Efficiency, waste heat and recovery: Rebound, Jevons and the limits of efficiency

3.1 Jevons saw that better steam engines raised Britain's coal use, not lowered it

In The Coal Question, published in 1865, William Stanley Jevons noticed something that seemed backwards. Steam engines had become far more efficient since Watt, yet Britain was burning more coal than ever. His explanation was that each gain in efficiency made steam power cheaper per unit of work, so it spread to new uses: more mills, more railways, more ships, more mines. Total coal use rose faster than the saving per engine. This has come to be called the Jevons paradox.

Economists now speak of rebound effects and divide them into kinds. Direct rebound happens when a cheaper service is used more: a more efficient car is driven further, a cheaper-to-run air conditioner is run longer, a better-insulated home is kept warmer. Indirect rebound happens when money saved on energy is spent on other things that also use energy, such as a holiday flight. Economy-wide rebound happens when efficiency makes whole industries more productive, which grows the economy and its energy demand.

How big are these effects? A widely cited review of the evidence for household energy services in rich countries, by Steve Sorrell and colleagues, concluded that direct rebound should generally be less than 30%, with many estimates for heating and car travel in the 10 to 30% range. Some of the expected saving is taken back, but most remains. Economy-wide effects are harder to measure and more disputed. Where a service is far from saturated, as with cooling or mobility in India, rebound can be large, because people use the savings to get more of something they wanted all along.

That last case is not a failure. A family that buys an efficient air conditioner and runs it more is better off. Rebound tells us that efficiency is a way to deliver more service per unit of energy, not a reliable way to cut total use on its own.

• Smil, Energy: A Beginner's Guide (2006) · via Smil Mastery Course, M1 'Conservation and efficiency'
• Smil, Energy and Civilization: A History (2017) · via Smil Mastery Course, M1 'Conservation and efficiency'
• Sorrell, S., Dimitropoulos, J. and Sommerville, M., 'Empirical estimates of the direct rebound effect: A review', Energy Policy 37(4) (2009) 1356-1371 · web

3.2 Relative savings are real; absolute savings need efficiency to outrun growth

The useful distinction is between relative and absolute savings. A relative saving means less energy per unit of service, per product or per dollar of GDP. An absolute saving means less energy in total. History is full of relative savings and short of absolute ones.

The arithmetic is unforgiving. If an economy grows at 6% a year, as India's has, and its energy intensity falls by 4% a year, its energy use still rises by about 2% a year. If intensity falls by only 2%, energy use rises by about 4%. Total use falls only when intensity drops faster than the economy grows, which has happened mainly in rich, slow-growing economies, and partly because they moved heavy industry abroad.

Smil's work on materials makes the same point. Aluminium cans became much lighter, but people drank from far more cans. Car engines became much lighter per watt, but cars became heavier and more powerful. Relative dematerialisation was real; absolute material use kept rising.

None of this makes efficiency pointless. For a fast-growing country, efficiency decides how much new supply must be built. Every air conditioner, motor and furnace that is twice as efficient halves the power plants, gas imports and grid capacity needed to serve it. It also cuts the cost of decarbonisation, because a smaller clean-energy system is needed to meet the same demand.

For an investor, the right question is not whether efficiency will cut total demand in India or Indonesia. It mostly will not, for decades. The question is how much of the growth in services it can deliver without new fuel, and at what cost per unit saved compared with building new supply.

• Smil, Energy at the Crossroads: Global Perspectives and Uncertainties (2003) · via Smil Mastery Course, M13 'Energy intensity'
• Smil, Making the Modern World (2014) / Materials and Dematerialization (2nd ed., 2023) · via Smil Mastery Course, M11 'Material flows, recycling and dematerialisation'

3.3 Smil's test for energy claims: check scale, speed and physics

In Energy Myths and Realities, Smil examines beliefs that misled energy policy for decades. Several bear directly on efficiency and recovery.

The first is that new technology will transform the system quickly. Electric cars were confidently called the future around 1900. Nuclear power was promised as too cheap to meter in the 1950s. In the 1970s, advocates of soft energy predicted that small decentralised renewables and efficiency would supply a large share of American energy by 2000. Each forecast underestimated how slowly huge, long-lived capital stocks turn over.

The second is that large add-on systems can be built at once. Smil calculated that capturing and storing just 15% of 2008's CO₂ emissions would need an industry handling 1.3 to 2.2 times the yearly volume of the entire world crude oil industry, built from scratch. Waste heat recovery faces a milder form of the same problem. Each project is bespoke, needs a nearby user for the heat, and must fit around a plant designed decades ago.

The third is that physics can be negotiated. Low-grade heat cannot become much electricity, whatever the marketing says. Biofuels and wind face low power densities. A device that claims to turn 60-degree water into large amounts of power should be checked against Carnot's limit before anyone invests.

The constructive lesson is to judge every efficiency or recovery proposal on three points. What is the scale of the opportunity in joules, not percentages? How fast can it be installed, given the turnover of the equipment it depends on? And does it respect the second law, especially the temperature of the heat it uses?

• Smil, Energy Myths and Realities: Bringing Science to the Energy Policy Debate (2010) · via Smil Mastery Course, M9 'Energy myths and realities'
• Smil, Numbers Don't Lie (2020) · via Smil Mastery Course, M9 'Energy myths and realities'

See it in the table

Check your understanding

What did Jevons observe about steam engines and coal in 1865?

An economy grows 6% a year and its energy intensity falls 4% a year. Roughly what happens to its energy use?

Which first check would Smil apply to a claim that a device can produce lots of electricity from 60-degree water?