How to Size Solar for a Fabrication Workshop

Most solar quotes for a fabrication shop start with the wrong question. They ask how big your roof is, fill it with panels, and hand you a number. That gives you an array sized to your rafters, not to your welders, and on a fabrication unit those two figures are rarely the same. Get the method right and a metal-fabrication or engineering workshop is one of the best-matched building types for solar in the country. Get it wrong and you either export cheap power you should have used yourself, or you buy a roof full of panels that never earns its keep.

This guide sets out how a fabrication array should actually be sized: to your daytime electrical load, anchored on the steady equipment that runs all day, and proven against your own meter data.

Size to the load, not the roof

The single most important idea in fabrication solar is that generation only pays well when you use it on site. Self-consumed solar replaces grid electricity at your full commercial import rate, currently around 25 to 30p per kWh. Anything you export instead earns the Smart Export Guarantee, which is supplier-set and typically sits around 12 to 16p per kWh in 2026. Every kWh you consume yourself is therefore worth roughly twice as much as every kWh you spill to the grid.

Fabrication has a structural advantage here that competitors routinely miss. Metal-bashing is overwhelmingly a single-shift, Monday-to-Friday, daytime operation, so your demand lands almost exactly on top of the solar generation curve. In practice that means 70 to 90 percent of everything a well-sized fabrication array generates is consumed on site at the full import rate. A home sits empty during the day and a 24/7 process plant exports more of its midday output than you will. Your single-shift pattern is not a limitation, it is the reason fabrication paybacks routinely land between three and seven years.

So the target is not “fill the roof”. The target is an array whose annual generation matches roughly 70 to 90 percent of your daytime consumption. Sometimes that fills the roof anyway. Often it does not, and the spare roof is better left for a future battery or a later expansion than filled with panels that only ever export.

Anchor the design on your baseload

A fabrication load looks alarming on paper: welders, plasma torches and fibre lasers are high-power, spiky, intermittent draws. It is tempting to conclude that solar cannot serve a load that jumps around like that. It can, because those spikes sit on top of a large, near-constant daytime baseload, and it is the baseload that anchors the array.

The steady loads that run all shift in a typical fab shop are:

• The rotary-screw compressor, usually the single biggest consumer, cycling continuously to hold line pressure for tools, clamps and actuators.
• LEV weld-fume extraction, which must run whenever anyone welds. This is not optional load: HSE Safety Bulletin STSU1-2019 classifies all welding fume, including mild-steel fume, as a carcinogen, so Local Exhaust Ventilation is legally required for all indoor welding under COSHH. That obligation turns your extraction into a guaranteed daytime baseload the solar offsets pound for pound.
• Fibre-laser chillers, servo drives and assist-gas compressors, which draw far more, and far more continuously, than the laser beam itself.
• CNC coolant pumps, hydraulics and chip conveyors, which run for the length of every machining cycle.

Solar feeds that smooth baseload cleanly, and the welding, plasma and laser peaks soak up the midday generation on top. Where demand charges or kVA peaks are especially heavy, a battery modelled alongside the array shaves the peaks and shifts surplus. But the array itself is sized to the baseload and the daytime average, not to the nameplate rating of your biggest machine. A fibre laser is the classic trap here: its chiller and drives set the real load, not the headline beam power, which is why you never size a fibre-laser shop from the laser’s spec sheet. There is more on this in our guide to the powder-coating and cure-oven load, where the same baseload-first logic applies to electric ovens.

Prove it with half-hourly meter data

The only honest way to size a fabrication array is against your own consumption, and every commercial meter records it. Pull 12 months of half-hourly (HH) data from your supplier or MOP, a single spreadsheet of your usage in 48 daily slots. That file tells you three things a roof survey never can:

1. Your daytime consumption shape - how much you actually draw between roughly 8am and 5pm, when solar generates.
2. Your baseload floor - the near-constant draw from compressor, extraction, chillers and coolant that never switches off during the shift.
3. Your genuine self-consumption ceiling - how much of a given array size you would use versus export, modelled slot by slot rather than guessed.

Size the array so annual generation covers around 70 to 90 percent of that daytime consumption, and you capture the maximum value at the full import rate while spilling little onto low-value export. Anyone quoting a fabrication shop without asking for the HH file is guessing. Ask for the modelled self-consumption figure and the assumptions behind it, then stress-test them.

Translate the load into panels and roof

Once the target kWp is set by the load, converting it to a physical array is straightforward. Two rules of thumb do most of the work in the UK:

• Roughly 5 to 6 square metres of unshaded roof per kWp of installed panels.
• Roughly 900 to 1,000 kWh generated per kWp per year at UK latitudes.

So a clean 1,000 square metre portal-frame workshop roof supports around 150 to 180 kWp, generating roughly 135,000 to 180,000 kWh a year. The table below shows how that scales across typical fabrication units.

Workshop roof (usable)	Indicative array	Annual generation	Rough fit
300 m²	50 to 60 kWp	45,000 to 60,000 kWh	Small jobbing or welding unit
700 m²	115 to 140 kWp	105,000 to 140,000 kWh	Mid-size sheet-metal and CNC shop
1,000 m²	150 to 180 kWp	135,000 to 180,000 kWh	Established fabrication shop
2,600 m²	400 to 500 kWp	360,000 to 500,000 kWh	Structural-steel or laser-profiling plant

Treat these as the roof ceiling, then let the meter data pull the real number down to whatever your daytime load can actually absorb. On many single-shift units the load-led figure is a little under the roof capacity, which is exactly right.

Fabrication-specific checks before you commit

Sizing to load is the core method, but a fabrication roof carries constraints a generic quote ignores. Design these in from the start:

• LEV and extraction penetrations. Your fume-extraction ductwork and discharge stacks pass through the same roof as the array. Panel layout, cable routing and maintenance walkways must be planned around existing and future extraction points so the PV never blocks a legally-required fume route.
• Structural and crane-rail loading. A framed array adds roughly 15 to 25 kg per square metre plus wind uplift. On heavy structural-steel shops, EOT overhead-crane rail and gantry dead loads must be deducted from the roof’s residual capacity first, with a structural engineer’s sign-off.
• Roof condition and asbestos. Any building from before 2000 needs an asbestos management survey, and asbestos-cement roofs cannot take rooftop PV directly, they need over-cladding or replacement first.
• Three-phase supply and G99. Any inverter output above about 11 kW three-phase needs a G99 application to your DNO, so a commercial fabrication array is effectively always G99. Submit it early, alongside the structural survey, so the connection clock starts on day one.
• DSEAR zoning. If you run a powder-coat or paint booth, DC cabling, isolators and inverters must not create an ignition source near a hazardous-area zone.

Bringing it together

A properly sized fabrication array follows a clear order: pull the half-hourly data, find the daytime baseload, size the array to cover 70 to 90 percent of daytime use, check the number against the roof at 5 to 6 square metres per kWp, then design around your extraction, structural and grid constraints. That sequence is what turns fabrication’s spiky, single-shift load into the sector’s strongest self-consumption case, and it is why a shop that sizes to load rather than roof sees a genuinely short payback.

If you want the detail behind the numbers, our cost breakdown shows current £/kWp and typical project values, the grants and funding guide covers the Annual Investment Allowance and the business-rates exemption that improve the return, and our laser and plasma cutting page works through the chiller-led baseload in more depth. When you are ready, send us your half-hourly file and we will model the self-consumption for your site. Start with a free feasibility quote.

For the underlying legal duty that makes your extraction a guaranteed daytime load, see the HSE Safety Bulletin on mild-steel welding fume.