Guides

Using smart-meter data to build a load profile

Published 8 October 2026. About 6 minutes to read.

Monthly bills tell you how much electricity a building uses. Interval data from a smart meter tells you when, and for solar with a battery, when is what decides the design.

What an export looks like

Most utilities and meter portals can export a CSV file with one row per interval: a date and time, and a reading. Intervals are usually 15, 30 or 60 minutes. A full year of 30-minute data is 17,520 rows.

Before using the file, check three things.

1. Is it energy or power?

Some exports give energy in kWh for each interval. Others give the average power in kW over the interval. They look similar but differ by a factor that depends on the interval length.

kWh per interval = average kW × interval in minutes ÷ 60

A reading of 2.0 in a 30-minute file is 2.0 kWh if the file is in energy, but only 1.0 kWh if it is in kW. Getting this wrong doubles or halves the household's consumption. The column header or the portal's help page usually says which it is. If in doubt, add up a month and compare it with the bill.

2. Which time does each row mean?

A row stamped 13:30 may cover 13:00 to 13:30 or 13:30 to 14:00. Some meters record in local time, others in UTC, and in countries with daylight saving there may be a missing hour in spring and a doubled hour in autumn. A half-hour shift barely matters for a load profile, but a whole timezone offset moves the evening peak into the afternoon and makes the battery look far more useful than it is.

3. Is there already solar on the building?

An import meter only sees what comes from the grid. If the building already has solar panels, the daytime readings are what was left after the existing system did its work. The household's real daytime use is higher. For an upgrade or an added battery, you need the existing system's production too, usually from its inverter's monitoring portal, and add it back in.

Gaps and odd days

Exports often have missing intervals from communication faults, and a year may include holidays or a month when the house was empty. A typical day for each month, averaged over all the days with data, smooths most of this out. Months with no data at all can borrow the average shape from the other months.

From profile to design

A typical day per month, at half-hour resolution, captures what matters for design: the evening peak, the overnight base load, and how both change between summer and winter. It is enough to simulate how solar covers daytime use and how a battery shifts surplus into the evening.

Importing data into Lekker Sun

On the Finance page, choose whether the values are kWh per interval or average kW, and upload the CSV. Lekker Sun reads timestamped rows at 15, 30 or 60-minute intervals in year-month-day or day/month/year format, builds a typical day for each month, and sets the monthly consumption from it. Two templates are available to download from the same page: one for meter intervals, and one for a typical day per month if you only have hourly averages.

This guide is general information to help you plan and compare options. It is not engineering, electrical or financial advice. Always follow the equipment manufacturer's instructions and your local wiring rules, and have a qualified installer check the design.