How to Read Your Half-Hourly Data Before Buying a Battery

Every honest battery proposal starts in the same place: a year of your own half-hourly consumption data. Everything else — the system size, the savings, the payback — is derived from it. The problem is that most businesses have never looked at theirs, so a proposal arrives with numbers that cannot be checked.

This guide covers how to get the data, the four numbers worth pulling out of it, and how to use those numbers to test whether a proposal is sized for your site or for the supplier’s stock. The six checks to run on it are in our buying guide to commercial battery storage.

Getting the data

If your site is half-hourly metered — most commercial sites above a modest size are — the readings already exist. Three routes to them:

• Your supplier. Ask for 12 months of half-hourly consumption data for your MPAN, as a CSV. Most suppliers provide it through an online portal or on request, and it is your data.
• Your meter operator or data collector. Named on your bill. Useful when the supplier’s portal only shows monthly totals.
• Your own monitoring. If the site already has sub-metering or a building management system, it may hold the same data at finer granularity, though a battery is sized against the settlement meter.

Ask for a full year. Winter and summer load shapes differ, and a battery sized on a summer sample will disappoint in January. If you have recently changed shift patterns, added plant or electrified a fleet, say so: the most recent months may matter more than the older ones.

The four numbers that decide the system

A spreadsheet is enough to get all four.

1. The peak half-hour (kW)

Find the single highest consumption half-hour in the year and double the kWh figure to get the average kW across that half-hour. This is what your agreed capacity has to cover. Compare it with the capacity you are paying for: sites frequently carry capacity they no longer need, or run uncomfortably close to a limit they have not revisited in years.

2. How often you approach that peak

Sort the half-hours by demand and look at the top 100. If the site touches its maximum in a handful of half-hours a year, a modest battery can shave them. If demand sits near maximum every weekday afternoon, cutting the same peak needs far more stored energy — the difference between a nice idea and a six-figure one.

3. Consumption inside the red band

Distribution charges are banded by time, and the red band is the expensive weekday peak. Total the kWh that land inside your area’s red hours. That pool is what a battery moves into cheaper hours, and it is the most predictable saving in the whole business case. Our DUoS charges guide explains how to find the hours that apply to your supply.

4. The overnight floor

Look at what the site draws between, say, midnight and 5am. That tells you two things: how much capacity is free for charging the battery, and whether you have a baseload worth attacking with efficiency measures before spending anything on storage.

Sanity-checking a proposal

With those four numbers, three questions expose most oversized proposals.

Does the power rating match the height of the peak? A battery rated well below the excess you want to remove cannot flatten it, whatever its capacity. One rated far above it is being sold on capacity you will not use.

Does the energy rating match the length of the peak? Power without energy runs flat partway through the expensive window, leaving the tail of the peak — and often the month’s capacity reading — untouched.

Has the model used your worst season or an annual average? An average conceals the day the system actually has to perform on.

If a proposal cannot show you which half-hours it modelled, it has not modelled your site. That applies as much to our own feasibility summaries as to anyone else’s: the arithmetic should be visible.

What the data cannot tell you

Half-hourly data shows consumption, not intent. It will not know about the second shift starting in March, the chiller being replaced, or the 20 vans arriving on fleet lease next year. Those change the answer more than any modelling assumption, so they belong in the conversation before a system is specified — see battery storage for EV fleet charging for the charging case specifically.

It also cannot tell you what the connection will allow. A site can have a textbook load shape and still face an export limitation or a curtailment condition; that question is settled by the G99 application, not by the meter.

Regional context still matters

Two sites with identical load shapes in different distribution areas do not have identical cases, because the charging bands and the local network’s headroom differ. Our guide to commercial battery storage in Northamptonshire works through that for the M1 distribution belt, where warehouse peaks and fleet charging tend to land in the same expensive hours.

Send us 12 months of half-hourly data and a postcode and the feasibility summary will show the four numbers above, the system they justify, and an honest payback range — or tell you plainly that storage does not stack up on your load.

Independent guidance published by SEO Dons Ltd. We are not an installer; if storage stacks up we introduce you to MCS-certified installers.