Business Load Profiles vs Solar Output: What Elexon's Profile Classes Show
Updated 23 October 2026 · SEO Dons Editorial
What do Elexon's profile classes describe?
Elexon lists eight profile classes for use in electricity settlement. Its profiling page says "Profile Classes 1 and 2 are for domestic premises and Profile Classes 3 to 8 are for non-domestic premises". Class 3 is non-domestic unrestricted and class 4 is non-domestic Economy 7. Classes 5 to 8 are maximum demand customers, split by peak load factor into below 20%, 20% to 30%, 30% to 40% and over 40%.
Elexon defines peak load factor as the ratio, expressed as a percentage, of consumption in kWh during a given period to the kWh that would have been supplied had the maximum demand been maintained throughout that period.
A profile is not a measurement of one site. Elexon says it "represents the pattern of electricity usage by day and by year for the average customer" in each class, built from a sample of "all types of customers (domestic, commercial, industrial) with a Maximum Demand below 100 kW". For classes 5 to 8 it adds that customers not yet half-hourly metered "will need to use historic Load Profiles as updated Load Profiles are only now updated for Profile Classes 1-4". A site's class shows in its supply number: the first two digits of the 21-digit MPAN indicate the profile class.
How does solar output line up with a summer day, a winter day and a weekend?
Solar follows the sun, not the class, so two things decide the overlap: how much of the load falls in daylight, and how large the array is against the load. Load factor helps with the first only loosely, since a round-the-clock load has a load factor of 100% and still spends much of its time in the dark. Elexon's profiles vary by five seasons and three day types, and solar has its own swing: PVGIS gives a Birmingham south-facing 35 degree array 1.2 kWh per kWp on an average January day against 3.9 kWh in June. The solar output by month page sets out the monthly shape.
The table averages PVGIS hourly output for 1 kWp on an east-west flat roof in Birmingham, 2005 to 2023, in UK clock time. Values are watt-hours in each hour. Solar is the same on a weekday and a weekend, so the weekend difference comes entirely from the load.
| Hour starting | July | January |
|---|---|---|
| 06:00 | 31 | 0 |
| 07:00 | 97 | 0 |
| 08:00 | 184 | 0 |
| 09:00 | 279 | 28 |
| 10:00 | 338 | 67 |
| 11:00 | 355 | 98 |
| 12:00 | 397 | 111 |
| 13:00 | 413 | 100 |
| 14:00 | 400 | 72 |
| 15:00 | 378 | 36 |
| 16:00 | 331 | 1 |
| 17:00 | 258 | 0 |
| 18:00 | 183 | 0 |
| 19:00 | 101 | 0 |
| 20:00 | 35 | 0 |
| Day total | 3,782 | 514 |
The sun produces from about 06:00 to 21:00 in July and from 09:00 to 16:00 in January. A south-facing 35 degree array makes 3,781 Wh in July but 1,160 Wh in January.
To compare with load, the tables below use five illustrative loads. Each is a constant level during its operating hours and zero outside them, all year. For a block load, peak load factor is operating hours divided by the 168 hours in a week. These are not Elexon's published profile coefficients; they are stand-ins whose load factor lands in each Elexon band.
| Shape | Operating hours | Load factor | Elexon band the load factor falls in |
|---|---|---|---|
| A | Monday to Friday, 09:00 to 15:00 | 18% | Below 20% (class 5 band) |
| B | Monday to Friday, 08:00 to 16:00 | 24% | 20% to 30% (class 6 band) |
| C | Monday to Friday, 07:00 to 19:00 | 36% | 30% to 40% (class 7 band) |
| D | Seven days, 07:00 to 19:00 | 50% | Over 40% (class 8 band) |
| E | All 24 hours, seven days | 100% | Over 40% (class 8 band) |
The share of a day's solar output that falls inside the operating hours is the most a small array can use.
| Shape | July weekday | January weekday | July Saturday or Sunday | January Saturday or Sunday |
|---|---|---|---|---|
| A | 58% | 93% | 0% | 0% |
| B | 73% | 100% | 0% | 0% |
| C | 96% | 100% | 0% | 0% |
| D | 96% | 100% | 96% | 100% |
| E | 100% | 100% | 100% | 100% |
Shape A catches 58% of a July weekday's output but 93% of a January weekday's, and nothing at weekends. Real sites rarely drop to zero at weekends, because standby equipment and refrigeration keep drawing power.
What share of the output would each load shape use?
For a site that uses 100,000 kWh a year, an array whose annual output equals 25%, 50% or 100% of that is about 32, 63 or 127 kWp on an east-west flat roof here. The table gives the share of the array's output that the load uses, hour by hour, across the 19 weather years.
| Shape | Array output = 25% of annual use | Array output = 50% of annual use | Array output = 100% of annual use |
|---|---|---|---|
| A | 46% | 46% | 45% |
| B | 56% | 56% | 49% |
| C | 69% | 67% | 51% |
| D | 98% | 86% | 59% |
| E | 88% | 61% | 38% |
- Small arrays are capped by daylight overlap. Shapes A and B use the same share at 25% and 50% (shape A: 46% at both), because the load exceeds the output in every operating hour.
- Bigger arrays push output above load at summer midday. The share used falls as the array grows, most for the round-the-clock load: 88% down to 38%.
- One Elexon band can hide very different results. Shapes D and E are both in the class 8 band, yet at the 50% array D uses 86% of its output and E uses 61%.
- At 100% the share used is the share of the site's electricity that solar supplies directly. It is 38% for shape E and 51% for shape C, so the grid supplies the rest.
A south-facing 35 degree array differs from these by up to 3 percentage points in every cell. How a designer picks a size from consumption is covered in sizing a commercial solar PV system.
Why does the site's own consumption data beat a profile class?
A class is an average built from sites with a maximum demand below 100 kW, on a shape that for classes 5 to 8 is historic. It cannot show your shutdown weeks, your night shift or your refrigeration base load. Your own readings in half-hour intervals show all of these, and a designer can lay them over the PVGIS hourly series.
A government consultation document says that from 1 December 2022 organisations of all sizes can request access to their smart meter data from their supplier. The earlier consultation's outcome described the on-request offer as covering up to 12 months of historic smart meter energy use data in a machine-readable format, for consumers and their nominated third parties. A useful request names each import meter by MPAN, any generation or export meter, the latest 12 months in half-hour intervals in a file a spreadsheet can open, and whether a designer should receive it as a nominated third party. Ask the supplier how to make the request.
If you want a designer to run your own readings against PVGIS, you can send an enquiry through the site's form. The wider specification sits in the commercial PV systems guide, and the guide to how much of its demand a roof can supply covers the area side of the question.
Which settings and limits sit behind these figures?
- PVGIS 5.3 hourly series for Birmingham, 1 kWp, crystalline silicon, 14% losses, PVGIS-SARAH3 radiation data for 2005 to 2023 with the calculated horizon. East-west is the mean of east-facing and west-facing 10 degree planes.
- PVGIS timestamps are shifted to UK clock time, with British Summer Time from the last Sunday in March to the last Sunday in October. Bank holidays count as ordinary days.
- Loads are hourly, constant in operating hours, with no seasonal change. Arrays are sized so average annual output equals the stated share of use. The hourly series sums to 951 kWh per kWp (south) and 787 (east-west), a little under the PVGIS annual runs (968 and 800); the shares hold because arrays are sized against use.
All arithmetic here is illustrative: it shows how load shape drives the answer, not what a particular site will use.
Sources
- Profiling and What is a Load Profile?, Elexon, profiling page and knowledge base, checked 22 October 2026.
- Non-domestic smart meter consultation document and earlier outcome, Department for Energy Security and Net Zero, post-2025 document and outcome, checked 22 October 2026.
- PVGIS user manual, European Commission Joint Research Centre, user manual, checked 22 October 2026.
- PVGIS 5.3 hourly series, Birmingham, 1 kWp, 14% losses, European Commission Joint Research Centre, checked 22 October 2026: south 35, east 10, west 10.
- PVGIS 5.3 PVcalc annual runs, Birmingham, European Commission Joint Research Centre, checked 22 October 2026: south 35, east 10, west 10.
Common questions
How can a business find out its profile class?
Look at the supply number on the electricity bill. Elexon says the first two digits of the full 21-digit MPAN indicate the profile class.
Does a high peak load factor always mean more solar is used on site?
No. Shape E has a load factor of 100% yet uses 61% of the output at the 50% array, against 86% for shape D. Load factor shows how evenly demand is spread, not whether it falls in daylight.
Why do shapes A to C show nothing used at weekends?
They are deliberately simple blocks with no standby load. Most real sites keep drawing power outside working hours, which lifts weekend use of solar.