# Efficiency, waste heat and recovery: Where energy is wasted Source page: https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p1/ [Understanding Energy](https://www.materialinsights.org/learn/energy/) / [Module 12: Efficiency, waste heat and recovery](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/) Part 1 of 5 ## Where energy is wasted Energy is never destroyed, but most of it does not end up doing what we wanted. Roughly seven-tenths of the world's primary energy leaves the system as unwanted heat. This part sets out how efficiency is measured, where along the chain the losses happen, and which losses physics makes unavoidable. ### 1.1 Efficiency is useful output divided by input, and converters range from 2% to 99% The first law of thermodynamics says that in a closed system energy is conserved. You can turn it from one form into another, but you cannot make or destroy it. So when engineers talk about losses, they do not mean energy vanishing. They mean energy leaving in a form nobody wanted: heat in a car radiator, warm water from a power station, noise, friction, light from a hot furnace wall. Efficiency is simply the useful output divided by the total input. A machine that takes in 100 units and delivers 35 as work is 35% efficient. The other 65 units are still there, almost always as heat at a temperature too low to be easily used. Smil's tables of converters span an astonishing range. An open cooking fire puts less than a tenth of the wood's energy into the pot. An incandescent bulb turns 2 to 5% of its electricity into light. A steam locomotive managed 6 to 10%. A petrol car engine reaches 20 to 30% at best, and less in stop-start traffic. Large marine diesels approach 50%. A modern combined-cycle gas plant reaches about 60%. Big electric motors exceed 95%, and large transformers exceed 99%. Two patterns stand out. First, converters that turn heat into work, such as engines and turbines, sit in the middle of the range, because physics caps them. Second, converters that work entirely with electricity or with direct chemical-to-electrical paths, such as motors, transformers, batteries and LEDs, sit near the top. This is the deep reason electrification saves energy: it replaces heat engines, which are fundamentally limited, with electrical machines, which are not. Efficiency also depends on how you draw the boundary. A heat pump can deliver more heat than the electricity it uses, which looks like over 100%, because it moves heat from outside rather than making it. Data Typical efficiencies of common converters | Converter | Conversion | Typical efficiency | |---|---|---| | Open cooking fire | chemical to useful heat | under 10% | | Incandescent bulb | electricity to light | 2 to 5% | | Steam locomotive | chemical to motion | 6 to 10% | | Petrol car engine | chemical to motion | 20 to 30% at best | | Large marine diesel | chemical to motion | about 50% | | Combined-cycle gas plant | chemical to electricity | about 60% | | Gas boiler | chemical to heat | 85 to 95% | | Large electric motor | electricity to motion | over 95% | | Large transformer | electricity to electricity | over 99% | Source: Smil, Energy: A Beginner's Guide (2006), via Smil Mastery Course, M1 'Conservation and efficiency' Common misreading #### Heat pumps do not break the first law A heat pump that gives 3.5 units of heat for 1 unit of electricity is not creating energy. The other 2.5 units are taken from outdoor air, the ground or a waste-heat stream. That is why engineers call its ratio a coefficient of performance rather than an efficiency. Counting all inputs, including the ambient heat, the books balance exactly. The same logic explains why a refrigerator can move more heat out of its cabinet than it draws as electricity. - Smil, Energy: A Beginner's Guide (2006) · via Smil Mastery Course, M1 'Conservation and efficiency' - Jaffe and Taylor, The Physics of Energy (2018) · via Smil Mastery Course, M1 'Conservation and efficiency' ### 1.2 About seven-tenths of primary energy ends up as waste heat Follow the world's 592 exajoules of primary energy in 2024 from source to service. The first big loss happens in conversion: thermal power stations, refineries and other energy plants. Most of the world's electricity still comes from burning coal and gas in plants that turn 35 to 60% of the fuel into electricity. The rest goes up the stack or into cooling water. Refineries, gas processing, coal washing and pipelines add smaller losses. Taken together, roughly a quarter to a third of primary energy is gone before anything reaches a user. The second big loss happens at the point of use. Engines in cars and lorries waste most of their fuel as heat. Boilers and furnaces lose heat up their flues and through their walls. Old motors, pumps and fans lose energy to friction and poor matching with the load. Lighting, appliances and electronics finish as heat in the room. A 2016 study by Clemens Forman and colleagues estimated, with a top-down method based on global energy statistics, that about 72% of the world's primary energy is lost after conversion, leaving under 30% as useful heat, motion and light. According to the study's summary, most of the waste heat below 100 degrees comes from electricity generation, followed by transport and industry. The study also found that 63% of the waste heat streams it considered were below 100 degrees Celsius. That detail matters a great deal, as later sections show. The result is that the world pays for about three units of primary energy for every unit of service it receives. This is both a problem and an opportunity. Every percentage point gained at the end of the chain saves more than a point at the start, and a large share of the waste is concentrated in a few places: power station condensers, vehicle exhausts and industrial furnaces. Process flowchart From world primary energy to useful energy Inputs **Primary energy, 2024**about 592 EJ Process steps 1 **Power stations, refineries and energy industry** roughly 160 EJ lost, mostly as low-temperature heat in cooling water and flue gas 2 **Final energy delivered** roughly 430 EJ of electricity, fuels and heat 3 **End-use machines: engines, boilers, motors, lamps** roughly 260 EJ lost, mostly as exhaust and surface heat Outputs **Useful heat, motion and light**roughly 165 EJ, about 28% **Waste heat**roughly 425 EJ, about 72% Illustrative split applying the Forman et al. 72% loss share to the 2024 primary total. Source: Forman, C., Muritala, I. K., Pardemann, R. and Meyer, B., 'Estimating the global waste heat potential', Renewable and Sustainable Energy Reviews 57 (2016) 1568-1579, https://ideas.repec.org/a/eee/rensus/v57y2016icp1568-1579.html (checked against the abstract only; full text is paywalled: about 72% of global primary energy lost after conversion; 63% of considered waste heat streams below 100 C, mostly from electricity generation, then transport and industry) In the data #### Rough global energy cascade Primary energy, 2024: about 592 EJ. Lost in conversion (power stations, refineries, own use): roughly 160 EJ. Final energy delivered: roughly 430 EJ. Lost at the point of use: roughly 260 EJ. Useful energy delivered: roughly 165 EJ, under 30% of the start. These are rounded figures that combine the Energy Institute total with the Forman share; they show orders of magnitude, not precise accounts. - [Forman, C., Muritala, I. K., Pardemann, R. and Meyer, B., 'Estimating the global waste heat potential', Renewable and Sustainable Energy Reviews 57 (2016) 1568-1579](https://ideas.repec.org/a/eee/rensus/v57y2016icp1568-1579.html (checked against the abstract only; full text is paywalled: about 72% of global primary energy lost after conversion; 63% of considered waste heat streams below 100 C, mostly from electricity generation, then transport and industry)) · web - Energy Institute, Statistical Review of World Energy 2025 (2024 data) [web ↗](https://www.energyinst.org/statistical-review) · web ### 1.3 The second law sets a floor under some losses, so not all waste can be recovered If energy is never destroyed, why can we not simply collect the waste heat and use it again? The answer is the second law of thermodynamics. Energy comes in different qualities. Electricity and mechanical work are the highest quality: they can be turned into almost anything with little loss. Heat is lower quality, and the lower its temperature relative to the surroundings, the less useful it is. The French engineer Sadi Carnot showed in 1824 that any engine turning heat into work must reject some heat to a colder place. The best possible share that can become work is one minus the ratio of the cold temperature to the hot temperature, both measured from absolute zero in kelvin. A steam turbine taking steam at 600 degrees Celsius (873 K) and rejecting heat to cooling water at 30 degrees (303 K) can at most turn about 65% of the heat into work. Real plants manage about 40%, because of friction, finite heat-transfer rates and boiler losses. This means a large part of the loss at a thermal power station is not a sign of bad engineering. It is the cost of turning heat into work. The 50-odd units of warm water leaving a coal plant's condenser at 30 to 40 degrees are vast in quantity but very low in quality. They cannot run another turbine. They can warm greenhouses, fish farms or, via a heat pump, a district heating network. Engineers use the word exergy for the share of energy that could in principle become work. Measured in exergy, the waste picture changes. A hot exhaust at 500 degrees carries a lot of exergy and is well worth recovering. Warm water at 35 degrees carries almost none, even if its energy content is enormous. Process flowchart Where the energy goes in a coal power station Inputs **Coal**100 units of chemical energy Process steps 1 **Boiler** about 88 to 90% to steam; about 10 units lost up the stack 2 **Steam turbine** high-temperature steam expands and drives the shaft 3 **Condenser** about 50 units rejected as warm cooling water at 30 to 40 degrees Celsius 4 **Generator and plant auxiliaries** a few units used by pumps, fans and mills Outputs **Electricity sent out**about 35 to 40 units **Low-grade waste heat**about 60 units Typical subcritical or supercritical plant; the condenser loss is largely set by the second law. Source: Jaffe and Taylor, The Physics of Energy (2018), via Smil Mastery Course, M1 'Conservation and efficiency' Worked example #### Why a cold sea helps a power station Carnot's limit depends on the cold side as well as the hot. A plant rejecting heat at 15 degrees Celsius instead of 35 can, in principle, turn a couple of percentage points more of its heat into work. That is one reason coastal plants with cold seawater cooling perform slightly better than inland plants with cooling towers on hot days, and why heatwaves cut thermal plant output. - Jaffe and Taylor, The Physics of Energy (2018) · via Smil Mastery Course, M1 'Conservation and efficiency' - Smil, Energy: A Beginner's Guide (2006) · via Smil Mastery Course, M1 'Conservation and efficiency' ### See it in the table [H₂O Water](https://www.materialinsights.org/commodities/water/)[Amb Ambient heat](https://www.materialinsights.org/energy/e_ambient/) ### Check your understanding About what share of world primary energy did Forman and colleagues estimate is lost after conversion? Their 2016 top-down estimate put losses at about 72%, leaving under 30% as useful energy. Power generation is the largest source. Why can a thermal power station not turn most of its heat into electricity, however well it is built? Carnot's limit means a turbine working between 600 and 30 degrees Celsius can at best convert about 65%; real plants reach about 40%. Generators themselves are over 98% efficient. Which waste heat stream is most valuable to recover, per unit of energy? Higher temperature means more exergy, the part that can do work or meet high-temperature needs. Low-temperature streams carry plenty of energy but little exergy. [← Module contents](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/)[Mark read · Part 2: Efficiency gains in history and energy intensity →](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p2/) ## Other sections - [Module overview](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/index.md) - [Efficiency gains in history and energy intensity](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p2/index.md) - [Rebound, Jevons and the limits of efficiency](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p3/index.md) - [Waste heat by temperature grade](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p4/index.md) - [Recovery technologies and what they cost](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/p5/index.md) - [Key terms, quiz and sources](https://www.materialinsights.org/learn/energy/efficiency-and-recovery/review/index.md)