E-commerce Fulfilment Operations: Solar panels for warehousing

Why an e-commerce fulfilment operation needs solar

A modern e-commerce fulfilment centre is not the low-load shed the rest of the warehousing sector is known for. Conveyors, sortation lines, print-and-apply labelling, goods-to-person robotics and a floor of charging MHE mean your electricity demand is steady, high and running straight through the working day. That is exactly the load profile solar is built for, and it is why a fulfilment operation is one of the strongest cases in the whole warehouse estate.

The pressure is coming from two directions at once. On cost, grid electricity sits at 25-45p/kWh and TNUoS network charges rise around 60% in April 2026, applied to every single unit you import, for years to come. For an operation already spending six or seven figures a year on power, that increase alone runs into tens or hundreds of thousands of pounds. On sustainability, the retailers and brands whose orders you fulfil are pushing Scope 3 targets down the supply chain through CDP Supply Chain, EcoVadis and their own net-zero pledges. Auditable on-site generation is now something that turns up in tender packs and account reviews, not a nice-to-have for the annual report.

Your roof is the largest unused asset on the site, and only around 5% of UK warehouses currently carry any solar at all. For a fulfilment operator with a real daytime load to consume against, self-consumption typically pushes toward 80% - meaning most of what you generate offsets power you would otherwise buy at the worst possible rate. That is a permanent, compounding hedge against a bill that is only going one way.

How we size it - load-led, from your half-hourly data

We size warehouse solar from the load, never from the roof. Roof area is rarely the binding constraint; your consumption profile and your grid connection are. The starting point is twelve months of your half-hourly (HH) meter data. For a fulfilment operation that data usually shows a firm, automation-driven daytime base-load with pronounced Q4 peaks - a very different shape from an ambient single-shift store.

That steady base-load is good news: it lets us design an array that matches your daytime consumption (typically 60-85% of it) with very little wasted export. As a planning rule, roughly 100-140 kWp fits per 1,000 m² of usable clear-span roof, though only about 40-60% of a gross roof is usable once rooflights, plant, walkways and fire setbacks are removed. UK generation runs at around 900 kWh per kWp per year. But we let your actual load - not the available square metres - set the system size. Where robotics and conveyors run beyond core hours, or your Q4 profile is especially peaky, we model a battery (BESS) to shift surplus generation and shave the demand peaks that drive your capacity and network charges. For the full method, see our guide on how to size warehouse solar from half-hourly data.

The defining blocker: grid import capacity, not roof area

Here is the constraint that decides most fulfilment projects, and the one competitors rarely mention: your grid import capacity, not your roof.

A fulfilment centre that has added automation, a mezzanine of robotics and a last-mile EV-van fleet is often running close to its agreed import limit already. Layer more charging on top and you can exceed the connection your Distribution Network Operator (DNO) has agreed to give you. When that happens, the question stops being “how many panels fit on the roof” and becomes “how do we work within, or around, the connection”. The roof could take 3 MW; the connection is the ceiling.

Solar actually helps here, because generating on-site behind your meter reduces the units you import at peak. But a large array also raises a second connection question - export. Following the 2026 grid-queue reforms (Gate 2 / TMO4+), a firm export connection is no longer guaranteed, and for systems approaching or above 1 MW, DNO timelines commonly stretch to 12-24 months. We solve this the same way on almost every fulfilment site:

• A full G99 application, submitted early, with an import- and export-capacity check as the very first design step - before anything is drawn up.
• Export limitation (G100) so the system is designed for high self-consumption and never pushes more onto the network than the connection allows. This routinely speeds the connection and sidesteps the wait for network reinforcement.
• A battery for peak-shaving, capturing surplus generation and discharging it into your Q4 demand peaks, so the combined draw of automation plus EV charging stays inside your agreed connection.
• Dynamic load management across the PV, the battery and the EV chargers, so the site’s total demand is actively kept within the connection rather than tripping it.

The outcome is a project that proceeds on your timeline rather than the network’s - the connection becomes a design input we work around, not a reason to shrink the array or shelve the scheme.

Compliance and technical: sprinklers, insurer, DNO and structure

Fulfilment centres are almost always sprinklered, high-value, and heavily insured, so the fire and insurance angle is designed in from the outset - not bolted on afterwards. We work to LPC / RISCAuthority RC62 guidance on clearances from sprinkler heads, zones and firewalls, DC isolation and rapid shutdown, and we obtain your insurer’s pre-design sign-off before anything is fabricated. Operators worry that rooftop PV will invalidate cover or foul the sprinkler system; insurer engagement is a standard step on every project precisely because that worry is legitimate and easily resolved when it is handled up front.

On the grid side, the G99 application and the import/export capacity check described above are the technical gate for any system over a few hundred kW. On the fabric, we commission a structural loading assessment for additional dead load and wind uplift (to BS EN 1991-1-4) on arrays over roughly 1,000 m², an asbestos survey on any roof built before 2000, and CDM 2015 governs the works. Mounting is non-penetrative clip-fix on standing-seam and trapezoidal metal roofs wherever possible, preserving the roof warranty. Most warehouse rooftop PV also falls under Permitted Development (Class A, Part 14), so planning is rarely the obstacle.

Illustrative scenario

The following is an illustrative example based on typical projects, not a named client. Figures are indicative planning-grade numbers.

An automated fulfilment operation runs from a 350,000 sq ft shed with conveyors, sortation and goods-to-person robotics, spending around £1.1m a year on electricity. The site’s grid import connection is already close to its agreed capacity, and management want to add EV-van charging without triggering a costly and slow DNO reinforcement.

Twelve months of HH data show a firm daytime base-load and sharp Q4 peaks. We design a 1.4 MW rooftop array paired with a 1 MWh battery, export-limited under G100 to sit inside the existing connection. The steady automation load drives 81% self-consumption, the battery shaves the seasonal demand peaks, and the design removes the need for network reinforcement altogether. Modelled annual generation is around 1,260,000 kWh and the annual saving lands near £280,000, for a payback of roughly 5.2 years - after which the array runs on for 15-20 years as near-free daytime power. On a self-funded basis a profitable company writes most of the capex off in year one via the Annual Investment Allowance, and the on-site solar and battery are exempt from business rates in England. See how the numbers scale on our cost page and the tax reliefs on grants and funding.

Frequently asked questions

Our grid connection is already near its limit - can we still install solar? Yes, and this is the most common situation we see on automated fulfilment sites. Solar reduces the units you import at peak, which eases the connection rather than straining it. Where adding the array or the EV fleet would push export or import beyond your agreed capacity, we design with export limitation (G100) and, where it helps, a battery to peak-shave, so the whole scheme fits inside your existing connection. We submit the G99 application and check your agreed capacity before we design, so there are no surprises - and on sites over 1 MW we plan around the 12-24 month DNO timelines and the post-2026 grid-queue reforms from day one.

How well does solar pair with our Q4 demand peaks and automation load? Very well. Steady conveyor, sortation and robotics load means you self-consume a high share of what the array generates - typically around 80% - instead of exporting it cheaply. The seasonal Q4 spike is where a battery earns its keep: it stores surplus daytime generation and discharges it into your peak demand, shaving the capacity and network charges that hit hardest at your busiest time of year. We model the battery alongside the PV against your actual half-hourly profile.

Can solar power our EV delivery vans and MHE without breaching the connection? Yes - daytime charging of EV vans and materials-handling equipment is one of the best ways to lift self-consumption, because it absorbs solar at close to 100%. We design the charging infrastructure alongside the PV with dynamic load management, so the combined draw of automation, MHE charging and the EV fleet stays within your grid connection. Growing your daytime load into the roof this way is the forward move: it turns more of your generation into avoided grid cost over time. A 3PL and contract logistics operation on a shorter lease might structure this through a PPA instead; a lighter-load ambient and general storage site would be sized more conservatively. Get an indicative system size, cost and payback for your own site in minutes with our instant quote.