Solar self-consumption on a commercial site

Solar self-consumption is the share of the electricity a solar array generates that a site uses on site as the power is produced, rather than sending the surplus out to the grid. It is expressed as a percentage of total generation, and it is worked out directly from half-hourly data by comparing what the array produces in each half-hour period against what the site is drawing at the same moment. Self-consumption is usually the figure that decides whether commercial solar is worth installing, because a unit of electricity the site uses itself is worth far more than a unit it exports.

Self-consumption and self-sufficiency are different ratios

Two ratios are easy to confuse, and a feasibility study needs both. Self-consumption measures generation: of all the electricity the array produces across the year, what proportion the site uses rather than exports. Self-sufficiency measures demand: of all the electricity the site consumes across the year, what proportion comes from the array rather than from the grid.

They share a numerator, the solar generation the site uses on itself, but they divide it by different totals. A small array on a large, busy site will have high self-consumption, because the site absorbs almost everything it makes, while its self-sufficiency stays low, because that generation covers only a fraction of a large demand. A large array on a small site shows the reverse. Knowing which ratio a quoted figure refers to matters, because a single headline percentage means very different things depending on which total it divides.

How it is worked out from half-hourly data

The calculation needs generation and demand at the same resolution, which is why half-hourly data is the starting point. For each half-hour period across the year, the model holds two numbers: the solar generation in that period and the site's demand in that period. The electricity used on site in that period is the smaller of the two. If the site is drawing more than the array is making, all of the generation is used and the rest of the demand comes from the grid. If the array is making more than the site is drawing, the demand is met in full and the surplus is available to export or store.

Add the used-on-site figure across every half hour of the year and divide by the total generation, and the result is the annual self-consumption. Because it is built half hour by half hour, it captures the timing that an annual total hides. A site can generate exactly as many units in a year as it consumes and still export most of its solar, if the generation lands in the middle of the day while the demand sits in the evening. Only data at half-hourly resolution shows that mismatch.

What raises or lowers self-consumption

Self-consumption rises when demand lines up with generation. A site that runs its heaviest load through daylight hours, a factory on day shifts or a distribution centre working through the afternoon, uses most of what its roof produces. It falls when demand and generation are out of step: a site that is quiet during the day and busy after dark exports much of its generation, because there is little load to absorb it while the sun is up.

Array size relative to demand is the other main driver. Size the array to the site's daytime baseload and almost every unit is used on site, so self-consumption is high, though the array is small and displaces less of the bill. Size it to fill the available roof and it will often generate more than the site can use on sunny days, so self-consumption falls and the surplus is exported at a lower price or lost to an export limit. The season has the same effect: a long summer day produces a midday peak well above what many sites draw, while the same array in winter may produce less than the site uses all day.

How a battery changes it

A battery raises self-consumption by removing the need for generation to be used at the instant it is produced. Surplus solar that would otherwise be exported charges the battery during the day, and the site draws that stored energy back in the evening when the array has stopped and demand is still there. Electricity that would have left the site at a low export price is used on site instead, displacing a unit that would have been bought at the full delivered price.

How much a battery lifts self-consumption depends on the size of the daily surplus and the shape of the evening demand that can absorb it. A battery adds little on a site that already uses nearly all its generation, since there is no surplus to store. It does most where a large midday surplus meets a substantial evening load.

Why self-consumption decides the economics

The reason self-consumption matters so much is the gap between two prices. A unit of solar used on site avoids the full delivered price of grid electricity: the wholesale energy cost plus the network and policy charges layered on top. A unit exported earns only the export price, which is usually well below the delivered price. So the same array on the same roof produces very different savings depending on how much of its output the site uses rather than exports, and self-consumption is the number that captures that split.

Self-consumption is one of the main things that sets a project's payback. A high self-consumption means most of the generation displaces expensive imported power, so the array saves more each year and pays back sooner. A low self-consumption means much of the output is sold cheaply, so the same capital saves less. It also affects the connection: a site exporting a large surplus may run into its export limit or its import and export capacity, which can cap how large an array is worth fitting.

Because self-consumption comes straight from a site's own demand and a modelled generation profile, it cannot be read off an average or a rule of thumb. Lupa works it out from the site's own half-hourly data and a modelled solar profile, and the platform reports self-consumption for each size of array and battery, so the choice of system rests on how much of its generation the site would actually use.