For most of the twentieth century a meter produced one number a month: total consumption since the last read. Advanced metering changed that. A modern smart meter records energy in short intervals and sends those reads back over a network, so a single account can generate thousands of data points a month instead of one. That shift is what makes interval analysis possible, but it also means the data you receive has a resolution, and resolution decides which questions you can answer. This piece explains what advanced metering infrastructure is, how widely it is deployed, and what each level of interval detail supports.
What AMI is
Advanced metering infrastructure (AMI) is the combination of smart meters, a communications network, and the systems that collect the data. The Federal Energy Regulatory Commission defines AMI as a metering system that records customer consumption at least hourly, and provides for daily or more frequent transmittal of those measurements over a communication network to a central collection point. Two properties define it: the meter records in intervals, and the reads travel back automatically rather than needing a person to visit the meter.
How widespread it is
Deployment is now the norm rather than the exception. The US Energy Information Administration reports that in 2022, US electric utilities had about 119 million AMI installations, roughly 72 percent of all electric meters. Residential customers accounted for about 88 percent of those installations, and about 73 percent of residential electric meters were AMI. The EIA marked the point where nearly half of all US electricity customers had smart meters several years earlier, and the count has kept climbing. FERC tracks the same trend in its annual Assessment of Demand Response and Advanced Metering.
The practical consequence is that for a large and growing share of accounts, interval data already exists at the meter. The question is usually not whether the data is being recorded, but whether you can get it out in a usable form and at a resolution that fits your analysis.
Resolution, and what each level supports
Interval data is meter data collected at a fixed cadence. In practice the common resolutions are 15 minute and hourly for interval reads, with monthly totals available from billing regardless. Each level answers a different class of question, and using a coarser resolution than your question requires is a frequent, avoidable mistake.
| Resolution | What it shows | Typical use |
|---|---|---|
| Monthly (billing) | Total consumption and cost per period | Benchmarking, cost tracking, energy intensity, emissions inputs |
| Hourly | The daily load shape and when energy is used | Time of use analysis, occupancy patterns, baseload versus peak |
| 15 minute | Short peaks and rapid changes in draw | Demand charge analysis, anomaly detection, load control |
A monthly total is enough to benchmark a building against its peers or to feed an emissions calculation, but it cannot tell you when energy was used. Hourly data reveals the load shape: whether a site has a sharp midday peak, a flat industrial profile, or overnight load that should not be there. Fifteen minute data is what you need for anything involving short peaks, most importantly the demand charge, which is typically billed on the single highest 15 minute interval of the month. Ask a 15 minute question of hourly data and the peak you care about is averaged away before you ever see it.
If you are not sure what resolution your accounts actually provide, that is worth checking before you scope any analysis. Book a walkthrough at ezgb.ca, and we will confirm what each utility exposes and get a standardized feed flowing at the resolution your work needs.
How the data reaches you
Green Button is the standard way interval data leaves the utility in a consistent shape. In the ESPI format, interval reads are expressed as interval blocks and interval readings with explicit start times and durations, so a 15 minute read and an hourly read are unambiguous about the window they cover. Because the schema is standardized, interval data from one utility parses like interval data from another, which is what lets you build one pipeline rather than one per utility.
EZGB (Easy Green Button Connector) collects that interval and billing data from authorized utility connections and from Green Button, standardizes it into consistent units and clean fields, and delivers it by REST API, CSV, or SFTP. Where a utility only exposes monthly billing detail, EZGB provides that; where it exposes 15 minute or hourly interval data, EZGB provides the finer resolution. For accounts whose only record is a PDF bill, BillScan turns the bill into structured monthly data in the same shape.
A few realities to plan for
Interval data is detailed, but it is not automatically perfect. Reads can have gaps when a meter loses communication, some values may be estimated rather than measured, and the finer the resolution, the larger the volume you store and process. None of these are reasons to avoid interval data; they are reasons to validate it and to match resolution to purpose. Pull monthly totals for benchmarking, hourly for load shape, and 15 minute for demand and anomaly work, and you get the detail you need without drowning in data you do not.