When a substation is at capacity, a renewable energy developer has several options: fund or share the cost of network reinforcement, accept a flexible connection that allows curtailment during peak constraint periods, install export limitation equipment, or wait for queue reform to free up headroom. The right choice depends on project size, site economics, and acceptable risk. None of these options automatically kills a project.
Introduction
Finding a promising site for a renewable energy project (good solar irradiance, accessible land, willing landowner) only to discover the local substation is constrained is one of the most common blockers developers encounter. DNO heat maps flag constrained areas in red or amber, and connection offer letters can quote six- or seven-figure reinforcement costs before a single panel goes in the ground.
But being told you can't get a grid connection, or that there's no grid capacity at your chosen site, doesn't necessarily mean the end of the road. Understanding the options, and the trade-offs between them, is essential to making a sound go/no-go decision on any site.
What "at capacity" actually means
A substation is "at capacity" when the existing network infrastructure cannot safely accept additional power flows at the voltage level where a new generator would connect. In practice, this is triggered when a Distribution Network Operator (DNO) runs a power flow study and finds that connecting a new generator would breach thermal limits on cables or transformers, push voltages outside statutory limits, or create fault level issues that could compromise network protection.
The DNO heat maps published by each of Great Britain's six DNO groups use red/amber/green (RAG) ratings to give an indicative picture of headroom. A red substation does not mean connection is impossible: it means the DNO will need to do more detailed modelling and is likely to return a connection offer that includes either reinforcement costs or a flexible connection option.
The statutory window for a DNO to return a connection offer is 45 working days for standard G99 applications. In 2026, real-world timelines are often longer where constraint analysis requires additional work.
Option 1: Wait for reinforcement
Network reinforcement means upgrading the physical assets (replacing a transformer, upsizing a cable, installing new switchgear) to increase the capacity of the constrained part of the network. The DNO carries out the work; the cost is shared between the connecting developer and, in some cases, the DNO itself.
How costs are shared
Reinforcement costs at distribution level are apportioned using a Cost Apportionment Factor (CAF). The CAF allocates costs in proportion to the share of upgraded capacity a connecting project will use. If a 5 MW solar farm triggers a 33kV circuit upgrade with a total cost of £500,000, and the upgraded circuit will have a total rating of 30 MW, the developer pays 5/30ths of the cost: roughly £83,000 (Roadnight Taylor, 2024).
Crucially, a rule change introduced from April 2023 means generators connecting at distribution voltage are no longer required to contribute to reinforcement costs at higher voltage levels. Reinforcement of bulk supply points or transmission assets is now funded by the DNO, not passed on to the connecting generator (UK Power Solutions, 2023).
When reinforcement makes sense
Reinforcement is typically the right choice where: the project is large enough to justify the cost; the developer is confident the site will proceed; and the reinforcement unlocks capacity for the long term rather than just for one project. It provides firm connection capacity, meaning the developer can export at full rated capacity without curtailment risk.
The downside is time and certainty. Reinforcement can add months or years to a project programme, depending on the extent of the works, planning requirements for new infrastructure, and supply chain constraints on equipment.
Option 2: Accept a flexible connection
A flexible connection, sometimes called a non-firm connection, allows a generator to connect without the network being reinforced first. In exchange, the developer agrees that the DNO can curtail (that is, reduce or halt) export from the site during periods when the network is genuinely constrained.
How curtailment is managed
Flexible connections are managed via Active Network Management (ANM) systems or Distributed Energy Resources Management Systems (DERMS). These are software-based platforms that monitor network conditions in near real-time and issue automated setpoint instructions to inverters or control systems at the generation site. The DNO's obligation is to curtail only as much as necessary to keep the network within safe limits (UK Power Networks, 2024).
Crucially, curtailment events are not continuous: they occur only when the network is under genuine stress. In many constrained areas, the actual number of curtailment hours per year is relatively low. UK Power Networks introduced an Advanced Curtailment Report in 2024, provided automatically with flexible connection quotes, which estimates the likely curtailment impact based on network analysis and the existing queue of connecting projects (Solar Power Portal, 2024). This gives developers a more transparent basis for financial modelling before they commit.
When flexible connections make sense
A flexible connection is often the faster and cheaper route to generation. There are no upfront reinforcement costs, connection timelines can be shorter, and for many sites the financial impact of curtailment is modest relative to the savings on connection costs. The trade-off is revenue uncertainty: if curtailment is heavier than projected (for example, because further generators connect to the same substation) the economics can deteriorate.
Flexible connections are most suitable for projects where speed to revenue is more important than certainty of full output, and where the curtailment estimate from the DNO is acceptably low, which for many solar and storage sites it is.
Option 3: Export limitation
Export limitation is a simpler form of constraint management. Rather than accepting dynamic curtailment managed by the DNO's ANM system, the developer installs a static export limiting device on site that caps the maximum power the site can export to the grid at a fixed level, typically below the site's installed generation capacity.
A 500 kW solar installation, for example, might accept a connection offer with an export limit of 200 kW if that is all the substation can accommodate without reinforcement. The remaining generation can be consumed on site (behind-the-meter use) or curtailed.
Export limitation is governed by the DNO's own technical standards and by Engineering Recommendation G100, which sets out the technical requirements for limitation systems. The limitation system must reduce exported current to the permitted level within 15 seconds, with a maximum reaction time of 60 seconds (National Grid, 2024).
When export limitation makes sense
This option is most appropriate for smaller projects where behind-the-meter consumption can absorb the generation that cannot be exported, or where the economics work even at the limited export level. It is often the quickest route to connection in heavily constrained areas and requires no ongoing relationship with a DERMS platform.
Option 4: Benefit from connection queue reform
The UK's grid connection queue grew to over 700 GW of contracted projects by mid-2025, far more than could connect to the network or was needed for energy system targets (NESO, 2025). A substantial portion of this queue consisted of speculative, stalled, or non-viable projects that were nonetheless occupying connection capacity at substations across the country, blocking viable projects behind them.
The Target Model Option 4+ (TMO4+) reforms, published in April 2025 and implemented through December 2025, replaced this system with a "first ready and needed, first connected" approach. NESO placed approximately 217 GW of projects that did not meet the new "ready and needed" criteria into "Gate 1", effectively removing them from the active queue (NESO, 2025). Many of these were projects assessed as "ready" but not strategically needed for the system, rather than purely speculative schemes. The total pipeline of generation and storage was cut to roughly 283 GW, down from over 700 GW before reform.
What this means for constrained substations
Where a substation appeared fully booked because of a deep queue of contracted projects, many of those projects are now no longer holding capacity. Developers screening sites in 2026 may find that substations that were shown as constrained 12 months ago now have available headroom, or that the reinforcement cost share has reduced because fewer projects are ahead of them in the queue.
This does not mean constraints have disappeared. But it does mean the picture at any given substation is more fluid than the DNO heat map alone suggests, and a direct conversation with the DNO, or a technical screening tool that queries live connection data, can give a more accurate current picture.
Comparing the options
| Option | Upfront cost | Time to connect | Revenue certainty | Best for |
|---|---|---|---|---|
| Reinforcement | High (CAF share) | Longest | Firm capacity | Larger projects with long horizons |
| Flexible connection | Low to none | Faster | Curtailment risk | Projects prioritising speed to revenue |
| Export limitation | Low (device cost) | Fastest | Firm at limited level | Smaller sites with on-site consumption |
| Queue reform benefit | None | Variable | Depends on outcome | All: worth checking before assuming constraint |
What this means for site selection
A constrained substation should not eliminate a site from consideration at the screening stage. The questions worth asking are:
- What is the actual reinforcement cost likely to be, and can the project absorb it?
- What curtailment estimate would the DNO provide for a flexible connection at this site?
- Is behind-the-meter consumption available to make export limitation viable?
- Has the TMO4+ reform changed the constraint picture at this substation?
Answering these questions accurately requires access to live DNO data (constraint maps, queue positions, and reinforcement cost signals) alongside the ability to run spatial analysis quickly enough that developers can screen multiple sites before committing to a detailed application. Manual checking across six DNO portals, each with different data formats and update frequencies, is impractical at scale.
This is precisely the bottleneck that Feasibly's platform is designed to remove. By aggregating live DNO capacity and constraint data alongside planning and geospatial datasets, and making all of it accessible via natural language, Feasibly gives developers and community energy groups the information they need to make a sound site decision in minutes, not weeks.
Conclusion
A constrained substation presents a genuine challenge, but rarely an insurmountable one. Reinforcement, flexible connections, export limitation, and the evolving picture created by queue reform each offer a different path forward, with different cost, time, and revenue profiles. The best choice depends on project specifics (size, timeline, on-site consumption, and risk appetite) and on a current, accurate picture of what the local network can actually accommodate.
Given the pace of change in the UK's connections landscape, that picture can shift significantly in a matter of months. The TMO4+ reforms removed a layer of phantom constraint from the queue; RIIO-ED2 funding is driving reinforcement investment across DNO networks; and new tools are making constraint data more accessible than it has ever been. For developers willing to look beyond the first heat map result, constrained substations increasingly contain viable projects.