For most of the twentieth century, a distribution utility had little direct visibility below the substation — nothing like what smart meter data would later provide. Billing was built around a monthly manual reading: a single number for total energy consumed since the previous visit, with no information about when consumption occurred, what voltage the customer experienced, or whether power had failed momentarily. Network visibility stopped at the feeder. Outages were usually discovered when customers called.
Smart meters changed part of that picture. The most commonly discussed application—remote billing—is only a narrow slice of what the technology makes possible. Once interval data exists at millions of endpoints, the more interesting question is how that data can support outage management, load forecasting, and the integration of distributed energy resources. This article explains what smart meters actually produce, and why the operational value is still being worked out.
The Data a Smart Meter Produces
A smart meter is a digital meter with two-way communication. It records energy consumption in short intervals—typically 15, 30, or 60 minutes, depending on local configuration—along with timestamps, voltage readings, and event flags for conditions such as power loss, restoration, or tampering. Two-way communication separates these devices from earlier automated meter reading systems, which generally transmitted a monthly total in one direction and provided little operational information.
Data flows from the meter through a local network—often radio frequency mesh, power line carrier, or cellular—to a head-end system, then into a meter data management system that validates, stores, and distributes the information. The communication and automation layer that carries this data is part of the wider build-out of digital grid infrastructure.
A constraint that receives less attention is that advanced metering networks are not normally designed for continuous streaming. Many utilities collect meter data in daily batches to reduce backhaul and communication costs. That architecture is adequate for billing, which tolerates delay, but it creates friction when the same data is expected to support operations that need near-real-time awareness.
The interval length itself matters. Shorter intervals capture fast ramps and voltage spikes but increase data volume and communication load. Longer intervals reduce cost but obscure short-duration events such as motor starts or momentary interruptions. Regulators in some jurisdictions specify minimum interval requirements, which is why meter configurations differ across markets.
Turning Meter Data into Outage Awareness
The earliest operational use of smart meters was outage detection. A meter can send a final message, often called a last gasp, when it loses power, and a first message when service returns. When a control centre correlates these signals across hundreds of meters on the same distribution transformer or feeder, it can distinguish a single-premise problem from a transformer-level or feeder-level event.
That capability changes repair dispatch. Instead of relying on customer calls and feeder-level alarms, operators can direct crews to a specific transformer or lateral, narrowing the search to a handful of premises. Grid modernization programs have made this capability more common outside large urban utilities, although implementation quality varies widely.
The trade-off is that a missing meter read is not the same as a power outage. Communication failure, battery depletion, or blocked radio paths can all produce silence. Operators therefore treat meter events as a probabilistic signal. A single silent meter may mean nothing; a cluster of silent meters on the same service area is a strong indicator of an outage, but even then a failed concentrator can mimic that pattern.
The same data can also indicate momentary interruptions—events shorter than a few minutes—that customers may not report but that can damage industrial equipment. A brief loss of supply can trip a production line or force a controller restart, and because the event does not escalate into a sustained outage it rarely generates a trouble ticket. Utilities can aggregate counts of these events by feeder or service area to identify segments with repeated problems and prioritise reliability spending accordingly.
Load Forecasting and Distribution Planning
Traditional load forecasting for distribution planning often relies on feeder measurements and engineering assumptions. Below the substation, transformers and secondary networks have historically been unmeasured. A planning engineer might assume a conservative coincidence factor when sizing a transformer, because no data exists to show how much load is actually simultaneous.
Interval meter data changes that. By aggregating reads from premises connected to the same transformer, a utility can estimate the transformer’s actual peak and how much headroom remains. That becomes especially relevant when electric vehicle charging or heat pumps concentrate load on a small number of premises. Without interval data, a utility may not know that several homes on one transformer adopted overnight charging and now share a similar charging window. With aggregated data, the pattern is visible before the transformer overheats.
Most utilities use this information in aggregated form rather than monitoring individual households. Readings are typically grouped by transformer or feeder, which is the level at which distribution planning actually operates. Privacy rules and customer expectations limit how granular consumption data can be used, and aggregated load profiles are usually sufficient for planning purposes.
Supporting Distributed Energy Resources
Smart meters have a less obvious role in integrating distributed generation. A standard rooftop solar installation typically causes the meter to report net consumption: imported energy minus exported energy. The meter does not directly show gross generation or gross load without a separate production meter or inverter telemetry. This creates a blind spot for distribution operators, because a midday net export could hide a large load behind an even larger solar output.
Even with that limitation, meter data contributes to voltage management. Reverse power flow from distributed solar can raise voltage at the ends of feeders, and meters that record voltage at customer locations give operators evidence of where excursions are actually occurring. Some utilities combine these measurements with distribution automation to adjust tap settings or reactive power resources.
Meter data also supports demand response and distributed energy aggregation. In the United States, FERC Order 2222 requires regional grid operators to allow aggregated distributed energy resources to participate in wholesale markets. Aggregators use interval meter data to demonstrate that a portfolio responded as dispatched, and to settle payments. The choice of baseline methodology—what a customer would have consumed without the event—remains a source of dispute, because small changes in baseline construction can shift payments materially.
Why Operational Use Still Lags Billing
Despite the available capabilities, many utilities still use smart meters primarily for billing and basic outage notification. Several constraints explain the gap. Some are technical, some organizational.
- Data systems are fragmented. Meter data management, outage management, geographic information, and planning systems often evolved separately and do not exchange data easily.
- Data quality is uneven. Missing intervals, clock drift, and firmware differences across meter fleets require cleaning before operational use.
- Latency matters. Daily batch collection is common and sufficient for billing, but it reduces the value of the data for near-real-time grid operations.
- Privacy and regulatory constraints limit how individual-level data can be used, which pushes many applications toward less granular aggregation.
- Analytics investment is uneven. Utilities may have the data without the tools and operational workflows to use it for planning and control.
IEA analysis of digitalisation in energy has noted that collecting data does not automatically create operational value. The institutions, integration work, and analytical capacity around the data are often the binding constraint rather than the meter itself. Raw interval readings only become useful once they are integrated into operational workflows, validated, and interpreted by staff with the training to act on them.
What Changes as Meter Data Becomes Operational
A gradual shift is underway from batch collection toward event-driven messaging for selected high-value locations. Instead of waiting for the daily batch, some utilities configure certain meters to report immediately on voltage excursions, outage events, or other triggers, while retaining periodic collection for the wider fleet. This selective approach balances communication cost with operational benefit.
Standards also shape how meter data connects to other systems. IEEE 2030.5, for example, defines a protocol for communication with distributed energy resources, including behind-the-meter devices, which allows meter infrastructure to participate in broader grid control ecosystems. These connections are part of the sensor and automation architecture that underpins emerging distribution operations.
Meter data complements feeder automation, substation telemetry, and dedicated power quality monitors rather than replacing them. Its value depends on how well the surrounding systems convert raw reads into decisions. The continuing work is less about installing more meters than about making the existing data available at the right time, in the right form, to the right operational process.
References
- IEA — Digitalisation and Energy (smart meter data, grid digitalisation, demand response context)
- FERC — Order No. 2222 (distributed energy resource aggregation and wholesale market participation)
- IEEE — IEEE 2030.5 (communication protocol for distributed energy resources and smart inverters)
