FORENSIC EVIDENCE AUDIT | 22 OCTOBER 2026 PUBLIC ACCOUNTS COMMITTEE
This is not merely a story of delays. It is a test of whether DESNZ, Ofgem and NESO exercise effective control over a national electricity programme whose components must work together on the same day. On 22 October, Parliament heard about coordination and reform. The independent audit and construction record, however, expose a much harder question: where is the integrated, costed, physically deliverable programme — and what is the authorised recovery plan when its critical dependencies slip?
THE CENTRAL FAILURE: POLICY DEADLINES ARE NOT PROGRAMME MANAGEMENT
A competent national delivery programme needs a critical path linking generation, grid connections, lines, substations, transformers, construction access, system protection, testing and commissioning. It needs one controlling timetable based on the date each asset is required by the electricity system, not merely the target date against which a company may receive a regulatory reward. It needs a named person with the authority to resolve conflicts, reorder work, rebaseline costs and implement contingency measures when essential assets will not arrive in time.
That is what should have been established before officials were asked to defend a programme involving up to £70 billion of transmission investment. Instead, the National Audit Office reported in September 2026 that the Government had not demonstrated the selected grid projects were the best way of achieving its objectives, because there was no integrated programme or whole-system appraisal. It also identified the absence of one accountable owner for portfolio progress and overall value for money.
The Committee’s own hearing briefing recorded that only three of 56 ongoing essential projects were expected to finish before their optimal delivery date and four were forecast beyond 2030. The NAO’s analysis also identified 28 essential projects facing significant risks and four forecast after 2030, as of March 2026. To be clear, missing an optimal system date is not necessarily missing a 2030 calendar deadline. But it can mean electricity-system costs rise even when a contractor technically meets its regulatory delivery date.
Look at the hard dates. Norwich–Tilbury and Sea Link have published licence delivery dates in December 2031. SSEN says the upgraded Peterhead–Kintore circuits will initially operate at 275 kV until 400 kV substation capacity becomes available in October 2032. These dates were not discovered after the hearing. They were recorded before it.
So what exactly does the 2030 model assume will be physically available? Where is the system-wide dependency register showing how each missing transmission element affects renewable output, replacement gas generation, constraint spending, peak security and the wider electricity investment case? Where is the calculation of the cost of each month of delay, and who has the authority to decide whether to speed up one investment, defer another or change the generation sequence?
Our criticism is not that the individual witnesses lack qualifications, or that every proposed upgrade is unnecessary. It is that the institutions have yet to demonstrate the programme controls expected for infrastructure of this scale. Co-ordination meetings, queue reform, regulatory incentives and reports about progress do not answer the controlling questions of engineering readiness, financial consequence or who takes the recovery decision.
The consequences are not hypothetical paperwork problems. If generation arrives before the network, consumers pay to manage congestion and may pay for replacement generation elsewhere. If network assets arrive but generation or demand assumptions change, expected financial benefits can be delayed or reduced. If supply security depends on assets unavailable at times of system stress, NESO must procure adequate backup and operating services. None of those risks alone establishes an inevitable national blackout, but all require transparent, costed contingency plans.
The question Parliament should have demanded an answer to is simple: Who can produce, today, the single integrated schedule and financial risk register proving that the electricity system planned for December 2030 can actually be built, operated securely and paid for at a demonstrably justified cost? Until that is shown, an assurance of confidence is not a programme baseline.
Britain is considering up to £70 billion of transmission-network investment from 2025 to 2031. Much of this investment will be financed through regulated charges recovered from electricity consumers. Yet the National Audit Office found no comprehensive programme-level appraisal comparing different configurations of generation and transmission, and no single accountable owner for the portfolio’s overall progress and value for money. NAO, September 2026
That is the central problem with the evidence session. Confidence was expressed in the programme, but confidence is not a substitute for an independently demonstrable comparison of costs, alternatives, construction dependencies and operating performance.
The hearing must be examined at two levels: what each witness actually answered, and what the underlying data can establish. Below, questions and answers are condensed summaries located from the hearing record, not purported verbatim quotations. They should be cross-checked against the corrected Parliamentary transcript before use in a formal allegation. The analysis is independent of the witnesses’ intentions.
1. CLIVE BETTS VERSUS DESNZ: WHO IS ACTUALLY IN CHARGE?
COMMITTEE QUESTION — approximately 10:07–10:08: With the department, system operator, regulator and private network owners all involved, who is ultimately accountable for delivering the full upgrade programme?
JONATHAN BREARLEY’S ANSWER: The department is ultimately accountable for strategy and oversight, while network companies are responsible for the works they undertake.
DOCUMENTED CONTRADICTION — a department can accept ultimate responsibility while still lacking effective portfolio control. The National Audit Office found that no single person or body was accountable for the progress and value for money of the entire transmission portfolio. The PAC’s own hearing briefing says only three of the 56 ongoing essential projects were expected to be delivered before their optimal date, while four were forecast beyond 2030. This is a management question, not a semantic argument about who sets policy: who can reorder dependencies and authorise a recovery plan when one late circuit prevents several generating projects operating at their assumed output?
THE ACCOUNTABILITY TEST: Publish a single named programme owner, the asset-by-asset delivery baseline, the difference between the regulatory incentive date and the date electricity is actually needed, and the additional consumer cost of every serious slippage. An assurance of oversight without those controls cannot establish that the programme is under command. PAC hearing briefing.
OUR EVIDENCE TEST: The NAO’s paragraph 3.3 identifies the absence of a single accountable organisation or individual for the complete portfolio’s progress and value for money. These two propositions can both be true in a narrow organisational sense. They cannot be allowed to obscure the missing portfolio-control mechanism. Who has the authority to reconcile a generating project’s completion date with a delayed circuit, a transformer procurement problem, an unavailable outage window and the resulting cost to consumers?
A general assertion of departmental responsibility is not the same as a named programme owner with a complete baseline, legally effective escalation arrangements and publishable financial consequences for missed milestones. If every organisation can point to its individual responsibilities while the combined system misses its date, the public is left with a costly accountability gap.
2. RUPERT LOWE VERSUS NESO: DOES QUEUE REFORM DELIVER ONE ADDITIONAL MEGAWATT?
COMMITTEE QUESTION — approximately 10:25–10:29: What has connections reform actually accelerated, and how many projects are moving earlier rather than simply changing position in the queue?
FINTAN SLYE’S ANSWER: NESO has removed speculative projects from the connections queue, but relatively few connection dates have actually been brought forward.
THE GRID EVIDENCE SLYE’S ANSWER DOES NOT RESOLVE: NESO’s queue reform has to be tested against the actual circuits, transformers, land rights, commissioning conditions and export restrictions behind each connection. Our October connection audit identifies developer-stated delay risks at Tween Bridge, a 275 kV-versus-400 kV description requiring reconciliation at Light Valley/Monk Fryston, and different customer-effective dates linked to the same works near Thorpe Marsh. These discrepancies do not prove every project is stranded; they prove that a connection agreement alone is not a verified energisation timetable.
THE NATIONAL COMPARISON: Scotland’s Peterhead–Kintore programme says 51 km of rebuilt line will initially operate at 275 kV until enabling substation upgrades allow 400 kV operation in October 2032. By contrast, Yorkshire GREEN reported pylons, conductors and transformers already installed in July 2026, with initial operation expected in 2027. We should therefore be shown a physically deliverable 2030 network, with constrained boundary transfer capabilities, not merely a cleaner connection queue. Yorkshire GREEN’s observed progress does not establish that every other network scheme is delayed.
OUR EVIDENCE TEST: Queue clearance is an administrative achievement. Physical electrical capacity requires consented circuits, suitable transformers, substations, land rights, operating conditions, construction and commissioning. An earlier offer on paper cannot do the work of an unbuilt 400kV line.
Our 29 October Electricity Delivery Audit compares connection records with physical project dependencies. Its results do not yet establish a definitive national percentage of stranded renewable generation. That would require every accepted connection offer to be reconciled with the relevant asset programme. But the records already expose why high-level assurances cannot establish deliverability.
UK Power Networks offers a particularly clear test. Against a regional 2030 solar allocation of 8,100 MW, its published connections-reform table shows 2,519 MW connected, another 3,004 MW in Phase 1 and a 2,577 MW Phase 1 undersupply. That shortfall is 31.8% of the regional allocation. Phase 1 is not equivalent to operational capacity. The figure concerns one distribution network and cannot be multiplied across Britain as a national shortfall. UKPN connections-reform figures
This is the question NESO should answer: Of the GW counted towards the 2030 pathway, how many will be physically energised, capable of full export, and supported by all mandatory network reinforcements? Without that reconciliation, queue statistics risk being mistaken for engineering progress.
3. THE BATTERY PIPELINE: A CONNECTION DATE IS NOT A COMPLETED STORAGE SYSTEM
Our earlier audit recorded an industry estimate of 8.7 GW / 14.1 GWh of operating battery storage — an implied average nominal discharge duration of approximately 1.62 hours. These are industry pipeline figures, not a fresh independent census of dispatch-ready capacity. They illustrate an important distinction between battery power, energy and reliable duration.
By October 2026, Modo Energy’s matched battery pipeline included 217 projects totalling 20.7 GW. It found that 6.5 GW of previously matched capacity now expected connections at least six months later. Modo expressly cautions that register matches may contain errors and that a delay does not by itself mean a project will miss 2030. Its updated buildout forecast projects 17.7 GW by the end of 2029, below the range envisaged for Clean Power 2030. Modo October 2026 pipeline and Modo October buildout forecast.
The appropriate challenge is not that batteries have no value. They are valuable for short-duration energy shifting and fast-response services. It is that 23–27 GW of nameplate battery power does not automatically establish adequate multi-day energy availability, usable connection capacity or commercial deliverability. The model must account for duration, charging access, cycling, state of charge, network location and the need for other reserves.
4. THE REAL GRID MAP: TWEEN BRIDGE, THORPE MARSH, MONK FRYSTON AND BRINSWORTH
This is the evidence that a national queue dashboard cannot answer.
Tween Bridge: RWE’s formal pause correspondence identified Gate 2 Phase 2 and a possible 2031–2035 connection window, materially affecting project viability. Our August 2026 TEC snapshot also contained phased entries extending beyond that window, including a later 2037 increment. These observations must be reconciled with any subsequently accepted offer: they do not justify claiming that all project capacity is unavailable until 2037. RWE pause request.
Thorpe Marsh and Fenwick: Different projects list different contractual effective dates even when they reference the same transmission-work identifier. The date assigned to one connection customer is not the construction-completion date of the shared asset. Our register comparison shows precisely why published contractual dates must be joined to commissioning evidence before anyone can claim the infrastructure will be ready.
Light Valley / Monk Fryston: The applicant’s published environmental material describes a 275kV feeder and cable while a later NESO TEC entry names Monk Fryston 400kV. The October 2026 TWR includes 22 entries associated with Light Valley, but those cannot automatically all be classified as mandatory prerequisites for first energisation. The physical connection point, any registry correction or redesign, and the exact mandatory works need written reconciliation. The already-running Yorkshire GREEN programme should not be described as unstarted.
Brinsworth / Long Lane: The position must be described with legal and engineering precision. Rotherham’s planning record shows conditional permission on 19 August 2026 for application RB2025/1468. But that is not proof that all pre-commencement conditions are discharged, the 400kV equipment is installed, the Brinsworth–Chesterfield uprating is complete or the substation is energised. A separate discharge-of-conditions application, RB2026/1266, had a determination deadline of 2 November 2026; this should not be mistaken for a fresh initial decision on the parent application. The connected line and land-rights work must be traced separately. Rotherham original application and National Grid project update.
The forensic point: a planning permission is not an energisation certificate, a compulsory purchase order being made is not proof that every land right has been secured, and a registered connection date is not evidence that every upstream asset has passed commissioning tests. A programme that treats these as interchangeable dates cannot give Parliament a credible delivery assurance.
5. EMILY BOURNE AND THE EAST ANGLIA DELAYS: THE 2030 DATE IS NOT AN ENGINEERING MILESTONE
COMMITTEE QUESTION — approximately 10:20–10:25: Which vital projects can actually be accelerated in time for 2030?
EMILY BOURNE’S ANSWER: Three East Anglia projects — two Norwich-to-Tilbury elements and Sea Link — had not proved capable of being accelerated to 2030.
ADDITIONAL DOCUMENTARY TEST — EAST ANGLIA: DESNZ gave a 31 December 2031 delivery date for Sea Link in September 2026. The Norwich–Tilbury delivery programme also extends beyond 2030. That is a documented conflict with the calendar year of the policy target, not a forecast invented by critics. Parliamentary answer on Sea Link. Scotland provides a further technical warning: SSEN expects the Peterhead–Kintore upgraded overhead line to operate initially at 275 kV until supporting substations allow 400 kV operation in October 2032. SSEN programme. The question for DESNZ is what network capacity, constraint costs and generation assumptions have been recalculated to reflect those available-in-2030 assets rather than their eventual full capabilities.
OUR EVIDENCE TEST: Of the NAO’s 56 ongoing essential projects, 32 were off track or at significant risk at the March 2026 assessment point, and four were forecast beyond 2030. The separate group of eight beneficial-to-accelerate projects must not be added to the four and called twelve essential failures. The correct question is which network capacity will actually exist in each year, not simply whether the model assumes eventual completion. NAO delivery assessment
Major links in Scotland illustrate the same problem: published programmes for Eastern Green Link 3 and 4 indicate operation in 2033, subject to their own conditions. These links should not automatically be represented as essential 2030 works without matching the original project categories. Peterhead-to-Kintore also demonstrates the difference between preparing an overhead line for 400kV and the capacity available while substations remain at earlier operating configurations. Our audit treats each stage distinctly.
6. CLIVE BETTS VERSUS FINTAN SLYE: WHY WAS THE £7.8 BILLION FORECAST NOT FULLY REBUILT?
COMMITTEE QUESTION — approximately 10:33–10:38: What could annual constraint costs reach if important grid upgrades arrive too late, and what is NESO doing to reduce them?
FINTAN SLYE’S ANSWER: The forecast discussed was around £7–8 billion; NESO was pursuing measures to reduce congestion, including more effective use of existing circuits. The figures under discussion originated in the 2024 Clean Power advice.
THE OPERATING DATA ARE ALREADY IN. NESO’s 2025 electricity-market figures show 14.37 TWh of thermal constraint actions and £1.82 billion of expenditure. Around 51% of the cost was attributed to replacement generation and 21% to wind curtailment. Those are distinct categories: constraint expenditure cannot accurately be described solely as paying wind farms to switch off. The total is a calendar-year outturn, not a forecast.
WHY THE 2030 PRESENTATION NEEDS REBUILDING: The familiar £6.58–£7.79 billion scenario assumes a particular renewable buildout combined with delayed East Anglia reinforcements. A different physical set of connections, different gas prices, altered transfer capacity and changed demand would produce a different result. NESO must publish the updated, locational, hour-by-hour comparison using assets actually expected to be commissioned, and specify which 2024 assumptions have been superseded. Otherwise Parliament is being shown the cost of a modelled mismatch without a reconciled forecast of the grid that will physically exist.
OUR EVIDENCE TEST: NESO’s own Electricity Markets Roadmap reports 14.37 TWh of thermal constraint actions and £1.82 billion of costs in calendar 2025, compared with 11 TWh and £1.48 billion in 2024. It attributes 51% of thermal costs to replacement generation and 21% to wind curtailment; these categories must not be confused with each other. These are actual aggregate outcomes, not 2030 forecasts. NESO Electricity Markets Roadmap
Separately, the NAO reproduces conditional 2030 scenarios: approximately £6.58–£7.79 billion annually without timely delivery of two key East Anglia projects, compared with £2.84–£3.57 billion if the essential programme is delivered, and £1.07–£1.86 billion if the additional beneficial projects are accelerated. The scenarios are mutually exclusive and assume the planned generation fleet arrives. They are not guaranteed bills, nor should they be summed.
Why is that qualification so important? Because a scenario assuming renewable generation is constructed while key transmission assets are late can produce very different outcomes from one in which renewable connections are themselves delayed. One situation involves substantial redispatch and curtailed generation. The other may involve reduced renewable output and greater gas generation but less measured wind curtailment. Neither outcome is automatically cheap. Both must be modelled against a like-for-like secure alternative.
7. ACTUAL POWER OPERATION: WHEN WIND FELL, GAS HAD TO RESPOND
Our independent June 2026 daily-data comparison illustrates why an annual capacity number cannot be treated as firm electricity.
Archived secondary operational summaries report 271 GWh of wind and 154 GWh of gas electricity on 28 June, but 81 GWh of wind and 353 GWh of gas on 30 June. Across those two days, wind output fell by 190 GWh while gas generation rose by 199 GWh. Grid demand, solar generation, imports and other operating conditions also changed. It would therefore be incorrect to attribute the entire gas increase to falling wind, or to claim these two days prove a nationwide 2030 forecast. 28 June archive; 30 June archive.
They nevertheless illustrate the engineering problem: renewable nameplate capacity tells us little about the exact output available in a demanding hour. Gas, interconnectors, storage, demand response and system services must collectively meet demand when variable generation is low.
We also recorded an October 2026 NESO emergency-assistance event in which a regional transmission constraint prevented available generation elsewhere from fully meeting a southern power requirement. The event illustrates a location and transfer-capacity problem; it is not proof that Britain had exhausted all generation nationally. The relevant test is secure deliverable power at the right location and time. NESO emergency-assistance explanation.
8. TIM JARVIS: DID GAS PRICES REALLY TREBLE?
COMMITTEE QUESTION — approximately 10:41–10:42: How do changing gas prices affect the forecast savings from accelerating the network?
TIM JARVIS’S ANSWER: During the response he referred to global gas prices having trebled over approximately nine months.
THE NUMBER NEEDS A SOURCE, NOT A RHETORICAL DEFENCE. ONS data showed UK wholesale gas averaging 3.064p/kWh in January 2026 and 6.433p/kWh in September — approximately 2.10 times, not three times, the January average. The seven-day average to 18 October was about 2.35 times January’s monthly average, although this mixes averaging periods. Jarvis spoke of global prices, so those UK figures cannot disprove every possible international benchmark. They do establish the need to publish his chosen index, start and end dates, averaging method and the effect of that assumption on network-benefit calculations.
THE REGULATORY QUESTION: How much of Ofgem’s projected financial benefit from faster transmission is attributable to a particular gas-price assumption? Would the conclusion survive the price range observed in other years? The regulator should publish the sensitivity analysis, not leave MPs with a dramatic, unqualified comparison.
OUR EVIDENCE TEST: ONS data put the UK wholesale System Average Price at 3.064 p/kWh in January 2026 and 6.433 p/kWh in September, approximately 2.10 times the January level. A seven-day average ending 18 October was 7.189 p/kWh, about 2.35 times the January monthly average, though the periods are not strictly like-for-like. These figures do not establish a threefold increase for UK wholesale gas. But Jarvis said global: European TTF, UK NBP and Asian LNG benchmarks need not move identically.
We should therefore demand the specific index, starting value, ending value, date range and calculation before calling the statement false. For a regulator whose calculations influence the costs borne by households, an unqualified dramatic price comparison is insufficient.
9. CHRIS KANE VERSUS OFGEM: CAN THE REGULATOR STOP A PROJECT FAILURE?
COMMITTEE QUESTION — approximately 10:11–10:22 and later delivery exchanges: When the regulator sees a major transmission project heading towards failure, what can it do before consumers incur the cost of delay?
OFgem’S ANSWER: It monitors, challenges, incentivises and, where applicable, penalises network companies. A consolidated public delivery report was discussed, with an initial publication anticipated in 2027.
REAL WORLD DELAYS EXPOSE THE LIMIT OF FINANCIAL PENALTIES. Ofgem’s own December 2025 decision set revised target delivery dates of December 2033 for Grimsby–Walpole and August 2034 for Eastern Green Links 3 and 4. These are not all essential Clean Power 2030 projects, and some dates reflect approved redesigns; they nevertheless demonstrate why a national investment story must distinguish today’s infrastructure from upgrades scheduled for years later.
THE EARLY-INTERVENTION TEST: What action can Ofgem compel before an asset misses the date NESO needs it? Name the intervention trigger, the authority to act, the revised commissioning programme and the quantified consumer exposure. Annual public reporting cannot replace contemporaneous control of the delivery programme.
OUR EVIDENCE TEST: A financial penalty cannot install a transformer, recover a lost planning window or manufacture an unavailable specialist component. The relevant test is how quickly an identified risk becomes a corrective decision, who holds authority to direct the response, and whether the public can see the revised date and financial exposure before the next annual report.
Ofgem also faces a regulatory trade-off recorded by the NAO: speedier approvals and early funding against the uncertainty of incomplete cost information. That makes transparent benefit-cost recalculation essential when projects change materially.
10. BEATRICE FILKIN: A FOUR-TO-ONE PROJECT APPRAISAL DOES NOT PROVE THE BEST NATIONAL PLAN
COMMITTEE QUESTION — approximately 10:39: What is the financial case for accelerating major grid investment?
BEATRICE FILKIN’S ANSWER: She referred to benefits of approximately four times costs in the appraisal of major accelerated schemes.
A FOUR-TO-ONE RESULT ONLY HOLDS INSIDE ITS MODEL. The NAO found that Britain lacks a comprehensive programme-level appraisal reconciling generation, transmission and credible system alternatives. Ofgem’s revised dates and redesigns show why apparently strong project economics must be re-tested against present construction costs and deliverable commissioning dates. A delay that postpones constraint savings changes the time profile of the claimed benefit, even if the project remains valuable over its lifetime.
THE VALUE-FOR-MONEY TEST: Put the original and updated benefit-to-cost calculations, gas-price sensitivity, generation-location assumptions, financing costs, delivery dates and system alternatives in the public record. An appraisal of selected projects does not prove the selected whole electricity system is the least-cost secure option.
OUR EVIDENCE TEST: Even a sound benefit-to-cost ratio for an individual project does not resolve the NAO’s complaint about the lack of a whole-system options appraisal. The proper request is for the actual appraisal, its base year, generation scenario, fuel-price assumptions, latest construction costs and sensitivity to delay. It must also be compared with alternative combinations of generation, transmission, dispatchable capacity and demand flexibility. A favourable project score is evidence about that project under specified assumptions, not proof that the entire portfolio is optimal.
11. THE DEMAND ASSUMPTIONS: WHAT IF ELECTRIFICATION ARRIVES MORE SLOWLY?
Our earlier work is not confined to the supply side. It tests the demand growth against actual adoption.
The UK had 34.888 million licensed cars in June 2026, including approximately 2.017 million zero-emission cars — about 5.8%. The Climate Change Committee’s scenario of a 29% battery-electric share by 2030 would mean roughly 10.12 million vehicles if the total fleet stayed the same. Under our deliberately simplified assumptions for the second half of 2026 and 2027, reaching that level would require about 2.36 million additional BEV registrations annually in 2028–2030. This is a conditional arithmetic challenge, not a prediction that the target cannot be met. Fleet changes, used imports and retirements matter. DfT vehicle licensing statistics; CCC Seventh Carbon Budget.
Our heat-pump reconciliation starts from 52,263 recorded retrofit installations in 2025 and a lower January–June 2026 reported count than the corresponding 2025 period. A flat-retrofit scenario against a high-growth CCC pathway produces an illustrative difference of 3.06 TWh of annual electricity demand in 2030 using our explicit consumption assumptions. It is not an observed 2030 demand shortfall. The Future Homes Standard’s transitional arrangements also mean it would be wrong to treat 2028 as an absolute end to all new-build homes completed with gas boilers.
These matters change system planning. If EV uptake and heat-pump deployment undershoot assumptions, electricity consumption, timing of peaks and the financial utilisation of new assets may differ. Equally, new data centres or electrified industry can increase demand. A credible business case must show multiple demand pathways rather than treating one forecast as destiny.
12. THE SOLAR AND FLEXIBILITY TARGETS: NAMEPLATE GW IS NOT DELIVERED GWh
Installed UK solar capacity was provisionally 23.249 GW by August 2026 in the dataset used in our audit, compared with a Government planning range of 45–47 GW in Great Britain by 2030. Because the geographical scopes differ slightly, the gap is approximate, and individual plant and rooftop commissioning records matter. Capacity can be added rapidly while still producing comparatively little electricity in winter afternoons and evenings. A planning model must use the output profile, not annual or nameplate totals alone. DESNZ renewables statistics.
Consumer-led flexibility offers another important accounting trap. DESNZ described a 2023 capability baseline of about 2.55 GW and envisaged 10–12 GW by 2030. Our audit also identifies approximately 14.94 GWh of energy accepted in one demand-flexibility market service over a defined part of 2026, based on a secondary reconstruction. These figures must not be compared directly as though they measure the same thing: GW is instantaneous capability; GWh is the amount of energy actually delivered; the Demand Flexibility Service is only one of several flexibility products. NESO Demand Flexibility Service.
The question for NESO is how much contracted flexibility can be called, for how long, where it is located, and how its response is validated during stressed system conditions. The existence of an ambition is not equivalent to demonstrated dispatchable capability.
13. THE MONEY IS BIGGER THAN THE £70 BILLION HEADLINE — BUT DO NOT ADD UNLIKE NUMBERS
Our separate four-ledger financial exercise identifies different obligations and costs that belong in a genuine whole-system affordability test. The categories overlap or use different valuation methods, so a sensational single grand total would mislead readers.
Ledger one: existing Contracts for Difference. DESNZ’s 2025–26 accounts recognise £110.923 billion of net CfD fair-value liabilities at 31 March 2026, up from £90.405 billion the year before. This is the discounted valuation of estimated future net payments, not an immediate cash debt or a figure for wind and solar alone: it includes the Hinkley Point C nuclear CfD. Recorded payments to CfD generators during 2025–26 were £2.924 billion. These are different measures. DESNZ financial statements, note 9.1
Ledger two: possible future CfDs. Our scenarios model additional potential commitments under specified strike prices, deployment assumptions and reference electricity prices. These are not contracts already signed; the central calculation cannot be presented as an audited taxpayer liability.
Ledger three: batteries, flexibility and system services. Battery construction, capacity support, ancillary services and demand response have different financing and charging routes. Some private battery capital expenditure can overlap with wider flexibility investment figures and cannot simply be added twice.
Ledger four: grid, networks and balancing. The transmission envelope of up to £70 billion is a multi-year programme with further approvals needed. The NAO separately records around £89 billion of further recommended transmission investment beyond 2030; this recommendation is not authorised expenditure and must be scoped to avoid overlap.
Ofgem’s household example makes the issue tangible: it projects the annual transmission component of a typical bill increasing from £44 to £104 by 2030–31, in constant prices, while advertising a possible £30 net saving relative to slower grid investment. That is not a £30 reduction compared with today’s household bill. The chosen counterfactual, assumed gas prices, generation delivery, financing terms and constraint-cost savings all matter. NAO Figure 10.
Our purpose is not to add every figure together and announce a fictitious multi-hundred-billion bill. It is to demand a reconciled lifetime ledger showing capital expenditure, financing and allowed returns, support contracts, operating costs, constraint costs, backup capacity, asset lives and the value of alternative systems — each assigned to the correct payer and time period.
THE DELIVERY RECORD: A 2030 SYSTEM WITH INFRASTRUCTURE DATED 2031, 2032, 2033 AND 2034
The decisive test is not how many projects have been designated essential. It is what the physical electricity system can do on 31 December 2030. This is where broad assurances about acceleration have to meet the published construction and licensing record.
Norwich to Tilbury. On 15 September 2026 DESNZ stated in a written Parliamentary answer that the licence delivery date is December 2031. National Grid says examination closed on 10 August 2026 and a development-consent decision is expected in early 2027. It says it is not yet building the pylons or line and would expect construction to begin in 2027 if consent is granted. This is not a project for which a completed 2030 transmission asset has been evidenced. Nor is 2031 a surprise uncovered by campaigners: it is in the Government’s own answer. Parliament, 15 September 2026; National Grid project FAQ.
Sea Link. In another Parliamentary answer dated 16 September, DESNZ confirmed the project’s licence delivery date as 31 December 2031. The Government therefore cannot present this particular reinforcement as already committed to physical service by 2030 without disclosing an alternative acceleration programme. Parliamentary written answer.
There is a financial consequence. The Government acknowledged in September 2025 that Norwich–Tilbury not arriving by 2030 could reduce the clean share of British electricity generation by around one percentage point and, under NESO’s modelled scenario, increase annual constraints by approximately £2.8 billion. This is a modelled consequence, not an observed 2030 bill. But it demonstrates that the strategic significance of a 2031 timetable has long been known. Parliamentary answer, 5 September 2025.
Scotland. The published record is not consistent with a blanket claim that all major north–south reinforcements will be fully operational by 2030. SSEN’s Peterhead–Persley–Kintore programme explains that 51 kilometres of overhead line will initially operate at 275 kV until associated substation upgrades enable 400 kV operation in October 2032. This is a specified staged technical limitation, not merely a vague risk. It does not prove that the existing 275 kV network has zero useful capacity. SSEN project programme.
SSEN’s proposed 186 km Beauly–Peterhead 400 kV line was still in the Scottish Government public-inquiry process in October 2026: the DPEA’s 20 October update recorded closing submissions and a February 2027 target for its report. That is a consent-stage risk, not yet proof that the developer’s ultimate operational date has formally slipped. The Spittal–Loch Buidhe–Beauly route likewise progressed towards a public inquiry. Scottish DPEA Beauly–Peterhead; SSEN Spittal–Loch Buidhe–Beauly.
Even the regulator’s own delivery targets extend beyond 2030. Ofgem has published revised dates of December 2033 for the Grimsby West–Walpole project and August 2034 for Eastern Green Links 3 and 4. These are future investments and must not all be misclassified as essential 2030 works; their importance is that public descriptions of a seamless acceleration programme must distinguish essential, beneficial and longer-term infrastructure. Ofgem’s revised target dates.
Yorkshire GREEN is a useful comparison because the physical milestones can be seen. National Grid says main construction began in January 2025. By July 2026, it reported 28 permanent pylons installed, 35 kilometres of overhead line and eight supergrid transformers installed, with initial operation expected in 2027 and final works in 2028. DESNZ separately identifies a 31 December 2028 licence delivery date. That supports a tangible delivery programme, though final commissioning has not occurred. National Grid, July 2026; DESNZ, September 2026.
It would be inaccurate to claim Yorkshire GREEN is the only national project on track without an exhaustive same-date review of every relevant scheme. But the contrast with unconsented, inquiry-stage or explicitly post-2030 projects is stark: actual foundations, transformers, pylons and planned commissioning carry a different evidential weight from a queue date or an aspiration to accelerate.
WHAT OUR OWN 29 OCTOBER AUDIT ADDS TO NESO’S ASSURANCES
Our 35-sheet 2030 Electricity Delivery Audit assesses connection dates, physical works, network balancing, demand assumptions, batteries, solar and financial exposure. It is a working evidence audit rather than a substitute for a fully commissioned power-system simulation. Its findings support a much sharper challenge to NESO: show the usable power-transfer capability year by year, not just the number of connections reformed.
First, constraint expenditure is already substantial. NESO’s calendar-2025 breakdown records 14.37 TWh of thermal constraint actions at a cost of £1.82 billion, of which approximately 51% of expenditure was associated with replacement generation and 21% with wind curtailment. The 14.37 TWh is action volume, not proof that an equivalent volume of wind electricity was wasted. The replacement-generation component helps explain why gas use and network congestion can coexist with increased installed renewable capacity. NESO Electricity Markets Roadmap evidence.
Second, battery connection reform is not the same as battery delivery. Our October industry cross-check records Modo Energy’s matched sample of 217 battery projects, approximately 20.7 GW, with 6.5 GW experiencing connection-date movement of at least six months. This does not establish that 6.5 GW will miss 2030, but it undermines any attempt to equate a reformed queue with a commissioned fleet. Battery capability also depends on its duration, charging availability and location relative to constraints. Modo Energy, October 2026.
Third, the future demand case remains sensitive to electrification assumptions. Our model tests actual licensed electric-car stocks, heat-pump retrofit numbers and the transitional exemption rules for new homes against future policy pathways. The scenarios do not prove that demand will be flat; they demonstrate how much the investment case can change if uptake differs from forecasts. A credible whole-system assessment must run both faster- and slower-electrification cases, including winter peaks rather than annual terawatt-hours alone.
Fourth, connection registers are not engineering completion certificates. Our Tween Bridge reconciliation finds developer warnings of a 2031–35 connection window and staged capacity dates that need reconciliation. Light Valley’s published environmental statement refers to a 275 kV feeder and grid cable, while a later NESO TEC record names Monk Fryston 400 kV; the physical point of connection needs written clarification. The NESO transmission works register contains linked works, but it is not proof that every listed item is a prerequisite for first energisation. These are precisely the questions a national operator with a complete physical delivery picture should be able to answer.
ENERGY SECURITY: WHAT IS PROVEN, AND WHAT IS NOT
The evidence demonstrates a material coordination and cost risk. It does not establish that Britain faces an inevitable nationwide electricity shortage or blackout. NESO’s winter 2026–27 outlook projected a de-rated capacity surplus of 5.9 GW and a 9.5% margin. Those operator forecasts deserve recognition; they do not validate a separate 2030 delivery case. Conversely, a healthy national margin does not remove the possibility of local transmission bottlenecks, increased balancing costs, voltage constraints and emergency operating action.
This distinction makes the challenge stronger, not weaker. Parliament should require DESNZ, NESO and Ofgem to publish an integrated 31 December 2030 physical-ready network: installed and tested circuits, transformer capability, operating voltage, available transfer capacity, commissioned generation, usable storage duration, constrained boundaries and realistic demand cases. Then it should demand the hourly reliability and annual cost results for that actual configuration, alongside a quantified alternative plan.
The conclusive finding from the evidence available is that the hearing did not demonstrate such a reconciled national physical-delivery and whole-system value-for-money case. That is the failure of proof that Parliament should now pursue. It is serious enough without inventing completion dates, counting unapproved expenditure as cash already spent or declaring an energy-security collapse that has not occurred.
14. THE CONCLUSION THE WITNESSES DID NOT DEMONSTRATE
None of the evidence above establishes that every grid upgrade is unnecessary, that Clean Power 2030 is mathematically impossible, or that an individual witness knowingly misled Parliament. The public hearing did, however, leave major questions unanswered that are measurable and of enormous financial importance.
The Government can reasonably say that speeding necessary transmission reinforcements reduces congestion costs. NESO can reasonably say queue reform removes speculative projects. Ofgem can reasonably say it scrutinises network expenditure. These statements may all be correct while the national programme still lacks the complete proof Parliament needs.
Here is the test that matters: show the electricity generation that will be operational each year, show the physical network capacity capable of carrying it, show the storage and reserves available during low-renewable periods, show the realistic demand ranges, then calculate the entire cost under credible alternatives.
The National Audit Office has already identified the absence of that integrated comparison. Our own project and operating ledgers show why the missing evidence matters. It is not enough to put another number into a connection queue and call it progress. It is not enough to quote an expected household saving without its counterfactual. It is not enough to cite high gas prices without naming the benchmark. It is not enough to point to planning permission where the essential electrical works have not yet been shown to be operational.
Britain needs an electricity system that works on a cold, still winter evening, not merely one that balances on an annual spreadsheet. Until the authorities publish the full physical, operational and financial reconciliation, the case for the precise programme and timetable they have chosen remains incompletely demonstrated.
MAKE THEM SHOW THE EVIDENCE.
The construction programme. The connections. The operating capacity. The backup. The complete cost.
FINAL CONCLUSION — ASSURANCES ARE NOT PROGRAMME MANAGEMENT
The Public Accounts Committee was presented with assurances about coordination, oversight and accelerated delivery. Yet the published construction schedules, the National Audit Office’s findings and our examination of grid connections reveal a more serious question: who is exercising effective control of the whole electricity programme, rather than simply reporting progress within separate organisations?
A genuine programme management system would connect each generating project to the lines, substations, transformers, land rights, network outages, protection systems and commissioning tests required before its output can be used. It would identify one accountable programme owner, a credible critical path, quantified financial consequences when that path slips, and an executable recovery plan. The evidence presented at the hearing did not demonstrate that those controls were in place across the complete programme.
These are not remote or imaginary risks. The Government’s own dates put Norwich–Tilbury and Sea Link in December 2031. Scottish reinforcement records include staged 400 kV capability extending into 2032. The NAO identified weaknesses in whole-system appraisal, accountability and delivery planning. NESO’s recorded constraint expenditure shows that misalignment between generation and transmission already has a cost. These facts require an updated, integrated 2030 delivery and operating assessment—not another assurance that the institutions are working together.
The Public Accounts Committee was presented with assurances about coordination, oversight and accelerating delivery. Yet the available evidence does not establish that DESNZ, Ofgem and NESO possess a fully reconciled, costed and physically deliverable programme for the electricity system they expect to operate in 2030.
The failure to demonstrate that control is not a minor administrative weakness. It is a fundamental question of national infrastructure governance, consumer protection and energy security.
The available evidence does not prove that Britain faces an inevitable national blackout, or that every planned investment is unnecessary. It does establish the need for these organisations to publish a single, independently testable statement of what will actually be operational by 2030, what the missing infrastructure will cost consumers, what capacity will remain available during system stress and who has the authority to intervene before further delays become unavoidable.
Parliament should insist on the critical path, the delivery dates, the full costs and the recovery plan. Confidence is not evidence. Coordination is not control. A policy deadline is not a commissioned electricity system.
PRIMARY SOURCES AND EVIDENCE STATUS
Public Accounts Committee — 22 October 2026 hearing | Parliamentary video | Unofficial searchable transcript | National Audit Office, September 2026 | NESO 2025 constraint-cost outturn | DESNZ 2025–26 accounts | Evidence CPU0002 submitted to Parliament
Research base: the 29 October 2026 35-tab Electricity Delivery Audit V5, the Stranded Renewables Audit, and the four-financial-ledgers workbook. Some figures are dated outturns, others developer or regulator forecasts, industry estimates or explicitly stated scenarios; they should not be merged. All hearing question/answer summaries must be reconciled with the corrected official transcript prior to formal allegations. Portraits are existing images from the campaign site’s media library; retain appropriate source credits and ensure reproduction rights where required.
Shane Oxer — Campaigner for fairer and affordable energy
Leave a Reply