# Energy transitions and climate: Decarbonisation so far, and the hard parts

Source page: https://www.materialinsights.org/learn/energy/energy-transitions-climate/p3/

[Understanding Energy](https://www.materialinsights.org/learn/energy/) / [Module 13: Energy transitions and climate](https://www.materialinsights.org/learn/energy/energy-transitions-climate/) Part 3 of 4

## Decarbonisation so far, and the hard parts

The world has cut the carbon intensity of its energy for two centuries, mostly without trying. Cutting total emissions is another matter. This part looks at what has been achieved, why electricity is leading, why four basic materials are so hard to clean up, and which popular beliefs about energy have repeatedly misled investors and governments.

### 3.1 Carbon intensity has fallen for two centuries while emissions kept rising

There are two ways to measure decarbonisation. Relative decarbonisation is a fall in the CO2 released per unit of energy, or per unit of GDP. Absolute decarbonisation is a fall in total emissions. The world has achieved a good deal of the first and very little of the second.

Relative decarbonisation has a long history. Each step in the fuel sequence moved towards fuels richer in hydrogen. In Smil's accounting, the ratio of hydrogen to carbon rises from about 0.1 for wood to about 1 for coal, 2 for oil products and 4 for methane. When a fuel burns, its hydrogen becomes water and its carbon becomes CO2, so the more hydrogen it carries, the less CO2 it releases per unit of heat. IPCC default factors put natural gas at about 56 kg of CO2 per GJ, crude oil at about 73 and bituminous coal at about 95. Hydro, nuclear, wind and solar electricity add no combustion CO2 at all. Add steady gains in efficiency, and the carbon intensity of the world economy has fallen for decades.

Absolute emissions behave differently because they are the product of several factors: the number of people, output per person, energy used per unit of output, and CO2 per unit of energy. The first two have grown faster than the last two have fallen. Global fossil CO2 has risen almost every year, apart from recessions and the pandemic, reaching the record 38.1 Gt projected for 2025.

There is a sign of change in the rate. The Global Carbon Budget finds that total CO2 emissions, including land use, grew by about 0.3% a year over the past decade, compared with 1.9% a year in the decade before. Emissions have not yet peaked, but the curve has flattened. Several rich economies have cut their emissions in absolute terms for 15 years or more, partly through efficiency and cleaner power, and partly by importing energy-intensive goods instead of making them at home.

Data The hydrogen ladder of fuels

| Fuel | Hydrogen to carbon ratio (Smil) | CO2 per GJ of heat (IPCC default) |
|---|---|---|
| Wood | about 0.1 | about 112 kg (counted as neutral if regrown) |
| Bituminous coal | about 1 | about 95 kg |
| Crude oil and products | about 2 | about 73 kg |
| Natural gas (methane) | 4 | about 56 kg |
| Hydro, nuclear, wind, solar | no carbon | zero at the point of generation |

 Source: IPCC, 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Vol. 2 Energy, Table 2.2 (2006), default CO2 emission factors for combustion, https://www.ipcc-nggip.iges.or.jp/public/2006gl/vol2.html  Worked example

#### The four levers of emissions

A simple identity helps: CO2 = people x GDP per person x energy per unit of GDP x CO2 per unit of energy. If world population grows about 1% a year and GDP per person about 2%, emissions fall only if energy intensity and carbon intensity together improve by more than 3% a year. For most of the past half-century the two together improved by around 2% a year, which is why emissions kept rising. Reaching net zero means pushing the last two levers far harder, especially carbon intensity, which has moved slowest.

- Smil, Energy Transitions: Global and National Perspectives (2017) · via Smil Mastery Course, M9 'Decarbonisation so far'
- Smil, Grand Transitions: How the Modern World Was Made (2021) · via Smil Mastery Course, M9 'Decarbonisation so far'
- IPCC, 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Vol. 2 Energy, Table 2.2 (2006) · default CO2 emission factors for combustion, https://www.ipcc-nggip.iges.or.jp/public/2006gl/vol2.html
- Global Carbon Project, Global Carbon Budget 2025 (November 2025) · 'Fossil fuel CO2 emissions hit record high in 2025', https://globalcarbonbudget.org/fossil-fuel-co2-emissions-hit-record-high-in-2025/

### 3.2 Electricity is where decarbonisation is furthest ahead

Electricity is the part of the energy system where clean technology competes most directly with fossil fuels, and it is where progress is clearest. Solar modules, wind turbines and batteries are manufactured products, and their costs have fallen with each doubling of production. The IEA's World Energy Outlook 2025 puts the decline since 2010 at about 90% for solar, 70% for wind and 90% for batteries.

The 2025 data compiled by Ember show the results. Renewables supplied 33.8% of the world's electricity, and with nuclear the low-carbon share was in the low forties per cent. China, which accounted for more than half of the world's new solar in 2025, saw its fossil generation fall for the first time since 2015, with wind and solar at 22% of its mix, above the OECD average of 20%. India's renewable generation rose by a record 24%, with solar up 37%, and its power-sector emissions fell by about 3%.

Germany's Energiewende is the most studied national example, and Smil uses it to show both sides of a policy-driven transition. Germany built enormous wind and solar capacity after 2000 and now gets more than half of its electricity from renewables. But it also closed its nuclear plants, which kept coal and lignite running longer than they otherwise would have, and its household electricity prices became among the highest in Europe. The lesson is not that the effort failed, but that it took two decades, cost a great deal, and was shaped by choices that went beyond cost per kWh.

Variable sources bring new tasks as their share grows. The system needs more transmission, storage such as Lithium iron phosphate batteries and Pumped hydro storage, flexible demand, and firm capacity for windless winter weeks or monsoon evenings. These integration costs rise as the share of wind and solar rises, and they belong in any honest comparison of technologies.
 Case

#### India's power sector in 2025

Ember's figures for India in 2025 show what a good year looks like. Renewable generation rose by a record 98 TWh, with solar up 53 TWh and wind up 22 TWh; rooftop and other distributed solar added an estimated 22 TWh. Demand grew only 2.4%, so fossil generation fell by 52 TWh and power-sector emissions fell by 42 Mt. India also installed more new solar capacity than the United States for the first time. A single year can flatter, though: mild weather and slow demand growth helped, and the test is whether clean supply keeps pace when demand returns to its usual growth of 6% or more.

- IEA, World Energy Outlook 2025 (2025) · Overview and key findings, https://www.iea.org/reports/world-energy-outlook-2025/overview-and-key-findings
- Carbon Brief, 'IEA: Fossil-fuel use will peak before 2030 unless stated policies are abandoned' (November 2025) · summary of IEA World Energy Outlook 2025 scenario results, https://www.carbonbrief.org/iea-fossil-fuel-use-will-peak-before-2030-unless-stated-policies-are-abandoned/
- Ember, Global Electricity Review 2026 (April 2026) · 2025 data, https://ember-energy.org/latest-insights/global-electricity-review-2026/ and https://ember-energy.org/chapter/2025-in-review/
- Smil, Energy Transitions: Global and National Perspectives (2017) · via Smil Mastery Course, M9 'Decarbonisation so far'

### 3.3 Ammonia, steel, cement and plastics are the hard core of emissions

Smil argues that modern civilisation rests on four materials. Ammonia, as nitrogen fertiliser, feeds a large share of humanity. Steel makes structures, machines and vehicles. Cement binds the concrete of buildings, dams, roads and even wind turbine foundations. Plastics run from packaging to medical devices. The world makes roughly 150 to 200 million tonnes of ammonia, close to 1.9 billion tonnes of steel, more than 4 billion tonnes of cement and several hundred million tonnes of plastics a year. In Smil's estimate their production takes about 17% of world primary energy and causes about a quarter of fossil CO2 emissions.

What makes them hard is chemistry, not just the source of heat. In a blast furnace, carbon from coke is the agent that strips oxygen from iron ore, so CO2 is an unavoidable product. In a cement kiln, more than half of the CO2 comes from the limestone itself as it splits into lime and CO2, whatever fuel heats the kiln. Ammonia needs hydrogen, today made mostly from natural gas or, in China, from coal. Plastics are built from carbon taken out of oil and gas, and that carbon is released if they are burned.

Alternatives exist. Hydrogen from Electrolysers can reduce iron ore directly; electric arc furnaces can melt scrap and directly reduced iron; cement plants can capture their CO2 or cut the clinker content of cement; ammonia can be made from hydrogen produced with clean electricity. Most of these work technically but cost more and need large amounts of clean power. Long-haul aviation and shipping face a parallel problem: batteries are too heavy for long journeys, so they need dense fuels such as e-fuels, ammonia or biofuels, which are scarce and expensive.

India makes this concrete. In the IEA's 2024 outlook its steel output was on track to grow about 70% and cement output nearly 55% by 2035. Most of the plants that will make that steel and cement have not been built yet, which makes the choice of process in the next decade decisive.

Process flowchart Why a cement kiln emits CO2 even with clean heat Inputs **Limestone (CaCO3) and clay**about 1.5 t of raw meal **Fuel heat**about 3 to 3.5 GJ Process steps 1 **Preheater and calciner**

CaCO3 → CaO + CO2, the process emission
 2 **Rotary kiln at about 1,450 C**

lime fuses with silica and alumina into clinker
 Outputs **Clinker**1 t, ground with gypsum into cement **Process CO2**about 0.5 t **Fuel CO2**about 0.3 t if coal is used

Switching the fuel removes only the smaller of the two CO2 streams.
 Source: Smil, How the World Really Works: A Scientist's Guide to Our Past, Present and Future (2022), via Smil Mastery Course, M9 'The four pillars and the hard parts of decarbonising'  Common misreading

#### Clean electricity alone cannot decarbonise cement

Even a cement kiln heated entirely by renewable electricity would still emit most of its CO2, because the largest source is the breakdown of limestone (CaCO3 → CaO + CO2), not the fuel. Each tonne of clinker releases roughly half a tonne of CO2 from the rock alone. The options are to use less clinker in each tonne of cement, to use less concrete through better design, to develop alternative binders, or to capture the CO2 at the kiln and store it permanently.

- Smil, How the World Really Works: A Scientist's Guide to Our Past, Present and Future (2022) · via Smil Mastery Course, M9 'The four pillars and the hard parts of decarbonising'
- Smil, Making the Modern World: Materials and Dematerialization (2014) · via Smil Mastery Course, M9 'The four pillars and the hard parts of decarbonising'
- Smil, Still the Iron Age: Iron and Steel in the Modern World (2016) · via Smil Mastery Course, M9 'The four pillars and the hard parts of decarbonising'
- IEA, World Energy Outlook 2024 (2024), India findings as reported by PTI · Outlook Business, 17 October 2024, https://www.outlookbusiness.com/sustainability/india-set-to-lead-global-energy-demand-growth-coal-will-remain-central-iea

### 3.4 Energy myths have repeatedly misled policy and investment

Smil devoted a whole book, Energy Myths and Realities, to beliefs that shaped energy policy and then failed to come true. They are worth knowing because the same patterns recur in every investment cycle.

The oldest is that the future belongs to electric cars. Electric vehicles were competitive around 1900, and their triumph was announced again and again through the century. This time the claim is coming true: the IEA counts more than 17 million electric cars sold in 2024, over a fifth of new car sales worldwide. Smil's caution still applies, though. An electric car's climate benefit depends on the electricity that charges it and on the emissions from making its battery.

A second myth was that nuclear electricity would be 'too cheap to meter'. Instead, reactors became some of the most capital-intensive and delay-prone projects in the energy business. A third was the 1970s vision of 'soft energy', in which small, decentralised renewables would supply a large share of US energy by 2000; the forecast missed by around 90%. A fourth was imminent and catastrophic peak oil, which kept failing to arrive as new techniques opened new resources, notably US shale.

Large-scale carbon capture and storage is a current example. Smil calculated that capturing just 15% of 2008 CO2 emissions would need a new industry to gather, compress, transport and bury 1.3 to 2.2 times the volume handled each year by the entire global crude oil industry, which took a century to build. Capture has a role at cement kilns and some industrial plants, but not at that scale soon.

The common thread is the myth of speed: the belief that a complex system can be transformed within a decade or two. Smil's scepticism is sometimes read as hostility to renewables. It is better read as a demand for arithmetic. The same arithmetic now shows solar growing faster than almost anyone forecast, a reminder that pessimistic beliefs can fail too.

Data Five energy beliefs and what happened

| Belief | When popular | What happened |
|---|---|---|
| Electric cars will soon dominate | around 1900, and repeatedly since | a fringe for a century; over a fifth of new car sales by 2024 |
| Nuclear power too cheap to meter | 1950s | costly, slow projects; useful but never cheap |
| Soft energy will power the US by 2000 | 1970s | missed by around 90% |
| Oil is about to run out | 1970s, and again around 2005 | new techniques such as shale raised supply |
| Carbon capture will clean up fossil fuels at scale | 2000s | small volumes; scale would rival the oil industry |

 Source: Smil, Energy Myths and Realities: Bringing Science to the Energy Policy Debate (2010), via Smil Mastery Course, M9 'Energy myths and realities'  Worked example

#### A quick test for any 'soon to dominate' claim

Before backing a claim that a technology will soon dominate, ask four questions. How big is it today in absolute terms, not in growth rates? What does it replace, and how long do those assets normally last? What materials, networks and skills does it need, and do they exist at scale? Who pays during the years when it is more expensive than the incumbent? A claim that survives all four deserves to be taken seriously, whoever makes it.

- Smil, Energy Myths and Realities: Bringing Science to the Energy Policy Debate (2010) · via Smil Mastery Course, M9 'Energy myths and realities'
- Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (2003) · via Smil Mastery Course, M9 'Energy myths and realities'
- IEA, Global EV Outlook 2025 (2025) · 2024 electric car sales above 17 million, share above 20%, https://www.iea.org/reports/global-ev-outlook-2025/executive-summary

### See it in the table
[H₂O Water](https://www.materialinsights.org/commodities/water/)[Oil Crude oil](https://www.materialinsights.org/energy/crude_oil/)[Gas Natural gas](https://www.materialinsights.org/energy/natural_gas/)[Fe Iron ore](https://www.materialinsights.org/commodities/iron_ore/)[St Steel](https://www.materialinsights.org/commodities/steel/)[Li Lithium](https://www.materialinsights.org/commodities/lithium/)[Cem Cement](https://www.materialinsights.org/commodities/cement/)[CaO Lime](https://www.materialinsights.org/commodities/lime/)[H2 Hydrogen](https://www.materialinsights.org/chemicals/hydrogen/)[O2 Oxygen](https://www.materialinsights.org/chemicals/oxygen/)[NH3 Ammonia](https://www.materialinsights.org/chemicals/ammonia/)[LFP Lithium iron phosphate batteries](https://www.materialinsights.org/energy/e_li_lfp/)[PHS Pumped hydro storage](https://www.materialinsights.org/energy/e_pumped_hydro/)[Elz Electrolysers](https://www.materialinsights.org/energy/e_electrolysers/)

### Check your understanding

Why does burning natural gas release less CO2 per unit of heat than burning coal?
 A higher hydrogen to carbon ratio means more of the heat comes from forming water. Gas is about 56 kg of CO2 per GJ against about 95 for coal.

What is the main reason cement is hard to decarbonise?
 Calcination, CaCO3 -> CaO + CO2, is chemistry, not combustion; it releases about half a tonne of CO2 per tonne of clinker.

Smil estimated that capturing 15% of 2008 CO2 emissions would require handling what volume each year?
 The comparison shows that capture at climate-relevant scale means building a new industry larger than the oil industry.

[← Part 2: Why transitions are slow](https://www.materialinsights.org/learn/energy/energy-transitions-climate/p2/)[Mark read · Part 4: Carbon dioxide, climate and the paths ahead →](https://www.materialinsights.org/learn/energy/energy-transitions-climate/p4/)

## Other sections

- [Module overview](https://www.materialinsights.org/learn/energy/energy-transitions-climate/index.md)
- [What a transition is and how fast transitions run](https://www.materialinsights.org/learn/energy/energy-transitions-climate/p1/index.md)
- [Why transitions are slow](https://www.materialinsights.org/learn/energy/energy-transitions-climate/p2/index.md)
- [Carbon dioxide, climate and the paths ahead](https://www.materialinsights.org/learn/energy/energy-transitions-climate/p4/index.md)
- [Key terms, quiz and sources](https://www.materialinsights.org/learn/energy/energy-transitions-climate/review/index.md)