Energy Efficiency: The Resource Nobody Can Photograph
Every discussion of energy supply carries a silent assumption: that the demand is fixed and the question is how to meet it. It is not fixed. Roughly a third of the energy entering a rich economy is wasted in ways that a paid-for piece of equipment would recover, and that recovery is generally cheaper than any way of generating the same amount.
The Arithmetic of Not Needing It
A kilowatt-hour that is never consumed needs no power plant, no transmission line, no storage and no fuel. It also incurs none of the losses along that chain, which in a thermal system are substantial: a coal plant discards roughly 60 percent of its fuel energy as heat before the electricity leaves the site, and the grid loses another 5 to 10 percent before it arrives.
That cascade means a unit saved at the point of use is worth considerably more than a unit generated. Saving one kilowatt-hour in a building fed from a coal grid avoids roughly three kilowatt-hours of primary energy. This is why efficiency measures often show costs per tonne of carbon avoided that are negative - the measure pays for itself and then continues saving.
The scale is larger than most estimates of any single generating technology. Studies of industrial economies consistently find 20 to 30 percent of final energy consumption recoverable with measures that pay back within their own lifetime. That is comparable to the entire output of the nuclear or hydro fleet, available without building anything that needs planning permission.
It also compounds with everything else. A building that needs half the heat can be served by a smaller heat pump at a lower flow temperature, which raises that heat pump's efficiency in turn. Efficiency is the one intervention that makes every other intervention cheaper.
What Has Already Happened
Lighting is the clearest case and the most recent. An incandescent bulb converts about 5 percent of its electricity into visible light and the rest into heat. A modern LED converts 40 to 50 percent, so the same illumination takes 80 to 90 percent less power. Global lighting electricity demand has fallen despite more light being used, and the transition took roughly a decade once the price fell.
Motors are the largest and least discussed. Electric motors consume around 45 percent of the world's electricity, driving fans, pumps, compressors and conveyors. Most were historically fixed-speed, throttled by valves and dampers - the equivalent of driving with the accelerator floored and controlling speed with the brake. Variable-speed drives cut consumption by 20 to 50 percent in variable-load applications, and the payback is often under two years.
Buildings hold the largest absolute saving and the slowest turnover. Insulation, airtightness and mechanical ventilation with heat recovery can reduce a building's heating demand by 50 to 90 percent, and the Passivhaus standard demonstrates the upper end routinely. None of it requires new technology; the difficulty is that buildings last 50 to 100 years, so the stock turns over at 1 to 2 percent a year.
Appliance standards have delivered quietly and at scale. A refrigerator sold today uses roughly a quarter of the electricity of its 1980 equivalent while being larger, and regulation rather than consumer choice drove most of that. The same pattern holds for washing machines, televisions and standby power.
The Rebound Effect, Honestly Stated
Making energy services cheaper increases demand for them. Efficient lighting leads people to light more rooms for longer; a warmer house is heated to a higher temperature than a cold one was. This is the rebound effect, and any honest account of efficiency has to include it.
The size is the contested part, and the evidence is narrower than the rhetoric. Direct rebound - using more of the same service - is typically measured at 10 to 30 percent in wealthy economies, meaning most of the saving persists. Indirect rebound, where money saved is spent on something else that uses energy, adds perhaps another 5 to 15 percent depending on what is bought.
Backfire, where the rebound exceeds 100 percent and efficiency increases total consumption, is occasionally claimed and rarely demonstrated at the level of an individual measure. It is more plausible historically at the level of a whole economy over a century, which is a different claim than the one usually being made in a policy argument.
In developing economies rebound is genuinely large, and it is not a problem. When efficient lighting lets a household light its home for the first time, the increased consumption is the point. Efficiency there delivers welfare rather than energy reduction, and treating that as a failure misreads what the measure is for.
Why It Is Chronically Under-Delivered
If efficiency is so cheap, the obvious question is why it is not already done. The answer is that the obstacles are structural rather than technical, and they are well catalogued.
The split incentive is the largest. A landlord buys the boiler and the tenant pays the heating bill, so neither party has both the ability and the motive to act. The same division runs through offices, rented equipment and speculative construction. No amount of technical improvement addresses it.
Capital and attention are the second. Efficiency requires spending now to save later, and the people best placed to benefit are often those least able to raise capital. It is also nobody's job. A factory manager is measured on output, and a 3 percent reduction in energy cost rarely competes for attention with a production problem, even when the return is excellent.
Invisibility is the third and perhaps the deepest. A wind farm is photographable, opens with a ceremony and is counted in a statistic. A building that was refurbished and now uses 60 percent less heat produces no such artefact - the saving shows up as an absence, in a counterfactual nobody measured. Political and financial systems reward things that can be pointed at, which is why efficiency is universally agreed to be the cheapest option and universally under-funded relative to generation.
Frequently asked questions
Why is efficiency called the first fuel?
Because energy not used needs no generation, transmission, storage or fuel, and avoids the losses along all of those steps. A kilowatt-hour saved in a building fed from a thermal grid avoids roughly three kilowatt-hours of primary energy, which is why efficiency measures often have negative net cost.
How much energy could efficiency actually save?
Studies of industrial economies consistently find 20 to 30 percent of final consumption recoverable with measures that pay for themselves within their own lifetime. That is comparable to the total output of the world's nuclear or hydro fleets, available without building new generation.
Does the rebound effect cancel out the savings?
No. Direct rebound is typically 10 to 30 percent in wealthy economies and indirect rebound adds perhaps 5 to 15 percent, so most of the saving persists. Backfire, where consumption rises overall, is occasionally claimed but rarely demonstrated for an individual measure. In developing economies rebound is large and is the intended outcome.
What is a split incentive?
When the party who would pay for an efficiency measure is not the party who would benefit. A landlord buys the boiler and the tenant pays the fuel bill, so neither has both the means and the motive. It is the single largest structural obstacle to efficiency and no technical improvement addresses it.
Why does efficiency get less attention than renewables?
Because it is invisible. A wind farm can be photographed and counted; a building that now uses 60 percent less heat produces no artefact, and the saving exists only as a counterfactual. Efficiency is also diffuse across millions of decisions rather than concentrated in projects that attract financing and political attention.
