Self-Consumption vs Export (SEG) for Manufacturers

Every kilowatt-hour your rooftop solar produces goes to one of two places. It is either used on site the moment it is generated, or it flows back out to the grid. Those two outcomes look identical on a generation report, but they are worth wildly different amounts of money. Getting this distinction right is the single most important decision in a manufacturing solar project, because it drives how big the system should be, whether a battery earns its keep, and how quickly the whole investment pays back.

For UK manufacturers, the maths is not close. A unit you use yourself is worth two to five times more than a unit you export. This guide explains why, and how that gap should shape the way you size your array.

The two unit values are not comparable

When solar electricity is consumed on site, it displaces a unit you would otherwise have bought from your supplier. So its value is your full industrial import rate, which currently sits at roughly 18 to 32p per kWh depending on your tariff, contract date and network region. Every self-consumed kWh is a unit you no longer pay for at that rate.

When solar electricity is exported, it earns whatever your supplier pays under the Smart Export Guarantee, the Ofgem-administered scheme that obliges larger licensed suppliers to offer an export tariff for MCS-certified installs up to 5 MW. Those tariffs are supplier-set and typically land in the region of 4 to 15p per kWh. The supplier then resells that same unit to another customer, which is exactly why they pay you less than they charge you.

So the two destinations for a single solar unit carry completely different values.

Comparison: the value of one solar kWh

Where the unit goes	What it is worth	Why
Self-consumed on site	Roughly 18 to 32p per kWh	Displaces a unit you would have bought at your full industrial import rate
Exported to the grid (SEG)	Roughly 4 to 15p per kWh	Supplier-set export tariff, well below retail because they resell it
Value gap	Around 2 to 5 times	The whole reason manufacturing solar is sized around self-consumption

The precise numbers move with your tariff and your export deal, but the direction never changes. Self-consumption always wins, often by a factor of three or more. Because the SEG tariff is set by each supplier, it does pay to shop around at contract stage, but even the best export rate rarely closes the gap with retail import.

Why this drives sizing, not roof area

The instinctive way to size a rooftop array is to fill the available roof. For a manufacturer this is usually the wrong instinct, because it optimises for the cheap outcome (export) rather than the valuable one (self-consumption).

The correct approach is to size the system to your daytime baseload. A well-designed manufacturing system installs 70 to 90 percent of peak daytime demand, so that almost everything the panels generate is soaked up by the plant as it is produced, and very little spills onto the low-value SEG tariff. Oversize beyond that point and you are effectively building extra capacity that only ever earns the export rate, which stretches payback rather than shortening it.

This is why we always pull at least 12 months of half-hourly meter data before final sizing, and model the load profile shift by shift rather than as an annual average. A site running a flat 24/5 pattern behaves very differently from one that ramps up mid-morning and winds down by mid-afternoon, and the ideal system size follows the shape of that curve, not the shape of the roof.

Manufacturing is close to the ideal profile for this. The typical site combines large unobstructed roof space with a strong, daytime-weighted electrical baseload from compressors, motors, process heat, refrigeration and machinery. That baseload lines up neatly with the midday generation peak, which is why correctly sized manufacturing arrays routinely self-supply 30 to 60 percent of annual demand while exporting very little.

The effect shows up clearly in real projects. Sites with a high, flat, around-the-clock baseload, such as food and beverage or pharmaceutical plants, reach the highest self-consumption of all, because there is always a load waiting to absorb the generation. A daytime-heavy engineering or metalworking workshop, running CNC machining, welding and compressed air, achieves strong self-consumption too, though slightly lower where the load tapers before the panels do.

What good self-consumption looks like

A typical well-sized manufacturing array lands in the region of 75 to 90 percent self-consumption on a healthy daytime baseload. The higher figure belongs to continuous-process sites where refrigeration, cleanroom HVAC or chilled water run close to 24/7; the lower end reflects single-shift daytime operations where some summer midday generation inevitably spills to export. Either way, the design goal is the same: keep as much generation on site as the load profile realistically allows, and treat export as the residual, not the plan.

You can dig into how this feeds the wider design decision in our guide to sizing a manufacturing solar system, and see how it flows through to the numbers in the full 2026 cost breakdown.

When a battery or a bigger daytime load changes the maths

The self-consumption argument is not static. Two things can change it: adding storage, or adding load. Both work by converting would-be exports into self-consumption, which is the same as upgrading those units from the SEG rate to the import rate.

Adding a battery

A battery captures surplus generation that would otherwise be exported at 4 to 15p and lets you discharge it later at times when you would otherwise import at 18 to 32p. Every unit shifted this way is upgraded from the export value to the import value, which is where the return comes from.

For most manufacturers with a daytime-heavy load profile, a battery is not the first move, because self-consumption is already close to maxed out and there is little surplus to store. Storage starts to pay above roughly 250 kW of PV in three situations:

• You run significant night shifts, so stored daytime solar can power evening or overnight production instead of being exported cheaply and re-bought expensively.
• You face heavy DUoS red-band charges, where discharging a battery during the expensive peak network window avoids the most costly grid units of the day.
• You want to trade flexibility in markets such as Dynamic Containment, adding a revenue stream on top of the bill savings.

Because the battery business case depends entirely on your specific load shape and tariff, we always model it alongside the PV so you can see whether it pays rather than buying storage on faith. There is a fuller treatment in our guide to battery storage with manufacturing solar.

Adding daytime load

The cheaper lever, and one manufacturers often overlook, is to move more consumption into daylight hours so the plant itself absorbs the generation. This needs no extra capital and every shifted unit jumps straight from export value to import value.

The clearest example is vehicle and plant charging. Daytime charging of electric forklift fleets, plant vehicles and staff or visitor EVs absorbs solar at 100 percent self-consumption, which is why it is one of the strongest economic cases on a manufacturing site. Automotive plants in particular frequently pair rooftop solar with staff and fleet charging for exactly this reason. Process electrification has the same effect: where an electric boiler or heater replaces a gas one, that new daytime electrical load soaks up solar that would otherwise have spilled to the grid, accelerating the return.

The general principle is simple. Anything that moves consumption into the middle of the day raises self-consumption, and every point of self-consumption you gain is worth several times more than the export unit it replaces.

What this means for your decision

The order of priority for a manufacturer is clear. Self-consumption is the prize, export is the consolation. Size the array to your real daytime baseload so that most generation is used on site at full import value; treat SEG income as a modest bonus on the residual, not as a reason to oversize; and reach for a battery or extra daytime load only where the modelling shows it converts enough cheap exports into valuable self-consumption to pay for itself.

This is also why financing and grants matter to the picture. Because self-consumption drives most of the return, the case holds up well under a PPA, asset finance or cash purchase, and it can be strengthened further with the reliefs and schemes set out on our grants and funding page, including capital allowances and, for eligible energy-intensive sectors, Climate Change Agreements. If you want the connection side handled properly alongside the sizing, our note on G99 grid connection for manufacturing explains how export capacity and self-consumption interact at the DNO application stage.

Different sub-sectors sit at different points on this curve. Continuous-process operations such as food and beverage manufacturing reach the highest self-consumption because their refrigeration and ovens run around the clock, while daytime-weighted operations such as engineering and metalworking still achieve strong self-consumption on a single-shift pattern. Whichever profile you run, the sizing logic is the same: build for the load you have, not the roof you have.

The only reliable way to know your own numbers is to model them from your actual meter data. If you would like us to size a system around your daytime baseload and show you the split between self-consumed and exported units before you commit a penny, request a quote and we will build the model from your half-hourly data.