Electricity is becoming the foundation of the modern economy.

It powers homes, hospitals, communications, manufacturing and public services. It will increasingly power transport, heating, industrial production, data centres and artificial intelligence. Almost every credible pathway to net zero therefore depends upon producing substantially more low-carbon electricity and using it to displace fossil fuels throughout the economy.

The Climate Change Committee concludes that the expansion of low-carbon electricity will be essential to reducing emissions from transport, buildings and industry. The National Audit Office has similarly described power-sector decarbonisation as the backbone of the wider transition to net zero.

Yet the electricity system is becoming simultaneously:

more important;
more expensive;
more complex;
more constrained;
more dependent on administrative intervention; and
more difficult to operate using its existing market and control architecture.

The individual symptoms are familiar:

unaffordable household and industrial electricity prices;
persistent fuel poverty;
large network reinforcement requirements;
increasing balancing and constraint-management costs;
renewable generation being curtailed;
uncertainty over security of supply;
fragmented investment incentives;
slow grid connections;
opaque levies and subsidies;
insufficient coordination between transmission, distribution and end users; and
decarbonisation targets that depend on electrification progressing far faster than it currently is.

These problems are usually addressed separately.

High bills are met with rebates. Renewable investment is supported through Contracts for Difference. Resource adequacy is supported through the Capacity Market. Congestion is managed through the Balancing Mechanism. Network investment is funded through regulated charges. Local problems are addressed through distribution flexibility markets. Fuel poverty is addressed through social programmes.

This project argues that these are not entirely separate problems.

They are interconnected symptoms of an electricity architecture that was designed for a fundamentally different physical system.

1. We are attempting to decarbonise through electrification while making electricity unaffordable

The central contradiction in current energy policy is straightforward:

We are attempting to decarbonise transport, heating and industry through electrification while maintaining an economic system that makes electricity expensive to consume.

For the period from July to September 2026, the Ofgem energy price cap corresponds to an annual bill of £1,862 for a representative dual-fuel household paying by Direct Debit.

The average capped rates are:

Component	Average capped charge
Electricity unit rate	26.11 p/kWh
Electricity standing charge	57.19 p/day
Gas unit rate	7.33 p/kWh
Gas standing charge	29.04 p/day

These figures include VAT and vary by region, payment method and actual consumption.

The problem is not simply that energy is expensive. It is that electricity remains substantially more expensive per unit than gas at the point of use.

That price relationship weakens the financial case for switching from:

petrol and diesel vehicles to electric vehicles;
gas boilers to heat pumps;
gas-fired industrial heat to electric processes; and
fossil-fuel equipment to flexible electrical alternatives.

Heat pumps and electric vehicles may use energy more efficiently than their fossil-fuel equivalents, but adoption still depends on capital cost, utilisation, tariffs, consumer confidence and the relative prices of electricity and fossil fuels.

A decarbonisation strategy built around electrification cannot be considered economically coherent if the commodity into which consumers are being encouraged to switch remains heavily burdened by network charges, policy costs, balancing costs and the cost of financing the transition itself.

The Climate Change Committee’s 2025 progress report states that more than 80% of the emissions savings required by 2030 must come from sectors outside energy supply, with much of that progress dependent on electrifying transport, buildings and industry.

The effectiveness of climate policy therefore increasingly depends not only on producing low-carbon electricity, but on making it economically attractive for people and businesses to use it.

2. Fuel poverty remains embedded in the system

High energy costs are not merely an abstract market inefficiency. They have direct human consequences.

According to official government statistics, approximately 11% of households in England were in fuel poverty in 2024, using the Low Income Low Energy Efficiency measure. The average fuel-poverty gap was approximately £407: the average reduction in required energy costs needed to move a fuel-poor household out of fuel poverty.

The official definition combines three factors:

household income;
the energy efficiency of the property; and
the cost of meeting the household’s energy needs.

Sources:

Governments respond through measures such as:

the Warm Home Discount;
targeted bill rebates;
cost-of-living payments;
supplier hardship funds;
insulation programmes; and
temporary interventions in retail prices.

Such support may be necessary to prevent immediate hardship. It does not, however, resolve the underlying cost structure.

The system generates unaffordable bills and then constructs an administrative apparatus to compensate some of the people least able to pay them.

This creates several problems:

support is often poorly targeted;
eligibility rules create cliff edges;
administrative costs increase;
households just outside eligibility thresholds remain exposed;
the underlying causes of high prices remain in place; and
the cost of support must ultimately be recovered from taxpayers, consumers or borrowing.

A better electricity architecture should not treat affordability solely as a welfare problem downstream of market operation. It should seek to reduce avoidable system costs at source and allocate unavoidable costs more fairly.

3. Cheap generation does not automatically produce cheap electricity

The cost of constructing wind and solar generation has fallen dramatically over recent decades. Yet household and industrial electricity prices have not fallen in proportion.

This is often presented as a paradox. It is not.

The cost of producing an additional megawatt-hour of renewable electricity is only one component of the cost of supplying a secure electricity service.

Consumers also pay for:

transmission networks;
distribution networks;
balancing actions;
reserve capacity;
system services;
policy programmes;
renewable support contracts;
supplier operating costs;
metering;
bad debt;
taxes;
legacy costs; and
the cost of maintaining reliability when renewable output is unavailable.

A generation technology can therefore be inexpensive in isolation while the wider system required to integrate, transport, balance and insure its output remains expensive.

The relevant question is not simply:

Which generator can produce annual energy at the lowest levelised cost?

It is:

Which combination of generation, storage, networks, flexibility and firm capacity can deliver electricity when and where it is required, at the required level of reliability, at the lowest total system cost?

Existing markets do not consistently answer that question.

They often reward annual energy production separately from:

time of delivery;
location;
congestion relief;
firmness;
flexibility;
restoration capability;
voltage support;
inertia;
contribution during scarcity; and
the ability to reduce future network investment.

The result is that a resource can appear cheap under one accounting framework while creating substantial costs elsewhere in the system.

4. The supply mix is changing faster than the operating architecture

Britain has made substantial progress in reducing the carbon intensity of electricity.

Coal generation has been eliminated, renewable capacity has increased rapidly and electricity-sector emissions have fallen dramatically relative to 1990.

For the three months from December 2025 to February 2026, renewable generators supplied 51.8% of generation from major power producers, while gas supplied 35.5% and nuclear supplied 11.9%.

The low-carbon share from major power producers was 63.7%.

Source: DESNZ Energy Trends and Prices, April 2026.

This is a major engineering and environmental achievement.

It also changes the physical and economic characteristics of the power system.

The traditional system was dominated by:

large synchronous generators;
controllable fuel inputs;
predictable one-way power flows;
passive consumers;
relatively small numbers of market participants; and
generation located around a transmission system designed to accommodate it.

The emerging system contains:

weather-dependent wind and solar generation;
utility-scale and domestic batteries;
electric vehicles;
heat pumps;
rooftop solar;
flexible industrial demand;
aggregators;
smart appliances;
prosumers capable of importing and exporting;
millions of software-controlled devices; and
bidirectional power flows throughout distribution networks.

The future electricity system is therefore not simply the old system with more renewable generators attached.

It is a distributed cyber-physical system in which the behaviour of communications, software, markets, consumers and physical infrastructure increasingly interact in real time.

The economic architecture has not evolved at the same rate as the physical system.

5. Renewable output is variable, but demand is being electrified as though supply were firm

Wind and solar generation are low-carbon but non-firm. Their available output depends on weather conditions rather than solely on the instructions of the system operator.

This does not make them undesirable. It does mean that energy capacity and dependable capacity are different products.

A system with high levels of variable generation still requires some combination of:

storage;
interconnection;
dispatchable generation;
demand flexibility;
reserve;
overbuilding;
geographic diversity;
curtailment;
long-duration energy storage; and
controlled reduction of demand during severe scarcity.

At the same time, electrification adds demand from transport, heating, industry and digital infrastructure.

The National Audit Office has previously noted that electricity demand could rise substantially as the economy electrifies and that decarbonising the power system requires new generation, networks and system-wide modernisation to be delivered coherently.

It is not sufficient to compare annual renewable production with annual consumption.

Reliability must hold:

during low-wind periods;
during winter peaks;
during network outages;
during generator outages;
during interconnector unavailability;
when storage is depleted;
during forecast errors; and
under credible combinations of adverse conditions.

The system must therefore distinguish clearly between:

energy produced over a year;
power available at a particular moment;
capacity available during scarcity;
flexibility available within a specified response time; and
the reliability of each commitment.

Existing economic arrangements often blur these distinctions or reward them through separate, overlapping mechanisms.

6. Network constraints make electricity locational

Electricity cannot be transported without limits.

Every part of the network is constrained by physical conditions including:

thermal capacity;
voltage;
stability;
fault levels;
inertia;
reactive power;
protection settings; and
operational security criteria.

A megawatt of additional generation is therefore not equally valuable everywhere.

Generation added behind an export constraint may increase curtailment. The same generation placed close to demand may reduce losses and congestion.

Similarly, flexible demand may be extremely valuable behind an overloaded transformer but provide little network benefit in an unconstrained area.

Despite this, many electricity payments remain weakly locational or entirely national.

Important mechanisms—including parts of wholesale trading, renewable support and capacity remuneration—do not consistently distinguish between resources according to their position relative to binding network constraints.

This means that investment can be encouraged without a sufficiently strong economic signal indicating:

where capacity is needed;
where additional generation creates congestion;
where demand should locate;
where storage provides greatest value;
where flexibility can defer reinforcement; or
where investment would improve resilience.

The physical network eventually resolves the inconsistency, but often only after the investment decision has already been made.

The cost then appears elsewhere as:

connection delays;
reinforcement expenditure;
curtailment;
balancing actions;
constraint payments; or
emergency operational measures.

7. Constraint and balancing costs reveal failures of coordination

When the market schedules an outcome that the physical network cannot accommodate, the system operator must intervene.

This can involve:

paying generators in constrained areas to reduce production;
paying generators elsewhere to increase production;
procuring reserves;
managing voltage and stability;
redispatching storage;
changing interconnector flows; and
taking actions through the Balancing Mechanism.

Consumers ultimately pay for these interventions through Balancing Services Use of System charges and related bill components.

NESO states that balancing costs have risen significantly over recent years because of multiple factors, including network constraints and changing system needs.

Sources:

Not every balancing action represents avoidable failure. A secure power system will always require reserves and corrective action.

The deeper concern is architectural.

The wholesale market, balancing arrangements, transmission constraints, distribution constraints and local flexibility mechanisms frequently operate through different processes, timeframes and objectives.

A simplified description of the current sequence is:

wholesale trading creates an initial commercial schedule;
network limitations make parts of that schedule physically infeasible;
the system operator procures balancing actions to correct it;
distribution operators may separately procure local flexibility;
suppliers and aggregators optimise portfolios against their own commercial positions; and
consumers ultimately fund the combined result.

The balancing mechanism is therefore often asked to repair inconsistencies produced elsewhere.

This is an expensive way to coordinate a physical system.

8. Britain needs enormous grid investment because investment has not kept pace with the transition

The electricity network must expand substantially to connect renewable generation, serve electrified demand and maintain reliability.

NESO’s Beyond 2030 plan proposed approximately £58 billion of electricity transmission investment to support the system through 2035, including the connection of substantial additional offshore wind capacity.

Source: NESO, Beyond 2030.

The scale of the programme reflects several realities:

generation is increasingly located far from demand;
electricity demand is expected to grow;
historic network development did not anticipate current policy objectives;
planning and consenting processes are slow;
supply chains are constrained;
network assets have long construction lead times; and
existing infrastructure must be maintained while the system is transformed.

The need for investment is not in itself evidence of failure. A larger, cleaner and more electrified economy will require more electricity infrastructure.

The failure lies in allowing the physical system, market design and planning framework to drift apart for so long that reinforcement becomes:

urgent;
expensive;
politically contentious;
difficult to deliver; and
increasingly recovered from consumers who have little influence over where generation and demand are located.

Network investment must also be judged alongside network utilisation.

Building infrastructure that is used only during a small number of extreme periods can be necessary for reliability, but it is expensive.

A modern system should therefore combine strategic reinforcement with:

dynamic network operation;
improved observability;
active voltage control;
topology optimisation;
flexible connections;
storage;
demand response;
intelligent scheduling; and
prices that reveal the time and location of scarcity.

The alternative is to build around unmanaged peaks and then socialise the resulting costs.

9. Distribution networks remain insufficiently observable and controllable

Much of the most consequential change is happening below the transmission system.

Electric vehicles, heat pumps, rooftop solar, batteries and flexible appliances connect predominantly to distribution networks.

Yet many low-voltage networks were designed and operated as passive systems.

Historically, distribution operators could estimate demand using diversified consumer profiles because:

power flowed mainly in one direction;
individual household behaviour was weakly correlated;
major loads were comparatively predictable; and
consumers did not respond collectively to real-time digital signals.

Those assumptions are weakening.

Large numbers of devices may now respond simultaneously to:

retail tariffs;
wholesale prices;
aggregator instructions;
flexibility-market dispatch;
weather forecasts;
software defaults; and
automated optimisation.

This introduces the possibility of correlated behaviour.

A tariff intended to move demand away from one peak can create another. A national price signal can increase demand in an area with a local constraint. An aggregator can alter thousands of devices without full visibility of their distribution-network location or the state of the relevant feeder.

The system is no longer managing passive demand. It is managing active software agents and the humans whose preferences sit behind them.

This requires:

more monitoring;
better state estimation;
secure communications;
interoperable device control;
visibility across organisational boundaries;
local constraint awareness;
auditable allocation rules; and
fail-safe operating modes.

Local flexibility markets alone do not provide this operating architecture.

They procure selected actions around specific anticipated constraints. They do not necessarily create continuous, system-wide coordination between all devices, networks and markets.

10. Grid connections have become an administrative substitute for operational coordination

When network capacity is scarce, access is commonly allocated through connection applications, studies, contractual queues and reinforcement plans.

This can create long delays between:

requesting a connection;
receiving an offer;
completing network studies;
obtaining planning consent;
building reinforcement; and
energisation.

Connection queues can also contain projects that are speculative, duplicated, delayed or unlikely to be delivered.

Administrative reform may improve queue management, but it does not remove the underlying scarcity.

A connection is not merely a bureaucratic permission. It is a long-term claim on network capacity and operating headroom.

The system therefore requires a more explicit method of deciding:

which projects receive capacity;
how firmly that capacity is guaranteed;
whether rights vary over time;
how unused capacity is released;
how flexible access is priced;
how competing users are prioritised; and
who pays for reinforcement.

Without such a mechanism, grid access becomes a mixture of queuing, negotiation, planning judgement and socialised investment.

11. We do not value electricity according to its full system contribution

The value of a resource depends on more than the number of megawatt-hours it produces.

Relevant characteristics include:

location;
time of availability;
predictability;
response speed;
ramping capability;
duration;
firmness;
contribution during scarcity;
voltage support;
reactive-power capability;
inertia or synthetic inertia;
black-start capability;
restoration value;
congestion relief;
reduction in losses;
diversity relative to other resources; and
the network investment it avoids or causes.

Current arrangements reward these characteristics through an assortment of markets, contracts, network charges and bilateral procurements.

A generator may receive income from:

the wholesale market;
a Contract for Difference;
the Capacity Market;
the Balancing Mechanism;
ancillary services;
constraint actions;
embedded benefits;
private power-purchase agreements; and
network-related payments.

A battery may earn revenue from:

frequency response;
wholesale arbitrage;
balancing;
capacity payments;
local flexibility;
constraint management; and
portfolio optimisation.

Each revenue stream may be rational in isolation.

Collectively, however, the system does not produce one transparent assessment of the resource’s net contribution to:

energy adequacy;
power adequacy;
system security;
network capacity; and
consumer reliability.

The result is investment driven by the available revenue stack rather than necessarily by total system value.

12. Marginal pricing is being asked to perform a role it was not designed to perform

Marginal pricing establishes the market price using the cost of serving an additional unit of demand.

In conventional electricity markets, generators submit offers, available resources are ordered economically and the marginal accepted resource influences the price paid to the wider market.

This framework has important theoretical and practical strengths. It can support efficient short-run dispatch and reveal scarcity under particular assumptions.

The problem is not that marginal reasoning is inherently invalid.

The problem is that a single wholesale marginal price is increasingly expected to coordinate a system containing:

non-convex investment costs;
weather-dependent output;
network constraints;
subsidised generation;
price-insensitive demand;
administrative contracts;
strategic bidding;
storage with intertemporal constraints;
flexible loads with deadlines;
millions of small devices;
human behavioural responses; and
reliability requirements that extend beyond the traded interval.

The wholesale price generally does not tell a household device:

whether its local transformer is constrained;
whether charging now will create a future peak;
whether renewable output is likely to disappear before charging completes;
whether its flexibility is uniquely valuable at its location;
whether another user has a more urgent need;
or what level of reliability has been promised to it.

Those decisions are instead handled through additional layers of tariffs, markets, controls, contracts and emergency interventions.

Marginal pricing may continue to play a role in electricity systems.

What should be questioned is the assumption that it can remain the central organising principle for a deeply distributed cyber-physical system without fundamental adaptation.

13. Human behaviour now sits inside the control loop

Traditional market models often assume that demand responds predictably to price or that deviations average out across a large population.

Digital automation changes this.

Consumers increasingly delegate decisions to:

vehicle-charging software;
smart thermostats;
home-energy-management systems;
battery controllers;
supplier algorithms;
aggregators; and
appliance manufacturers.

At the same time, human preferences remain central.

People care about:

comfort;
convenience;
mobility;
fairness;
risk;
trust;
bill certainty;
privacy; and
control.

A consumer may reject an economically optimal schedule if it creates uncertainty over whether their car will be charged. They may override a thermostat after becoming uncomfortable. They may stop participating after receiving an unexpectedly high bill. They may respond differently to the same financial incentive depending on how it is presented.

This means the consumer is not merely a passive price-taking demand curve.

The consumer and their software are part of the operating system.

Mechanisms that rely on demand flexibility must therefore account for:

bounded rationality;
loss aversion;
default effects;
unequal ability to automate;
different tolerances for interruption;
distrust of complex tariffs;
unequal access to capital;
digital exclusion; and
the possibility that many devices respond to the same signal simultaneously.

A market that requires millions of people to behave like perfectly rational real-time traders is not a credible control architecture.

14. The costs and benefits of flexibility are distributed unfairly

Demand flexibility is often presented as universally beneficial.

It can reduce peaks, lower balancing requirements, absorb renewable output and defer network reinforcement.

But the ability to provide flexibility is not equally distributed.

A wealthy household may own:

an electric vehicle;
a home battery;
rooftop solar;
a heat pump;
smart appliances;
a well-insulated home; and
automation capable of responding to prices.

A lower-income household may:

rent its home;
use inefficient electric heating;
have little discretionary demand;
be unable to shift cooking or heating;
lack access to capital;
lack a smart meter or reliable connectivity; and
need electricity at fixed times.

Poorly designed dynamic tariffs may therefore reward consumers who already possess flexible assets while exposing inflexible consumers to higher residual costs.

There is also a cost-allocation problem.

When some consumers reduce their contribution to shared network and policy costs, those costs may be redistributed across the remaining customer base.

Fairness cannot therefore be reduced to offering everyone the same tariff.

A fair system must consider:

ability to respond;
responsibility for causing costs;
benefit received;
essential energy needs;
historical investment decisions;
reliability expectations; and
the distributional consequences of market participation.

15. Contracts for Difference solve financing risk, not system coordination

The Contract for Difference scheme is the government’s principal mechanism for supporting low-carbon generation.

A CfD provides a generator with a contractual strike price.

When the relevant market price is below the strike price, the generator receives a top-up. When it is above the strike price, the generator generally pays back the difference according to the contract.

This reduces exposure to wholesale-price volatility and can lower financing costs for capital-intensive projects.

The government’s 2025 CfD and Capacity Market update describes CfDs as a means of providing greater certainty and revenue stability while protecting consumers from elevated support costs when wholesale prices are high.

This is useful.

It does not, however, answer several other questions:

Is the generator located where its output is most valuable?
Does it increase or reduce congestion?
Is its output correlated with existing generation?
What firm capacity is required alongside it?
What network reinforcement does it cause?
What system services can it provide?
Is it available during the periods of greatest scarcity?
Does the contract encourage efficient behaviour during negative prices or constraints?
How should its capital cost be allocated across consumers?
Does the contract reward total system value or merely eligible energy production?

CfDs principally address revenue certainty and financing risk.

They are not a complete market design, dispatch mechanism, network-access regime or measure of system value.

A system can therefore procure generation at an apparently attractive strike price while subsequently incurring substantial costs to connect, balance and constrain it.

Those costs may appear in different institutions and different parts of the bill, obscuring the total cost of the investment decision.

16. The Capacity Market exists because the energy market does not provide sufficient assurance of capacity

Britain also operates a Capacity Market to procure sufficient dependable capacity for future periods of system stress.

The existence of the Capacity Market reflects a real problem.

An energy-only price may not provide adequate or sufficiently predictable revenue for resources that:

run infrequently;
are maintained primarily for scarcity periods;
face politically constrained scarcity prices;
provide insurance rather than large annual energy volumes; or
must recover high fixed costs from a small number of operating hours.

The Capacity Market addresses this by paying resources for being available, subject to scheme rules and performance obligations.

This improves resource adequacy but also illustrates the fragmentation of the underlying economic architecture.

The system pays separately for:

energy;
capacity;
balancing;
ancillary services;
network access;
renewable investment support; and
local flexibility.

Each market exists because another mechanism does not fully capture the relevant value.

The challenge is not that separate products can never be justified.

It is that the interactions between them are difficult to understand, can produce conflicting incentives and do not necessarily yield a coherent assessment of total system contribution.

17. The transition is funded through electricity bills, taxation, private finance and public borrowing without a coherent allocation principle

The energy transition requires very large capital investment in:

renewable generation;
nuclear generation;
transmission;
distribution;
storage;
flexibility;
energy efficiency;
electric-vehicle charging;
low-carbon heating;
system digitalisation; and
firm low-carbon capacity.

These investments must ultimately be funded by some combination of:

consumers;
taxpayers;
investors;
asset owners;
landowners;
developers;
future users; and
future taxpayers through public borrowing.

At present, costs are distributed through a mixture of:

regulated network charges;
electricity levies;
CfD payments;
Capacity Market charges;
supplier obligations;
direct public spending;
government guarantees;
tax incentives;
regulated asset bases;
consumer capital expenditure; and
public borrowing.

The allocation is rarely based on one transparent principle.

Funding policy objectives through electricity bills can be regressive because lower-income households spend a greater share of income on essential energy.

It can also undermine electrification by raising the price of electricity relative to gas and other fossil fuels.

Funding everything through general taxation is not costless either. It shifts expenditure onto the public finances and may weaken the connection between cost causation and payment.

The relevant questions are therefore:

Which costs represent private consumption?
Which costs are caused by particular users?
Which investments create public goods?
Which costs support national industrial or climate policy?
Which risks should be borne by investors?
Which risks should be socialised?
Over what period should infrastructure be paid for?
How should costs be divided between current and future users?
When should borrowing be used?
How should those who create network costs be distinguished from those who reduce them?

The current framework answers these questions inconsistently across technologies and institutions.

18. Consumers cannot see what they are paying for

A household electricity bill combines costs arising from several distinct functions.

These include:

wholesale energy;
transmission;
distribution;
balancing;
metering;
supplier operations;
environmental and social programmes;
renewable-support mechanisms;
bad debt;
taxes; and
regulatory allowances.

Ofgem publishes the detailed methodologies used to calculate the energy price cap, including separate allowances for wholesale, network and policy costs.

However, consumers generally experience the result as:

a unit rate; and
a standing charge.

The bill does not clearly communicate:

which costs vary with consumption;
which arise from peak demand;
which arise from location;
which are fixed;
which fund social policy;
which finance generation;
which reflect historical investment;
and which actions by the consumer could genuinely reduce future system costs.

This weakens both accountability and incentives.

A consumer cannot respond intelligently to costs that are bundled together and poorly explained.

19. Standing charges expose unresolved questions about fairness

Standing charges recover costs that do not vary directly with each additional kilowatt-hour consumed.

These can include elements of:

network infrastructure;
metering;
supplier administration;
policy obligations;
industry charges; and
residual costs.

A fixed charge can be economically defensible where costs genuinely arise from maintaining a connection rather than from energy consumption.

However, high standing charges create distributional problems.

They mean that:

low-consumption households cannot avoid a significant portion of the bill;
households reducing consumption receive limited benefit;
vulnerable consumers may face substantial costs before consuming any energy;
small properties may contribute similarly to much larger users for some cost categories; and
the relationship between cost causation and payment becomes unclear.

Moving all costs into the unit rate would create different distortions, including stronger incentives for high users to avoid shared costs through self-generation.

The underlying problem is therefore not simply whether charges should be fixed or volumetric.

It is the absence of a sufficiently transparent allocation method distinguishing:

customer-specific costs;
energy-related costs;
capacity-related costs;
network-location costs;
reliability costs;
policy costs; and
public-interest expenditure.

20. Industrial electricity prices weaken competitiveness and electrification

Energy affordability is also an industrial-policy issue.

Electricity-intensive businesses compare locations internationally.

Persistently high electricity prices can affect:

manufacturing investment;
data-centre development;
steel;
chemicals;
ceramics;
glass;
food production;
hydrogen production;
electrification of industrial heat; and
the development of energy-intensive digital industries.

Where electricity is expensive relative to competing jurisdictions, firms may:

defer investment;
continue using fossil fuels;
reduce production;
relocate;
demand exemptions; or
rely on government compensation.

This creates another cycle of intervention.

The system establishes high costs, identifies industries unable to bear them and then introduces relief schemes that redistribute costs elsewhere.

Official UK industrial energy-price data are published through the government’s:

The relevant measure is not merely the wholesale price.

Businesses pay the total delivered cost, including networks, policy costs, taxes and the effect of exemptions or relief programmes.

21. We are not yet delivering all the changes needed to meet climate objectives

The UK has made substantial progress in reducing emissions.

The Climate Change Committee’s 2026 progress report estimates that UK greenhouse-gas emissions in 2025 were approximately 50% below 1990 levels.

However, progress is uneven.

The Committee assessed:

15 of 31 delivery indicators as on track;
three as almost on track; and
six as off track or significantly off track.

Electricity-sector decarbonisation has historically delivered a large share of UK emissions reductions. The next stage requires low-carbon electricity to displace fossil fuels in more difficult sectors.

That means success increasingly depends on:

electric-vehicle adoption;
heat-pump deployment;
industrial electrification;
renewable and nuclear construction;
network delivery;
storage;
planning reform;
consumer acceptance; and
electricity affordability.

The transition can therefore fail even if renewable generation continues to grow.

It can fail because:

networks are not delivered;
electricity remains too expensive;
consumers do not adopt electrical technologies;
supply chains cannot deliver;
firm capacity is unavailable;
planning delays persist;
public consent weakens; or
the costs are distributed in a politically unsustainable way.

Decarbonisation is not simply a generation-procurement exercise.

It is a system transformation.

22. Reliability cannot be assumed merely because it has historically been delivered

Britain has historically operated a highly reliable electricity system.

That achievement should not be confused with proof that future reliability is automatic.

Reliability is produced through:

engineering standards;
reserve procurement;
operational experience;
network redundancy;
dispatchable generation;
interconnection;
balancing actions;
emergency procedures;
demand forecasting; and
continuous intervention by system operators.

As the system changes, the sources of reliability must change with it.

A future system with:

greater weather dependency;
increased electricity demand;
more power-electronic generation;
more distributed resources;
more active demand;
greater communications dependency; and
tighter network constraints

will require different tools and operating practices.

It is not sufficient to state that the system has always been reliable.

The relevant question is whether the future combination of generation, networks, storage, flexibility, control and market incentives can deliver the same reliability under a different set of physical conditions.

The National Audit Office has warned that decarbonising the power sector requires system-wide modernisation and a coherent delivery plan, and that delay increases the risk of missing objectives or delivering them at greater cost.

23. Markets and system operation are institutionally fragmented

Responsibility for the electricity system is distributed across:

central government;
Ofgem;
NESO;
transmission owners;
distribution network operators;
suppliers;
generators;
aggregators;
the Low Carbon Contracts Company;
the Electricity Settlements Company;
code administrators;
local authorities;
planning bodies; and
numerous market platforms.

Each organisation has a defined role.

The physical system, however, does not respect institutional boundaries.

A dispatch decision at transmission level can affect:

a distribution constraint;
local voltage;
an aggregator’s portfolio;
a supplier’s imbalance position;
battery schedules;
consumer tariffs; and
the availability of flexibility for another service.

A distribution operator may procure flexibility without complete visibility of the actions being taken by NESO.

An aggregator may offer a portfolio without the system operator having precise visibility of every asset’s network location.

A supplier may send a national tariff signal that causes a synchronised response across locally constrained networks.

Each participant may act rationally according to its own objective while the overall result remains inefficient or insecure.

This is a coordination problem created by institutional and informational fragmentation.

24. Local flexibility markets are a patch, not a complete operating system

Distribution flexibility markets attempt to procure changes in demand or generation at locations where the network expects a constraint.

They can provide useful services and may defer conventional reinforcement.

However, they also reveal the shortcomings of the wider architecture.

A separate flexibility market is needed because the main energy market does not adequately represent the relevant local network state.

This creates further questions:

Does the wholesale signal conflict with the local flexibility instruction?
Is the flexibility still available to NESO?
Can the asset be committed simultaneously in multiple markets?
Does the distribution operator know what the aggregator will do next?
Does the customer understand which signal takes priority?
Is the local procurement cheaper than reinforcement?
Does the payment reflect enduring system value or a temporary procurement rule?
What happens outside the contracted flexibility window?

A flexibility market can procure an action.

It does not automatically coordinate the entire network.

The longer-term objective should be to embed network feasibility and local value into the main resource-allocation architecture rather than continually creating new markets around failures of the existing one.

25. Smart grids remain treated as an optional programme rather than the core operating architecture

A modern electricity system requires more than smart meters and isolated flexibility trials.

A genuine smart grid requires:

widespread measurement;
accurate state estimation;
interoperable communications;
automated control;
device-level flexibility descriptions;
dynamic operating limits;
secure identity and permissions;
local and system-wide coordination;
real-time or near-real-time settlement;
auditable decision-making;
consumer protections;
cybersecurity;
and mechanisms for operating safely when communications fail.

Without these capabilities, the system remains dependent on:

conservative network assumptions;
manual intervention;
broad national signals;
administrative connection limits;
expensive reinforcement; and
fragmented markets.

The value of observability is not merely operational.

Better information can reveal:

where capacity actually exists;
which users create peaks;
where flexibility is effective;
whether reinforcement is necessary;
how costs should be allocated;
and which resources provide genuine system value.

A market architecture that values local capacity would create an economic incentive to measure it properly.

Smart-grid investment should therefore not be viewed only as an additional cost. It is part of the infrastructure required to operate the emerging electricity system efficiently.

26. Carbon accounting and certificate trading can become detached from physical emissions

Environmental certificates can help track attributes, support disclosure and create additional revenue for low-carbon generation.

However, a certificate is a financial or accounting instrument. It is not itself a physical reduction in emissions.

The environmental effect depends on questions such as:

Did the purchase cause additional low-carbon generation?
Did it change dispatch?
Did it alter investment?
Was the electricity delivered at the relevant time and location?
Was the same environmental claim counted elsewhere?
What counterfactual generation was displaced?
Were network and balancing effects considered?

Where certificates are used primarily to relabel existing production, the accounting claim can become detached from causal emissions reduction.

This is particularly important for claims of “100% renewable electricity” based on annual matching.

Annual certificate matching does not demonstrate that consumption was physically supplied by renewable generation in every hour, nor that the purchase caused new renewable capacity to be built.

A credible environmental mechanism should distinguish between:

ownership of an attribute;
temporal matching;
locational matching;
investment additionality;
operational emissions reduction; and
system-wide consequences.

The purpose of climate policy should be to reduce physical emissions, not merely to create tradable claims over them.

27. Contract reform alone cannot solve an architectural problem

Many proposed reforms focus on changing contracts:

changing CfD terms;
changing auction duration;
introducing new capacity products;
modifying balancing contracts;
changing network-access agreements;
redesigning flexibility procurement;
or creating new long-term agreements.

Contracts matter.

But a contract defines rights, obligations and payments within an architecture. It does not by itself determine whether the architecture is coherent.

Changing contractual terms does not automatically solve:

inadequate observability;
disconnected markets;
conflicting control signals;
poor locational incentives;
weak cost allocation;
missing reliability products;
network infeasibility;
slow coordination;
or the absence of a unified operating state.

If the underlying resource-allocation mechanism remains fragmented, contract reform may simply change how participants are paid within the same fragmented system.

The central question is therefore not only:

How should we contract with generators, networks and consumers?

It is:

What operating architecture should those contracts support?

28. The system has accumulated mechanisms instead of being redesigned

The present architecture was not designed as a single coherent whole.

It evolved incrementally.

When the energy market did not provide sufficient investment certainty, Contracts for Difference were introduced.

When energy revenues did not provide sufficient capacity assurance, the Capacity Market was introduced.

When wholesale schedules did not reflect network feasibility, the Balancing Mechanism corrected them.

When local constraints became material, distribution flexibility markets were introduced.

When consumers could not afford the resulting bills, rebates and social programmes were expanded.

When connection demand exceeded available capacity, queue-management reforms and flexible connections were introduced.

Each intervention may address a real problem.

Collectively, however, they produce a complex web of:

auctions;
contracts;
levies;
tariffs;
charges;
markets;
controls;
redispatch actions;
regulatory settlements; and
administrative decisions.

Complexity is not automatically bad. Electricity systems are intrinsically complex.

The concern is complexity without integration.

The system has added economic mechanisms faster than it has developed a common operational architecture capable of coordinating them.

29. The underlying problem is a mismatch between the market and the machine

The electricity system is a physical machine.

At every moment:

supply and demand must balance;
network limits must be respected;
voltage must remain within bounds;
frequency must be controlled;
reserves must be available;
credible faults must be survivable; and
resources must be coordinated across time.

The market is an information and allocation system layered over that machine.

Its purpose should be to:

reveal scarcity;
coordinate competing uses;
attract investment;
reward valuable behaviour;
allocate costs fairly;
and keep the physical system within its feasible operating region.

When the market schedules outcomes that the machine cannot deliver, operators intervene.

When interventions become structural rather than exceptional, the problem is no longer merely one of forecasting or execution.

It suggests that the market representation is diverging from the physical system it is intended to coordinate.

That divergence can be seen in:

increasing redispatch;
constraint payments;
connection queues;
separate flexibility markets;
dynamic operating envelopes;
emergency controls;
growing network investment;
and increasingly complex revenue stacks.

The diagnosis

Britain does not have a single broken electricity market.

It has a collection of individually rational mechanisms operating across a physical system that has fundamentally changed.

The current architecture:

makes electricity expensive while requiring widespread electrification;
does not consistently value energy by time and location;
pays separately for energy, capacity, flexibility and system services without a unified measure of system contribution;
encourages investment without fully internalising network consequences;
relies on balancing actions to repair commercially scheduled outcomes;
funds major public-policy objectives through opaque bill components;
places increasingly active consumers inside the control loop without adequately accounting for behaviour and fairness;
uses administrative queues and contracts to allocate scarce physical capacity;
treats smart-grid operation as an additional programme rather than the foundation of the future system; and
continues layering new mechanisms onto an architecture designed for centralised, dispatchable generation and passive demand.

The result is not simply high prices or slow decarbonisation.

It is a system in which:

physical value;
financial value;
consumer value;
environmental value; and
investment value

are no longer consistently aligned.

The central question

This project begins with a simple question:

If we were designing an electricity system today for millions of distributed generators, batteries, vehicles, heat pumps, flexible loads and intelligent devices, would we recreate the present architecture from scratch?

Would we create:

a national wholesale price largely detached from local network conditions;
a balancing mechanism to correct infeasible schedules;
separate local flexibility markets;
separate capacity payments;
separate renewable-investment contracts;
administratively allocated grid connections;
opaque network and policy charges;
and consumer tariffs that only weakly communicate the costs their behaviour creates?

Or would we design a system in which:

every resource can express what it needs and what it can provide;
allocation remains physically feasible;
prices reflect scarcity in time and location;
reliability is explicitly defined;
flexibility is coordinated continuously;
investment is rewarded according to system contribution;
costs are allocated according to causation and benefit;
and essential electricity remains affordable?

The problem — What’s broken today