The Consumption Baseline: Why Half-Hourly Data Should Come Before Capital

Somewhere in your building, equipment is running through the night and the weekend that nobody decided to leave on. Your site is probably paying for more supply capacity than it has drawn in years. And the solar quote sitting on someone's desk was most likely sized to your roof, not to the way your site actually uses power. None of that shows up on a monthly bill. All of it shows up in the half-hourly data your meter already produces, which in most businesses nobody has read.

Reading that data is the cheapest and most useful step in any energy project, and it belongs at the start, before a panel, a battery or a heat pump gets chosen. It costs little. It carries no commercial agenda, because it happens before any technology is on the table. And it turns a supplier's quote into a case your finance team can actually rely on. The order usually runs the other way. A business sets out to cut its energy bills, gets quotes from installers, picks one and installs. Both the quote and the returns it promises are built on monthly consumption figures. The year of half-hourly data that would test those numbers goes unread. That data is already paid for, and putting it to work is the highest-return decision in the whole project.

Your Delivered Rate Is Not the Wholesale Price

UK businesses pay somewhere around 24p to 26p per kWh for electricity. DESNZ put the average across all non-domestic users at 24.3p in mid-2025, including the Climate Change Levy. The wholesale cost of the power itself, the part that gets quoted in the headlines, is often less than half of that.

The rest is non-commodity charges: the cost of moving electricity across the network, balancing the system, funding capacity and paying for policy. You don't need the full breakdown to draw the useful conclusion. These charges make up a large share of the bill, they're rising, and a better contract won't protect against them. Transmission network charges for larger users rose sharply in April 2026, with the main residual element widely reported as up around 64% after the system operator published its final tariffs in January. A new levy to fund nuclear build was added from late 2025. Procurement does nothing about any of this. Using less, and using it at the right time, does.

Prices also vary widely by size, so the average is close to meaningless for your site. DESNZ had large industrial users down around 16p per kWh before the levy at the end of 2025, while the smallest sites sit closer to 30p. Your own delivered rate is the only one that matters to a business case, and most organisations cannot break theirs down. You cannot manage a cost you cannot see, and a monthly bill is built to hide it.

The Data You Already Generate

Half-hourly metering records your consumption every 30 minutes. That's 17,520 readings a year, against the handful of estimated figures behind an older profile-class meter. It's the difference between knowing what your site did and guessing at it.

That data is about to become universal. Every electricity meter in Great Britain is moving onto half-hourly settlement under a programme run by Elexon for Ofgem. Migration began in October 2025 and runs to May 2027, with the new settlement model taking effect in July 2027. The practical effect is that the cost of when you use power, not only how much, gets sharper and harder to ignore. The government's Clean Power 2030 Action Plan treats half-hourly settlement as a critical enabler of flexibility, so understanding the consumption profile of a business creates more value, not less.

What the Profile Actually Tells You

A year of half-hourly data is not a spreadsheet to file. Read properly, it answers commercial questions a tariff comparison never will, and two of them decide more than the rest.

The first is how your demand is spread across the day. This is the reading that determines the technology. A warehouse running long day shifts has a profile that tracks solar generation almost hour for hour, so most of what a roof array makes gets used on site. A site that does its real work after dark does not, and no amount of roof space changes that. Whether solar, battery storage or load-shifting fits your site is written into this curve, and you cannot read it off a monthly total. It only appears at half-hourly resolution.

The second is your agreed capacity. A site pays for the capacity it has agreed with its network operator, in kVA, whether it draws that capacity or not. Plenty of sites carry a ceiling they last came near years ago, often set when the building did something it no longer does. A site sitting well below its agreed capacity can usually reduce it, save in the region of £10,000 a year for a larger site, and sometimes drop into a lower charging band as well. The exact figure depends on the site and the operator, though the data shows you the headroom in an afternoon, and the saving needs no capital at all.

Underneath those two, the same dataset surfaces a handful of smaller things worth money. Every site has a base load, the floor below which consumption never drops, and a surprising amount of it is usually equipment left running overnight that nobody chose to leave on. The relationship between your average demand and your peak, the load factor, shows whether the site is paying for short sharp spikes you could flatten by staggering start-up, often for nothing. On certain metering arrangements a poor power factor inflates your measured demand and attracts a charge that correction pays back quickly. And distribution charges still vary by time of day, with the priciest window on weekday late afternoons, so shifting flexible load out of it saves money too. None of these needs a site visit to find. They're all sitting in the half-hourly record.

Why the Baseline Decides the Technology

Take that shape-of-the-day reading and follow it into the most common decision a business faces, a rooftop solar proposal. Self-consumption is the single biggest driver of whether that project pays, and the half-hourly profile is the only thing that reveals it.

The logic is simple. Power you use on site as it's generated displaces electricity you would have bought at 25p or more. Power you can't use at the moment it's made gets exported, and export earns a fraction of that, with some commercial rates as low as 8p to 9p per kWh against premium rates near 15p. So the same solar panel is worth two or three times as much when its output is consumed by equipment behind the meter, compared to when the power is exported to the grid. A system sized to fill your roof generates plenty and sells a large share of it back cheaply. A system sized to match your load keeps the value on site. Those are very different investments, and only one of them is built around how the site actually runs.

The gap is large enough to decide the case on its own. A daytime operation might use 70% or more of what its panels generate. A poorly matched site might use a third. At 25p of displaced grid cost against 9p of export, that's the difference between a project that pays back inside the decade and one that limps. You cannot see which you're looking at without a year of half-hourly consumption sitting under the design. A solar proposal built without that overlay isn't a design. It's a quote, sized to the roof because the roof is the easy thing to measure.

Monthly data gets you nowhere on the number that matters most. A monthly total can tell you roughly how much a system might generate across a year. It cannot tell you how much of that generation you'll actually use, which is the number the entire return depends on. Get self-consumption wrong by 20 percentage points and the payback you took to the board was fiction. The half-hourly profile is what turns a generation estimate into an investment-grade case, with a self-consumption figure you can defend and a payback that survives scrutiny.

The same data protects you on the way into electrification. Add EV charging or a heat pump without checking your headroom and you can push a site past its agreed capacity, which means excess charges at best and a reinforcement bill at worst. The profile tells you whether the capacity is there, and whether moving charging into the small hours keeps you under the ceiling. Same principle as the solar case: the design follows the data, not the other way round.

The Wins That Come Before Capital

The same dataset pays its way before you build anything at all. The Carbon Trust has held the same line for years: good housekeeping, better controls and a little staff engagement take around 10% off a typical energy bill at little or no capital cost. Assessments under the Energy Savings Opportunity Scheme tend to bear that out, with most of the savings they identify coming from low-cost operational changes rather than big installs.

Right-sizing your agreed capacity, correcting a poor power factor, rescheduling to flatten a peak: none of it needs a panel, a battery or a charger, and all of it falls straight out of the half-hourly data. These wins matter twice. They cut the bill now, and they leave you sizing any future kit against a leaner, better-understood load, which makes that investment case stronger too.

The Public Sector Has the Same Problem, with More Scrutiny

Councils, NHS trusts and universities face the same question with an extra layer of accountability over it. The Public Sector Decarbonisation Scheme, delivered by Salix, now expects evidenced baselines and credible heat decarbonisation plans rather than applications filed on a first-come basis. The Greening Government Commitments rest on measured baselines too. For a public body, the consumption baseline is not only good practice. It's the audit trail that justifies spending public money, and the thing that survives committee scrutiny when the business case is challenged.

How Tipio Energy Approaches This

We start with your data, not with a pre-set answer. The Energy Baseline and Opportunities Review takes a year of your half-hourly consumption, your tariff history and your supply agreements, and turns them into a clear picture of where your money goes and what would change it.

We separate the operational wins that cost nothing from the measures that need capital. Any technology gets sized against your actual load rather than a generic assumption, with self-consumption, payback and the effect on your agreed capacity modelled before anyone mentions procurement.

Because we do not sell panels, batteries or chargers, the baseline is the product. The data decides what's worth doing, not a supplier's catalogue. We present it the way a board or a finance team needs to see it: clear numbers, clear assumptions, a clear route to whatever comes next.

Start with the Data

The order matters. Read the data, then decide what to build. A consumption baseline is essential, it informs the technology choice, and it's the difference between a capital decision you can defend and one you're hoping comes good.

Get a baseline and see where your energy spend is actually going before you commit a pound of capital.