Cogeneration
A power station throws away two-thirds of its energy as waste heat, up a cooling tower, into the sky. Cogeneration is simply the decision to use it.
The effect compounds within years. Put it in place and it keeps working.
Origins
Every power station is a heat engine, and every heat engine obeys a rule laid down by a French officer in 1824.
Sadi Carnot worked out that you cannot convert heat into work with perfect efficiency. There is a hard thermodynamic ceiling, and no engineering can beat it. A conventional power station converts perhaps a third of the energy in its fuel into electricity, and the other two-thirds leaves as heat — up the cooling towers, into a river, into the sky.
Those enormous cooling towers, the ones that appear in every photograph of a power station and that people mistake for chimneys, are not emitting smoke. They are emitting the majority of the energy in the fuel, as warm water vapour, for nothing.
The insight of cogeneration — combined heat and power — is almost embarrassingly obvious once stated: put the power station near something that needs heat, and give it the heat.
Edison did this at Pearl Street in 1882, selling both electricity and steam to his neighbours. Then the grid centralised, the plants moved far from cities to where the coal and the cooling water were, and the heat became too far from anyone to use. We built the waste into the geography.
Denmark, after the 1973 oil shock, did the opposite, and rebuilt its energy system around using the heat. Danish plants routinely reach total efficiencies far above anything achievable by generating electricity alone, and Danish cities are warmed by it.
What it actually is
Cogeneration means generating electricity and capturing the waste heat for a useful purpose — heating buildings, providing industrial process steam, running absorption chillers for cooling.
The arithmetic is stark. A conventional plant might convert 35–40% of its fuel into electricity and lose the rest. A combined heat and power plant can reach total efficiencies of 70–90%, because the heat is no longer waste, it is a product. You have not made the thermodynamics better — Carnot still wins — you have simply stopped throwing away the part he told you you could not convert.
The two dominant forms:
District heating. Hot water piped from a central plant to the buildings of a town. Common across Scandinavia, Russia, and much of northern Europe; almost unknown in Britain and North America. It requires laying pipes, which is expensive, disruptive, and a genuinely long-lived asset.
Industrial cogeneration. A factory that needs process steam anyway generates its own electricity on the way. This is the single most economically obvious application and it is still substantially undeployed.
And the honest problem, which is now the important one: cogeneration is usually burning something, and it is usually gas. It is dramatically better than burning gas for electricity alone and throwing away the heat. But it is still fossil fuel, and a plant built today has a thirty-year life, and it can become a lock-in. The future of district heating is heat pumps and waste heat, not gas.
The numbers
The waste. A conventional thermal power station converts roughly a third of its fuel energy into electricity. The remaining two-thirds leaves as heat and is discarded. This is not a scandal; it is thermodynamics. Discarding it, however, is a choice.
The gain. Combined heat and power can reach total fuel efficiencies of 70–90% by using that heat. The same fuel does two jobs.
The heat demand nobody counts. Space heating and industrial process heat are an enormous share of global final energy use — comparable to, and in cold countries exceeding, electricity. It is chronically underdiscussed relative to electricity because it is less visible and less glamorous.
Denmark, again. Danish district heating serves a very large proportion of the country’s buildings, built out deliberately after 1973, and is now being progressively converted from fossil fuel to heat pumps, waste heat and renewables — which is exactly the right sequence: build the network, then clean the source.
The lock-in risk. A gas CHP plant is a thirty-year asset. Built today, it may still be burning gas in 2056. That is the strongest argument against building new ones, and it is a good one.
Why it matters
Two-thirds of the energy in the fuel goes up the cooling tower, and we have been watching it happen for a hundred years.
Stand near a power station on a cold morning and look at the plume. That is not pollution, exactly — it is water vapour. But it is also the majority of the energy in the coal or the gas, warming the sky, while a mile away somebody is burning more fuel in a boiler to heat their house.
There is a particular kind of waste that offends people regardless of their politics, and this is it. It is not a moral failing or an environmental abstraction. It is a thing being thrown away that somebody, very nearby, wants.
Every thrifty person who ever lived would recognise the problem immediately. You do not let the oven heat go out the window while you shiver in the next room. You open the door. You put the bread in while it is still hot. Your grandmother did not know the word cogeneration and she would have understood the principle in four seconds, and been irritated by the alternative.
The reason we do not do it is geography — we put the power stations far away from the people, because that made sense for the coal and the cooling water. It made sense in 1930. It has been quietly costing us two-thirds of the fuel ever since, and we simply stopped noticing, the way you stop noticing a sound that has always been there.
What it actually takes
Proximity, which is a planning decision. You cannot move heat far. The plant has to be near the buildings or the factory, which means this is fundamentally a question of urban planning and industrial siting, decided decades before anybody thinks about the energy system.
The pipes. District heating requires a network, and networks are expensive, disruptive to build, and last for fifty years. Nobody wants to dig up a city. But the asset, once built, outlives several generations of the plant feeding it — which is why the sequence matters: build the network, then clean the source.
Not locking in gas. This is the crucial modern caveat. A gas CHP plant built today is a fossil asset until 2056. The right move in most developed countries now is to build the heat network and feed it with heat pumps, industrial waste heat, geothermal or solar thermal — not to build new gas cogeneration and call it green.
Using the waste heat that already exists. Data centres, steel works, chemical plants and sewage works all reject enormous quantities of low-grade heat. Most cities have several sources of free heat within a few kilometres of buildings that are burning gas to keep warm. This is close to free and almost nobody does it.
Where it matters most
Denmark and Scandinavia are the model. District heating built out deliberately after the 1973 oil shock, serving most buildings, and now being converted from fossil fuel to heat pumps and waste heat. They built the network first, which turned out to be the decisive move.
The North German Plain, Poland and Russia have extensive district heating, much of it old, inefficient and coal-fired — and therefore representing an enormous, unglamorous opportunity to decarbonise heat at scale simply by changing what feeds the pipes.
Britain and North America are the great omission. Almost no district heating, individual gas boilers in nearly every building, and power stations far away venting their heat into the sky. It is one of the largest wasted opportunities in the developed world’s energy system.
Industrial corridors everywhere. Any factory that needs process steam should be generating its own electricity on the way, and a great many still do not.
And the new one: data centres. They produce vast quantities of low-grade waste heat and are frequently sited near cities. A handful of places — notably in the Nordics — now pipe it into district heating. Almost everywhere else simply blows it into the air.
How to tell it’s being done well
Is the heat actually being used? A CHP plant with no heat customer is just a power plant with a good story. Ask who takes the heat and how far it travels.
What fuels it, and for how long? A gas CHP plant is a thirty-year fossil asset. Building the heat network is the durable move; feeding it with gas is a temporary one that must be honestly labelled as such.
Could a heat pump do this instead? In most developed contexts now, the honest answer is increasingly yes, and it is getting more yes every year as electricity cleans up.
Is there free waste heat nearby being ignored? Data centres, sewage works, steel plants and chemical works reject enormous quantities of heat within a few kilometres of buildings burning gas. This is close to free and it is mostly not done.
What you can do
Anyone
- Those enormous cooling towers at a power station are not chimneys. They are venting the majority of the energy in the fuel into the sky, as heat, for nothing.
- In most of northern Europe, homes are heated by hot water piped from a central plant. In Britain and North America they are heated by burning gas in every individual building, which is a remarkable and unnecessary difference.
Communities and cities
- Build the heat network. It lasts fifty years and it outlives several generations of whatever plant is feeding it, which means you can clean up the source later.
- Find the waste heat you already have. Data centres, sewage works and factories reject enormous quantities of it within a few kilometres of buildings burning gas to keep warm.
Policymakers
- Build heat networks and feed them with heat pumps, waste heat and geothermal. Do not build new gas cogeneration and call it green: it is a thirty-year fossil asset.
- Require large heat sources to offer their waste heat. Data centres in particular are being built at scale right now and most of the heat is simply discarded.
Industry
- If your process needs steam, you should be generating your own electricity on the way. It is the most economically obvious application of cogeneration and it remains substantially undeployed.
- Your waste heat is somebody else's heating bill. In several countries it can now be sold.
Who is working on this
We are researching which organizations in our directory of 8,493 actively work on this solution, and we only list an organization once we have verified it. That research is ongoing. In the meantime, search the directory yourself:
Questions
What is cogeneration?
Generating electricity and capturing the waste heat for a useful purpose, such as heating buildings or providing industrial process steam. A conventional power station converts about a third of its fuel into electricity and throws the rest away as heat. Combined heat and power uses that heat, reaching total fuel efficiencies of 70 to 90%.
Why is so much energy wasted in a normal power station?
Because of thermodynamics. Carnot showed in 1824 that you cannot convert heat into work with perfect efficiency, and a thermal power station converts roughly a third of its fuel into electricity. The remaining two-thirds leaves as heat. That is not a scandal; it is physics. Throwing it away, however, is a choice.
What are those cooling towers actually emitting?
Water vapour, not smoke. But that plume is carrying the majority of the energy in the fuel, warming the sky, while nearby buildings burn more fuel to heat themselves. It is one of the most visible and least noticed wastes in the entire energy system.
What is district heating?
Hot water piped from a central plant to the buildings of a town. It is standard across Scandinavia, Russia and much of northern Europe, and almost unknown in Britain and North America, where each building burns its own gas instead. It requires an expensive pipe network, which then lasts fifty years.
Is cogeneration still a good idea if it burns gas?
It is far better than burning gas for electricity alone and discarding the heat. But it is still a fossil asset with a thirty-year life, and a gas CHP plant built today may still be burning gas in 2056. The right modern move in most developed countries is to build the heat network and feed it with heat pumps, waste heat and geothermal rather than new gas.
What about data centres?
They produce enormous quantities of low-grade waste heat and are frequently built near cities. A few places, mainly in the Nordics, pipe that heat into district heating networks. Almost everywhere else simply blows it into the air while buildings a kilometre away burn gas to stay warm.
Sources
- Project Drawdown - Deploy Cogeneration (Drawdown Explorer) Framework and classification. Cited, not reproduced.
- IEA - District Heating and Combined Heat and Power
- Danish Energy Agency - district heating and CHP
- IPCC (2022), AR6 Working Group III - Energy Systems and Buildings
- Euroheat & Power
The solution taxonomy follows the framework popularised by Project Drawdown. The analysis above is our own; for their carbon modeling and rankings, visit them directly.