How do you read half-hourly electricity interval data?

Start by matching the meter point, dates, units and import or export direction to your bill. Then check missing intervals and reading quality before plotting your operating pattern. NESO describes a normal settlement day as 48 half-hour periods [S1]. Convert interval energy into average power before assessing extra equipment, and keep fixed bill charges separate. Checked October 2026.

Last checked 18 October 2026

What do the columns in half-hourly data mean?

Keep the original download and work on a copy. Ask whoever supplied it for a field description before changing column names or deleting flags. Your first task is to establish what each value measures, rather than immediately adding up the spreadsheet.

A Meter Point Administration Number, or MPAN, identifies the metering point. Elexon describes its core as a 13-digit identifier [S2]. Compare it with your bill and use the MPAN explainer if the bill shows a larger supply-number box. Preserve identifiers as text when importing the file.

Use this checklist to ask for clarification. It describes fields to look for, rather than a compulsory CSV layout.

The government energy glossary defines kW as power and kWh as energy [S3]. Do not treat those labels as interchangeable. Ofgem also distinguishes import and export MPANs [S5]; retain that distinction when joining files.

How do you convert interval kWh into average kW?

Divide the interval’s energy by its duration in hours. This follows directly from the definition of a kilowatt-hour [S3]. For an illustration, not a quote, assume the following readings from a workshop. They are invented inputs solely to show the calculation, not measured customer data.

Add the energy column to find total kWh. Find the largest converted average to locate the highest interval demand. Avoid summing the kW column and calling the result energy. Keep the interval length alongside every conversion so the calculation remains auditable.

An interval average cannot reveal every shorter surge. Treat it as a starting point for your electrical designer, who should decide whether faster monitoring is needed. See what a load profile shows for the connection between timing and site demand.

Why can clock-change days have different numbers of periods?

NESO’s settlement-period definition describes 48 half-hour segments on a typical day, 46 on a short day and 50 on a long day [S1]. A local-clock file therefore needs special handling around daylight-saving changes. A file using Coordinated Universal Time, or UTC, has a different day boundary: ask the exporter to confirm its convention.

Do not force every local day into the same row length. On the autumn change, repeated local clock labels need a time offset or another distinguishing identifier. Otherwise a spreadsheet may remove valid records as duplicates. On the spring change, check the convention before treating the shorter day as missing consumption.

Retain both the original period identifier and its timestamp in your working copy. Ask for a corrected export if the file cannot distinguish repeated times. Align generation, charger and meter records to the same time basis before combining them.

How do you spot baseload, peaks, weekends and gaps?

Plot average power through the day, then separate trading days, weekends and shutdowns. Annotate opening hours, cleaning, refrigeration and production schedules. Look for the recurring level when the main business activity has stopped: that is your working estimate of baseload, not automatic evidence of waste.

Investigate unexpected overnight steps or peaks with the person running the site. Compare similar operating days rather than treating an exceptional delivery or maintenance day as representative. Keep shutdown periods visible and explain them instead of silently deleting them.

Prepare a quality log. Mark blanks, repeated timestamps, unexplained flat runs and unusually large readings. Ask the data supplier what each quality code means; do not assume a single-letter code has the same meaning in every export. Ofgem distinguishes actual half-hourly consumption from estimates in its metering guidance [S4]. Record where your assessment still relies on estimates.

Never replace blanks with zero without evidence. Request missing intervals, identify the affected date range and ask the designer to show how unresolved gaps affect the conclusion.

What should you compare with your electricity bill?

Match the MPAN, direction and exact billing window first. Sum the corresponding import kWh, then reconcile them with the bill’s energy quantities. Government guidance identifies kWh as the usage unit shown on electricity bills [S3]. Check meter changes and revised data before deciding that a difference is a billing error.

Separate energy quantities from charges. A capacity line concerns reserved network capacity, as Ofgem explains [S4]; it is not another batch of consumed kWh. Read the metering and DUoS guide for charging components, and use business electricity price data for market context rather than reconstructing your contract from a national average.

What does a solar or EV assessment need from the file?

Prepare a full seasonal cycle where available, your operating calendar, bills and a list of planned equipment. State explicitly which months are missing. For existing solar or batteries, provide generation and operating data alongside grid import and export. Export metering measures flows at the grid connection [S5]; import alone does not describe all the electricity used behind that connection.

For EV charging, add expected arrivals, departures and charging requirements. Ask the designer to overlay those loads rather than assume that historic demand already includes them. Use the interval-data request template to assemble the records and the charging-cost tool to explore energy-cost assumptions separately.

What should you do next?

Start with the cut energy costs hub and site check, then get matched with an installer that covers your area. Ask for an assessment that identifies data gaps and shows its assumptions.

Piggy Pineapple Ltd operates this referral service. We refer enquiries to installer partners and may be paid a fee. Our owner holds shares in at least one partner. Before passing on details, we name the proposed partner and disclose any shareholding our owner holds in it. Read how we work.