Energy use in Industry

About 1/3 of global GHG emissions come from making material things.  The broad term for this is “industry.”  It includes not just what we do inside factories, but also mining, construction, and food processing – almost everything other than growing food and producing energy.  Some of these emissions don’t come from burning fossil fuels; instead, they’re byproducts of the chemical process by which we make things (especially cement and fertilizer).  On the pie chart, those are the emissions in the grey sector called just “industry.”  We discuss those on a separate industry page.  But most emissions in industry are produced by using fossil fuels for energy.  That’s what this page is about.

Most of the energy used in industry comes from burning fuels directly to produce heat where it is needed.1 About ¼ of emissions from industry are “indirect” – they come from burning fossil fuels in power plants to generate electricity that industry uses. Eliminating these “indirect” emissions is just the relatively easy task of cleaning up electricity.    About a third of this heat is for processes, like drying food, that require low or medium temperatures (less than 400° C).  These are applications that we can electrify with current technology. They are easy decarbonization wins.  

Credit: McKinsey and Co.

Another 1/3 of industrial energy is used in processes, like cement and steel production, that not only require very high temperatures (above 1000° C), but also require a high, steady, flow of heat (“heat flux”).  We do not have proven electrification technologies for these demanding applications – although several start-ups are hard at work developing them.  And another 16% is used in processes between 400° C and 1000° C – for instance, distilling petrochemicals in oil refineries.  We have technologies to electrify these processes when we’re designing and building new plants from scratch (as the graph above indicates); but it’s often not easy to swap them into existing plants.  So, if we want rapidly to decarbonize high-temperature and very-high-temperature processes, we need to look at several, possible pathways.  Here are some of the main alternatives.

Carbon Capture, Utilization, and Storage (CCUS)

The path to decarbonizing that involves least change to existing systems is to keep burning the same fossil fuels in the same equipment, but to separate out and capture the carbon dioxide in the exhaust stream before it enters the atmosphere. The captured carbon dioxide can then be transported via pipeline to a location with a suitable geological formation underground, and then pumped into that formation, where it may be stored (or “sequestered”) for thousands of years.  This is Carbon Capture and Storage (CCS).2“CCS” refers specifically to capturing carbon dioxide from a point source, like an exhaust stream, where it is already concentrated. By contrast, capturing carbon dioxide from ambient air, where it is very dilute, is called “Carbon Dioxide Removal” (CDR); and doing this by engineered, mechanical means is called “Direct Air Capture” (DAC). We discuss DAC and other non-mechanical forms of CDR in the section on Building Carbon Sinks. If we add the possibility that some of the carbon dioxide might be used in other industrial processes, we have Carbon Capture, Utilization and Storage (CCUS).  The fossil fuel industry has promoted CCUS for years, for electricity generation as well as for industry. 

In practice, CCUS has not worked well – or at all – in many of the uses for which it has been promoted.  This is for two, connected reasons: 

  1. In many exhaust streams, CO2 is dilute (often less than 5%).  The more dilute it is, the more difficult it is to separate CO2 from other gasses in the exhaust stream, and the more energy is required for the electrochemical process that does this.
  2. When CCUS is simply “bolted on” to an existing facility, it adds extra complexity, extra energy demand, and extra cost, without producing anything of significant value.  (There is a market for CO2, but it is much smaller than possible supply, and so CO2 is inexpensive.  Most captured carbon must simply be stored underground, which is a further cost.)  In cases like power generation, where coal-fired or gas-fired power plants are already struggling to compete with renewables on price, adding CCUS makes it even harder for them to do so.

Thus several CCUS demonstration projects (especially those promising “clean coal”) have proved uneconomical, or have failed outright.

There are some exhaust streams where CO2 is more concentrated, and in these cases, CCUS is already being implemented successfully on a small scale.3In the US, most captured carbon dioxide is currently used for “Enhanced Oil Recovery” (EOR) – that is, to help extract more oil from existing oil wells. Proponents argue that this use still has a climate benefit, but this is contestable. In fact, in processes where CO2 is separated prior to combustion (or where nothing is combusted), the exhaust stream can consist of pure CO2, so that no separation is needed, and the only added cost from CCUS comes from compressing the CO2, transporting it, and pumping it underground.  (This is true in Allam Cycle gas power plants, which we discuss in the page on cleaning up electricity.) In other processes, such as cement manufacture, carbon dioxide is not pure, but it may be concentrated enough to make CCUS feasible and cost-effective.  This technology has yet to be proven on a commercial scale, but governments and companies are investing in it, and many demonstration projects are now being built around the world.

Two downsides of CCUS are worth noting:

  1. Except in cases where the exhaust stream consists in pure CO2, carbon capture technologies will reduce emissions but not eliminate them, because it will capture only a percentage of the CO2.
  2. When methane is extracted, compressed, transported, and used, leaks at each step in the process create fugitive emissions. Uncombusted methane is a very powerful greenhouse gas (84 times more powerful than carbon dioxide over a 20-year period).  This means that even a perfectly functioning CCUS facility can still be responsible for upstream emissions from the methane that supplies it.


Hydrogen emits no CO2 when burned, and it burns at over 2000° C – hot enough for the most energy-intensive industrial processes.  It can be transported cheaply through pipes, and can deliver a high flow of heat energy.  In many industrial applications, hydrogen can be substituted for methane as a fuel source without extensive modification of existing equipment.  Right now, most hydrogen is produced in ways that emit large quantities of CO2; but as we discuss in the separate page on non-energy emissions from industry, it can be produced without emitting any GHGs at all.

In addition to providing heat, hydrogen can play a crucial role in making steel that is currently played mostly by coal.  When iron is mined, much of it is in the form of iron oxide – otherwise known as rust.  To make steel, the oxygen atom need to be stripped from the iron oxide in a reduction reaction. This is usually done by burning coke – a fuel made from coal – to heat the iron ore.  When the coke is burned, it not only releases heat, it releases its carbon atoms – and these then bond with the oxygen atoms from the iron ore, producing CO2, and leaving the iron pure.  Hydrogen can be used to “reduce” iron ore in a similar way: when hydrogen is burned, oxygen atoms from iron oxide will bond with hydrogen atoms, producing water vapor and leaving pure iron behind.   

In 2020, Swedish steelmaker SSAB opened Hybrit, a first-in-the-world pilot plant using hydrogen to make “green steel.” 

Watch their videos on the process here

Read this Canary Media explainer

Most iron is currently made in blast furnaces.  These cannot be converted to use hydrogen.  In order to decarbonize the sector, these furnaces, and the enormous plants that are built around them, will need to be replaced entirely.  However, about 7% of iron is currently made using “shaft furnaces,” which can be converted.  This could provide a relatively quick way to jump-start the process of decarbonizing steel.


There are many possible technologies with which electricity can produce high temperature heat, but none of them has yet been demonstrated at commercial scale.  The main reason for this is economic rather than technological: burning fossil fuels on site has long been the cheapest way to produce high temperatures, and so before climate became an urgent concern, there was no reason for businesses to develop these technologies.  That is now changing.  Here are two of the most promising technologies that companies are currently developing.

1.Direct Electrification of Steelmaking through Molten Oxide Electrolysis

Boston Metals is working to apply a technology already used in aluminum production to steel.  In Molten Oxide Electrolysis, an electric current is passed through iron ore.  In a single step, this simultaneously heats the ore to 1600C and “reduces” it, by stripping the oxygen atoms, leaving pure iron.   Boston Metals aims to have a commercial-scale demonstration plant operating by 2024-25.

Read this MIT Technology Review piece on Boston Metals

2. Indirect Electrification.

Directly electrifying high-temperature, high-energy industrial processes requires a high flow rate of enormous quantities of electricity.  This can be very expensive, because it can require large upgrades to the electricity grid, and because utilities charge industrial customers extremely high rates at times of peak demand.  Industries that are sited near wind or solar resources can often build their own, renewable generation capacity, which will supply electricity at a much lower cost than they get from the grid; but industries need firm, consistent energy, and the power that renewables supply is intermittent. 

Rondo Energy and Antora Energy are both developing systems that solve this problem.  Both of their systems use resistive electric heating (like a toaster) to heat bricks made of very low-cost materials, like graphite, to very high temperatures (2000C, in Antora’s case).  In both cases, the bricks can be heated with cheap, intermittent renewable energy, when it is available in excess (e.g., in the middle of the day, when solar panels produce more electricity than the grid needs).  The bricks are insulated and so act as a thermal battery, able to release heat at a steady rate for industrial processes whenever it is needed.  (In the case of Antora, the heat can also be converted back to electricity by means of thermophotovoltaic cells.)  Both companies are developing pilot projects.

Listen to this Watt it Takes podcast interview with Andrew Ponac, co-founder of Antora Energy
⇒Read this Canary Media profile of Rondo Energy

Small Modular Reactors (SMRs)

A new generation of advanced nuclear reactor designs are small enough to be sited on the grounds of a factory or refinery.  These small modular reactors (SMRs) generate heat that can be used directly in high-temperature industrial processes, in addition to being used to generate electricity. 

Companies in the US and around the world are developing a wide range of SMR technologies.  So far, only one has been fully approved by the US Nuclear Regulatory Commission, but others are far along in the process.  X-energy and Dow have announced plans to deploy an X-energy SMR to provide heat and power at one of Dow’s chemical facilities by around 2030.

Reducing the need for industrial energy

So far, we’ve been exploring ways in which industry can produce the enormous quantity of energy that it currently uses without also producing GHG emissions.  Another (complementary) decarbonization strategy is available: we can find ways for industry to use less energy.

To understand how this might be possible, we need a better understanding of why industry uses so much energy now.  In this great podcast interview, Dr. Rebecca Dell notes that just four kinds of product – cement, steel, plastic, and fertilizer – are together responsible two thirds of all GHGs from industry.4This figure includes not only emissions from energy use, but also “process” emissions that are listed in the grey sector labeled “industry” on the pie chart.

[A]ll of these industries are a variation on the following theme: you dig something out of the ground and the first thing you do with it transforms a raw material into a useful molecule; everything that’s downstream of that in your supply chain is arranging your useful molecules in different combinations and sizes and ratios. But all of that rearranging takes a lot less energy and emits a lot less greenhouse gas than making the useful molecule in the first place.

Dell gives an illuminating example: the German chemical manufacturer BASF makes about 100,000 different products.  But 80% of their GHG emissions come from making just twenty products.  These are basic chemicals, which can then be rearranged with a little extra energy to make the vast variety of distinct chemicals that the world uses.

These observations point to a number of strategies for reducing energy use in industry.

Recycling as an energy strategy
In theory, recycling should not only allow us to extract fewer resources from the earth and produce less waste; it should also allow us to use less energy, because the materials that we recycle have already undergone the initial, energy-intensive transformation from raw materials into useful molecules.  It does work this way when we recycle metals like steel and aluminum.  But we do not recycle concrete at all.  And even when we do recycle plastics, it does not work this way. Because we mix many different kinds of plastics together into a single stream for recycling, melting down the mix downcycles them into a very low-quality plastic, without the desirable qualities that we get by turning raw materials into new plastics. 

We could solve this problem for plastics with public policy.  For instance, as Dr. Dell suggests, we could require manufacturers to choose among a few types of plastics (rather than the huge variety, often mixed with other materials, that they use now), so that their products could then be easily segregated and recycled in separate streams.  Such pure streams of single plastics would be repeatedly recyclable into high-quality products.  

Eventually, we might also be able to solve some recycling problems with technology. For instance, AMP Robotics is building robots that use AI to sort recyclable materials in a mixed waste stream. Epoch Biodesign is working to produce enzymes that can break plastics down into pure component chemicals, which can then be used to make plastics or other products. And Carbon Crusher even recycles the asphalt or concrete in roads, by crushing it on the spot, and then binding it with plant lignen to create a new surface.  

Non-thermal separations
10 to 15 percent of the world’s total energy use – roughly half the energy used in industry – is used to produce heat in order to separate chemicals from one another through distillation, drying, or evaporation. For instance, refineries heat petroleum in 50-meter-tall columns in order to distill the complex mix of molecules that make it up into individual products like diesel, kerosene, and naphtha (which is used to make plastics). These components boil at different temperatures, so as the petroleum is heated, the lighter molecules with lower boiling points are first to rise and exit the column as gasses, with progressively heavier molecules following as the temperature increases.   

Dr. Shreya Dave, the co-founder of the startup Via Separations, compares heat-based separation processes to separating pasta from water in a pot by boiling all the water away. The energy-efficient alternative is to use a strainer. That is what Via Separations is trying to do: they are developing membranes that can separate chemicals without high temperatures. Their initial membranes are for food processing; but they aim to extend their work to other industrial processes, including petroleum refining, as well.  If it could be made to work, membrane-based separation could use 90% less energy than heat-based separation.

Alternative chemical feedstocks
The vast majority of chemicals, including plastics, are made from petroleum, which has to be separated into component molecules through the distillation process we’ve discussed.  However, the same chemicals can be produced from feedstocks other than petroleum, in ways that do not require an energy-intensive separation process.  

One approach is to use plant biomass (often sugar) as a feedstock.  The tools of synthetic biology can be used to turn this feedstock into useful chemicals in one of two ways:

  1. Synthetic biology can be used to engineer microorganisms that can produce the desired chemical from the feedstock.  The microorganisms are fermented in large bioreactors, with the feedstock as “food,” and the desired chemicals as their output. This is a variation on the familiar fermentation process by which bacteria turn sugars into alcohol when beer or wine is made – except that the microorganisms are specially engineered to produce the molecule desired.  Start-ups such as Lygos are trying this approach.
  2. Synthetic biology can be used to engineer microorganisms to produce desirable enzymes from the feedstock.  These enzymes can then be used, apart from the microorganisms that produced them, to catalyze chemical reactions turning a feedstock into the desired chemical.  Because this second reaction does not require living microorganisms, it can be much simpler and more efficient.  Solugen is the first company successfully to use this approach at commercial scale.  Epoch Biodesign is at an earlier stage in developing a similar approach.

This approach is promising, but it is limited by the availability of biomass as a feedstock.  There is not enough biomass to serve all the purposes for which different industries would like to use it, even if one takes into account waste products such as corn stover, which can be used for “second-generation” biofuels and as a vehicle for carbon dioxide removal.  Using feedstocks such as corn syrup (which is what Solugen uses) is much worse, because these compete directly with food for scarce cropland, driving up food prices and increasing pressure to clear rainforests.

A second approach avoids this difficulty by using carbon dioxide as its main feedstock.  This can be captured from a point-source, such as a factory exhaust stream, with CCUS; or it could be captured directly from ambient air through Direct Air Capture (DAC).  Using CO2 as a feedstock may work with some synthetic biology approaches.  (This article says that Solugen could use captured carbon dioxide as a feedstock in the future.)  And other companies, such as Twelve, are working to use electrochemistry to turn captured carbon dioxide directly into useful chemicals, using only electricity.  This possibility is exciting; but the technology is still in early stages, and has not yet proven that it can work economically at commercial scale.

“Ok, all this is pretty interesting. But it seems like this is all work that needs to be done by engineers working in industry, or maybe biologists working to make useful microbes. Is there other work that needs doing?”

Yes!  The solutions we’ve discussed so far are all ways of meeting our demand for materials without emitting GHGs.  But there’s another (complementary) way to reduce GHG emissions from industry: we can reduce our demand for materials.  That’s something that people outside of industry can help with.

  • People can work to pass public policy, at national, state, or local levels, to reduce demand for materials.  Bans on single-use plastics are a starting point.  So are right-to-repair laws, which require manufacturers to make their products easily repairable – so that you don’t need to throw your cell phone away, for instance, just because the battery has worn out or a single part has broken.
  • People can also work to pass public policies that help to implement the other solutions we described above.  Buy Clean laws, for instance, require governments (local, state, or national) to use only low-carbon steel and concrete.  This gives manufacturers a strong incentive to invest in cleaning up their industrial processes. And we can work to make Rebecca Dell’s idea real, by requiring manufacturers to choose from among just a few types of plastic which can be easily separated, so that plastic of each type will be infinitely recyclable. 
  • People can build businesses that meet some need while using fewer materials.  For instance, Grove Collaborative has created an entire line of cleaning products which do not require plastic packaging.

Listen to this great My Climate Journey podcast interview with Stuart Landesberg, co-founder of Grove Collaborative.

  • People can build businesses that create one part of a circular economy.  For instance, Dispatch Goods provides reusable metal and silicon take-out containers to restaurants, then collects them from homes, washes and returns them for reuse. And companies like Patagonia collect, repair, and resell “worn wear” clothing. 

Listen to this great My Climate Journey podcast interview with Lindsey Hoell and Maia Tekle, co-founders of Dispatch Goods

  • Take your broken stuff to a Repair Café – or volunteer at one! 

⇒For an inspiring book on the ideas behind the repair café movement, read Sandra Goldmark’s book, Fixation: How to Have Stuff Without Breaking the Planet.

Even apart from climate considerations, there are many good reasons to reduce our demand for material things.  In 2020, for the first time, the weight of human-made things exceeded the weight of all living things (animals, plants, fungi, bacteria…) on earth. The world’s plastics alone now weigh twice as much as all terrestrial and marine animals.  We need to solve the climate crisis, but we also need to find ways to make our footprint lighter on the earth.  That’s work for everyone.

Further Resources