Energy Security and Resilience

Judge the system, not just the target

Energy policy is climate policy, industrial policy and national resilience at the same time. A credible settlement must explain how cleaner supply, affordability and reliability will be delivered together.

The distinction in The Hollow State is between announcing an outcome and building the machinery that makes it possible. Energy makes that distinction tangible: a project on a list cannot power a factory until it is built, connected and capable of operating when needed.

The question is not whether a particular technology is virtuous. It is whether the whole system can meet demand at an acceptable cost, including during stress.

A case study: the connections queue

In an August 2026 account of its connections reforms, the National Energy System Operator described how ready projects had waited behind schemes making little progress. It reported that the queue at its peak was around four times the capacity needed for Britain's 2030 energy requirements. Its reform prioritises readiness and strategic need. This is the system operator's account of the problem and its response. [1]

That is a useful example of a delivery constraint: interest in developing projects was not the same as the ability to connect them. It also shows why criticism must recognise corrective work. Reform is under way; the reasonable test is whether it produces working connections, not whether it produces another announcement. [1]

For an investor, a credible connection date affects when capital can be committed and operations can begin. My proposed public test would follow projects from an offer through construction to energisation, explaining delay and responsibility at each stage. An offer should not be counted as completed infrastructure.

Three costs that should not be confused

A generation-cost estimate asks what it costs to produce electricity from a particular asset under specified assumptions. A system-cost assessment asks what else is needed to make the combined network work. A consumer bill is the amount charged under a particular tariff and policy framework. These answer different questions.

NESO's Clean Power 2030 advice, published in November 2024, explicitly brings together demand and supply, networks and connections, operability, and costs and benefits. Its pathways are conditional analyses of what could be delivered, with substantial enabling work required. [2]

A cheap unit of generation is valuable. It is not sufficient evidence that a whole-system transition will be cheap. Equally, the need for network investment is not sufficient evidence that a transition is uneconomic. The comparison needs consistent boundaries, financing assumptions and alternatives. [2]

That discipline should apply to all technologies. A fair comparison should include the relevant network and flexibility requirements, fuel exposure, construction and financing risks, maintenance and environmental effects. Selecting a favourable measure for one technology and an unfavourable one for another does not help citizens choose.

The strongest counterargument

There is a strong case that cleaner domestic generation can reduce exposure to imported fuel prices and support energy security. NESO's advice identifies potential security and consumer benefits alongside the delivery challenge. An argument for resilience should engage with those benefits rather than imply that preserving the existing system is risk-free. [2]

But replacing one dependency can create others. The question is how those risks are managed, not whether dependence can be abolished. Imported equipment, domestic construction capacity, balancing capability and reliable financing all belong in the assessment.

My position is that neither climate ambition nor concern about energy security should exempt a project from evidence. The public needs to understand the cost of acting, the cost of delay and the uncertainty in both.

A practical delivery scorecard

I propose that major energy commitments be reported through a small, consistent set of measures:

• Capacity actually commissioned and connected, rather than announced.
• Delivery against agreed milestones, with reasons for material changes.
• Full-system cost assumptions and the effect of alternative scenarios.
• Reliability under defined stress cases, including how shortfalls are handled.
• Distribution of costs and benefits between households, businesses and taxpayers.

These are proposals for public accountability, not a claim that no relevant reporting already exists. The improvement would be to connect technical reporting to the promises made to citizens, so the same measure follows a commitment from announcement to delivery.

Accountability without false certainty

Not every delay is incompetence. Infrastructure involves engineering, environmental consent, supply chains and local consequences. Some projects should change when new evidence appears. A credible system distinguishes justified revision from avoidable drift.

It should also explain who has authority to resolve a constraint. A developer cannot repair a policy inconsistency merely by working faster; a minister cannot solve a network constraint by repeating a deadline.

Clean energy that is reliable and affordable can strengthen national capability. Making that happen requires a delivery system citizens can inspect. Targets are the starting point. Working infrastructure is the test.