Commercial solar for food & drink processing.

Food and drink processing covers the sites that turn raw ingredients into finished, packed product: bakeries and confectioners, dairies and creameries, meat and fish processing, ready-meal and prepared-food plants, breweries, distilleries and soft-drink bottlers. These are some of the most electricity-hungry buildings in the region, running mixing, cooking, chilling, filling and packing lines through long production shifts.

That demand profile is exactly what makes commercial solar worth modelling here. The return on a commercial array is governed by how much of the generation you use on site, and a processing plant pulls heavy, steady load straight through the daylight hours. York and the Humber hold some of the UK's largest food-manufacturing sites, so this is a sector we build for often.

Solar earns most when generation is consumed on site. Every unit you use offsets an expensive import unit, while a unit exported to the grid is paid far less. So the value of a system tracks how well its output lines up with the building's demand, not how many panels the roof can hold. A processing plant lines up unusually well. Refrigeration and chilled storage run continuously, process heat and steam plant draw through every shift, and the mixing, filling and packing lines carry a high, sustained daytime base load. That combination tends to self-consume a high share of what the array makes.

The match is rarely flat across the year, and it is worth being honest about that. Production peaks, seasonal product runs and cleaning-down cycles all move the demand curve, and a chilled or frozen plant typically lifts its refrigeration load in summer, which lines up with the strongest generation. Where shift patterns leave gaps, or where a site runs lighter at weekends, we look at whether battery storage shifts generation into the evening run or the overnight chill load. A site built mainly around chilled and frozen stores reads more like our cold storage case, where refrigeration is the dominant continuous load. None of this is a rule of thumb. We size to the demand the building actually carries.

Processing plants tend to occupy large single-occupier sheds, often a mix of high-bay production hall, chilled or frozen stores and ancillary blocks, which between them present a substantial run of flat or shallow-pitch roof. For orientation only, sites in this sector tend to fall in an indicative band of 150 kWp to 1 MWp and above, which spans our commercial system-size guidance from mid-scale roofs up to the largest single-occupier sheds. Treat that as a sense of scale rather than a quote. The real figure comes from the on-site survey and the half-hourly load model, and on a food site the roof needs particular care: heavy rooftop plant, condensers, flues and extract runs all reduce the usable array area and shade parts of it, so the survey works around them before anything is sized.

The capital decision on a processing plant usually runs through a finance director who wants the route and the tax position settled before the array is sized. The right structure depends on how the business already funds plant and machinery. A capital purchase keeps the asset and the full generation benefit on your own balance sheet; a lease or operating agreement spreads the cost against the energy it saves and can be arranged to sit off balance sheet; and a power purchase agreement lets a third party fund and own the array while you simply buy the output, which suits operators who would rather not commit capital to a roof. Our finance comparison sets out how each one fits a manufacturing balance sheet, and the survey can be paired with whichever route your board prefers.

On a site of this size the tax treatment moves the numbers as much as the headline cost. A rooftop array is plant and machinery, so the capital-allowances position, including the Annual Investment Allowance and any first-year allowances in force, can pull a large share of the spend forward against taxable profit in the year you install. The qualifying treatment and the limits change between Budgets, so they have to be verified for the year the work lands rather than assumed, and a processing business with steady taxable profits is usually well placed to use them. We set the mechanism out in the capital-allowances guide, and where a figure is needed for a board paper we model it on the conservative end rather than the best case.

Cost itself is never published from a page, because no two processing plants present the same problem. A creamery running deep refrigeration around the clock, a bakery with a bank of ovens on a single day shift, and a bottling line drawing in bursts all carry a different load and a different roof, so a figure pulled from a sector average would mislead more than it helped. The per-kWp bands on our commercial finance pages let you sense-check a budget against system size, and the on-site survey and PV*SOL model settle the firm figure for your plant before any number is committed.

We do not put a single payback figure against the sector, because a continuous-load chilled plant and a single-shift bakery return very differently, and an honest number has to come from your own half-hourly data. What we can state plainly is the mechanism: the more of the generation a plant self-consumes against its own import price, the faster the array pays back, and the heavy daytime refrigeration and process load typical of food and drink processing is one of the stronger self-consumption profiles we model. Any payback we present is built on the conservative end of the modelled self-consumption split rather than the optimistic one, so the case holds if a year runs lighter than planned, and it is always a model confirmed against your meter, not a promise. The generation we line against it is itself a modelling starting point we settle in PV*SOL for your exact roof, after shading from rooftop plant and the array losses are applied, never a headline yield carried over from another site. The payback-by-sector guide walks through how those figures are constructed before any survey.

A live food factory also brings statutory duties that shape how the work is delivered, beyond the question of whether it pays. A system over 50 kWp is delivered under CDM 2015 with a Principal Designer and Principal Contractor appointed, and on a food site that programme is built around your production schedule, your hygiene zones and the food-safety controls that govern roof and high-level access above an open line. The electrical work is wired to BS 7671 and commissioned to IEC 62446-1, the connection is agreed with Northern Powergrid under a G99 application, and where the model shows generation that would otherwise spill cheaply to export we look at G100 export limitation so the design fits what the local network will accept. None of that touches the MCS scheme, which stops at 50 kWp and is the domestic certification route rather than a trust signal for a plant of this scale; the assurance here is the engineering stack and the commissioning record, not a domestic certificate.

The load shape itself is what we test hardest, because it is what the whole case turns on. A plant that runs deep refrigeration and process heat through the night exports very little but may want battery storage to shift midday generation into the overnight chill load, whereas a day-shift operation closer to the cold storage profile self-consumes almost everything it makes while the line runs. Seasonal product runs, deep-clean shutdowns and the summer lift in refrigeration duty all move that curve through the year, and the summer peak in chilling load tends to fall where generation is strongest. We read which of these patterns your meter shows before sizing anything, so the array is matched to the demand the plant genuinely carries through its shifts rather than to the area on the roof.