Life after the zombies: Our roundtable discussion on the future of grid capacity
At a private London roundtable late last year, Aurora Utilities convened a small group of the UK’s most senior energy leaders and infrastructure investors to answer one question:
What happens after the zombies are gone?
And when we talk about “zombies”, we don’t mean the ghouls from your Saturday night horror movie, but rather stalled or speculative infrastructure developments - from housing schemes to data centres and EV charging hubs - sitting in the queue and tying up grid capacity without progressing to the build stage.
Hosted by our CEO, Simon Reilly, and CCO, Jamie McAinsh, the discussion brought together developers, investors, demand users, and policy influencers at the sharp end of the UK’s grid constraint crisis.
As an Independent Distribution Network Operator (IDNO), Aurora works across the UK to make grid connections simpler, faster and more predictable - helping projects move from planning to energisation with less friction, so we wanted to hear directly from these important stakeholders.
The November timing of our event was deliberate. The National Energy Systems Operator (NESO) and Ofgem had promised reform through Gate 2, with the aim of removing more than 600GW of “zombie” projects from a queue that has swollen beyond 700G.
On paper, it’s a reset.
In the room, confidence was more cautious. Capital is committed, but delivery is at risk. To coincide with our roundtable, we launched a survey of 800 senior infrastructure leaders, and shared the insights exclusively that evening. The study revealed the scale of exposure:
As one senior leader put it: “Without connections, we do not have economic growth in the UK. It really is that simple.”
Grid access is no longer just an infrastructure issue, it’s an economic constraint.
Gate 2: solution or reshuffle?
There was broad agreement: Gate 2 is necessary - but not enough. “The risk is we just replace the zombies with a different set of zombies,” commented one developer.
While another warned: “The risk is that Gate 2 offers are issued and they think ‘job done’. But what will surprise them is how many investors simply won’t sign.”
Our survey reflects that tension. While 74% of leaders expect Gate 2 to reduce delays, nearly half believe the benefits will favour larger players with deeper balance sheets.
As one participant put it: “Three to four weeks to sign a Gate 2 offer, with no certainty on timing or cost - that’s not a bankable position.”
Several guests suggested this is only the beginning: “Gate 2 shouldn’t be the end of the story. At this rate you can already see Gate 3 and Gate 4 coming down the track.”
From infrastructure issues to political risk
A clear shift emerged: grid capacity is becoming political. “If the system doesn’t connect, the economy doesn’t grow. That’s about to become a political problem, not just a technical one,” said one investor.
At the same time, capital is increasingly global: “You’re not competing with UK projects - you’re competing with Singapore,” they added.
And yet, despite the friction: “We’re still here, still trying to build,” shared a developer.
Delay is the real cost
If Gate 2 is about discipline, today’s system is still defined by delay - and its consequences.
One developer said, “Every month there’s a delay, our stakeholders still demand every penny. We’re being invoiced for projects everyone knows won’t happen.”
Even completed sites aren’t immune: “Fully built but waiting a year for connection. Every month wipes out profit.”
And “protection” offers limited comfort: “Protected doesn’t mean when, or at what cost.”
For smaller developers: “This has become a poker game, and the winner has the biggest wallet.”
Guests also reflected on how the system reached this point: “Only now is it coming to light that we may be losing the very investment the UK needs.”
Are IDNOs part of the answer?
One of the more consistent themes was the role of Independent Distribution Network Operators (IDNOs).
One developer who regularly partners with IDNOs to secure connections on new projects said, “The UK has a model that’s designed to deliver, and we’re underusing it.”
IDNOs are aligned to deliver energisation, only generating revenue once a network is live so speed to connection is incentivised.
At distribution level, delivery speaks for itself: “IDNOs connect the vast majority of new projects in the UK now.”
The message wasn’t that IDNOs are the solution - but that they are an underutilised lever for improving delivery speed and certainty in a system under pressure.
New demand, same constraints
Pressure is only increasing. With UK electricity demand expected to double by 2050, driven by electrification and digital infrastructure, the strain on connections is accelerating.
AI and data centres are already reshaping demand: “Single sites now want a gigawatt of power.” But delivery isn’t keeping up: “We’re promising AI growth - but can’t guarantee the power.”
International comparisons were blunt: “Other countries are building at speed - while we debate process.”
What’s changed since?
Months on from those discussions, the direction of travel is clearer:
In short: the queue is being cleaned up - but delivery risk hasn’t disappeared, it’s moved downstream.
As several guests noted, progress will depend not just on reform - but on more joined-up delivery across design, legal and construction, where delays often compound.
A turning point or just the next phase?
Gate 2 will improve discipline. But it won’t, on its own, unlock delivery at scale.
As one participant summarised: “We are not short of capital or projects - we’re missing a system that can join the dots.”
At Aurora, we see this first-hand: the challenge isn’t a lack of ambition - it’s connecting the last mile. If that doesn’t change, the industry may find itself asking the same question again: just with a different set of zombies.
A really interesting discussion. Are traditional infrastructure planning cycles are now fundamentally struggling to keep pace with the speed of change across both AI data centres and renewable energy infrastructure. The conversation is often framed around queue reform and “zombie” projects, but the deeper challenge may actually be delivery synchronisation. AI compute architectures, renewable generation growth, BESS deployment and electrification demands are all evolving at extraordinary speed, while transmission reinforcement, planning, HV infrastructure procurement and energisation strategies still operate on multi-year timelines. It increasingly feels like many of the biggest risks are no longer emerging during planning itself, but much later during concept and detailed electrical design, when grid realities, power quality constraints, system strength limitations and energisation uncertainty become unavoidable. “Delivery risk has moved downstream.”...feels like a very important observation. Particularly as both AI infrastructure and renewable generation projects become increasingly dependent on flexible, adaptable and resilient electrical infrastructure strategies from the earliest stages of development.
A really interesting discussion. Are traditional infrastructure planning cycles are now fundamentally struggling to keep pace with the speed of change across both AI data centres and renewable energy infrastructure. The conversation is often framed around queue reform and “zombie” projects, but the deeper challenge may actually be delivery synchronisation. AI compute architectures, renewable generation growth, BESS deployment and electrification demands are all evolving at extraordinary speed, while transmission reinforcement, planning, HV infrastructure procurement and energisation strategies still operate on multi-year timelines. It increasingly feels like many of the biggest risks are no longer emerging during planning itself, but much later during concept and detailed electrical design, when grid realities, power quality constraints, system strength limitations and energisation uncertainty become unavoidable.“Delivery risk has moved downstream.”...feels like a very important observation. Particularly as both AI infrastructure and renewable generation projects become increasingly dependent on flexible, adaptable and resilient electrical infrastructure strategies from the earliest stages of development.