charging for interval data access last year too. $15/month per meter for "enhanced data services." Total scam considering we're the ones paying for the AMI infrastructure through rate riders. I've been telling clients to factor these fees into their audit cost-benefit analysis. For small accounts, the data fees might exceed potential savings. The utilities know exactly what they're doing.
Pacific Gas & Electric doesn't charge for basic interval data yet, but their Green Button downloads are limited to 13 months of history. Need older data? That's a $150 "research fee" plus $0.10 per day requested. Used to get 5+ years of billing history for free with a simple records request. AMI was supposed to make data more accessible, not create new revenue streams for utilities. This industry is going backwards on customer data rights.
Dan, that's exactly the problem. PG&E's 13-month limit killed a major audit I was doing for a food processing plant. Needed three years of data to establish baseline usage patterns after equipment upgrades. Had to pay their research fee and wait six weeks for historical records. Meanwhile, the billing errors were accumulating. Smart meters generate terabytes of data but utilities want to charge us for access to our own usage information.
Here's a positive AMI story: Eversource Connecticut's interval data helped me identify a power factor issue that was costing a machine shop $890/month in penalties. The 15-minute PF readings showed the problem was isolated to specific equipment cycling patterns. Fixed it with a $3,200 capacitor bank installation. Client is saving over $10K annually now. Old demand meters would have just shown monthly averages - couldn't have pinpointed the source this precisely.
Vince, that's the kind of analysis that makes AMI worthwhile. Arizona Public Service provides similar PF interval data, and I've found multiple cases where brief equipment faults were causing ongoing penalty charges. One hotel had a faulty elevator motor controller that was pulling leading power factor for maybe 3 minutes per day, but enough to trigger monthly penalties under Schedule E-32. $1,800 in avoided charges just by identifying the exact time intervals when PF went bad.
The diagnostic capabilities are impressive when they work. Xcel Minnesota's AMI caught a failing voltage regulator that was causing low voltage conditions during peak hours. My client's equipment was drawing 8-12% more current to compensate, inflating both energy and demand charges. Utility fixed the regulator and bills dropped by $420/month immediately. Problem had been going on for over a year with the old mechanical meter showing nothing unusual.
Bottom line: AMI interval data is incredibly powerful for identifying billing errors and system issues, but utilities are making it harder and more expensive to access. We need federal regulations requiring free customer access to their own usage data. Until then, factor data access costs into your audit proposals and keep fighting the good fight. The technology is there - we just need to keep the utilities honest about providing reasonable access.