1. Introduction
Climate change is one of the major global challenges of the 21st century. Anthropogenic emissions of greenhouse gases (GHG), primarily in the form of carbon dioxide (CO
2) and methane (CH
4), have been identified as a driving force for global warming. The majority of these emissions are connected to the combustion of fossil fuels such as coal, oil or natural gas, which provided more than 80% of the global primary energy demand in 2023 [
1]. Decarbonization, i.e., the reduction of these GHG emissions by phasing out the use of fossil fuels with their inevitable CO
2 emissions, is considered one of the most important strategies to combat climate change and its consequences. At the same time, readily available energy is a prerequisite for modern society. Given a growing world population and improving standards of living, global energy demand is expected to increase [
2].
Energy is required in different forms for different applications, ranging from electricity in household appliances or data centers, to heat in buildings and industrial manufacturing processes, or mechanical energy for the propulsion of vehicles. For heat, in particular, the demand is substantial, especially in advanced economies such as the European Union or the United States: in Germany, for example, residential heating and industrial process heat account for more than 50% of the country’s final energy demand [
3]. Fossil fuels are still the dominant energy source, both for residential and industrial heating. On a global scale, almost 30% of the world’s GHG emissions are caused by the industrial sector: the vast majority (>80%) are due to the direct use of energy, while the rest is accounted for by product-related so-called process emissions, e.g., CO
2 formation due to calcination processes in the cement industry [
4].
Heat is generally used directly by the consumer in the residential sector, but industrial process heat is an indirect form of heat utilization: heat is required as part of the manufacturing process of a product, which may range from foodstuff and beverages to textiles and paper and to essential materials like steel, glass, ceramics, non-ferrous metals, cement and the many products of the chemical industry. Industrial process heating is a means to an end, and product quality is of paramount importance.
Industrial process heat demand accounts for almost 70% of Germany’s final energy demand in the industrial sector, which translates into more than 20% of the entire country’s final energy demand [
3] (cf.
Figure 1). Process heating is also clearly dominated by fossil fuels, natural gas in particular. Almost 50% of the process heat demand in various industries is supplied by natural gas, while coal contributes another 24%. Renewable energies or electricity play only relatively minor roles today in the context of process heating, as is visualized in the pie chart in
Figure 2. The situation is rather similar in other advanced economies around the world such as the United States [
5], the European Union [
6] and the United Kingdom [
7]. While the industrial sector actually uses a significant amount of electricity (according to [
3], almost 30% of its total energy demand, second only to natural gas in terms of energy carriers), electric power is mostly needed for drive applications, e.g., pumps or compressors, not for process heating.
In absolute terms, industrial process heat accounted for almost 500 TWh in Germany in 2021, which is about the same as its entire annual electricity generation [
3]. These figures underline the economic importance of industrial process heating, but also the scale of the challenge to decarbonize this sector. However, decarbonizing process heating is essential since the global demand for products of all kinds is likely to increase. While energy demand for residential heating is expected to decrease due to efficiency gains, estimates indicate that demand for industrial process heating is likely to rise: the growing world population and the corresponding increasing demand for energy-intensive materials is forecast to more than compensate for technological improvements in process efficiencies [
8].
In the following sections, the peculiarities of industrial process heating in comparison to other end-use sectors such as residential heating or transportation will be discussed, as the specific requirements of industrial process heating have an impact on potential decarbonization options. Then, the main pathways to decarbonize industrial process heat will be briefly summarized before two options, the use of hydrogen and direct electrification, are explored in more detail with their respective benefits and drawbacks, particularly in the context of high-temperature process heat. The considerations of whether electrification or carbon-free combustion with hydrogen best serve the needs of a specific process will be illustrated by a number of examples from energy-intensive industries that already use both electric heating and natural gas combustion to provide high-temperature process heat today. This review will then be concluded by a brief discussion of a scenario analysis from [
6] and what its results mean both for industrial process heating and for the required energy infrastructures.
2. Industrial Process Heat
It is generally accepted that industrial process heating is one of the hard-to-abate sectors in the context of decarbonization [
9,
10]. There are several reasons why: first of all, it is a very diverse and heterogeneous sector. Given the wide variety of products, ranging from paper and textiles to energy-intensive basic materials such as steel, cement and glass, the processes, requirements and technologies are highly specialized and optimized for their specific purposes [
11,
12]. This heterogeneity of industrial process heating is highlighted by
Figure 3, which shows process heat demand in Germany as a function of its respective industries as well as required process temperatures [
11]. While there are many processes in the low- and medium-temperature range, e.g., for drying or steam generation, about 50% of industrial process heat is needed at temperatures of more than 400 °C [
5,
11]. This is particularly relevant in the context of decarbonization, as high-temperature process heat is generally harder to decarbonize than low-temperature heat, especially when high energy densities are required as well [
5,
6,
10]. The focus of this article will therefore be on the decarbonization of high-temperature heat, unless stated otherwise.
Process heating is often also not about heat transfer alone. In many manufacturing processes, additional phenomena like the interaction of the product with the furnace atmosphere and the exposure of the product to well-specified temperature-time profiles are crucial for achieving the desired physical properties in the materials [
13].
Some products are more sensitive than others, and in some instances, even small deviations in process operation can cause unacceptable changes in product quality.
There are also structural differences between industrial process heating and other energy-intensive end-use sectors such as buildings or transportation. Those sectors are characterized by millions of small, independent and use applications that, in total, generate substantial energy demand (cf.
Figure 1). For example, it is estimated that there are more than 200 million gas-fired residential and commercial appliances in the EU alone, mostly for heating and food preparation [
14]. In the industrial sector, it is just the opposite: far fewer installations, which are, however, much larger. Industrial furnaces and kilns can be rated in the double- or triple-digit MW range, with production rates that can reach more than 100 t/h [
15]. They also are frequently part of complex, integrated production chains. Kilns and furnaces are often operated continuously, both for reasons of efficiency and productivity, but also because in many cases, high-temperature systems cannot easily be shut down without lengthy preparation, lest they take damage. Consequently, energy efficiency has always been the focus of energy-intensive industries, as has security of supply. Combining decarbonized power generation from decentralized, intermittent sources like wind and solar with the localized, high, continuous energy demand typical for energy-intensive industrial sites will be a major challenge for the global energy transition [
13]. The continuous operation of industrial furnaces and kilns at high temperatures also limits the use of thermal storage options on-site, which can have significant decarbonization potential in other applications.
There are additional specifics to consider in the industrial sector, ranging from strict environmental requirements, e.g., in terms of air-quality-relevant pollutants such as nitrogen oxides (NOX), to cost considerations and international competition. Energy costs, but also the security of energy supply, are locational factors, and in contrast to private citizens, industrial operators are more likely to shift their production elsewhere if local conditions are no longer tenable.
These specifics of the industrial sector have to be taken into account when considering decarbonization strategies for this sector, particularly in the context of high-temperature process heating. Just like the products themselves, solutions will be highly specialized and tailored to the various industries and their specific demands. But given the huge amount of energy involved and the critical need for continuous security of supply in many cases, the decarbonization of energy-intensive industries will also have consequences for the expansion of necessary low-carbon energy infrastructures.
3. Pathways to Decarbonized Industrial (High-Temperature) Process Heat
In principle, there are several options available to decarbonize industrial process heat, starting with improving efficiency. Efficiency has always been a key optimization criterion for the industrial sector, and especially in the energy-intensive industries where a major part of operational cost is directly related to energy. In the context of minimizing GHG emissions, an improved process efficiency also translates into reduced CO
2 emissions when using fossil fuels. Many technologies have been developed and implemented to reduce energy demand, from waste heat recovery for combustion air preheating and oxy-fuel combustion (i.e., the combustion with pure oxygen instead of air [
16]) to electrification and recycling, e.g., in the glass and aluminum industries. However, the potential for further efficiency increases is limited, and some industries already operate close to theoretical limits with their given technologies [
17]. At the same time, efficiency has to be seen in an economic context. For example, while electric melting furnaces in the glass industry are more efficient than equivalent gas-fired furnaces, natural gas is also currently much cheaper than electricity per unit of energy, so the more efficient solution may not necessarily be the most economic one [
18]. While efficiency is a crucial step toward decarbonization, it alone will be insufficient to completely eliminate GHG emissions from process heating, as energy will still be required to produce steam or melt glass. It is also not the only criterion to consider in the context of industrial process heating: aspects such as product quality and productivity may have priority, despite a possible efficiency penalty.
One option to theoretically almost fully decarbonize industrial process heat is Carbon Capture, Utilization and Storage (CCUS). Greenhouse gases are removed from flue gas and then either used as feedstock (e.g., in the chemical industry) or permanently stored in suitable geological formations. Originally envisioned as a way to decarbonize coal-fired power plants [
19], it is today often considered as an option to deal with industries with large shares of process emissions. One example is the cement industry, where 50–70% of all CO
2 emissions are not energy-related, but caused by calcination processes in the cement [
13,
20]. However, CCUS also faces serious challenges: the costs for implementing CCUS at scale are high, both in terms of investment and additional energy demand. CO
2 quantities are so substantial that dedicated pipelines would be required to convey the CO
2 either to a chemical plant to be used as feedstock or to a permanent storage facility in most cases. Thus, CCUS will likely only be relevant for applications where no alternative decarbonization option can be found.
Electrification (using low-carbon electricity) is widely considered a powerful tool to decarbonize industrial process heat in many industries [
6,
21,
22,
23], with some studies claiming that more than 90% of industrial process heat could potentially be electrified [
22]. Electric heating comprises a range of different technologies such as electric arc furnaces (EAF), resistive or inductive heating and the use of plasma- and microwave-based technologies [
5], to name just a few. Each technology has its specific advantages and drawbacks. For low-temperature heat, technologies such as electric boilers or industrial high-temperature heat pumps [
24] are commercially available.
Some are mature technologies that have long been established in various industries (e.g., EAF or resistance heating in the steel and glass industries, respectively). Others are still relatively early in their development cycle and have yet to be implemented on an industrial scale. Prior to concerns about climate change, electric process heating was only used where it was beneficial for the manufacturing process, as electricity tends to be more expensive than natural gas per unit of energy; efficiency gains due to electrification of high-temperature process heating are often far smaller than in sectors such as mobility [
5,
25]. While electric heating today only plays a minor role in process heating (cf.
Figure 2), its share is expected to increase drastically [
6].
Another pathway to decarbonized process heating is to replace fossil fuels with carbon-neutral or carbon-free fuels. Biogenic fuels such as biomass, biogas and biomethane fall into the first category, while fuels such as hydrogen and ammonia make up the second. The use of biogenic fuels is already common practice, for example in the cement or paper industries [
9], while investigations showed that even sensitive manufacturing processes, e.g., in the glass industry, can be converted to partially using biogas, without loss of product quality or even process efficiency [
26]. Biomethane is a direct drop-in fuel for natural gas, as it has to comply with the same quality specifications for feed-in into public gas grids. The challenge of biogenic fuels is less technological and more one of supply: availability is limited, and other hard-to-abate applications, e.g., the production of synthetic aviation fuels, will likely compete with the available supply.
Carbon-free fuels such as hydrogen (and to some extent, ammonia) are therefore often seen as the main alternative to electrification in the context of decarbonizing industrial process heat, primarily for high-temperature applications across many industries that currently require both high energy densities and temperatures and rely on natural gas, e.g., [
6,
10,
25,
27]. However, only low-carbon hydrogen can serve to actually reduce GHG emissions. The vast majority of hydrogen today is used in the (petro-)chemical industry and produced locally from natural gas or other hydrocarbons without CCS [
28]. It is thus unsuited as a decarbonization option as it has a substantial carbon footprint of its own. Hydrogen, whose production process does not lead to a significant overall reduction of GHG emissions when used, will therefore not be considered further in this article. The effects of fugitive hydrogen on climate change must be considered as well in this context, as hydrogen does have a Global Warming Potential of about 12, when considering a 100-year time span [
29].
In recent years, many nations around the world have published hydrogen strategies, committing to ramp up both local hydrogen production from sustainable sources as well as imports of low-carbon hydrogen.
The list of decarbonization pathways for process heating given here is not exclusive. Other options, e.g., solar or geothermal technologies, exist, but are likely to be niche applications only. Also, hybrid systems which combine the use of different energy carriers to maximize flexibility can be an interesting option for some industries.
4. Hydrogen as a Fuel
Compared to natural gas, hydrogen is a very different fuel. These differences have to be accounted for when considering a fuel switch in any combustion application. Some of the most relevant fuel properties of hydrogen can be found in
Table 1, with methane, the main component of natural gas, for comparison. A more comprehensive and detailed discussion of the differences between hydrogen and methane from the combustion perspective can be found for reference in [
30,
31].
From an operational perspective, the main difference is likely that properties such as net calorific values and minimum oxygen or air requirements change. The volumetric net calorific value of hydrogen is about one third of that of methane, while the minimum oxygen or air requirements of hydrogen are 25% of their respective values for methane. This means that the necessary volume flows of both fuel and oxidizer will change if the two main operational parameters of the combustion process, firing rate and air excess ratio, are kept constant. These changes can have implications particularly for the non-premixed burner systems commonly employed for industrial process heating, as flow patterns and mixing of fuel, oxidizer and recirculated flue gas may change, resulting in different temperature and heat distributions as well as changes in pollutant formation. Interestingly, the mass-based net calorific value of hydrogen is much higher than that of methane due to its extremely low density, but for gaseous fuels, the volumetric calorific value is generally more relevant. The Wobbe Index, a classical criterion for the interchangeability of fuel gases, is only of limited suitability when comparing hydrogen and methane, as these fuels are too chemically different [
30,
31].
The stoichiometric combustion of hydrogen with air (i.e., an idealized combustion process where only as much oxidizer is provided as is theoretically needed to completely consume the fuel) results in a much higher adiabatic combustion temperature than the combustion of methane, the main component of natural gas. While the adiabatic temperature is a theoretical maximum value that can be achieved in a combustion process, this indicates that hydrogen flames tend to have higher peak temperatures than comparable natural gas flames. As a consequence, there is potential for local overheating when using hydrogen in an unsuitable burner, but possibly more importantly, there are implications for the emissions of nitrogen oxides (NOX), a strictly regulated pollutant. With gaseous fuels such as natural gas or hydrogen, NOX formation predominantly follows the thermal formation pathway, where high local temperatures are needed to break up the triple bond of nitrogen molecules and thus initiate thermal NOX formation. Given the higher adiabatic combustion temperature of hydrogen (especially in non-premixed flames), hydrogen combustion with air can potentially lead to much higher NOX emissions. However, as the formation mechanism for NOX is the same both for natural gas and hydrogen, many of the primary measures to reduce NOX formation in technical combustion processes work with both fuels.
Interestingly, the adiabatic combustion temperatures of both methane and hydrogen with oxygen are almost the same, albeit significantly higher than those in combustion with air. As a consequence, NO
X emissions in oxy-fuel combustion processes are less sensitive to local temperature peaks or excess oxygen but are more affected by the availability of nitrogen in the system. Due to the differences in flue gas compositions and particularly the higher water vapor concentrations when burning hydrogen instead of natural gas, conventional units to quantify pollutant emissions such as [ppm] or [mg/m
3] are not suitable to compare emissions adequately. Other metrics, e.g., NO
X emission mass flows referenced to the firing rate in metrics such as [mg/MJ], offer a better and more consistent comparability between emissions from different fuels and are being discussed [
33,
34,
35].
One of the most striking differences between methane and hydrogen can be found in their laminar combustion velocity, which is the measure of a fuel’s reactivity and has a significant impact on flame stabilization. In fact, the extremely high reactivity of hydrogen compared to other fuels is one of the major technological challenges for the use of hydrogen for power generation in gas turbines where burners have to be re-designed in order to avoid phenomena such as flame flash-backs [
35,
36]. The usually non-premixed combustion processes found in industrial process heating, however, are far less sensitive to this effect, because in non-premixed combustion, flame stabilization is less dependent on the balance of flow and combustion velocities. Instead, flame stability is controlled by the mixing of fuel, oxidizer and flue gas in the combustion space [
12]. Thus, many industrial non-premixed burners that were originally designed for natural gas can safely be operated with hydrogen, with no or only minor modifications. Both experimental and simulation-based investigations of industrial combustion processes show that, in terms of heat transfer, both natural gas and hydrogen combustion result in very similar heat flux and temperature distributions, as long as main operational parameters, such as firing rates and air excess ratios, are maintained constant. Examples for investigations in aluminum recycling oxy-fuel furnaces, burner test rigs and regenerative glass melting furnaces can be found in [
32,
37,
38], respectively.
In addition to the combustion-related differences between hydrogen and natural gas, there are also safety- and handling-related aspects to consider. The minimum ignition energy of hydrogen is lower by about one order of magnitude compared to natural gas, while the flammability range of hydrogen is far wider than that of natural gas. The lower flammability limits (which are usually more critical in terms of leaks) are not so far apart, however (4 vol.-% for H
2 in air vs. 5 vol.-% for CH
4 in air). Hydrogen is also a significantly smaller and lighter molecule than methane, making leakage a bigger concern. Flame monitoring by flame ionization is unsuitable, but UV sensors work if calibrated appropriately [
39]. The higher water vapor content due to hydrogen combustion leads to somewhat higher condensation temperatures, which is important for operators of industrial kilns and furnaces who usually avoid condensation in the flue gas ducts due to concerns about corrosion. The change of the heat capacity of flue gas will also have an impact on the performance of waste-heat recovery equipment, if it is not modified as well.
While the fuel switch may require some modifications in current practices and equipment, these challenges do not appear to be unsurmountable: hydrogen and hydrogen-rich fuels have long been used in the chemical and in the steel industries as fuels, for example, and there is a lot of experience among the respective equipment manufacturers.
5. Electrification vs. Hydrogen for Process Heating: Benefits and Drawbacks
Many studies point to direct electric heating and hydrogen (and to some degree hydrogen derivates such as ammonia) as the two main pathways toward decarbonizing industrial process heat. While some, such as Madeddu et al. [
21] and Rehfeldt et al. [
22], claim that almost all industrial process heat demand across all industries and temperature levels could be electrified, others expect a more differentiated actual evolution of process heating, especially in the context of high-temperature processes [
6,
10,
13,
40]. For process temperatures below 400 °C, there is a general consensus that most of the heat demand will be supplied electrically in the future. While electric boilers by themselves do not offer significantly better heating efficiencies than the gas-fired boilers used today (or likely hydrogen-fired boilers in the future), low-carbon hydrogen has to be produced from low-carbon electricity, resulting in lower round-trip efficiencies overall for hydrogen boilers or steam generators. High-temperature heat pumps, on the other hand, can harness ambient energy into the process, resulting in much higher process efficiencies overall, with possible process temperatures of up to about 200 °C [
24,
41].
Table 2 gives a general overview of some of the benefits and drawbacks of electric high-temperature process heating, based on [
18,
22,
40,
42]. Given the heterogeneity of industrial process heating, a generalized quantitative comparison using criteria such as key performance indicators is impossible, even within a given industry or application. However, some examples are given to illustrate matters.
In terms of efficiency, two different aspects need to be considered, the efficiency of the heating process itself and the round-trip efficiency, which, in the context of “green” hydrogen, also includes the production of hydrogen from renewable electricity via electrolysis. Electric heating technologies can often achieve very high heat transfer efficiencies in high-temperature applications [
5,
18], especially when electrically conductive materials like metals or glass are being heated. With other materials, especially granular bulk materials in packed beds, electric heat transfer efficiency can be severely reduced due to the presence of insulating air. The efficiency improvements of electrification in comparison to incumbent technologies are, however, not as drastic as in other sectors such as residential heating or light-duty vehicles [
25]. The heat transfer efficiencies of hydrogen-fired systems are usually comparable to conventional gas-fired furnaces and boiler applications (see, for example [
37,
43]). In regenerative glass melting furnaces, for example, CFD studies (CFD: computational fluid dynamics) show heat transfer efficiencies for both gas- and hydrogen- fired furnaces to be around 66% [
38]. While fully electric melters achieve significantly higher efficiencies (up to 85%, according to [
18]), they are also technologically limited in their production rates (<250 t/d, compared to about 1200 t/d for the biggest gas-fired furnaces). This, and the fact that gas-fired furnaces can accept larger fractions of cullet (both positive factors in terms of process efficiency), make a quantitative comparison difficult, even within a single application such as glass melting.
Additionally, the round-trip efficiency of direct electric process heating has to be compared to that of a hydrogen-based approach. As low-carbon hydrogen is supposed to be produced from water electrolysis and renewable electricity (some nations consider hydrogen production from nuclear power an interesting option), and electrolyzers have their own conversion losses (current electrolyzers tend to operate at efficiencies of 70–75%), the round-trip efficiency of process heating via hydrogen has to be lower than with direct electrification, even without considering losses due to transportation. However, this is based on the assumption that there will always be sufficient low-carbon electricity available to cover the demand. If, as it is for example planned in the German “Kraftwerksstrategie” (Power Station Strategy) [
44], hydrogen is to be burned in gas turbine power plants to compensate for shortfalls in wind and solar power generation, the round-trip efficiency advantage of electric process heating will be reduced or may even disappear completely, depending on the share of electricity generated from hydrogen combustion in those thermal power plants present in the overall power mix [
42]. Given the importance of continuous energy supply for many industries, this is of particular relevance. This consideration also does not take into account alternative routes for low-carbon hydrogen production. It is worth noting that other ways to produce low-carbon hydrogen at scale are being explored, for example, methane pyrolysis [
45,
46].
The efficacy of electric heating for decarbonization is directly tied to the carbon intensity of the power grid. Without low-carbon electricity, electric high-temperature process heating will not reduce GHG emissions much or at all, and may even result in higher overall emissions, as emissions are only shifted from a thermal processing plant to a conventional coal- or gas-fired power plant [
25,
42,
47,
48]. In the same way, any hydrogen used has to come from low-carbon sources to be considered a viable decarbonization option. Hydrogen produced via natural gas and conventional steam reforming without CCS would have a higher CO
2 footprint, as if the natural gas were burned directly.
In terms of investment costs, electric furnaces often have the advantage as they are simpler and require less infrastructure, such as fuel lines, heat recovery systems or, in many cases, flue gas treatment equipment. They usually incur, however, much higher operating costs than conventional natural-gas-fired furnaces used currently, as natural gas is much cheaper than electricity in many places. This makes the use of electric process heating uneconomical in many locations, despite the higher process efficiencies that can often be achieved. How operating costs will develop when using hydrogen remains to be seen, as there are no reliable price estimates available since the entire market is still ramping up. It can be expected, however, that hydrogen will be significantly more expensive than natural gas in terms of EUR/kWh, despite plans to import large quantities of sustainably sourced hydrogen into the EU.
One major technological constraint for many high-temperature electric heating systems is that heating elements are often limited in the energy densities that can be achieved at high temperatures [
5,
10,
18,
22,
40]. This means that electric furnaces tend to be bigger than their gas-fired counterparts, or operators have to accept lower production rates. With hydrogen (as with natural gas), this limitation does not exist. This also has an impact on potential retrofitting of existing plants, an important consideration as industrial equipment tends to be very long-lived. Converting an existing furnace or kiln from natural gas to hydrogen is relatively easy and straightforward and may only involve replacing or modifying certain components, such as burners, fuel-related equipment and refractory linings. First examples of “H2-ready” furnaces have already been commissioned [
49]. Switching to electric heating, on the other hand, will generally require replacing in the entire furnace [
40,
50].
As was mentioned previously, industrial process heating often involves more than just heat transfer. The interaction of the product with the furnace atmosphere can be important in order to achieve desired product qualities (e.g., certain colors in glass or ceramics), as can well-defined temperature-time profiles that a piece of metal has to be exposed to in order to produce the material properties needed. In many furnaces, burners also serve to create convection patterns in the furnace space. With electric furnaces, such patterns may have to be created with fans or blowers that require additional electrical energy. Extensive waste heat recovery is quite common, especially in energy-intensive industries, where flue gas may serve as a heat source for other processes, e.g., product preheating, steam and electric power generation, or may even be used for municipal district heating. Switching from a combustion-based furnace to an electric furnace, which will not produce hot flue gases (at least not in similar quantities), will affect these subsequent processes as well, which needs to be considered. Electric on-site infrastructure in many industrial sites will also have to be significantly upgraded if more process heat is to be generated electrically [
18].
Finally, there is the question of the maturity of new technologies. Electric high-temperature process heating is currently used for certain applications where the advantages for the manufacturing process outweigh the economic penalties of using electricity instead of a fossil fuel such as natural gas or coal. Some mature electric heating technologies exist, such as electric arc furnaces, resistance or induction heating, which can be adapted for new fields of application. Other electric heating technologies such as plasma-based technologies are still in the early stages of development, without widespread experience on an industrial scale. Considering the investments involved in new furnaces or kilns, many industries are hesitant to commit to technologies they do not have experience with [
51].
The use of hydrogen combustion for process heating is formally at a low Technological Readiness Level (TRL), as there are hardly any industrial examples in various industries yet. It is, however, generally expected to be relatively easy to implement in existing manufacturing processes [
40,
42,
50]. Many industries have had extensive experience with fuel changes. Industrial burner and equipment manufacturers have been testing their products with both natural gas/hydrogen blends or pure hydrogen for several years [
52,
53,
54,
55]. It was shown that many burner designs work with hydrogen or can at least easily be adapted. Challenges remain, for example in the context of NO
X emissions and how to fairly compare them with emissions from other fuels [
33,
34]. The impact of hydrogen combustion on product quality or refractory materials is also being investigated in different industries: while there are some effects, no “show stoppers” have been found so far that rule out hydrogen combustion [
52,
56,
57]. Interestingly, in many cases it is not so much the hydrogen itself which causes changes in the product or refractory. Instead, the observed changes are often induced by the higher water vapor content in the furnace atmosphere [
43,
56,
57]. Many of these findings have already been corroborated by short-term tests on the full industrial scale (see, e.g., [
52,
58]). Häggström et al. [
50] give the use of hydrogen in the context of a steel reheating furnace a technological readiness level of 8–9. They also emphasize that even for a single application such as steel reheating, there is no clear optimum technology, be it fuel- or electricity-based. Instead, on-site opportunities and infrastructures must be considered.
The decarbonization of industrial process heating must not be seen as a challenge for the industrial sector alone. Given the amounts of energy involved and the economic importance of the sector, industrial decarbonization strategies have repercussions for the development of sustainable energy infrastructures. In most countries, power grids were never designed to continuously supply large quantities of electric power to industrial end-users. This has traditionally been the task of natural gas grids: the German natural gas grid, for example, transports more than double the amount of energy than the German power grid, and, as stated previously, the demand for industrial process heat alone in the country is about the same as the entire annual electric power generation [
3]. Transitioning existing energy infrastructures into decarbonized systems primarily powered by intermittent energy sources like wind and solar will require huge expansions of transmission and storage capacities. Synthetic fuels such as hydrogen, despite their inherent inefficiencies, are expected to play an important role, not only to ensure security of supply and power grid stability [
13,
59], but also to enable the global trade of renewable energies. Plans are currently underway to convert parts of the European natural gas grid to transport hydrogen [
60]. Hydrogen is also particularly relevant for global regions which will continue to rely on energy imports, such as the EU, but also China, Japan and South Korea [
28,
61].
6. Industrial Examples
The complexities of mapping a way toward decarbonized industrial process heat can best be illustrated by a few examples from different energy-intensive basic materials industries. It is worth pointing out that these examples can just give an insight into the task and cannot cover all aspects or even industries. For some industries, such as the cement and lime industries, the pathway to decarbonization will likely look completely different, given their predominant current use of solid fuels and their large share of process emissions. Biomass, refuse-derived fuels and CCS will probably play a much bigger role there than in other industries. In all cases presented here, electric heating is already being used to some degree while these industries also actively investigate hydrogen as a potential pathway for their decarbonization strategies.
The first example is the glass industry. Most glass melting furnaces today are fired with natural gas. The German glass industry, for example, covers about 80% of its energy demand with natural gas [
43]. Smaller electric furnaces do exist, primarily for specialty glasses for optical purposes or high-priced products such as perfume flacons. Electric furnaces are more efficient than gas-fired furnaces, but limited in production rates [
18,
62], as there is a technological limit on how much glass can be molten electrically, at least with current furnace designs. This limit is at about 250 t/d, while the biggest gas-fired float glass furnaces for flat glass products such as windowpanes can produce up to 1200 t/d, with firing rates up to 100 MW. Process temperatures are very high, up to 1600 °C and more. Some glasses, especially those with darker colorations, are difficult to produce electrically, and the use of large fractions of recycled glass, the so-called cullet, can also be challenging in electric furnaces [
62]. The latter is especially relevant as the use of cullet significantly reduces the amount of energy needed to melt glass, compared to glass being molten purely from raw materials. Thus, using recycled glass is an effective way to improve process efficiency. The glass industry aims both at developing bigger electrical furnaces and at hybrid solutions which use both electricity and hydrogen. Auxiliary electric heating, so-called “boosting”, is not new, and many modern gas-fired furnaces can increase their energy input by up to around 10% using these booster electrodes. Large furnaces, especially in the flat glass industry, will likely be hybrid systems with electricity providing somewhere between 50–80% of required energy, with the rest probably being supplied by hydrogen–oxygen combustion. Many glassmakers and their respective associations investigate the use of hydrogen (cf. [
43,
58]). In addition to energy-related GHG emissions, the glass industry also has to consider process emissions, as about 25–30% of a gas-fired melting furnace’s CO
2 emissions are caused by chemical reactions in the glass melt.
The aluminum recycling industry is another industry which today uses both electric- and gas-based process heating. Aluminum recycling is by itself a potent decarbonization measure as recycled aluminum (so-called “secondary aluminum”) requires up to 95% less energy than aluminum produced via the primary aluminum route [
63]. Electric aluminum recycling furnaces tend to be smaller than their gas-fired counterparts (max. tonnage 16 t vs. 120 t [
64]). There are other constraints to electric aluminum recycling as well: aluminum scrap is often contaminated with organic residues from glue, paint or lubricants which leads to material loss and reduced product quality in electric furnaces, while these organics are burned off in gas-fired furnaces. In fact, the combustion of these organic residues also serves to provide heat to the furnace, reducing fuel consumption. The aluminum industry is investigating the use of hydrogen to decarbonize the secondary aluminum route, particularly in combination with oxy-fuel combustion [
52,
65].
The final example is the electric arc furnace used for steel scrap recycling. These furnaces are often highlighted in the context of electric high-temperature process heat where the scrap is molten using electric arcs. However, modern EAFs are actually hybrid systems where 30–40% of the energy input is supplied by fuels [
66]. Supersonic natural gas–oxygen burners are used primarily in the early stages of the production cycle when the scrap is still solid and electric heat transfer is poor. Once the scrap is molten, the majority of the energy input is provided by the electric arcs since electric heat transfer is much better in the liquid metal. The burners then serve to inject oxygen into the melt for metallurgical reasons. This hybrid operation allows for better overall efficiency and reduces the consumption of electrodes, which can be a significant cost factor. The positioning of burners can also be used to homogenize temperature distributions in the melt by compensating for “cold spots”. Investigations are currently underway on how to adapt these burner systems for the use of hydrogen [
55,
67]. Additionally, the steel industry is also actively investigating the use of hydrogen as a fuel in other parts of the manufacturing process, e.g., in reheating furnaces, where some burners and furnaces have already been proven in an industrial setting [
49].
These examples serve to underline the complexities involved in the decarbonization of industrial high-temperature process heat. The situation is similar in many other industries, such as the ceramics industry, while industries like the cement and lime industries face additional challenges such as significant process-related emissions. They also highlight that in the context of industrial process heating, it is usually insufficient to just consider energy quantities and efficiencies, but that it is necessary to consider these complex processes as a whole.
7. The Future of Industrial Process Heating
Given the predominance of fossil fuels and in particular natural gas in industrial process heating all around the world, the decarbonization of industrial process heating is one of the major challenges of the global energy transition. Part of the challenge lies in the inherent complexities of the often highly integrated manufacturing processes themselves, but other factors have to be considered as well, e.g., the huge, often continuous energy demand, the reliance on (decarbonized) energy infrastructures and the generally long lifespans of industrial equipment. The impact of the large-scale electrification of industrial process heat on power grids must not be underestimated, especially since the demand for secure, uninterrupted power supply from other end-users, such as data centers, is expected to increase drastically [
68,
69,
70]. Global process heat demand is also likely to increase, due to a growing world population and increasing quality of life in many parts of the world, despite efficiency gains in some industries [
8].
Considering the need for a drastic reduction of greenhouse gas emissions, industrial process heating has to change.
Figure 4 and
Figure 5 give an estimate for sources for industrial process heat in the European Union (and the UK) till 2050 [
6]. The study differentiates between low-temperature and high-temperature heat, with process temperatures above 300 °C considered to be high-temperature in this study. It is estimated that roughly 50% of an industrialized nation’s process heat demand is required at high temperatures [
5,
9,
11]. The forecasts in [
6] are based on three different policy scenarios: the “Elec+” scenario focuses on the electrification of process heat as the main decarbonization pathway, while the “H2+” scenario favors the use of hydrogen as the main pathway. Finally, the “Elec+_VC” scenario is similar to “Elec+”, but additionally assumes that many energy-intensive intermediate products, such as sponge iron or certain base chemicals, will be imported into the EU in the future.
Looking at low-temperature industrial process heat in
Figure 4 first, the expected share of energy carriers emerges, with electricity being the main source of process heat by 2050, although biomass and ambient heat also contribute. Hydrogen only plays a role in the H2+ scenario, which specifically favors hydrogen as a decarbonization option, though it is the biggest contributor to low-temperature process heat in this specific scenario. More interesting, however, is
Figure 5, which gives the study’s estimated evolution for high-temperature process heat, again based on the different policy scenarios.
For high-temperature process heat the scenarios do not seem to matter that much. In each scenario, the respective estimated contributions of both electricity and hydrogen to process heat are roughly similar by 2050. This underlines the conclusion that a full electrification of process heating in energy-intensive industries appears unlikely, particularly in the industries where both high energy densities and temperatures are required. As one would expect, the overall high-temperature process heat demand decreases in the Elec+_VC scenario.
These findings match analyses from other sources, which also point to the important role that fuels, and hence combustion, will play in a future net-zero energy system, not only in the industrial sector [
10,
40,
42], but for power generation, long-term energy storage and transportation as well [
13,
59,
71]. Lieuwen et al. [
59] emphasize that in a decarbonized energy system, the role of fuels changes: they are no longer a source of primary energy, but instead, synthetic fuels such as hydrogen or ammonia serve as energy vectors as well as long-term, large-scale storage media, complementing the bigger role of electricity in a net-zero energy system, despite the inherent inefficiencies. A part of this change will likely also happen in the context of high-temperature process heat.
Also, hybrid systems that can make use of a variety of energy carriers can be expected to play a much bigger role in the future and may even serve to stabilize power grids via industrial demand-side management, if the specific manufacturing process allows for it [
72]. Some industries, e.g., primary aluminum production, are better suited for this aspect of a sector coupling approach, while processes in the glass industry tend to offer far less flexibility in this regard [
73].
8. Conclusions
Most anthropogenic greenhouse gas emissions are directly related to the use of energy, across all sectors of human activity. In the context of industrial manufacturing processes, process heat accounts for roughly two thirds of the entire sector’s energy demand, and thus, for the majority of industrial GHG emissions since process heating is dominated by natural gas, and to a lesser degree, coal and oil. At the same time, the products and materials provided by the industrial sector are essential for modern society, and global industrial energy demand is likely to increase in the coming decades. Thus, decarbonizing industrial process heat is one of the bigger challenges of the global energy transition, especially since this sector is characterized by a large degree of heterogeneity and has some distinct technological, economic and structural features, compared to other energy-intensive end-use sectors such as buildings and transportation. High-temperature process heat in particular is generally acknowledged to be one of the hard-to-abate applications in terms of decarbonization. Several options exist, ranging from further increasing process efficiency to CCUS strategies and the use of biogenic fuels such as biomass and biogas, but the two most prominent decarbonization options in the context of high-temperature process heat are direct electrification and the use of hydrogen. In both cases, the energy carrier has to be produced in a low-carbon way in order to produce a benefit in terms of GHG emissions reduction, which is in itself a significant scaling challenge in both cases. Low-carbon electricity is not yet available in the quantities required, and the global production of low-carbon hydrogen is also far from sufficient to cover the demand, though both are expected to increase in the near future.
As decarbonization options for high-temperature process heat, both energy carriers have their specific advantages and disadvantages. Electric heating is usually more efficient and offers better round-trip efficiencies compared to “green” hydrogen, while existing manufacturing processes can much more easily be retrofitted for the use of hydrogen than for electric heating. Electric furnaces may be simpler and require less capital investment, but can be limited in production rates, as energy densities at high temperatures are a constraint for many electric heating technologies. Recent studies indicate that while heat at temperatures below 300–400 °C will likely be mostly decarbonized by direct electrification or the use of biomass, both electricity and hydrogen are expected to contribute to the decarbonization of high-temperature process heat with roughly equal shares. The choice of the best energy carrier will be highly specific and has to weigh the respective advantages and drawbacks of both hydrogen and electricity for a given application. In some cases, alternative approaches, such as biomass combustion in combination with CCS, may be the best option, e.g., in the cement and lime industries. While today natural gas and coal are the most relevant energy carriers by far in the context of industrial process heating, energy supply for manufacturing processes will become more diverse, just like the sector itself.
For both low-carbon electricity and hydrogen, the impact of the decarbonization of the industrial sector on energy infrastructures will be profound, given the huge energy demand and the high requirements for security of supply in high-temperature industries. At the same time, decarbonized energy infrastructures are crucial for the supply of energy-intensive industries across the globe.