Interval data vs billing data: which one your use case needs

When a new team starts collecting utility data, the first request is usually "everything." The second conversation, a few weeks later, is about why the dashboard has hourly kilowatt-hours but no dollar figures, or why the finance report has bills but cannot say when the building peaks. Interval data and billing data are different datasets that answer different questions, and choosing the wrong one, or assuming one contains the other, is the most common early mistake.

This guide explains what each dataset contains, which use cases need which, what can go wrong with each, and how to obtain both through the same connection. If you are new to interval reads, the smart meters, AMI, and interval resolution post is a good primer on where the data originates.

What interval data is

Interval data is consumption recorded at a fixed cadence by the meter: a reading for every 15 minutes, every hour, or every day, each with a start time and a duration. In Green Button, the Green Button Alliance notes that intervals can be 5-minute, hourly, daily, or monthly, depending on what an energy provider decides. Ontario's Green Button page describes the province's offering as natural gas and hourly electricity usage data. Hydro Ottawa, as one example, says its Green Button data uses the same intervals as captured by your smart meter (e.g., 15-minute, hourly, daily) and covers up to 24 months.

What interval data tells you is shape: when energy is used, how high the peaks are, what the overnight baseload looks like, and how a site responds to weather or occupancy. What it does not tell you, on its own, is what any of it cost.

What billing data is

Billing data is what the utility charged for a billing period: the period start and end dates, the consumption billed, and the charges. In Green Button, the usage summary carries fields such as payments, generation and distribution charges, tariff name, demand charges, and third-party charges. On a PDF bill, the same information sits in the line items and the meter-read table, which is what bill digitization extracts.

Billing data tells you cost, rate class, demand as billed, and the periods that finance reconciles against. It does not tell you what happened inside the month. A 31-day total cannot show a peak, a schedule problem, or an overnight load.

The two side by side

	Interval data	Billing data
Granularity	15-minute, hourly, or daily reads	One record per billing period (roughly monthly)
Source	Smart meter via the utility's Green Button feed or portal	Utility billing system, via Green Button usage summary or the PDF bill
Contains cost	No	Yes: charges, taxes, demand, rate class
Shows timing of use	Yes	No
Typical history	Up to 24 months at many Ontario utilities	Up to 24 months via Green Button; longer from bill archives
Typical consumers	Engineers, analysts, demand response, load studies	Finance, benchmarking, emissions reporting, budgeting

Which use cases need which

Use cases that need interval data

Demand charge management and peak analysis: you cannot see a 15-minute peak in a monthly total. The using interval data to manage commercial demand charges post shows the method.
Load profiling for solar, storage, or electrification sizing, where the hour-by-hour shape drives the design.
Scheduling and operations checks: equipment running overnight or on weekends shows up immediately in hourly reads.
Time-of-use analysis, where the split between on-peak and off-peak consumption is the question.
Anomaly detection and fault finding, which depend on comparing short intervals against a baseline.

Use cases that need billing data

Benchmarking in ENERGY STAR Portfolio Manager. Natural Resources Canada requires at least 12 full consecutive calendar months of energy data covering all fuel types, and the tool works from period totals, not intervals.
Ontario public sector and large building reporting, which asks for consumption by reporting year and is submitted through Portfolio Manager.
Emissions reporting, where consumption by fuel and period is multiplied by an emission factor.
Budgeting, bill validation, and accruals, which need charges and periods that reconcile to invoices.
Cost allocation across tenants or departments.

Use cases that need both

Many programs need both from the start. A retrofit business case uses billing data to establish baseline cost and interval data to model what changes. An energy manager uses billing totals for the monthly report and interval reads to explain why the total moved. A property team uses billing data for reporting and interval data to find the site that runs its chillers all night. Treating the two as one dataset from the beginning, with the same site and meter identifiers, avoids a reconciliation project later.

What goes wrong with each

Interval data has gaps. Meters lose communication, reads arrive late, and some values are estimated rather than measured. A feed that is 99 percent complete still has hours missing, and those hours are rarely random. Billing data has different failure modes: estimated reads that are trued up later, re-issued bills, billing periods that do not align to calendar months, and multi-meter accounts where the bill total does not match the sum of the meters. The Green Button data quality gaps and validation post catalogues these in detail.

Destination tools notice. ENERGY STAR reports that Portfolio Manager runs more than 70 data quality checks, and Ontario's reporting guide tells submitters to run the Data Quality Checker before filing. Finding those problems on arrival, rather than at the reporting deadline, is the difference between a routine submission and a scramble.

Checks to run on arrival

Coverage: every expected interval or billing period is present, with no gaps and no overlaps.
Continuity: each billing period ends where the next begins, and interval timestamps are contiguous.
Reasonableness: totals fall within the range of the same period last year and of neighbouring periods.
Reconciliation: for accounts with both datasets, the sum of intervals in a billing period approximates the billed consumption.
Estimates: estimated reads are flagged, and trued-up values replace them rather than duplicating.
Units: kWh, cubic metres, gigajoules, and cost are consistent across every site and source.

How to get both through one connection

In Ontario, both datasets travel through Green Button. Download My Data produces a file with interval reads and billing summaries; Connect My Data delivers the same on an ongoing, authorized basis. The Green Button data format explained post describes how interval blocks and usage summaries are laid out in the XML. Utilities also publish electricity, natural gas, and water through the standard, so a mixed portfolio can arrive the same way. Where a utility does not offer a feed, or a site only has paper bills, PDF bill digitization supplies the billing dataset and sometimes the monthly totals that stand in for intervals.

EZGB collects consumption and billing data from utilities through authorized utility connections, Green Button (Download My Data and Connect My Data), and PDF bill digitization. It standardizes interval reads and billing periods into consistent units and fields, with shared site and meter identifiers, and delivers both by REST API (JSON), CSV, or SFTP, plus embeddable components, into ENERGY STAR Portfolio Manager, RETScreen, Power BI, CRMs, ERPs, and ESG platforms. EZGB is part of MartinAI / Screaming Power, and MartinAI builds custom interfaces where a use case needs one.

If you are deciding what to collect for a new program, book a walkthrough at ezgb.ca. We will confirm which of your accounts provide interval reads, which only provide billing data, and how to deliver each to the tool that needs it.

Frequently asked questions

Can I calculate cost from interval data?

Only approximately. You can apply a published rate schedule to hourly consumption to estimate energy charges, but the actual bill includes demand charges, delivery charges, adjustments, and taxes that depend on the utility's billing rules. For anything finance will rely on, use billing data.

Can I get interval data from a PDF bill?

No. A bill shows period totals and sometimes the billed peak demand, not the underlying reads. Interval data comes from the meter through the utility's feed or portal. Bill digitization is the right tool for the billing dataset and for sites with no digital feed.

How much history is available?

It varies by utility. Hydro Ottawa and Enbridge Gas each describe up to 24 months through their Green Button downloads; EPCOR's Ontario page cites historical usage up to a maximum of 2 years. Older billing history usually has to come from your own bill archive, which is another reason to digitize it.

Does interval data exist for natural gas and water?

Sometimes. The standard supports it, and utilities such as London Hydro publish electricity, gas, and water usage data through Green Button. In practice gas and water are often daily or monthly rather than hourly. Check what each utility provides before designing an analysis around it.

Which should I start with?

Start with the question you need to answer. Reporting, benchmarking, and budgeting: billing data. Peaks, load shape, and operations: interval data. If both are on the roadmap, collect both from day one under the same identifiers so they reconcile.

Sources

Source: greenbuttonalliance.org