The Energy Transition Is a Transition
Decarbonisation is not a generator-replacement programme. It is the controlled re-engineering of an energy system while that system remains fully operational.
There is an uncomfortable truth at the centre of the energy transition, we cannot decarbonise an energy system by declaration, and we cannot defend the incumbent system simply because it is still needed today.
For a period, we have to build and pay for much of the new system while continuing to operate enough of the old one to keep society functioning whilst deliberately making the old system redundant.
That is what transition means. The first proposition is uncomfortable for those who assume political ambition can somehow accelerate beyond physical system constraints.
The second is equally uncomfortable for those who use today's dependence upon hydrocarbons as evidence of their permanent necessity and neither argument survives serious engineering examination.
Targets can establish direction and stimulate investment. They cannot waive the requirements of energy balance, system stability, materials, infrastructure, capital or time.
Conversely, present necessity is not evidence of future indispensability. If lower-carbon technologies can progressively assume the functions currently provided by higher-carbon systems, continuing to defend those incumbent technologies purely because they were once essential, becomes equally difficult to justify.
The engineering task is therefore not to choose between the old system and the new one, It is to transfer function from one to the other without losing the services upon which society depends.
Deployment is not yet displacement
The scale at which renewable technology is now being deployed is extraordinary.
IRENA reports that 692 GW of renewable generating capacity was added worldwide during 2025, bringing global renewable power capacity to 5,149 GW. Renewables accounted for 85.6% of all new generating capacity installed that year.[1] IRENA
At the same time, however, the Energy Institute's 2026 Statistical Review records something equally important.
Global total energy supply increased by 1.7% during 2025. Every major energy source reached a record high for the second consecutive year. Renewables provided the largest contribution to growth in total energy supply for the first time outside a recession — but the incumbent sources continued to grow as well.[2] Energy Institute
This neither demonstrates that the energy transition has failed, nor does it demonstrate that renewable deployment is ineffective.
It demonstrates that, at global level, we are still substantially adding new energy supply while attempting to change its composition.
Transition cannot ultimately be measured only in gigawatts installed; it must be measured by the higher-impact energy and infrastructure those gigawatts eventually allow us to stop using.
Physics comes before policy
Whatever our preferred technology, the underlying physics remains inconveniently neutral.
The first law of thermodynamics expresses conservation of energy. Energy is not created or destroyed; it is transferred or converted from one form to another.[3] NASA
The second law reminds us that real processes involve irreversibility. In practical engineering systems, conversion, transmission and storage are therefore not perfectly lossless.[4] Glenn Research Center
It is surprisingly easy to lose sight of these fundamental laws when discussing energy.
A wind turbine does not create energy. It extracts part of the kinetic energy contained within moving air and converts it into electrical energy.
A photovoltaic module converts part of the incident solar radiation into electricity.
A battery does not generate energy. It stores energy through one process and releases it through another.
Hydrogen produced through electrolysis is not a primary energy source. It is an energy carrier: energy is consumed producing it so that some of that energy can subsequently be transported, stored or converted again.
None of those observations represents an argument against those technologies, they simply mean we must evaluate them as engineering systems rather than convenient political and ideological labels.
And this leads to another distinction that matters, renewable describes the resource. It does not describe the absence of consequence.
Renewable does not mean impact-free
Wind and sunlight are naturally replenished resources whilst wind turbines and photovoltaic modules are not.
Neither are batteries, substations, export cables, transmission networks or electrolysers.
They must be manufactured, steel must be produced, copper must be mined and refined. Concrete requires raw materials and energy whilst polymers and composites require manufacturing.
Permanent magnets require minerals and processing. Subsea and terrestrial electrical networks require enormous quantities of physical infrastructure and battery systems require extraction, refining, manufacturing and eventually recycling or disposal.
The IEA has described the change explicitly as a shift from a historically fuel-intensive energy system towards one that is increasingly material-intensive. Its more recent critical-minerals work shows continuing rapid growth in demand for copper, lithium, graphite, rare earths and other materials as grids, renewables, EVs and energy storage expand.[5] IEA
Those supply chains have environmental consequences of their own: emissions, water consumption, land disturbance, waste, biodiversity impacts and social risks all require management. The IEA specifically identifies the environmental and social performance of mineral extraction and processing as an important component of a responsible energy transition. IEA
But intellectual honesty must work in both directions, recognising those impacts does not make renewable electricity environmentally equivalent to unabated fossil generation.
Lifecycle analysis consistently finds the greenhouse-gas intensity of wind, solar and nuclear generation substantially below fossil-fuel generation. NREL's harmonised lifecycle analysis, for example, found median cradle-to-grave greenhouse-gas emissions from coal-fired electricity to be around twenty times those associated with wind, solar and nuclear generation.[6] NREL
So, neither extreme survives examination.
“Renewables have no environmental impact” is incorrect but so is: “Renewables require mining and energy to manufacture; therefore, they provide no environmental benefit.”
The meaningful engineering question is: What is the complete lifecycle consequence of providing the energy service we require, and how do we progressively reduce it?
That means examining extraction, manufacture, construction, operation, transmission, storage, maintenance, replacement and eventual decommissioning.
The transition is not a search for impact-free energy, no engineering system provides that.
It is a search for progressively lower-impact ways of supplying the energy services society requires.
Generation is only one part of an energy system
Another distinction gets lost surprisingly easily, Installed capacity is not delivered energy.
Delivered energy is not dependable capacity and dependable capacity is not flexibility.
None of those, in isolation, guarantees system stability. A power system has to perform multiple functions continuously.
Supply and demand must balance; frequency must remain within limits.
Voltage must be controlled and faults managed.
Reserves must exist but networks have finite capacity whilst generation and demand are geographically distributed.
And all the time, systems must continue operating when individual components fail.
As increasing proportions of electricity come from variable and inverter-connected generation, some of those functions must be provided differently.
Although this is entirely achievable, it also has to be engineered.
The IEA now identifies grids and flexibility as central constraints on the next phase of electricity-system development. More than 2,500 GW of renewable generation, storage and large-load projects are currently stalled in grid-connection queues worldwide. Meeting forecast electricity demand through 2030 would require annual grid investment to rise by approximately 50% from today's roughly USD 400 billion level.[7] IEA
We are becoming increasingly capable of generating low-carbon electricity, the harder problem is understanding how to transmit it, balance it, store it and depend upon it.
Flexibility is not one problem, and storage is not one answer
“Storage” is frequently discussed as though it solves intermittency as a single engineering problem. Sadly, it does not, because system flexibility is required across very different timescales.
Some requirements occur in fractions of a second or seconds whilst others occur over minutes and hours. Then there are those that extend across days or longer periods of changing demand and renewable availability.
IRENA evaluates flexibility requirements at daily, weekly and monthly scales and estimates that, under its 1.5°C scenario, global daily flexibility requirements in 2030 would be around three times their 2019 level. Weekly and monthly requirements also increase materially.[8] IRENA
The appropriate response is therefore unlikely to be one technology but a portfolio.
Batteries already provide fast response, short-duration balancing and energy shifting.
The IEA's current electricity analysis explicitly identifies batteries, demand response, grid expansion and technologies such as grid-forming inverters and synchronous condensers among the tools available for managing systems with increasing proportions of wind and solar.[9] IEA
The challenge is their architecture, scale, location, commercial model and sequencing.
For a period, we need both systems
This may be the most politically inconvenient part of transition; major engineering systems are not normally replaced by shutting down the incumbent system and subsequently constructing something else.
The replacement is designed, infrastructure built and interfaces created. As equipment is designed and commissioned.
Performance is then demonstrated before load is progressively transferred. Only then is the equipment whose function is no longer required can be retired.
The energy system is unusual primarily in its scale and in the fact that it cannot be taken offline while we replace it as the demand domestically and industrially continues.
Consequently, there is an unavoidable period of overlap and for a period we will have to finance and construct renewable generation, networks, storage, flexibility, firm low-carbon generation and associated infrastructure while retaining enough incumbent capacity to keep the existing system secure.
That creates a cost that neither side of the debate benefits from pretending does not exist.
The IEA frames the wider energy challenge around precisely this need to balance security, affordability and sustainability, noting that while some choices improve all three simultaneously, others involve genuine trade-offs.[10] IEA
Overlap is therefore not necessarily evidence of policy failure; it is often the cost of changing a critical system while keeping it operational. But this is an argument cannot be allowed to run indefinitely.
“We still need it” is not an energy strategy
The present requirement for hydrocarbons is real, it is also insufficient justification for their permanent retention.
The IPCC's assessment is clear that energy systems consistent with deep decarbonisation involve substantially lower overall fossil-fuel use, widespread electrification, greater efficiency, greater integration of energy systems and alternative energy carriers for applications that are difficult to electrify directly. It also identifies continued investment in unabated fossil infrastructure as a source of potential carbon lock-in.[11] IPCC
That means calling a technology transitional has consequences. A transitional technology is only transitional if its role diminishes as replacement capability becomes available.
Natural gas, for example, can perform valuable functions within a changing power system but if storage, networks, demand flexibility, firm low-carbon generation and other resources progressively acquire those functions, then the technical justification for continuing to use gas for them should correspondingly diminish.
Equally, closing an existing source of firm energy before its functions have been credibly replaced is not transition either, it is simply the premature removal of system capability.
The correct test is not ideology but function.
Retire function, not nameplate capacity
This leads to one of the most important distinctions in the whole discussion.
If a thermal generating unit is removed from the system, we are not simply removing megawatts. Depending upon its technology and operating regime, it may be contributing energy, dependable capacity, frequency response, reserve, ramping capability, voltage support, short-circuit strength, or other system services.
Replacing its annual energy production with an equivalent quantity of renewable electricity does not necessarily replace all functions.
That does not mean those functions have to continue to be supplied by fossil-fuelled plant.
Nuclear, hydro and other firm low-carbon sources can contribute dependable energy where technically, geographically and economically appropriate.
What matters is that the function has genuinely transferred before we declare the incumbent redundant.
This is why I would resist measuring transition principally through installed generation capacity.
A better question is: What dependency has this investment enabled us to remove?
Offshore wind makes the system problem visible
Offshore wind provides a particularly useful illustration. An offshore wind farm is not simply a collection of turbines.
The operating energy asset includes foundations, array cables, offshore substations, export cables, landfall infrastructure, onshore transmission, balancing arrangements, ports, vessels, maintenance systems, spares, competent personnel and ultimately a grid capable of accepting and distributing its output.
A relatively small failure or constraint at one interface can restrict the economic value of billions of pounds of otherwise operational infrastructure.
That is familiar territory for anyone who has worked on complex offshore projects in that systems frequently fail at interfaces. The energy transition contains an extraordinary number of them.
We should therefore be cautious about celebrating optimisation of individual technologies if the overall system remains poorly integrated.
The objective is not the most efficient turbine, battery, gas plant or interconnector considered independently.
The objective is the most effective energy system.
The solutions are available — but they must be assembled
None of this suggests that the transition is technically impossible, quite the opposite.
We know how to build renewable generation at enormous scale, battery deployment is accelerating quickly and costs have fallen substantially.
Demand response remains underutilised but offers significant system potential whilst grid-forming inverter technologies are developing rapidly.
Transmission and interconnection are well understood technically even where planning and delivery remain difficult.
Hydro, nuclear and geothermal can provide firm low-carbon energy where circumstances support them.
Long-duration storage technologies can increasingly address requirements beyond conventional battery timescales.
Electrification can eliminate direct combustion from many end uses.
Efficiency reduces the quantity of energy the whole system has to provide.
Hydrogen and derivatives may have important roles where direct electrification is impractical.
Carbon capture may contribute in industrial processes and residual applications where emissions are difficult to eliminate by other means.
Recycling, material efficiency and improved supply-chain design can progressively reduce the environmental burden of the infrastructure we construct.
The question is therefore becoming less:
“Do we possess technologies capable of transition?”
and increasingly:
“Can we integrate, finance and deploy the correct technologies in the correct sequence while keeping the system reliable?”
Which is fundamentally an engineering problem to solve.
The objective is controlled redundancy
This brings us back to the uncomfortable truth.
The hydrocarbon argument cannot ultimately remain: We are still required; therefore, we must always be required.
If that were accepted, the transition would never occur.
The appropriate response is: You are required today. So let us engineer the system until you are not.
Conversely the renewable argument cannot simply become: We have installed equivalent generating capacity, therefore the incumbent system can now close.
The appropriate response is equally straightforward: Demonstrate that the energy and system functions have been replaced.
Neither proposition is anti-renewable and neither is anti-hydrocarbon.
They simply apply the same engineering test to both.
The transition succeeds when lower-impact energy systems become sufficiently available, affordable, controllable and dependable that higher-impact alternatives lose their technical justification.
Not because we have renamed the system, not because a target date has arrived and not because one side has won an ideological argument. But because we no longer need them.
The energy transition will be measured by how much of the old system the new one allows us to stop needing.
References
[1] International Renewable Energy Agency (IRENA), Renewable Capacity Statistics 2026 and “Near-700 GW Surge in 2025 Proves Renewable Energy Resilience”, 2026. Renewable capacity reached 5,149 GW after 692 GW of additions in 2025; renewables represented 85.6% of capacity additions. IRENA — Renewable Capacity Statistics 2026
[2] Energy Institute, Statistical Review of World Energy 2026. Total energy supply rose 1.7% in 2025; all major energy sources reached record highs, while renewables provided the largest contribution to supply growth. Energy Institute — Statistical Review of World Energy 2026
[3] NASA Glenn Research Center, “First Law — Conservation of Energy”. The first law expresses conservation of energy and conversion between forms. NASA — First Law of Thermodynamics
[4] NASA Glenn Research Center, “Second Law of Thermodynamics”; Adkins, C.J., Equilibrium Thermodynamics. The second law introduces entropy and irreversibility into real thermodynamic processes. NASA — Second Law of Thermodynamics
[5] International Energy Agency, The Role of Critical Minerals in Clean Energy Transitions and Global Critical Minerals Outlook 2025/2026. Clean-energy systems shift demand towards materials and critical-mineral supply chains; energy technologies are a major driver of recent mineral-demand growth. IEA — Global Critical Minerals Outlook 2025
[6] National Renewable Energy Laboratory, Life Cycle Assessment Harmonization. Harmonised lifecycle estimates show substantially lower median greenhouse-gas emissions for wind, solar and nuclear generation than coal-fired electricity. NREL — Life Cycle Assessment of Energy Systems
[7] International Energy Agency, Electricity 2026 — Grids. More than 2,500 GW of generation, storage and large-load projects are in connection queues; annual grid investment needs to rise substantially through 2030. IEA — Electricity 2026: Grids
[8] International Renewable Energy Agency, Flexibility for a Secure and Affordable Power Sector Transformation, 2026. Assesses daily, weekly and monthly flexibility requirements under increasing variable-renewable penetration. IRENA — Power System Flexibility 2026
[9] International Energy Agency, Electricity 2026 — Flexibility. Reviews batteries, demand response, grid-forming inverters, synchronous condensers and other sources of system flexibility. IEA — Electricity 2026: Flexibility
[10] International Energy Agency, World Energy Outlook 2024 — Security, Affordability and Sustainability. Examines the trade-offs and interactions between energy security, affordability and emissions reduction. IEA — Security, Affordability and Sustainability
[11] Intergovernmental Panel on Climate Change, AR6 WGIII, Summary for Policymakers and Chapter 6, Energy Systems. Assesses fossil-fuel reduction, electrification, low-emission energy, efficiency, system integration and risks of carbon lock-in. IPCC — AR6 WGIII Energy Systems