Why does the grid connection queue hold up commercial solar?

The local distribution network was built to deliver power to buildings, not to receive it back from many generators. When a solar system exports, it pushes power the other way, and the DNO has to confirm the local cables, transformers and protection can handle that without pushing voltage or fault levels outside safe limits. That assessment takes engineering time, and it has to consider other generation already connected or already accepted in the area.

Where there is plenty of spare network capacity, the DNO can usually offer a connection quickly. Where the local network is already carrying a lot of generation, the assessment is more involved, and the offer may depend on work the DNO needs to do first. The timescale is set by the network operator and the state of the local network, not by the installer, so the honest answer to “how long” is that it depends on where you are connecting and what is already connected near you.

Connection applications are assessed broadly in the order they are made, against the capacity that is physically available on that part of the network. If earlier applications have already taken the headroom, a later project can find that the cheap, fast connection is gone and a larger piece of network work would be needed to accommodate full export. That is what people mean by the connection queue: you are waiting your turn for capacity, and your position depends on what sits ahead of you.

When the available capacity is used up, the DNO may quote network reinforcement, meaning an upgrade to cables, a transformer or a substation so the network can carry the extra generation. Reinforcement can carry a real cost and a real lead time, both of which the DNO sets. The key point for a buyer is that a constrained connection is not a dead end. It is a fork in the design, and the route you take is a commercial decision as much as a technical one.

A full export connection lets the system send any surplus generation back to the grid without a cap. It is the right answer where the network has the headroom and where exported units add enough value to justify any wait or reinforcement cost. The trade-off is that it is the design most exposed to the queue, because it asks the most of the network.

An export-limited design caps how much the system can send back, often to a set figure or to zero, using an export limitation device that throttles the inverters so the agreed limit is never breached. Because it asks little or nothing of the network on the export side, it can sidestep the capacity assessment that drives reinforcement and shorten the path to a connection. For many commercial sites this fits the economics anyway, because a well-matched system is sized to be used on site rather than exported, and self-consumed power is usually worth more than an exported unit. Where exported units would only ever be a small, low-value tail, capping export to protect the programme is often the stronger commercial call.

• Full export: no cap, best where network capacity exists and exported value is worth the offer.
• Export-limited or zero-export: a capped or no-export design that can avoid reinforcement, suited to sites that self-consume most of their generation.

We treat the grid connection as a design input that shapes the system from the start. The starting point is your half-hourly consumption data modelled against generation for your exact roof, which tells us how much of the array’s output would be used on site and how much would ever look for an export route. That self-consumption figure is what makes the export decision concrete rather than guessed.

From there we frame the connection route honestly. Where the network has headroom and full export earns its keep, we design for it. Where the queue or a reinforcement quote would stall the project, we look at an export-limited or zero-export design that holds the same on-site benefit while keeping the connection straightforward. The application itself follows the relevant connection process for the system size, with the protection and export limitation specified to the DNO’s requirements. We tell you the trade-off in plain terms before anything is committed, and we confirm the actual connection position with your network operator for your specific site rather than assuming it.

The connection queue is not only a local question of who applied first on your street. Across Great Britain the transmission and distribution queues had grown to a backlog measured in hundreds of gigawatts of accepted-but-not-yet-built projects, far more than the network can carry, and the rules for managing that backlog have been overhauled. The National Energy System Operator (NESO), working with the energy regulator Ofgem, has moved the process away from a pure date-order queue toward a model that prioritises projects that are ready to proceed and that align with strategic network need. The reform package commonly referenced as the connections queue reform, and the associated grid code modification taken through Ofgem, reorders the basis on which capacity is allocated.

The practical shift for a buyer is the move from first come, first served toward first ready, first connected. Under the older model a position in the queue could be held by a project that had stalled, blocking capacity for schemes that were genuinely ready to build. The reformed approach asks projects to demonstrate readiness, for example evidence of land rights and planning progress, before they keep their place. For a commercial rooftop array this is usually a help rather than a hindrance, because a real building with a real load and a funded scheme is, by definition, a ready project. It also means a connection position cannot be treated as a permanent reservation, so the timing of your application against your wider programme matters more than it used to. Because the detail of these reforms continues to move, we confirm the current position with your network operator rather than stating a reform milestone as settled, in keeping with the orientation purpose of this guide.

For a site in the Yorkshire, north and north-east Lincolnshire region the distribution network operator is Northern Powergrid, and a commercial array above the light-touch notification threshold connects under ENA Engineering Recommendation G99, the framework published by the Energy Networks Association that governs how generation connects to and exports onto the network. The application is not a single form. It is a technical submission describing the generation you intend to connect, the inverters and their protection settings, the proposed point of connection and the export arrangement, assessed by the operator against the state of the local network. The depth of that assessment scales with the size and connection voltage of the scheme, a point we set out in our G99 application guide.

What comes back is a connection offer: a document that states the capacity the operator will permit, any conditions attached to it, any network reinforcement it considers necessary, and the associated cost and indicative timescale. The offer carries a fixed acceptance window and a price that holds only for that window, so it has to be acted on rather than left open while other decisions catch up. Before signing, the figures worth checking are the permitted export capacity against your modelled surplus, whether any reinforcement is attributed to your project, the protection and metering obligations, and any commissioning conditions that gate switch-on. Who the operator is, and why their permission sits ahead of your supplier arrangements, is explained in our DNO explained guide, and the metering that underpins the export side is covered in half-hourly metering. We read the offer with you and translate its conditions into plain commercial terms before any capital is committed.

Where the network has no export headroom, the lever that most often frees a stalled scheme is to change what the site asks of the grid rather than to wait for the grid to change. An export-limited or zero-export design caps what leaves the boundary, and the device that enforces that cap is governed by ENA Engineering Recommendation G100, the standard for export limitation schemes. Because a G100 scheme demonstrates to the operator that the agreed limit cannot be breached, it can let a larger array connect on a constrained network than a full-export application would allow. The detail of how that limitation is engineered and witnessed is set out in our export limitation and G100 guide.

Battery storage strengthens the same argument. A battery lets the site absorb generation that would otherwise have to be exported, shifting solar produced at midday into the late-afternoon and evening load, which raises self-consumption and shrinks the surplus the connection ever has to carry. On a capacity-constrained part of the network that can be the difference between a quick connection and a reinforcement quote, and it does so while keeping the energy on site, where a self-consumed unit is generally worth more than an exported one. The way storage reshapes the load profile is covered across our commercial battery storage pages. None of this is a blanket recommendation, because a battery has its own capital cost and its own payback case; whether it earns its place on a given site is a question we settle from your half-hourly data, not from a rule of thumb, and the wider numbers are framed under commercial finance.

On a very large roof, building the whole array at once is a choice rather than a requirement. Capital gets committed across more than one financial year, load is expected to grow as a fleet electrifies, or the building is live and the operation cannot release the whole roof at once. The rule that matters is to phase the build and not the connection.

Apply once, for the capacity you intend to reach. A second application later is assessed against the network as it stands on that later day, including everything else that has connected in the meantime, so a phase two that looked trivial can come back carrying reinforcement the first phase never faced. Holding a slightly larger connection from the start almost always costs less than re-entering the queue.

The design follows from that. Size the containment, switchgear ways, metering position and any transformer for the final capacity at phase one, and leave the spare ways physically there rather than notionally available. Zone the roof by inverter so each phase is a contiguous area with its own strings and isolation, which stops a later phase taking the earlier one offline to commission. Where the scheme runs behind an export limitation, budget to re-prove and re-witness it each time capacity is added, because the limitation is measured at the site boundary and the boundary has changed.