1 Introduction
As the world's largest energy consumer, China is short of high-quality energy such as oil and natural gas, and has a high degree of dependence on foreign countries, which seriously threatens China's energy security. In 2018, China imported 460 million tons of crude oil, and its external dependence was as high as 71%. The import volume of natural gas is 120 billion cubic meters, and the external dependence is as high as 43%
[1]. Coal resources are abundant, and its proven reserves rank second in the world, but its extensive development and heavy pollution are not conducive to sustainable development in the future. Faced with the increasingly serious problem of carbon emissions, China put forward the goal of "carbon peak" and "carbon neutralization" in 2020
[2]. Hydrogen energy is a kind of clean energy with zero carbon emission and various application forms. It is an ideal energy storage medium to realize the low-carbon transformation of energy system, which can accelerate the deep decarbonization of power, industry, transportation, construction and other fields. The development and utilization of hydrogen energy is an important way to promote the transformation of China's energy structure, ensure national energy security, and achieve the goal of "carbon peak" and "carbon neutrality"
[3,4]. In order to promote the high-quality development of China's hydrogen energy industry, Sinopec invited experts to hold the "Hydrogen Energy Development Strategy Seminar" in July 2020. Gan Yong, academician of the Chinese Academy of Engineering, pointed out at the meeting that in recent years, China's hydrogen energy industry has developed rapidly, basically completed the layout of the whole hydrogen energy industry chain, and formed six hydrogen energy industrial clusters in Beijing, Tianjin and Hebei, East China, South China, Central China, North China and Northeast China
[5]. At present, the technological gap between China and foreign countries in the production, storage, transportation, filling and utilization of hydrogen energy is narrowing day by day, and some areas have been surpassed. Especially in the fields of clean coal hydrogen production, by-product hydrogen purification, alkaline cell electrolysis hydrogen production, solid hydrogen storage and other sub-areas, China has reached the international leading level, and has initially formed a complete technical chain from upstream hydrogen production to downstream hydrogen use (as shown in Figure 1), and the proportion of hydrogen energy in China's energy system is increasing year by year
[6].
To use hydrogen on a large scale, we must first solve the problem of large-scale hydrogen production. At present, China has become the world's largest hydrogen producer, with a preliminary assessment of industrial hydrogen production capacity exceeding 20 million tons
[6]. There are many ways to prepare hydrogen (as shown in Figure 1).
The main hydrogen production routes in China are divided into four categories: (1) hydrogen production based on reforming of fossil fuels such as coal and natural gas
[7,8]; (2) Industrial by-product hydrogen represented by coke oven gas, chlor-alkali tail gas and propane dehydrogenation; (3) Hydrogen production by water electrolysis based on new clean energy; (4) New hydrogen production technologies based on clean energy, such as solar water photolysis, biomass hydrogen production and thermochemical cycle hydrogen production.
Academician Guo Liejin of Xi'an Jiaotong University believes that China's coal-based energy resources endowment is difficult to improve in the short term, and the realistic energy optimization development strategy based on the combination of short-term, medium-term and long-term is bound to change the form of coal utilization and realize low-carbon coal utilization.At the same time, we will vigorously develop renewable energy, improve the energy consumption structure, promote the development of hydrogen-electricity exchange and complementary secondary energy subjects, and ultimately achieve the development goal of a completely clean, low-carbon, efficient and safe energy system from primary energy to terminal energy
[9]. Peng Suping, academician of China University of Mining and Technology, believes that fossil fuel reforming + CCS (carbon capture and storage) technology is an effective way to achieve clean coal hydrogen production
[10]. Hydrogen production from reforming of fossil energy has the largest scale and the lowest cost, but carbon emissions are serious. If it is combined with CCS technology, clean hydrogen production from fossil energy can be realized, but the cost will increase significantly
[6]. Yi Baolian, academician of Dalian Institute of Chemical Physics, Chinese Academy of Sciences, believes that industrial by-product hydrogen and hydrogen production by electrolysis of water are two effective ways of clean hydrogen production, and points out that hydrogen production by electrolysis of water can solve the problem of renewable energy volatility. With the increasing proportion of renewable energy in the energy system, hydrogen production from electrolysis of water will enter a development climax
[11]. The annual output of industrial by-product hydrogen in China is huge, and the utilization rate of resources is low at present. Data show that the annual by-product hydrogen venting rate of chlor-alkali enterprises in China is as high as 30%, of which 250000 tons of industrial by-product hydrogen were vented in 2017. If the industrial by-product hydrogen can be fully recovered, purified and utilized, it is expected to become one of the main sources of hydrogen in China in recent decades
[12].
From the perspective of sustainable development, renewable energy, as the main primary energy donor, has become an inevitable trend in the transformation of energy structure. The storage and conversion of renewable energy based on large-scale water electrolysis for hydrogen production is an ideal mode to achieve efficient utilization of renewable energy and "carbon neutrality" of energy systems. Academician Ouyang Minggao of Tsinghua University believes that hydrogen production from renewable energy is an ideal gas source for fuel cell vehicles, and proposes a silver combination of intelligent energy ecological development of "centralized wind power and photovoltaic + hydrogen + fuel cell vehicles + Internet of Things + block chains"
[13]. Mogensen et al. Calculated the basic requirements for deep decarburization of European transportation systems by electrolysis technology: 1600 GW of electrolysis devices and 7500 TW · H of chemical energy storage media are needed to achieve complete deep decarburization of heavy vehicles such as trucks, ships and airplanes throughout Europe
[4]. However, there are still many problems in the renewable energy water electrolysis technology, which is difficult to achieve in the short term. In the long run, hydrogen production from water electrolysis by renewable energy is the key to realize the energy transformation with hydrogen energy as the medium, and is also the main way of large-scale hydrogen production in the future. Other new technologies for hydrogen production from clean energy, such as water photolysis by solar energy, biomass fermentation/thermochemical conversion, and thermochemical cycle, can be used as effective supplements.
At present, the review of hydrogen production technology mainly focuses on the introduction of one or a limited number of hydrogen production methods. In order to further compare and evaluate the development of existing hydrogen production technology routes, this paper summarizes more than 10 kinds of hydrogen production methods in four categories, and comprehensively reviews the latest research progress and commercialization status of hydrogen production technology based on the latest research reports and research reports at home and abroad. On this basis, through the comparative analysis of technology maturity, environmental protection and economy, the future route selection of hydrogen production technology in China is discussed in three stages: short-term (~ 2025), medium-term (~ 2030) and long-term (after 2050). In addition, considering the good application potential of artificial intelligence in the prediction, evaluation and optimization of hydrogen production system and the relative lack of relevant research in China, this paper reviews the international research progress of artificial intelligence technology in hydrogen production system, in order to provide new ideas for the future research and development of hydrogen production technology in China.
2 Hydrogen production from conventional fossil fuel reforming
Fossil fuels, including coal, oil and natural gas, are the main energy carriers in the world today. The traditional use of fossil fuels has brought serious environmental pollution problems. Developing a cleaner form of fossil fuel utilization is an urgent need for the development of the current energy system. Hydrogen production from reforming of fossil fuels is an important way to realize the clean and efficient utilization of fossil fuels, and it is also the most mature hydrogen production technology in the world.
2.1 Hydrogen production from coal
China's resource endowment is characterized by rich coal, poor oil and little gas. Although the proportion of clean energy in China's energy structure is increasing year by year, the proportion of coal in the total energy is still as high as 70%. Facing the increasingly serious environmental problems, vigorously promoting the clean utilization technology of coal is an important development direction of China's energy industry.
The technology of hydrogen production from coal is one of the important ways of clean utilization of coal, which mainly includes three processes: coal gasification, coal coking and supercritical water gasification
[14]. The first two processes have been developed for a long time and are mature. At present, they have been industrialized on a large scale. The main product of coal coking is coke, and the hydrogen comes from the purification of by-product coke oven gas, so they are classified as industrial by-product hydrogen. Supercritical water gasification of coal is a new process for coal gasification, which has been pilot-tested in China.
2.1.1 Hydrogen production from coal gasification
Coal gasification refers to the process in which coal reacts with water vapor or oxygen (air) at high temperature and normal pressure or high temperature and high pressure to convert into synthesis gas mainly composed of hydrogen and CO, and then CO is converted into hydrogen and CO2 by water-gas shift reaction. The main reactions are shown in Table 1.
The process of hydrogen production by coal gasification is shown in Figure 2. The synthesis gas after coal gasification is purified by desulfurization, water gas shift and pressure swing adsorption (PSA) to produce high-purity hydrogen with a purity of 99.9%
[15]. PSA is the most important part of hydrogen production by coal gasification technology, which ensures the purity of the final product hydrogen.
Gasifier is the key device of coal gasification. The main gasifier is divided into three types: fixed bed, fluidized bed and entrained flow bed
[16,17]. The fixed bed is mainly composed of Lurgi furnace and improved Lurgi furnace (the fourth generation of Lurgi Furnace, BGL) of German company. The main difference between the two is that the slag discharge system of BGL is improved from solid slag discharge to liquid slag discharge
[18,19]. Fluidized beds mainly include Winker furnace in Germany, U-GAS furnace in the United States and ash agglomeration gasification (ICC) process developed by Institute of Coal Chemistry, Chinese Academy of Sciences
[17]. The coal used in the fluidized bed gasifier is crushed coal with small particle size and large specific surface area, and the gasifying agent contacts with the coal more fully, which greatly improves the production capacity of the gasifier, which is nearly 6 times larger than that of the fixed bed gasifier with the same diameter, but the capacity of the fluidized bed is small, and the content of unburned coal in the fly ash is high. Compared with the former two gasifiers, the entrained flow gasifier has stronger coal adaptability and better gasification index, which is an important direction of coal gasification development
[20]. According to the different feeding methods, the entrained flow bed is divided into dry powder gasification and coal-water slurry gasification. The gasifiers of dry powder gasification include K-T furnace, prenflo furnace, shell furnace and GSP furnace. The gasifiers of coal-water slurry gasification include Texaco furnace, E-gas furnace and Tsinghua furnace. Table 2 lists the gasification methods, advantages and disadvantages of several representative entrained flow beds
[20⇓⇓⇓~24].
At present, the main development direction of coal gasifier in China is the entrained flow bed with coal-water slurry feeding. Tsinghua gasifier is a coal-water slurry feeding entrained flow bed independently developed in China. The second generation of Tsinghua gasifier first applies the water wall technology to the coal-water slurry feeding entrained flow bed, which improves the service life of the entrained flow. For the first time, the third generation of Tsinghua gasifier combined coal-water slurry, water-cooled wall and radiant steam generator, and successfully ignited, fed and combined gas at one time, which successfully realized industrialization and pioneered the transformation of new coal gasification technology
[26].
Hydrogen production from coal gasification is the lowest cost hydrogen production technology at present, the cost is less than 10 yuan/kg, but in addition to considering the economic cost, the environmental benefits of hydrogen source should also be considered
[15][27]. The coal gasification process emits 22 kg of carbon dioxide for every 1 kg of hydrogen produced
[28]. Carbon dioxide sequestration (CCS) technology is an effective way to solve the large amount of carbon dioxide emissions.If the "coal gasification + CCS" technology is used to collect and store the CO
2 produced, the overall cost of hydrogen production may be greatly increased, up to ~ 16 yuan/kg
[6]. With the development of CO
2 capture technology, the cost of hydrogen production from clean coal gasification is expected to be reduced
[6].
2.1.2 Coal supercritical water gasification hydrogen production
In 1978, MIT's Modell first proposed the concept of using coal to react in supercritical water to produce high calorific value gases. It was not until the beginning of the 21st century that people realized that hydrogen production from coal by supercritical water gasification was another promising method for hydrogen production from coal
[29]. Supercritical water gasification is a process of coal gasification above the critical point of water (temperature > 647 K, pressure > 22 MPa), which mainly includes three transformation processes: gasification, water-gas transformation and methanation. Supercritical water (SCW) has a high diffusion coefficient, about 100 times that of water at normal temperature and pressure, and has good mass transfer performance. It is completely soluble with non-polar gases and hydrocarbons, and almost insoluble with inorganic salts. The process can effectively and cleanly convert coal into H
2 and pure carbon dioxide, and has the following advantages compared with other coal hydrogen production processes: (1) the gas making reaction and the water gas shift reaction are carried out in the same reactor, hydrogen in supercritical water and coal is completely converted into hydrogen, and the yield of hydrogen is high
[30,31]; (2) during the gasification process, the coal reacts in the supercritical water, S and N in the coal are precipitated in the form of inorganic salts, and pollutants such as SO
x, NO
x, PM2.5 and the like are not discharged into the air; (3) Supercritical water gasification of coal can provide high pressure hydrogen above 20 MPa, which can be used as raw material for ammonia synthesis and hydrogenation reaction without pressurization step; (4) the energy of the supercritical water is easy to recover, and the supercritical water can be used for turbine power generation; (5) strong adaptability to different types of coal; (6) The synthesis gas is composed of hydrogen and carbon dioxide, which have large solubility difference and can be separated by absorption.
KOH, K
2CO
3, Ca(OH)
2, CaO and other alkaline catalysts are often used to promote the water-gas shift reaction in the process of coal gasification in supercritical water to improve the yield of hydrogen, and some scholars also use supported metal catalysts, such as Ni/CNT, Ni/AC, Ni/γ-Al
2O
3 and so on
[32][33]. At present, the reported reaction devices include a batch autoclave, a pipeline flow reactor and a fluidized bed reactor
[31]. Supercritical water gasification technology can be used for turbine power generation, thereby greatly reducing the production cost of hydrogen. Guo Liejin, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, has successfully developed "supercritical water steaming coal technology" after more than 20 years of scientific research
[34]. They have successfully completed a pilot project with a coal processing capacity of 1 t/H, and the gasification rate of various coal types in China can reach 100% above 700 ℃
[35]. Through the technical and economic analysis, when the coal processing capacity reaches 83. 3 t/H, the hydrogen price can be as low as 0. 7 yuan /Nm
3(1 kg H
2 is about 11.24 Nm
3)
[36].
At present, the cost of hydrogen production from coal gasification is the lowest, but there are a lot of carbon emissions, which seriously pollute the environment. If CCS technology is coupled to capture CO2 to achieve clean hydrogen production, the cost will increase by about 2 times. The technology of hydrogen production by supercritical water gasification is environmentally friendly, and its cost is equivalent to that of hydrogen production by coal gasification. This technology has entered the stage of demonstration project in China. Considering the characteristics of China's existing resource endowment and the development status of renewable energy, coal hydrogen will still be the main source of hydrogen in China in recent years. With the continuous maturity of CCS and supercritical water gasification technology, "coal gasification + CCS" and "supercritical water coal gasification hydrogen production" are expected to provide hydrogen sources with lower cost and better environmental protection for China.
2.2 Hydrogen production from natural gas
The main component of natural gas is methane with high hydrogen ratio, which is a high-quality raw material for hydrogen production. The United States and other western countries are rich in natural gas resources with low cost, so hydrogen production from natural gas is generally favored by western countries at present. Although China's natural gas reserves are relatively small nationwide, there are also excess natural gas resources in provinces such as Sichuan, which have the advantage of developing hydrogen production from natural gas
[5]. At present, hydrogen production from natural gas has become the second largest source of hydrogen in China. In addition, worldwide, more than 50% of organic carbon is stored on the seabed in the form of natural gas hydrate (combustible ice). The South China Sea and the East China Sea are rich in combustible ice reserves. The reserves of natural gas hydrate in the South China Sea alone are equivalent to 7 billion tons of oil equivalent, which can provide raw materials for the future development of hydrogen production from natural gas in China
[37].
Hydrogen production reactor is the key device for hydrogen production from natural gas. As shown in Table 3, it includes not only the traditional fixed bed and fluidized bed reactors, but also the membrane reactor, microchannel reactor and plasma reactor, which have been developed rapidly in recent years
[37]. There are various processes for hydrogen production from natural gas, including steam methane reforming (SMR), partial oxidation of methane (POM), autothermal reforming of methane (ARM), dry methane reforming (MDR) and methane cracking
[38][39][40][41⇓~43][44,45]. At present, SMR technology has been developed for nearly a hundred years, which has technical advantages and cost advantages, and is the leading technology of hydrogen production from natural gas.
2.2.1 Fundamentals of SMR
In the process of steam reforming of methane, methane and steam react with a catalyst in a high-temperature environment to be converted into synthesis gas (CO+H
2). In order to improve the yield of hydrogen, CO in the synthesis gas can further react with water to produce H
2 and CO
2 by water gas shift reaction. The main chemical reactions involved in this process are: (1) steam reforming reaction
[46]; (2) water-gas shift reaction; (3) Total reaction.
The process conditions such as temperature, pressure, steam-carbon ratio and space velocity have significant effects on the SMR process. The total reactions of steam reforming and hydrogen production are strongly endothermic reactions, and the reaction can be carried out smoothly only when the temperature is above 800 ℃. Considering the service life of the reactor and other factors, the temperature is generally controlled at 800 ~ 900 ℃, and the heating cost accounts for a high proportion (accounting for more than 50% of the total cost)
[47]. The overall reaction of SMR is a molecule-increasing reaction. From the point of view of thermodynamics, low pressure is conducive to increasing the conversion rate of methane, but from the point of view of engineering, high pressure is conducive to heat transfer, and the pressure required by the subsequent process is higher. Considering comprehensively, the pressure is generally controlled at about 3 MPa. High water-carbon ratio can obviously promote the conversion of methane, but it will also increase the cost of industrial production. In order to balance the relationship between the two, the water-carbon ratio is generally controlled between 3.0 and 4.0. Increasing the space velocity is beneficial to heat transfer and can improve the production efficiency.However, too high space velocity will lead to excessive bed pressure drop and serious power loss, so the gas hourly space velocity (GHSV) of methane is generally controlled at the 1000~2000 h
-1[48]. The final synthesis gas is a mixture of CO
2 and H
2, which is often separated by pressure swing adsorption in industry.
2.2.2 Catalyst for SMR
In addition to the above main reactions, there are also many side reactions in the SMR process, which produce carbon deposition and affect the activity of the catalyst
[46]. Therefore, the development of catalysts with high activity and strong resistance to carbon deposition is an important and difficult task in SMR process.
In the SMR process, the catalyst plays a role in activating CH4 molecules, reducing the activation energy of the reaction, and inhibiting carbon deposition. The catalyst for SMR consists of an active catalytic component and an inactive carrier. The active components of the catalyst may be noble metals such as Ru, Rh, Pd, Ir and Pt, or non-noble metals such as Fe, Co and Ni. The catalytic activity sequence of these metals under different conditions is shown in Table 4.
It can be seen from Table 4 that the activity of noble metal catalysts is generally higher than that of non-noble metal elements, and the activity of nickel-based catalysts is the highest among non-noble metal elements. The cost of precious metals is high, and it is difficult to realize industrial application. In contrast, nickel-based catalysts with low cost and moderate activity have become the first choice for large-scale industrial applications. At present, many studies have been devoted to the modification of Ni-based catalysts to further enhance their catalytic activity and operational stability. Sebai et al. Prepared a Ni/γ-Al
2O
3 supported catalyst with NiO loading of 5 wt%, and explored the effect of the size of the aliphatic structure in the ligand on its performance. The results showed that the large aliphatic structure could promote the dispersion of NiO particle size.The highly dispersed NiO nanoparticles with an average particle size of 9 ~ 15 nm were formed, which significantly improved the catalytic activity.At 700 ℃, the CH
4 conversion was as high as 99%, which was about 80% higher than that of the traditional Ni/Al catalyst, and about 10% higher than that of the traditional Ni/Al catalyst
[44].
The inactive carrier in the catalyst plays the role of physical support and dispersion of the active component, and sometimes affects the catalytic activity through the interaction with the active component. As SMR is generally a high temperature process, the support materials need to have good thermal stability, and the commonly used supports are Al
2O
3, CeO
2, ZrO
2, CaO
2, MgO and so on
[38]. A large number of studies have been devoted to the modification of catalyst supports in order to improve the catalytic efficiency, selectivity and anti-carbon deposition ability of catalysts
[46]. Wei et al. studied the activity and stability of a new nickel-based catalyst prepared by loading NiO on a new carrier Ce
xZr
yO
2-Sil-1, and explored the effects of different Ce/Zr ratios on its activity and stability
[49]. The results show that when the Ce/Zr ratio is 1, the catalyst performance is the best, and the methane conversion is as high as 99.5%; The methane conversion rate was still as high as 97. 8% after continuous operation for 16 H, showing good stability.
China's SMR process technology is mature, and many large and medium-sized enterprises in China have related equipment operation, so there are advantages in technology and equipment
[60]. However, China is short of natural gas resources as a whole, and the cost of SMR hydrogen production is relatively high, which does not have cost advantages. In areas with abundant natural gas resources in China (Sichuan and other places), low-cost SMR hydrogen production can be carried out relying on technological advantages.
3 Industrial by-product hydrogen
Industrial by-product hydrogen is another major source of hydrogen in China. Coking enterprises, iron and steel enterprises and chlor-alkali industry produce millions of tons of hydrogen every year. China is the largest coke producer in the world. 350~450 m
3 coke oven gas (cog) can be produced when one ton of coke is produced, and the hydrogen content in cog is about 54% ~ 59%. Each ton of caustic soda produced by chlor-alkali industry can produce about 0.025 ton of hydrogen as a by-product. From 2010 to 2017, the output of hydrogen as a by-product of coke enterprises and chlor-alkali industry in China and the changes of the total amount of the two over the years are shown in Figure 3
[61].
It can be seen from Figure 3 that the amount of by-product hydrogen of coke oven gas in China is maintained at more than 7 million tons all the year round; The by-product hydrogen of chlor-alkali industry has been basically maintained between 700000 and 800000 tons since 2014. Since 2012, the output of by-product hydrogen in China has remained above 8 million tons. In addition, the syngas of methanol, synthetic ammonia and propane dehydrogenation (HPD) projects all contain 60% -95% hydrogen, and the annual purge gas of methanol reaches tens of billions of cubic meters, and the hydrogen content is as high as billions of cubic meters.The annual output of synthetic ammonia industry is 150 million tons, each ton of ammonia produces 150~250 m
3[6] of purge gas, the total amount of hydrogen that can be recovered every year is about 1 million tons, and the amount of by-product hydrogen produced by 17 HPD projects built and under construction is about 370,000 tons per year. The total annual output of industrial by-product hydrogen in China is between 9 million and 10 million tons
[6].
Although the amount of industrial by-product hydrogen is huge, it contains a large number of impurity gases, such as CH
4, CO, CO
2, SO
2, etc. (Table 5)
[62⇓⇓~65]. Industry, transportation and other fields have higher requirements for hydrogen purity, such as the petrochemical industry requires hydrogen purity greater than 99.9%, the electronic industry requires hydrogen purity greater than 99.999%, and the proton exchange membrane fuel cell requires CO content in hydrogen less than 100 ppm
[62]. Therefore, the separation and purification of by-product hydrogen is a necessary way to realize its large-scale utilization. The composition of the synthetic ammonia purge gas and the chlor-alkali industrial tail gas is simple, the purification process is easier, the synthetic ammonia purge gas only needs to remove methane, and the chlor-alkali industrial tail gas only needs to remove chlorine, and can be used for hydrogen fuel cell vehicles without denitrification treatment. In contrast, the composition of coke oven gas is the most complex, and the purification process is more cumbersome
[11]. The main purification methods of hydrogen include pressure swing adsorption (PSA), cryogenic separation, membrane separation and metal hydride separation
[62].
3.1 Pressure swing adsorption method
Pressure swing adsorption (PSA) is based on the characteristics of selective adsorption of different gases on solid adsorbent and the change of gas adsorption capacity on adsorbent with the change of pressure.Hydrogen separation is achieved by periodically changing the pressure of the adsorption bed and regulating the adsorption capacity of different components on the adsorbent, which is the most widely used and technically mature hydrogen separation method in industry at present. PSA technology requires that the feed gas to be treated is hydrogen-rich gas, and the volume content of hydrogen should be at least higher than 20%
[63].
The PSA process typically consists of four steps: high pressure adsorption, low pressure desorption, reduced pressure flushing, and increased pressure recovery. The simplest device consists of two towers, one for adsorption and the other for the remaining three processes. Hydrogen recovery rate can be improved by increasing the number of towers and increasing the number of pressure equalizations. At present, a four-tower process has been developed
[64].
The selection of adsorbents is the key factor to determine the success of gas separation. Generally, adsorbents are required to have large adsorption capacity and excellent selective adsorption. Adsorbents for separating hydrogen by pressure swing adsorption in industry are generally solid particles with large specific surface area and sufficient strength and wear resistance. Commonly used adsorbents are alumina, activated carbon, zeolite molecular sieve, activated carbon fiber and silica gel. Alumina, zeolite molecular sieve and silica are hydrophilic adsorbents, which are often used for water absorption and drying. Activated carbon and carbon fiber are hydrophobic adsorbents, which are often used to adsorb weak polar or non-polar organic molecules
[66]. In addition, silica gel can also adsorb polar molecules such as CO
2. Zeolite molecular sieves have regular pore structure, have strong selectivity to molecular structure, and can make specific selective adsorption to molecules through pore design. Structural modification is an effective way to improve the adsorption capacity of activated carbon, such as the introduction of carboxyl and metal ions on the surface of activated carbon, which can enhance the adsorption capacity of polar molecules such as CO
2[66]. Cerium ions were introduced into activated carbon by Shamsudin et al., and PSA test was carried out with the syngas prepared by SMR as the gas source. The results showed that the adsorption performance of the modified activated carbon for CO
2 was much better than that for H
2. When the adsorption time was 5 min, the purity of the recovered H
2 was close to 100%, and the recovery rate was 88. 43%
[67].
For the feed gas with complex composition, the combination of various adsorbents is often used to separate the complex gas in sections
[68]. In the process of separation, specific adsorption materials are often developed for specific components and specific environment, such as the separation of specific raw gas in Daqing Petrochemical Refinery, in order to improve the hydrogen production rate, the hydrogen production unit is reformed.HX5A-10H and HXNA-CO/10 new adsorbents with high strength, large adsorption capacity and high bulk density were used for combined packing, which significantly improved the purity of hydrogen and hydrogen production rate. The purity of hydrogen at the outlet was 99. 9%, and the yield was more than 90%
[62].
PSA has the advantages of low energy consumption, high product purity, flexible adjustment, simple process, high automation, simple operation, long service life of adsorbent and high reliability of equipment, but its biggest disadvantage is low product recovery rate, which is generally only about 75%. In order to improve the recovery rate of hydrogen, the PSA process was improved, and the vacuum pressure swing adsorption (VPSA) was developed. The main improvement was to regenerate the adsorbent by vacuum pumping, so that the strong adsorptive impurities were forcibly desorbed under negative pressure. The regeneration effect of the adsorbent was good, and the product yield was high. But the disadvantage is that this process needs to add a vacuum pump, which consumes a lot of energy and has a high maintenance cost. When the pressure of raw gas is low and the requirement of recovery rate is high, the VPSA process will be adopted.
3.2 Cryogenic separation
Low temperature separation (cryogenic separation) can be divided into low temperature condensation and low temperature adsorption. Among them, cryogenic condensation is a cryogenic separation method which uses the relative difference of volatility of different components of syngas to achieve hydrogen concentration through gas expansion refrigeration, distillation and other operations. The critical temperature of hydrogen is 20 K. Affected by its critical temperature, when the hydrogen content in the feed gas is high, the critical temperature of the mixture is too low, and the difficulty of liquefaction increases. Generally, only the gas mixture with hydrogen volume fraction of 30%~80%,CH
4 and C
3H
6 volume fraction of about 40% is suitable for cryogenic condensation separation, the hydrogen volume fraction after concentration is 90% ~ 95%, and the highest yield can reach 98%
[63]. Cryogenic separation requires more compressors and cryogenic equipment, and the equipment investment and maintenance costs are high, so the production and use are limited. It is usually used for hydrogen purification of catalytic cracking gas, hydrocracking gas, coking gas and other gases with appropriate hydrogen volume, high C
3~C
5 content and low impurity content in raw gas
[63].
Cryogenic adsorption is a method to produce ultra-high purity hydrogen with a purity of more than 6 N (99.9999%) at low temperature (usually at the temperature of liquid nitrogen) by using the selective adsorption of low boiling point gas impurities in hydrogen source by adsorbent
[63]. Activated carbon, silica gel and molecular sieve are generally used as the adsorbent. After the adsorbent is saturated, the adsorbent can be regenerated by temperature rise, decompression desorption or desorption operation. The method has high requirements on the feed gas, the hydrogen content in the feed gas is generally required to be more than 95%, and impurities such as CO
2, H
2S, H
2O and the like need to be finely removed, so the method is usually used in combination with other separation methods to prepare ultra-high purity hydrogen
[69]. The advantage of this process is that the product purity is high, but the equipment investment is large, the energy consumption is high, and the operation is complex, so it is not suitable for large-scale production.
3.3 Membrane separation method
Gas membrane separation technology is considered to be the third generation of new gas separation technology with the most development prospects after cryogenic separation and PSA because of its economic, convenient, efficient and clean characteristics
[64]. In the 1960s, DuPont Company of the United States initiated the research on the separation of hydrogen and helium by polyester hollow fiber membrane for the first time. In 1990, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, successfully developed a polysulfone composite hollow fiber membrane, which can be used to recover hydrogen from FCC dry gas with a hydrogen content of only 40% ~ 60%, and the recovery rate can reach more than 85%
[62,70,71].
The process of the gas membrane separation technology is shown in fig. 4A, which realizes the gas separation and purification technology by utilizing the different gas permselectivity and diffusion performance of the membrane to different components.The driving force of mass transfer is the partial pressure difference between the two sides of the membrane. There is no phase change in the separation process, and the energy consumption is low. The separation process is easy to achieve. If the gas itself has pressure, the economy of the separation process will be more obvious. There are two kinds of membrane separation mechanisms: (1) micropore diffusion mechanism and (2) solution-diffusion mechanism. The microporous diffusion mechanism is mainly for porous films, as shown in Figure 4B
[72]. Its separation principle is that molecules with different molecular dynamics diameters have different diffusion rates in micropores. The selectivity of the membrane can be improved by controlling the pore size of the membrane. The dissolution-diffusion mechanism is mainly applicable to the gas separation of dense non-porous membranes. The principle is shown in Figure 4C. Under the action of pressure, gas molecules first contact the membrane at the high-pressure side, and then separate specific gases through the steps of adsorption, dissolution, diffusion, desolvation and escape
[73]. According to the different membrane materials, it can be divided into organic membrane and inorganic membrane.
Commonly used organic membranes for gas separation include polydimethylsiloxane (PDMS), polysulfone (PSF), cellulose acetate (CA), polyimide (PI), polyphenylene oxide (PEO), polycarbonate (PC), and polytetrafluoroethylene (PTFE). Organic membrane has good chemical stability to hydrogen, oxygen, nitrogen, methane and carbon dioxide, but when the concentration of hydrogen sulfide, ammonia and hydrocarbons in the feed gas is high, it needs to be pretreated. Polymer membranes are limited by Robseon's upper limit in gas separation, that is, there is a contradiction between permeability and selectivity: the higher the permeability, the lower the selectivity, and vice versa
[74,75]. The introduction of new structure, surface modification, blending modification and mixed matrix modification are effective means to prepare organic membranes with excellent separation effect
[75][75][75][76,77].
Inorganic membrane materials have good chemical stability and thermal stability, and can work in harsh environments such as high temperature and strong acid. When some inorganic membranes are used for gas separation, they can greatly exceed the Robseon limit, so they are a kind of potential membrane separation materials. Common inorganic membranes used for hydrogen separation are silica membranes, zeolite membranes, and metal membranes
[78][64][79]. Among them, the metal membrane has high permeability and diffusion coefficient, good thermal stability and mechanical properties, which is very suitable for the preparation of high-purity hydrogen. Hydrogen separation metal membranes are generally referred to as palladium membranes or palladium alloy membranes. Palladium has a high apparent permeability coefficient and a strong catalytic effect on the dissociation of hydrogen molecules, and hydrogen with a purity of 99.999% can be prepared by palladium membrane separation. However, the pure palladium membrane is prone to wrinkling, distortion and cracking after several thermal cycles in hydrogen separation, resulting in hydrogen embrittlement, which seriously affects the industrial service life. In order to further improve the hydrogen permeability and mechanical strength and reduce the cost, palladium alloy membranes have received a lot of attention and research. At present, many palladium alloy membranes with excellent performance have been prepared, such as Pd-Ag, Pd-Cu, Pd-Y and other alloy membranes with better performance
[79]. The preparation of supported metal palladium membranes or palladium alloy membranes is an effective way to improve the mechanical strength and hydrogen permeability and reduce the cost. At present, the main porous materials that can be used as the substrate of supported palladium composite membranes are porous alumina, porous ceramics, porous glass and porous stainless steel
[80][81,82].
The advantages of the palladium alloy membrane method for hydrogen purification are significant, such as high purification efficiency, high product purity, extremely high hydrogen recovery rate (up to 99%), and strong anti-poisoning ability of the palladium alloy membrane against impurity gases, so that the palladium membrane method can be applied to hydrogen purification of various feed gases
[83]. However, the disadvantages of using palladium alloy membrane to purify hydrogen are also obvious, for example, the hydrogen permeation rate of palladium membrane is not high, resulting in very small output, and palladium metal membrane is extremely expensive, so the production cost is high, and large-scale industrial application can not be realized. Recent studies have shown that liquid metal membranes are more effective than palladium in the separation of hydrogen and other gases. For example, the hydrogen permeability of liquid gallium at high temperature is higher than that of palladium, and its price is lower than that of palladium. It is expected to become a new generation of metal separation membrane materials to replace palladium
[84,85].
3.4 Metal hydride separation method
The metal hydride separation method uses hydrogen storage materials to react with hydrogen to store hydrogen, which has a high selectivity for hydrogen. When the hydrogen gas is in contact with the hydrogen storage material, only the hydrogen gas can be hydrogenated with the hydrogen storage material, and other impurity gases cannot be reacted with it. Hydrogen storage materials can absorb hydrogen when they are cooled and pressurized, and release hydrogen when they are heated and decompressed, which is helpful for their application in hydrogen purification
[62]. During the adsorption of hydrogen, hydrogen molecules dissociate into hydrogen atoms under the catalysis of the alloy, diffuse into the metal, and finally fix in the metal lattice. When the hydrogen storage alloy is heated, hydrogen is discharged from the metal lattice, and the purity of hydrogen can be as high as 99.9999%. Therefore, the metal hydride method is usually used in the field of hydrogen storage and purification
[62]. Hydrogen storage materials mainly include rare earth series, magnesium series, Zr series and Ti series hydrogen storage alloys, such as LaNi
5, Mg
2Ni, ZrMnFe, TiVMn and so on
[86].
Metal hydride separation method is one of the most suitable technologies to obtain high purity hydrogen, which has the characteristics of high purity hydrogen, simple operation, low energy consumption and low material price. However, the disadvantages are also obvious: in the process of hydrogen recovery, the material is easy to react with the impurity gas, causing the purification material to be poisoned and lose its activity, thus reducing the purification efficiency; There is hydrogen retention when hydrogen is released. The hydrogen processing capacity is relatively smaller, and the method is more suitable for small-scale hydrogen production in laboratories and the like
[62,87].
Among the above four hydrogen purification methods, PSA, cryogenic separation and membrane separation can be used to purify large-scale industrial by-product hydrogen, while metal hydride separation has low production capacity and limited industrial application. The technical and economic characteristics of the three main large-scale industrial by-product hydrogen purification methods are shown in Table 6. The three methods have their own advantages and limitations
[63,87].
PSA is the most widely used hydrogen recovery method in industry at present, which is widely used in the purification of coke oven gas, coal hydrogen synthesis gas, SMR synthesis gas and so on. In 2008, Shenhua Ordos built a large PSA hydrogen production plant with 280000 Nm
3/h
[88]. In 2017, most of the fifteen integrated refining and chemical projects planned by the state adopted coal gasification hydrogen production technology, and all of them had PSA hydrogen production capacity of tens of thousands to hundreds of thousands of Nm
3/h
[89]. Shenhua Group is developing a PSA unit with a maximum hydrogen production capacity of 500 thousand Nm
3/h. The hydrogen production unit with an operating pressure of 5. 0 MPa has been in normal operation, and the hydrogen production unit with an operating pressure of 6. 0 MPa is under construction
[90].
As early as 1823, cryogenic separation method was applied in the cooling of liquid chlorine, liquid ammonia, liquid carbon dioxide and other gases
[91]. In recent years, the cryogenic separation method is mainly used to separate hydrogen and CO in China. The 55,000 tons/year CO separation unit supporting the 100,000 tons/year acetic acid project of British BP Company introduced by Shanghai Wujing General Chemical Plant and the 500,000 tons/year Yangzi BP acetic acid project jointly invested by Sinopec Yangzi Petrochemical CO., Ltd. are all used to separate CO by cryogenic separation unit
[91].
Pd-Ag alloy is the only commercialized metal membrane in membrane separation technology. In the early 1960s, Johnson developed a Pd-Ag (Pd 77 wt%, Ag 23 wt%) alloy membrane tube as an independent high-purity hydrogen production unit, which was successfully pilot-tested with a maximum hydrogen yield of 56 Nm
3/h and an impurity content of less than 0.1 ppm
[92,93]. Palladium alloy materials are all precious metals with high cost, which limits their large-scale application. In industry, almost all the membranes used to recover and purify hydrogen from refinery gas are organic membranes. In 1979, the "Prism" polysulfone hollow fiber membrane introduced by Monsanto Company successfully realized the commercial application of membrane separation in hydrogen recovery from refinery gas
[94]. In the 1980s, the cellulose acetate Separex membrane separator developed by Air Product was put into use in the United States and Japan, and a set of separation unit with a processing capacity of 64 900 Nm
3/h was built in Fawlay Refinery in the United Kingdom in 1988
[94]. The polyimide membrane separator developed by Ube Industries of Japan has a capacity of 12 000 Nm
3/h, a hydrogen recovery rate of 80%, and a purity of 97%
[94]. China's initial hydrogen purification membrane separators were mainly imported. In 1985, Dalian Institute of Chemical Physics, Chinese Academy of Sciences successfully developed the first polysulfone hollow fiber N
2/H
2 separator, which was successfully industrialized in Zhenhai Refining and Chemical Company of Sinopec, filling the domestic technical gap
[94]. A hydrogen separation unit with Prism membrane was put into operation in Changqing Petrochemical Company in 2010, with a treatment capacity of ~7000 m
3/h and a hydrogen recovery rate of more than 80%, and the unit has been running steadily so far.
Different feed gas composition has different separation requirements, and only one separation method may be difficult to achieve the separation goal. Combining different separation methods and giving full play to their respective advantages to separate hydrogen is an effective way to reduce costs and meet the separation requirements. Membrane separation + PSA, cryogenic separation + PSA and temperature swing adsorption + PSA are common combinations.
4 Clean energy electrolyze water to produce hydrogen
In order to solve the problems of environmental pollution and energy depletion, it has become the main trend of the development of the world energy system to vigorously develop clean energy such as renewable energy and nuclear energy. The annual availability of renewable energy resources in China is more than 187 billion tons of oil equivalent, but the utilization rate of renewable energy is low at present, and the future development potential is huge
[95]. Hydrogen production from renewable energy is expected to become an important source of hydrogen production in China in the future, in which solar energy and biomass energy can not only generate electricity, but also produce hydrogen directly. Other renewable energy sources can not directly produce hydrogen, but can only generate electricity first, and then use the electricity that the grid can not absorb to produce hydrogen
[96,97].
Nuclear energy is a clean primary energy, which is developing well in China at present, especially the high-temperature gas-cooled reactor with the characteristics of the fourth generation nuclear energy system independently developed by the Institute of Nuclear Research of Tsinghua University, which greatly promotes the development of China's nuclear energy industry
[98]. While actively developing nuclear power, nuclear power or nuclear heat can also be used to produce hydrogen to expand the application scenarios of nuclear energy and improve the economy of nuclear energy. Hydrogen production by thermochemical cycle and high temperature steam electrolysis, which use water as raw material and utilize nuclear heat completely or partially, is an important development direction of nuclear hydrogen production. Compared with traditional methods, nuclear hydrogen production has the advantages of high efficiency, large scale and no greenhouse gas emissions.
Hydrogen production by water electrolysis has a long history of development and has gradually realized industrialization, accounting for about 3% of the total hydrogen production in China
[6]. Hydrogen production by water electrolysis can use nuclear power, and can also effectively absorb renewable energy power such as wind power, photovoltaic power generation, hydropower and so on, which can meet the large-scale energy storage demand of the future power system with a high proportion of renewable energy. In addition, the electrolytic cell stack can realize modular and flexible design, which can not only centralize hydrogen production and regional hydrogen supply, but also build a small electrolytic water hydrogen production device in a single hydrogenation station to realize the intelligent interconnection of different energy networks. Hydrogen production by water electrolysis provides new possibilities for large-scale, long-term storage and long-distance transportation of electric energy, provides new ideas for solving the problem of "three abandonment" (light abandonment, water abandonment and wind abandonment), and provides a new way for "peak shaving and valley filling" of power grid. The core equipment of water electrolysis is electrolytic cell. At present, the common electrolytic cells are alkaline electrolytic cell (AEC), proton exchange membrane electrolytic cell (PEMEC) and solid oxide electrolytic cell (SOEC)
[99][100,101][102][103].
4.1 Alkaline electrolytic cell
Alkaline electrolytic cell (AEC) is the most mature electrolytic cell at present, which has been industrialized as early as the mid-20th century. The basic principle of hydrogen production by AEC water electrolysis is shown in Figure 5. Under the action of current, water is decomposed into hydrogen and oxygen through electrochemical reaction, which are separated out at the cathode and anode of the electrolytic cell. Water is a weak electrolyte. In order to enhance its conductivity, alkaline substances such as NaOH and KOH are usually added to water to form an alkaline electrolyte, and asbestos is used as a diaphragm
[104]. The electrode reaction is as follows:
The reversible potential of electrolyzed water is calculated as follows:
[99]
Where
is the equilibrium potential of the cell under standard conditions,
is the electrodeable potential of the oxygen and hydrogen electrodes under standard conditions, respectively,
is the Gibbs free energy, n is the number of moles of electrons transferred in the reaction, and F is Faraday constant.
at 25 ℃
[99]. In actual electrolysis, there are ohmic impedance and polarization impedance in the circuit, which cause the increase of electrolysis voltage. The expression of actual electrolysis voltage is:
[99]
The
is the overpotential caused by polarization, I is the current of the circuit, and the
is the ohmic resistance in the circuit. In practical industrial water electrolysis, E is usually 1.8 to 2.0 V at a current density of 1000~300 mA/cm
2[105]. Fig. 6 shows the relationship between temperature and electrolysis voltage. The equilibrium voltage line and the hot neutral voltage line divide the potential-temperature plane into three parts. The equilibrium voltage line represents the theoretical minimum voltage (
) of electrolysis, and the hot neutral voltage line represents the actual minimum voltage of electrolysis without external energy input
[99]. Below the equilibrium voltage, the electrolysis cannot be carried out; Between the equilibrium voltage and the hot neutral voltage, external heat supply is required for electrolysis; When electrolyzing above the hot neutral voltage, heat is released to the outside.
Electrolysis efficiency The energy utilization efficiency of the electrolytic cell can be evaluated. Table 7 summarizes the expression and meaning of the electrolysis efficiency under different criteria
[99]. There are two main ways to improve the electrolysis efficiency: (1) from the thermodynamic point of view, reduce the electric energy required for electrolysis of water (reduce the overpotential), that is, increase the temperature or pressure; (2) reduce energy losses in electrolysis by minimize that ohmic impedance of the system.
In addition to electrolysis efficiency, electrolysis rate is another important parameter to describe the performance of the electrolytic cell. The rate of the electrolysis reaction can be expressed as:
Where n is the number of electrons consumed in the electrode reaction, A is the surface area where the electrode reaction occurs, and J is the current density. The relationship between the overpotential and the current density of the hydrogen electrode can be described by the Tafel equation:
A and B are empirical constants, a is related to electrode material, electrode surface state, solution concentration and temperature, and B is called Tafel slope, indicating the sensitivity of overpotential to current. B and J are the main parameters to describe the kinetics, which are often used to reflect the kinetic performance of different electrodes.
4.1.1 Key electrode material
Most transition metals and noble metals can be used as electrode materials of AEC. Noble metals have high cost, and Ni, Co, Cu, Fe and other metals with moderate cost and reactivity have received the most attention. Electrode materials used in industry are Ni-Co and other alloy materials. Electrode materials have a great impact on the electrolytic performance. A large number of scholars have done a full study on the modification of electrode materials. Figs. 7A and 7B show the overpotential of different hydrogen electrode materials as a function of current density, and the porous Ra-Ni electrode with a large internal specific surface area has the highest activity; Compared with pure Ni, Ni-Co-Cu alloy can effectively improve the electrode activity. And that overpotential of the electrode decrease obviously with the increase of the temperature
[106]. The electrolysis parameters of various Ni alloy electrodes at the current density of 100 mA/cm
2 are summarized in Table 8. By comparison, the electrolysis performance of Ni alloy is significantly better than that of pure metal Ni, but its performance is much worse than that of Ra-Ni. This is due to the porous structure of Ra-Ni, which provides a large number of active sites for hydrogen adsorption and greatly improves the hydrogen production rate.
The irreversibility of the oxygen electrode in the electrolysis reaction is the main reason for the decrease of the efficiency of the electrolytic cell. The results show that the porous or smooth structure of the electrode has no significant effect on the performance of the oxygen electrode. The main reason is that the large bubbles produced by the oxygen electrode block the pore structure and the internal pores can not be fully utilized
[106]. Fig. 7 C and d show the variation of overpotential of different oxygen electrodes with current density, and SS316L and NiFe(OH)
2 show good electrolysis activity, which are better than pure metal Ni. These results indicate that Ni-Fe and Ni-Co alloy materials can effectively improve the activity of oxygen electrode
[106].
4.1.2 Structural design of electrolytic cell
Besides the electrode material, the electrode configuration of the electrolytic cell is also an important factor affecting the electrolytic performance. The battery configuration is divided into a unipolar configuration and a multipolar configuration, wherein the unipolar configuration is the parallel connection of a plurality of single batteries, and the multipolar configuration is the series connection of a plurality of single batteries. The unipolar configuration is simple in structure and easy to manufacture and maintain, but the ohmic loss in the circuit is large; Ohmic losses are greatly reduced in multistage configuration circuits, but greater precision is required in design and manufacture to prevent electrolyte and gas leakage between cells. Another important design issue is that the smaller the gap between the design electrode and the separator, the lower the ohmic cell resistance caused by the electrolyte and the bubbles generated, and the smaller the ohmic losses. The aim of the "zero-gap system" studied so far is to reduce or eliminate the bubbles in the critical electrode area. The disadvantage of this design, however, is that it increases the likelihood of sparking and is more expensive to manufacture. For the unipolar configuration, the general industrial electrolysis voltage is 2.2 V, the multistage configuration voltage is generally 2.2 × (n − 1) V (n is the number of electrodes), and the current density is about 0.4 A/cm
2[106].
4.1.3 Development Status of AEC Reactor
The advantage of AEC is that its technology is mature, durable and low cost. At present, AEC has been industrialized in China. The installation of AEC devices in China is about 1500 ~ 2000 sets, and the total amount of hydrogen production is about 8×10
4t/ years
[113]. The maximum hydrogen production capacity of a single AEC unit in China can reach 1000 Nm
3/h
[6]. The representative units of AEC research in China include Tianjin Continental Hydrogen Production Equipment Co., Ltd., Suzhou Jingli Hydrogen Production Equipment Co., Ltd., Dalian Institute of Physical Chemistry, Chinese Academy of Sciences, 718 Institute of China Shipbuilding Heavy Industry, etc. The 718 Institute has built the largest production base of hydrogen production equipment in China, and developed new electrodes and diaphragms that can make the current density of alkaline electrolyzers reach more than 6000 A/m
2, so that the electrolysis efficiency is 2 to 3 times higher than that of traditional AEC
[114]. In 2018, Suzhou Jingli developed a water electrolysis hydrogen production equipment with 1000 Nm
3/h hydrogen production capacity
[115]. In 2019, the Dalian Institute of Chemical Physics of the Chinese Academy of Sciences developed a new generation of AEC catalyst with the highest efficiency in hydrogen production by large-scale electrolysis of water, and achieved stable operation in the pilot demonstration project in Suzhou.The catalyst will be assembled in the 1000 Nm
3/h of water electrolysis hydrogen production equipment produced by Suzhou Jingli and applied to the first demonstration project of liquid solar fuel synthesis in China, which is expected to reduce the energy consumption of large-scale industrial water electrolysis hydrogen production equipment by more than 15%
[116].
However, the electrolyte of AEC is strong alkali, if leakage occurs in production or improper post-treatment, it is easy to cause environmental pollution, asbestos diaphragm has carcinogenic risk, at the same time, the products prepared by AEC need to be dealkalized, which has the disadvantages of environmental pollution and large equipment volume. In addition, the alkaline electrolytic cell cannot be quickly closed or opened, and the speed of hydrogen production cannot be well controlled. Therefore, the pressure balance between the anode and cathode of the electrolytic cell must be maintained at all times to prevent the hydrogen and oxygen gas from passing through the porous asbestos membrane and mixing to cause safety accidents. This also makes it difficult for AEC to work with renewable energy sources with rapid fluctuations.
4.2 Proton exchange membrane electrolytic cell
Proton exchange membrane electrolysis cell (PEMEC) rose in the 1960s and is now in the early stage of marketization. Its main components include a polymer membrane with proton exchange capacity and a cathode and anode catalyst layer closely connected with both sides of the electrolyte membrane. The working principle of PEMEC is shown in Figure 8. Water is catalytically decomposed into oxygen and H
+,H
+ at the anode, which pass through the electrolyte membrane to the cathode, where electrons are obtained to generate hydrogen. The electrode reaction is as follows
[117].
4.2.1 Key electrode material
The catalytic activity of PEMEC electrodes is essentially dependent on noble metals or noble metal oxides. Miles et al. investigated the hydrogen and oxygen electrode catalytic activity of various metals in a H
2SO
4 solution at 80 ° C and 0.1 mol/L
[118]. The results show that for hydrogen electrode, the order of activity is Pd > Pt > Rh > Ir > Re > Os > Ru > Ni; For the oxygen electrode, the order of activity is Ir ≈ Ru > Pd > Rh > Pt > Au > Nb. At present, the commonly used hydrogen electrode material is Pt-based catalyst, and the commonly used oxygen electrode catalyst is Ir or Ru-based catalyst
[119]. The electrode materials with high catalytic activity are noble metals, which have high cost. Therefore, how to reduce the catalyst load without significantly affecting the catalytic activity has attracted wide attention. The results show that the hydrogen electrode of PEMEC exhibits good kinetics at the current density of 1 A/cm
2, and the polarization loss mainly comes from the oxygen electrode
[120]. The change of the catalyst loading of hydrogen and oxygen electrodes in recent years also shows this point: the loading of hydrogen electrode has decreased from 5 mg/cm
2 to 0.5 mg/cm
2[119]; The loading capacity of the oxygen electrode is basically maintained at a ~0.2 mg/cm
2, and the decrease is not obvious.
Using Ta, Ni, Co and other non-noble metal elements to reduce the content of noble metals, reduce the cost and maintain the catalytic activity is an important direction of oxygen electrode research. Marshall et al. Prepared a series of Ta-doped Ir
xRu
yTa
zO
2 oxide powders with different compositions by the method of hydrolysis followed by drying, and used them as anodes for water electrolysis test. The results showed that 7 of the 16 synthesized oxides met the following conditions: when the temperature was 80 ℃ and the electrolysis current density was 1 A/cm
2, the electrolysis voltage was lower than 1.6 V
[120]. Alia et al. Prepared Ir-Ni and Ir-Co nanowires doped with Ni and Co by current replacement method, and compared the catalytic activity of oxygen electrode with commercial Ir and IrO
2. The results showed that the catalytic activity of commercial Ir and IrO
2 nanoparticles was basically the same, and the activity of Ir-Ni (9.3) 90.5 and Ir-Co (4.2) 93.6 nanowires after acid leaching was 4 to 5 times higher than that of commercial nanoparticles in the kinetic region
[121].
It is another important direction for the development of oxygen electrode to improve the catalytic activity by using new preparation technology to realize the nano-catalyst. Yi Baolian et al. Prepared a porous nano-gold /IrO
2 composite ultrathin film oxygen electrode by thermal decomposition method, in which the loading of Au and IrO
2 was as low as 0.1 mg/cm
2 and 0.086 mg/cm
2, respectively, while ensuring good electrolytic performance. When the temperature was 80 ℃ and the electrolytic current density was 2 A/cm
2, the electrolytic voltage was as low as 1.728 V
[122].
4.2.2 Lectrolytic cell key structures
Proton exchange membrane and bipolar plate are the key structures of PEMEC. PEMEC uses proton exchange membrane to replace the traditional strong alkaline electrolyte solution for proton conduction, and isolates the possibility of contact between two gases through a dense diaphragm, which avoids the potential environmental and safety hazards of AEC, and also increases the operating pressure of the electrolytic cell (up to 85 bar)
[99]; However, the polymer membrane also limits the operating temperature of the PEMEC, which is typically below 80 ° C. In addition, the PEMEC adopts a zero-gap structure, so that the structure of the electrolytic cell is more compact and concise, the ohmic resistance of the electrolytic cell is reduced, and the overall performance of the electrolysis cell is greatly improved; the operating current density of the PEMEC can be as high as more than four times that of an AEC (1.6 A/m
2); and the purity of a prepared product can be as high as 99.99%
[99,104]. Because the PEMEC system does not require dealkalization treatment, the pressure control range is also larger. Another advantage of PEMEC is that its protons are transmitted through polymer membranes and respond quickly to power fluctuations, which makes it possible to produce hydrogen by coupling PEMEC with renewable energy sources with large fluctuations.
At present, Nafion series membranes with perfluorinated polymer as the main chain and sulfonic acid group (—SO
3H) as the side chain developed by DuPont Company are commonly used in commercial proton exchange membranes. The —SO
3H can be easily decomposed into a — S
and a H
+, and the latter is combined with a water molecule to generate a H
3O
+, passes through the polymer film under the action of an electric field, is reduced into a hydrogen atom at the cathode and is combined to generate hydrogen, so that the proton transmission is promoted, and the electrolysis performance of the PEMEC is improved. On this basis, other companies have developed several successful membrane materials, such as Dow membrane developed by Dow Company in the United States, Aciplex membrane developed by Asahi Chemical Company in Japan, Flemion membrane developed by Asahi Glass Company, etc
[123]. By changing the length of the perfluorosulfonic acid membrane branch chain, the sulfonic acid radical content of the membrane can be regulated, and then the membrane performance can be regulated
[123,124]. Dow membranes have relatively short branched chains and high sulfonate content, so the relative molecular mass of the membranes is lower and the conductivity is higher. However, Dow membranes are more difficult to synthesize and more expensive than Nafion membranes with longer side chains.
Perfluorosulfonic acid membrane has excellent performance, but its cost is high. In recent years, many studies have shown that other polymer membranes can also be used as proton exchange membranes for PEMEC, such as polybenzimidazole (SPBI), polyethersulfone (PES), polyetherketone (PEEK) and sulfonated polyphenylquinoxaline (SPPQ)
[123,125,126].
At present, the domestic proton exchange membranes are mainly imported, and the cost is high, which limits the commercialization of PEMEC. Shandong Dongyue Group is building a production line of perfluorosulfonic acid proton exchange membrane with an annual capacity of 1.5 million m
2. After the production line is put into operation, it will provide China with low-cost proton exchange membranes that can meet the needs of low-end production
[5].
The bipolar plate plays a role in separating the gas of the two poles and collecting and conducting the current. Bipolar plate materials have high requirements for corrosion resistance and electrical conductivity, especially in the oxygen electrode side of PEMEC, the strong acidic environment puts forward extremely high requirements for the corrosion resistance of the bipolar plate
[127,128]. At present, the commonly used bipolar plate material is titanium alloy, and the surface of titanium plate must be modified to meet the requirements of PEMEC
[127]. Titanium itself is a precious metal material with high cost, and the cost is further increased after surface modification. The extremely high cost of bipolar plates has become another important factor limiting the mass production of PEMEC.
Developing new membrane materials and bipolar plate materials with high performance and low cost or reducing the cost by improving the preparation methods of existing materials is a necessary condition to promote the large-scale commercialization of PEMEC in China.
4.2.3 Development Status of PEMEC Reactor
At present, PEMEC has achieved small-scale industrial application with maximum hydrogen production of about 30 Nm
3/h and power consumption of 174 kW
[104]. Overseas PEMEC technology is mature. ProtonOnsite is the world's leading hydrogen supplier for PEMEC hydrogen production, providing more than 2000 sets of PEMEC devices for 72 countries around the world. The hydrogen production capacity of HOGENC-C series PEMECs developed by ProtonOnsite is 30 Nm
3/h, which is mainly used for hydrogen refueling stations. The HOGEN-H series can be coupled with a 65 kW wind turbine to produce 12 kg/day of hydrogen; In 2015, the world's first megawatt PEMEC was launched, with hydrogen production up to 400 Nm
3/h, which is expected to meet the future demand for large-scale energy storage
[113]. In 2012, the German Power to Gas project used surplus electricity to produce hydrogen and injected it into the natural Gas pipeline in 2013, which improved the utilization rate of electricity and reduced the waste of peak and valley electricity
[113]. The Energiepark project in Mainz, Germany, based on a 6 MW PEMEC stack and an 8 MW wind farm, can produce up to 1000 Nm
3[129] of hydrogen per hour. In 2014, a PEMEC hydrogen production plant with a total capacity of 2 MW was deployed in Toronto, Canada, realizing hydrogen energy storage
[113]. Domestic PEMEC is in the transition stage from scientific research to industrialization, and the main research and development units are China Shipbuilding Heavy Industry 718 Institute, China Aerospace 507 Institute, Chunhua Hydrogen Energy, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, etc
[130]. At present, the PEMEC products sold in the market are mainly agents of foreign products, and the hydrogen production is 0.3~2.0 Nm
3/h. In 2008, Dalian Institute of Chemical Industry developed an electrolytic cell stack and system with hydrogen production of 8 Nm
3/h, output pressure of 4.0 MPa and purity of 99.99%
[113]. In 2010, the energy consumption index of PEMEC hydrogen generator developed by Dalian Chemical Institute was superior to that of similar products in the world
[113]. In 2015, Hebei Guyuan began to build a 10MW demonstration project of hydrogen production from wind power, using advanced German hydrogen production technology and equipment. The project will realize the local conversion of wind power into hydrogen and explore new ways of local consumption of wind power
[131].
4.3 Olid oxide electrolytic cell
The development of solid oxide electrolytic cell (SOEC) started in the 1970s, which is mainly used for hydrogen production by high-temperature electrolysis of steam (HTSE). Compared with the other two electrolytic cells, SOEC has the lowest technical maturity and has not yet been industrialized. However, due to its high electrolysis efficiency and other advantages, it has been developed rapidly in recent years. The working principle of SOEC is shown in Figure 9. At high temperature (800 ~ 1000 ℃), water vapor enters from the cathode and is electrolyzed at the cathode to generate H
2 and O
2-,O
2-, which reach the anode through the electrolyte layer and lose electrons at the anode to generate O
2[104]. The electrode reaction is as follows:
The basic composition of the SOEC is shown in Figure 9, with a dense electrolyte layer in the middle, which is mainly used to separate oxygen and hydrogen and conduct cations and protons. Therefore, the electrolyte structure is generally required to be dense and have high ionic conduction and negligible proton conduction
[132]. And that two side are respectively a porous hydrogen electrode and an oxygen electrode, and the porou structure is beneficial to the transmission of the hydrogen and the oxygen.
HTSE is carried out at high temperature, so hydrogen production energy includes two parts of electric energy and heat energy, which can be expressed as:
[133]
is the reaction enthalpy change, representing the total energy required for electrolysis,
is the Gibbs free energy change of the reaction, corresponding to the electrical energy required for electrolysis,
is the reaction entropy change,
corresponds to the high temperature heat energy, E is the theoretical decomposition voltage of water at different temperatures, and
is the total isobaric molar heat capacity of different gases. According to the above formula, the energy of each part at different temperatures can be calculated, and the results are shown in Figure 10
[134]. It can be seen from fig. 10 that the
of electric energy required for electrolysis decreases as the temperature increases, the theoretical decomposition voltage of water also decreases, and the proportion of heat energy increases. From the point of view of thermodynamics, high temperature can improve the efficiency of hydrogen production and reduce energy loss. In addition, the electrolyte material of SOEC is oxygen ion conductor, and its ionic conductivity increases with the increase of temperature, which further reduces the ohmic loss of electrolysis. At present, the electrolysis efficiency of SOEC in the laboratory is close to 100%
[134]. From the kinetic point of view, high operating temperature can increase the electrode reaction rate and reduce the electrode reaction overpotential, thus realizing hydrogen production at low voltage and high current and increasing the hydrogen production.
4.3.1 Key Material
The key materials of SOEC include electrolyte, oxygen electrode and hydrogen electrode. The operating temperature of SOEC is high, so the structural stability and chemical stability of key materials at high temperature have a great influence on the performance of SOEC.
Electrolyte is an important part of SOEC. Its properties not only directly affect the performance of the cell, but also determine the choice of electrode materials and preparation technology. In addition to ion conduction and gas barrier mentioned above, SOEC electrolyte materials also need to have good mechanical properties in different ranges of room temperature, working temperature and manufacturing temperature, as well as good compatibility with cathode and anode materials. At present, the electrolyte materials are mainly divided into four types: ZrO
2 based, CeO
2 based, Bi
2O
3 based and ABO
3 based perovskites, and the relationship between conductivity and temperature of different electrolytes is shown in Figure 11
[135]. ZrO
2 is the most studied electrolyte material in solid oxide electrolysis cells. In the YSZ(Y
2O
3stable ZrO
2) system, the YSZ exhibits the largest ionic conductivity with 8 mol% (8YSZ) doping amount of Y
2O
3, and is chemically stable
[136]. Bi
2O
3 is a solid electrolyte material with the highest oxygen ion conductivity developed so far, and its conductivity reaches 1 S·cm
-1 at 825 ℃, but its phase transition occurs at low temperature, which easily leads to material fracture and performance deterioration, and it is easily reduced to metal Bi at low oxygen partial pressure
[137].
The resistivity of electrolyte is several orders of magnitude higher than that of cathode and anode materials, which is the main source of ohmic impedance of electrolytic cell. The preparation of electrolyte film is an important research direction to reduce its ohmic resistance and improve the performance of the electrolytic cell. Liang Mingde of Northeastern University realized the thin film preparation of YSZ electrolyte by screen printing method. By controlling the printing times, the thickness of the electrolyte can be controlled below 10 μm, and the thinnest can be 4 μm. At this time, it can still maintain good electrolytic performance
[132]. Thinning of the electrolyte can also be achieved using the plasma spraying process. Marr et al. Prepared a ~ 35 μm thick YSZ electrolyte by solution precursor plasma spraying technology, and tested its full cell performance, with an open circuit voltage of ~ 1 V
[138].
The main function of oxygen electrode is to provide O
2 generation sites and diffusion channels, so the oxygen electrode is required to have high electronic conductivity and porous structure, and to have high catalytic activity for oxygen evolution reaction at high temperature. LSM(La
1-xSr
xMnO
3-δ) is a good oxygen electrode material widely used in the early stage, but it is a pure electronic conductor with low ionic conductivity, which limits its application. Mixed ionic electronic conductor (MIEC) is the most widely used and studied oxygen electrode material nowadays, which can be divided into three categories according to the crystal structure (Figure 12): perovskite-type oxide (La
0.8Sr
0.2CoO
3), perovskite-like oxide (RP-type oxide, La
2NiO
4+δ, Pr
2NiO
4+δ) and double perovskite structure (PrBaCo
2O
5+δ, GdBaCo
2O
5+δ)
[139].
Suntivich et al. studied the catalytic activity law of oxygen evolution reaction (OER) of perovskite oxygen electrode by using molecular orbital theory, and the results showed that the OER specific activity of perovskite changed with the filling of e
g orbital on the surface of B-site anion in a volcanic distribution (as shown in Figure 13), in which the e
g of Ba
0.5Sr
0.5Co
0.8Fe
0.2O
3-δ(BSCF) was 1.3, showing extremely high catalytic activity, which was one order of magnitude higher than that of noble metal oxide IrO
2[140].
RP-type oxides have attracted much attention and research because of their high oxygen ion conductivity. Tsvetkov et al. Prepared Nd
2NiO
4+δ oxygen electrode and found that the oxygen reduction reaction (ORR) activity could be improved by strain-induced oxygen defects
[141]. Double perovskite is also an oxygen electrode material which has been studied more in recent years. Using PrBa
0.5Sr
0.5Co
1.5Fe
0.5O
5+δ(PBSCF) as a model material, Zhu et al. Generated different concentrations of oxygen vacancies in the material by Ar and H
2 plasma treatments
[142]. Electrochemical tests showed that the OER activity of PBSCF was significantly improved with the increase of oxygen vacancies, and the overpotential of the oxygen vacancy generating PBSCF powder was 90 mV lower than that of the commercial IrO
2 catalyst at a current density of 35 mA/cm
2, indicating its high OER reaction activity at a large current density.
Nanocrystallization of oxygen electrode materials or preparation of nanocomposite structures is an important method to improve the performance limit of SOEC. The National Institute of Advanced Industrial Science and Technology of Japan reported that a novel nano-film composite oxygen electrode was fabricated by spray pyrolysis method, and the SOEC prepared by which reached 3.13 A/cm
2 at 750 ℃ and 4.08 A/cm
2 at 800 ℃, corresponding to a hydrogen yield of 1.31 and 1.71 L·h
-1·cm
-2[143], respectively, showing good stability for high-current operation. Chen et al. First loaded a layer of PrNi
0.5M
O
3(PNM) film on the surface of LSCF, and then prepared PrO
x nanoparticles by in-situ desolvation technology. The polarization impedance of oxygen electrode was as low as ~0.022Ω·cm
2, which was only 1/6 of that of pure LSCF electrode
[144].
The main function of the hydrogen electrode is to provide a decomposition reaction site for water vapor and a channel for electron conduction and hydrogen diffusion, which requires good electronic conductivity and electrocatalytic activity, as well as sufficient pores to ensure the smooth entry and exit of water vapor and hydrogen. Ni-YSZ, a porous ceramic-based electrode prepared by dispersing Ni in YSZ, is the most commonly used hydrogen electrode material for SOEC due to its high catalytic activity and low price
[145]. Compared with SOFC, the water vapor content of hydrogen electrode in SOEC mode is higher, and the migration ability of water vapor is weaker than that of hydrogen, which puts forward higher requirements for the pore structure and stability of the electrode. A study by Riso Laboratory in Denmark found that SOEC will have a short-term "passivation phenomenon" in the first few hundred hours of operation
[134]. In-depth study shows that the performance degradation of the battery mainly comes from the hydrogen electrode. The trace impurities such as S, C and Si in the electrolytic cell materials can form passivation layers on the surface of Ni particles, Ni/YSZ interface and battery/electrolyte three-phase interface (TPB), which reduces the active area of the electrode and reduces the performance
[146]. And that passivation phenomenon can be effectively inhibit by improving the purity of the hydrogen electrode raw material. In addition, by means of impregnation and desolvation, the active component of Ni can be nanocrystallized, the active area of hydrogen electrode reaction can be increased, and the activity and stability of hydrogen electrode can be improved
[4]. Neagu et al. prepared a hydrogen electrode with Ni nanoparticles removed in situ by reducing La
0.52Sr0
.28Ni
0.06Ti
0.94O
3 electrode materials for 20 H at 930 ℃ in a 5%H
2/95%Ar reducing atmosphere, and treated it for 4 H in a 20%CH
4/80%H
2 at 800 ℃. The detection results showed that there was almost no carbon deposition on Ni particles with particle size of 25 ~ 60 nm, showing good coking resistance
[147].
4.3.2 Optimization Design of Electrolytic Cell Structure
During the long-term operation of the traditional SOEC, the interface between the oxygen electrode and the electrolyte is prone to delamination, and the activity of the oxygen electrode decreases, resulting in a sharp decline in the performance of the SOEC. This is mainly caused by two reasons: on the one hand, oxygen is generated at the oxygen electrode-electrolyte interface, and under the long-term operation of large current, the local oxygen partial pressure at the interface is easy to be too high, which destroys the interface structure; On the other hand, because the thermal expansion coefficient of oxygen electrode and electrolyte material does not match at high temperature, it is easy to delaminate in long-term operation.
It is an effective way to solve the problem of interface delamination by optimizing the preparation process to realize the structural integration of electrolyte/oxygen electrode. Wu et al. prepared an electrolyte/oxygen electrode integrated YSZ support skeleton with a microchannel structure by a freeze-drying method, and loaded the oxygen electrode active material LSC on the inner wall of the microchannel by an infiltration method to prepare a composite oxygen electrode with high catalytic activity
[148]. The polarization impedance of this new LSC-YSZ composite electrode is 0.0094Ω·cm
2 at 800 ℃, and the polarization impedance is basically unchanged after applying a large current of 2 A/cm
2 for 6 H, which verifies the good stability of the new oxygen electrode. The addition of a contact layer at the electrolyte/oxygen electrode interface is another important means to solve the delamination problem. Kim et al. added a layer of Ce
0.43Zr
0.43Gd
0.1Y
0.04O
2-δ as a contact layer between the oxygen electrode and the electrolyte, and the ohmic resistance of the SOEC only increased by ~0.02Ω·cm
2 after operating at 800 ℃ for 100 H, and its performance decay rate was significantly lower than that of the SOEC without a contact layer
[149].
4.3.3 Development Status of SOEC Reactor
It is generally believed that SOEC is still in the stage of laboratory research and development, but from a technical point of view, it has the potential to achieve industrialization. From the single cell level, the initial performance of electrolyzed water has increased by about 2.5 times in the past 15 years, and the surface specific resistance (ASR) has decreased from 0.71Ω/cm
2 to 0.27Ω/cm
2. Durability is also greatly improved, with the decay rate reduced from 40%/1000 H to 0.4%/1000 H (Figure 14b). From the cell stack level, the long cycle test of SOEC is increasing (Fig. 14C), and the decay rate has been reduced to less than 1%/1000 H (Fig. 14d); Stability has also been demonstrated, with ceramic-based SOECs withstanding 150 cold/hot cycles and metal-based SOECs more than 2,500 cold/hot cycles
[4]. The main research institutes abroad include Japan Atomic Energy Research Institute, Mitsubishi Heavy Industries, Toshiba and Kyocera. Idaho National Laboratory, Bloom Energy, USA; Topsoe Fuel Cell Company, Denmark; Korea Energy Research Institute, etc. Idaho Laboratory in the United States has made great progress in SOEC stack and hydrogen production system in recent years. In 2004, the results of hydrogen production by single cell were reported for the first time. In 2006, a stack with hydrogen production rate > 90 L/H was prepared. In 2012, a 15 kW high-temperature electrolytic hydrogen production system was built. In 2018, a 25 kW stack and a 250 kW high-temperature electrolytic hydrogen production test system were built
[150][151][152][153]. In 2017, Julich, Germany, constructed two stacks with four single cells based on two air electrodes of La
0.6Sr
0.4CoO
3-δ(LSC) and two oxygen electrodes of La
0.6S
Co
0.2Fe
0.8O
3-δ(LSCF), and carried out 1000 H electrolysis test under the electrolysis conditions of 800 ℃ and -0.5 A/cm
2 to analyze the reasons for the electrochemical performance and long-term operation performance degradation of the stack
[154]. In China, Shanghai Silicate Research Institute of Chinese Academy of Sciences, University of Science and Technology of China, China University of Mining and Technology, Nanjing University of Technology, South China University of Technology and Tsinghua University have made great progress in the research of SOEC
[155][156][157][158][159][160]. The Institute of Nuclear Energy and New Energy Technology of Tsinghua University has made a series of progress in the research and development of key materials, such as the new nano-micro structure design of SOEC oxygen electrode, the precise regulation of active sites on the surface and interface of materials, and the design and preparation of high-efficiency energy materials
[148,159][161][162]. The SOEC stack (as shown in Fig. 15) was independently developed, and two continuous operation experiments were successfully carried out on the hydrogen production system. The hydrogen production was operated for 100 hours, the hydrogen production rate was 105 L/H, and the single electrolysis voltage was less than 1.3 V
[62]. It solves the problems of stable supply of water vapor, precise control and line diagnosis in high temperature hydrogen production system
[163]. The development of SOEC in China has gradually entered the stage of demonstration project.
Currently, it is believed that the two biggest problems facing SOECs are that ceramic-based SOECs are not suitable for operation at high voltages and that SOECs are not suitable for operation at dynamic power outputs. However, studies have shown that the maximum operating pressure of SOEC stacks can be as high as 25 bar
[4]. Operational studies under dynamic power output show that the SOEC can be ramped up from zero to 80% load in milliseconds without compromising battery performance
[164]; Another study showed that the performance of SOEC was not significantly degraded when it was operated for 1200 H in an environment where the current was changed every 5 min
[165]. Based on the progress of electrolytic cell monomer and stack, Mogensen and others believe that SOEC technology has been basically mature, there is no fundamental technical obstacle to the large-scale application of SOEC stack, and the unreliability of the matching system components (compressor, heater, etc.) is the main reason for the unexpected shutdown of SOEC system
[4].
Electrolysis efficiency and hydrogen production rate are important parameters to measure the performance of an electrolytic device, and the higher electrolytic current density at a lower potential proves that the electrolytic performance of the electrolytic device is better. Fig. 14a shows the current density as a function of electrolysis potential during AEC, PEMEC, and SOEC electrolysis
[166]. It can be seen from Figure 14 a that the current density of AEC at high electrolysis potential (> 1.5 V) is still very low (<0.5 A/cm
2), indicating that its electrolysis efficiency and hydrogen production rate are both low
[167]; At higher potential (> 1.5 V), PEMEC can obtain higher current density, up to 2 A/cm
2, and higher hydrogen production rate, but its electrolysis potential is higher, which indicates that the total impedance of the electrolysis device is higher, and the efficiency of power utilization needs to be improved
[168]. SOEC has the advantages of high hydrogen production rate and high electrolysis efficiency.When the electrolysis voltage is lower than 1. 5 V, the highest current density can reach the 2 A/cm
2[169]. Under the same electrolysis voltage, the greater the current density, the greater the hydrogen yield, the less the number of electrolysis equipment required to produce the same yield of hydrogen, and the lower the fixed investment cost (assuming that the unit price of different electrolysis equipment is basically the same); Under the same electrolysis current, the smaller the voltage is, the less the electricity consumption is, and the lower the equipment operation cost is. SOEC has potential advantages in reducing investment costs and operating costs.
In the future, the proportion of renewable energy in the energy system will continue to increase, and the role of electrolytic energy storage and energy conversion technology will become increasingly important. PEMEC and SOEC are two promising technologies for electrolytic energy storage and conversion. Although PEMEC has entered the stage of industrialization, its cost is still high, and the high cost of precious metal catalysts is the key to restrict its cost reduction. The development of low noble metal supported catalysts with high activity is an important development direction to promote PEMEC to expand production scale and reduce costs. SOEC has the potential of industrialization, but its main obstacle is that it does not have an advantage in cost. With the further maturity of technology, the continuous expansion of industrial scale and the further reduction of cost, SOEC is expected to become an important way of large-scale hydrogen production by water electrolysis in the future.
5 Other new technologies for hydrogen production
In addition to water electrolysis, other new hydrogen production technologies based on clean energy, such as water photolysis by solar energy, hydrogen production by biomass fermentation, hydrogen production by biomass thermochemical conversion and hydrogen production by nuclear thermochemical cycle, have also been widely concerned and studied.It provides more choices for improving the utilization efficiency of clean energy.
5.1 Hydrogen production by solar energy photolysis
Since Fujishima and Honda first reported the realization of hydrogen production by photolysis of water using n-type TiO
2 single crystal electrode in 1972, the technology of direct hydrogen production by photolysis of water using semiconductor catalysis has attracted extensive research
[170].
Photolysis of water to produce hydrogen is actually based on the photoelectric conversion effect of semiconductors. The principle is shown in Figure 16. The photolysis process can be divided into three steps: (1) The semiconductor is excited under illumination
[171]. Semiconductors absorb photons whose energy is equal to or greater than their band gap. Electrons in the valence band jump to the conduction band after being excited. Free electrons appear in the conduction band, leaving holes in the valence band. Electron-hole pairs are called photogenerated carriers. (2) The recombination and migration of photogenerated carriers. Due to thermal vibration and other reasons, most of the photogenerated electrons will jump back to the valence band, electron-hole pairs will be recombined, and the remaining small number of photogenerated carriers will migrate to the semiconductor surface. (3) Surface reaction. Some of the carriers reaching the surface will recombine on the surface, and the rest will be captured by water molecules adsorbed on the semiconductor surface, thus causing the decomposition of water molecules.
All semiconductors can complete the above steps (1) and (2), but only some semiconductors can complete step (3), which mainly depends on whether the reduction/oxidation ability of electrons/holes can meet the requirements of water decomposition. The theoretical decomposition voltage of water is 1.23 V under the standard condition. Due to the influence of overpotential, the band gap of semiconductor should be larger than the theoretical decomposition voltage of water. The position of the bottom of the conduction band is higher (negative) than the electrode potential of hydrogen, and the position of the top of the valence band is lower (positive) than the electrode potential of oxygen
[172]. Fig. 17 lists the band structures of some semiconductors and compares them with the hydrogen and oxygen electrode potentials. It is found that only some semiconductors meet the water splitting conditions
[172].
TiO
2 is a good photocatalytic material, but it has a large band gap (3.2 eV) and can only absorb the ultraviolet part of sunlight, so its utilization efficiency for sunlight is low. In order to improve the utilization efficiency of sunlight, on the one hand, the existing materials can be modified to make the absorption band red shift to the visible region; On the other hand, new materials that respond in the visible region can be developed. Khan et al. prepared a chemically modified n-type TiO
2 by pyrolyzing Ti metal with natural gas flame. Part of the lattice oxygen atoms in the modified material were replaced by carbon atoms, which reduced the band gap of the TiO
2 and increased the maximum optical conversion efficiency from 1% to 8.35%
[173]. In addition, materials such as metal oxides (ZnO, etc.), metal sulfides (CdS, etc.), metal complexes (Ru [Bopr
, etc.]), and titanates (PbTi
) have shown good photoresponsiveness in the visible region and have attracted more attention
[174][175][176]. Anion doping (S, N, P, Br, etc.) Can effectively reduce the overpotential and improve the separation and transfer of photogenerated carriers, which is an important means to improve the activity of catalysts
[172].
The serious electron-hole recombination is another major factor affecting the low optical conversion efficiency. In order to reduce the recombination of electrons and holes, scientists introduced the Z-reaction mechanism of plant photosynthesis into the photocatalytic decomposition hydrogen production system, and constructed a Z-photocatalyst with heterostructure
[179]. Different components of the heterostructure have different potentials, and some components show good synergistic effect on the utilization of visible light, which can improve the separation and transmission of photogenerated carriers
[172]. The composition principle of the Z-type catalyst is shown in Figure 18. The whole catalytic system is like the English letter "Z", consisting of a p-type semiconductor and an n-type semiconductor, and the conduction band bottom of the n-type semiconductor (positive electrode catalyst) is greater than or equal to the valence band top of the p-type semiconductor (negative electrode catalyst).Both semiconductors can absorb sunlight and excite their respective electrons from the valence band to the conduction band, and the two catalysts are connected by a suitable electron transport mechanism to achieve electron transport between the two catalysts
[178][180]. In this way, the utilization efficiency of sunlight can be effectively improved, and electrons with strong reducing ability are generated in one catalyst, and holes with strong oxidizing ability are generated in the other catalyst.
Yang et al. designed a graphene-carbon nitride-graphene "sandwich" structure composed of modified graphene and two-dimensional carbon nitrogen materials, which can realize the integration of hydrogen preparation and storage
[181]. The system can absorb ultraviolet light and visible light at the same time, and has high light utilization efficiency. Under the irradiation of sunlight, the system generates excitons, which quickly separate into high-energy electrons and holes, and migrate to the carbon nitride layer and the graphene layer, respectively. Water molecules adsorbed on the outer surface of graphene are cracked under the action of holes to generate hydrogen protons and oxygen at the same time, and the hydrogen protons pass through the graphene layer to enter the carbon and nitrogen compound layer and combine with electrons to generate hydrogen. Because of the characteristics of graphene, the hydrogen generated inside can not diffuse to the outside, while the oxygen and water from the outside can not diffuse to the inside, thus achieving the safe storage of hydrogen.
It is generally believed that the technology of hydrogen production by solar photolysis has industrial application value only when the light conversion rate reaches more than 10%, and the current system of hydrogen production by solar photolysis is still difficult to meet this requirement
[182].
5.2 Hydrogen production by biomass fermentation
Biomass fermentative hydrogen production is to convert the energy stored in natural organic matter into hydrogen energy through hydrogen-producing bacteria and other microorganisms. Biological hydrogen production is the result of microbial metabolism
[183,184]. According to the differences of microorganisms, hydrogen-containing substrates and hydrogen production mechanisms, the hydrogen production methods can be divided into three categories: (1) photolysis of water to produce hydrogen
[185]. Bacteria such as green algae and cyanobacteria decompose water under light and anaerobic conditions to produce hydrogen; (2) hydrogen production by photofermentation. Photosynthetic bacteria decompose organic matters to generate hydrogen under the conditions of illumination and anaerobic; (3) dark fermentation hydrogen production. Anaerobic fermentation bacteria decompose organic matter to produce hydrogen under dark, anaerobic conditions.
Compared with the other two methods, dark fermentation has obvious advantages. Dark fermentation does not require light, and anaerobic fermentation bacteria can decompose almost all organic matter into small molecular organic acids and produce hydrogen at the same time. However, in the absence of external energy input, the reaction can only proceed in the direction of decreasing free energy, and other organic acids are difficult to be decomposed by anaerobic bacteria except formic acid, so the hydrogen production rate of dark fermentation is low
[186]. Photosynthetic bacteria can use solar energy to decompose various small molecular organic acids into hydrogen and CO
2, but they can not directly use starch, cellulose and other complex organic matter
[186]. Therefore, the combination of dark fermentation and photo fermentation can combine the advantages of the two methods and effectively improve the hydrogen production efficiency. Compared with dark fermentation, the combination of photo fermentation and dark fermentation can increase the hydrogen production by two times
[186].
Microbial electrolytic cell (MEC) is a new type of biological hydrogen production technology, which can produce hydrogen by using organic matter in wastewater. The MEC uses a proton exchange membrane as an electrolyte diaphragm, the anode contains electrogenic bacteria, and the electrogenic bacteria decompose organic matters in water and release electrons at the same time; The MEC cathode maintains an anaerobic environment, where hydrogen protons combine with electrons to produce hydrogen
[186]. If a strain (such as Spirulina platensis) capable of producing hydrogen by fermentation is used at the cathode, hydrogen can be produced at both the cathode and the anode of the MEC, and the hydrogen production device at this time is called an electrochemical photobioreactor (EPBR). Both anode and cathode of EPBR can produce hydrogen. The anode is produced by microbial fermentation (biohydrogen), while the cathode is produced by the combination of protons and electrons (electrochemical hydrogen). Its hydrogen production efficiency is higher than that of traditional MEC, which is a more efficient way of hydrogen production
[187].
China's bio-fermentation hydrogen production technology is in the forefront of international development. The continuous flow hydrogen production technology developed by Ren Nanqi of Harbin University of Technology is in the leading international level, and the hydrogen production rate of this technology can reach 1.93 mol H
2/mol glucose
[188]. Based on this technology, the demonstration project of hydrogen production by biological fermentation was started in 2002, and the first phase of the project was completed in 2005, with an annual hydrogen production capacity of 12×10
4Nm
3. The project has successfully carried out linkage research with the fuel cell development group of Tsinghua University, laying a foundation for the application of distributed energy. However, at present, there is a lack of research on new technologies of biological hydrogen production in China, such as MEC and EPBR. In the future development of biological hydrogen production technology in China, on the one hand, it is necessary to continue to explore and cultivate new strains with high hydrogen production rate to improve the hydrogen production capacity of biomass, and at the same time, it is necessary to devote appropriate energy to fill the gaps in frontier technology to avoid the problem of "sticking neck". On the other hand, relying on the existing mature technology, industrialization and commercialization are carried out to prepare for the large-scale application of biomass.
5.3 Thermochemical conversion of biomass for hydrogen production
Biomass thermochemical conversion processes can be divided into steam gasification, supercritical water gasification and rapid pyrolysis, and only steam gasification process has the prospect of large-scale application
[189].
Hydrogen production by biomass steam gasification uses biomass as raw material, oxygen (air), steam or hydrogen as gasifying agent, and converts the combustible part of biomass into combustible gas such as hydrogen through thermochemical reaction at high temperature. The process of hydrogen production takes place in the biomass gasifier. The thermochemical reaction mainly includes the thermal decomposition reaction of biomass, the oxidation reaction of biomass carbon and oxygen, and the reduction reaction of carbon, carbon dioxide and water. The main effective components of the gas produced by gasification are H2, CO, CH4, CO2, etc. Hydrogen can be obtained after separation and purification.
The technology of hydrogen production from biomass steam gasification is relatively mature, but it is difficult to collect raw materials, and the logistics cost accounts for 50% -70% of the production cost of biomass energy products, which is not dominant in economic cost
[190]. With the progress of technology in the future, the efficiency of biomass collection, storage and transportation is expected to improve, the cost is expected to decrease, and the hydrogen production from biomass thermochemical conversion has the possibility of large-scale development.
5.4 Thermochemical cycle hydrogen production
Thermochemical cycle hydrogen production is one of the nuclear hydrogen production technologies. The fourth generation nuclear energy system has the characteristics of safety, economy and sustainability, and its outlet temperature can reach up to 950 ℃, which can be used as an efficient heat source for hydrogen production processes such as SMR, high temperature electrolysis and thermochemical cycle
[191]. The development of China's fourth-generation high-temperature gas-cooled reactor (HTGR) with nuclear energy characteristics ranks among the world's advanced level, laying a solid foundation for the development of hydrogen production by thermochemical cycle in China.
The direct thermal decomposition temperature of water is as high as 2000 ℃, which is difficult to achieve under ordinary conditions
[191]. However, if two or more thermally driven chemical reactions are used to form a closed cycle through a thermochemical cycle process, and then the decomposition of water at a lower temperature is realized, the required heat source temperature can be reduced to 800 to 900 deg C. The thermal efficiency of thermochemical cycle is similar to that of Carnot cycle, and high temperature can improve the conversion efficiency.
In the 1970s, the Ispra Joint Research Center in Italy carried out research on hydrogen production by thermochemical cycle, and Marcheltti and Beni proposed the Mark1 cycle based on Ca, Br and Hg elements
[191]. Subsequently, iodine-sulfur cycle, hybrid sulfur cycle, bromine-sulfur mixed cycle, metal oxide cycle and other processes appeared.
In 1998, the nuclear hydrogen production program was launched in the United States, and all the published thermochemical cycle processes were evaluated and screened. The evaluation results showed that the iodine-sulfur (IS) cycle invented by General Atomic Energy Corporation (GA) of the United States and the UT-3 cycle invented by Tokyo University of Japan were the best, and the I-S cycle was selected for research and development
[191]. Since then, Japan, France and China have chosen the I-S cycle process for research.
The iodine-sulfur circulation process is divided into three steps of reactions:
[192]② Sulfuric acid decomposition reaction
③ Hydroiodic acid decomposition reaction
The reaction of sulfuric acid decomposition takes place above 750 ℃, which is a strong endothermic reaction, and the required high temperature can be provided by HTGR. The whole process is a closed cycle. The raw material is only water, and other materials are recycled. It can be operated continuously, and the expected efficiency can reach ~ 38%
[191].
Since the 1970s, hydrogen production by iodine-sulfur cycle has been widely valued by the United States, Japan, France, the European Union, South Korea, Canada, China and other countries and organizations. In 2004, Japan successfully built a cycle bench with a hydrogen production scale of 50 L/H, and successfully carried out closed cycle. In cooperation with GA, Sandia National Laboratory and the French Atomic Energy Commission, a small I-S cycle bench was built in 2009 and experiments were carried out. A loop with a hydrogen production rate of 50 L/H has been built in Korea, and a closed cycle experiment is currently under way. The Institute of Nuclear Energy Technology of Tsinghua University in China has carried out systematic research on the chemical reaction and separation process of the I-S cycle, and developed a high-performance sulfuric acid and hydroiodic acid decomposition catalyst that can be used in high temperature and strong corrosive environment, with no significant performance degradation in the 100 H life test experiment.A hydrogen production capacity of 100 L/H has been built in the laboratory, and the continuous and stable operation for 60 H has been successfully realized under the hydrogen production rate of 60 L/H, which proves the technical reliability of hydrogen production by I-S cycle
[62,193,194].
The proportion of clean energy in primary energy will be higher and higher. In order to solve the problems of randomness and temporal and spatial differences of renewable energy, it is an inevitable choice to use water electrolysis hydrogen production system to achieve energy storage and peak shaving, and large-scale electrolysis hydrogen production technology will have broad prospects for development in the future. With the development of the fourth generation of HTGR, SOEC high-temperature electrolysis hydrogen production and I-S cycle hydrogen production process will have great potential for development. However, from the current industrial development status and economic point of view, AEC electrolytic hydrogen production is still the mainstream way of new energy hydrogen production.
6 Comparison of different hydrogen production method
Table 9 compares 14 hydrogen production methods in terms of cost, environmental benefits, and technology maturity. From the point of view of technology maturity, coal gasification hydrogen production, SMR, industrial by-product hydrogen and AEC are the most mature hydrogen production processes, which have already achieved large-scale commercial application. Hydrogen production from water electrolysis and other new hydrogen production technologies are new clean hydrogen production methods developed in the face of the future clean energy system dominated by renewable energy, and the technology maturity is low.
From the cost point of view, the traditional fossil energy hydrogen production route has obvious advantages in the more mature technology. The cost of coal gasification hydrogen production is the lowest, the lowest is 6 yuan/kg, and the cost of coal gasification + CCS, SMR, industrial by-product hydrogen and other technologies is less than 20 yuan/kg. The cost of hydrogen production by AEC water electrolysis is the highest, and the cost is about 45 yuan/kg when the grid electricity price is 0.6 yuan/kWh. Among the emerging technologies, the cost of hydrogen production by coal supercritical water gasification is relatively low, and the research results of Guo Liejin and others for many years show that the cost of hydrogen production by this technology is about ~ 8 yuan/kg. The overall cost of renewable energy water electrolysis hydrogen production technology is relatively high, which is mainly affected by electricity price and electrolytic cell cost, and the cost of hydrogen production by abandoning electricity can be reduced to ~ 10 yuan/kg. According to calculation, when the price of filling hydrogen is reduced to less than 40 yuan/kg, the cost is equivalent to the cost of gasoline, and hydrogen fuel vehicles will no longer be at a disadvantage in terms of fuel price. Because the cost of hydrogen production accounts for about 30% -50% of the total cost of filling hydrogen, the cost of hydrogen production is at least less than 20 yuan/kg to meet the needs of large-scale development of hydrogen fuel cell vehicles
[195]. At present, traditional fossil fuel hydrogen production and industrial by-product hydrogen can basically meet the requirements. Hydrogen production from renewable energy has an advantage in economy, but the maturity of electrolytic cell technology is low, which makes it impossible to achieve large-scale electrolytic hydrogen supply.
From the perspective of environmental benefits, the CO2 emission of coal gasification hydrogen production is the largest, which can be as high as 25 kg CO2/kg H2, and is the most harmful to the environment. Coal gasification combined with CCS technology can reduce carbon emissions to 2~7 kg CO2/kg H2, which is an effective way to solve the problem of carbon emissions from hydrogen production from coal. The technology of supercritical water gasification of coal realizes the recovery of CO2 and other gases in the form of inorganic salts, achieves near-zero emission of greenhouse gases, and the technology has entered the pilot stage, which is very in line with the development direction of China's resource endowment. Carbon emissions from SMR technologies are also relatively high (8~16 kg CO2/kg H2). Industrial by-product hydrogen is the recovery and utilization of industrial purge gas, and there is no carbon emission problem in essence. Hydrogen production by AEC electrolysis itself does not have the problem of carbon emissions, however, the power grid in China is mainly thermal power, and the carbon emissions caused by power generation are high, which makes the overall carbon emissions of hydrogen production by electrolysis as high as ~45 kg CO2/kg H2. Clean energy water electrolysis and other new hydrogen production processes have obvious advantages in reducing carbon emissions.
It is predicted that in 2050, China's hydrogen demand will be close to 60 million tons, and the proportion of hydrogen energy in the energy system will increase to 10%, of which nearly half will be used in the field of transportation
[6]. Considering the cost, technology maturity, carbon emissions and other factors, the future hydrogen production route and hydrogen energy development path in China can be divided into three stages: short-term (~ 2025), medium-term (~ 2030) and long-term (after 2050).
The hydrogen fuel cell vehicle industry is in the transition stage from the demonstration stage to the initial commercial development. The demand for hydrogen as a vehicle fuel is small, and most of the hydrogen is mainly used as raw materials for petroleum refining, synthetic ammonia, synthetic methanol and other chemical industries. At this stage, the main route of hydrogen production is coal hydrogen production, which provides sufficient raw materials for the chemical industry. Industrial by-product hydrogen and small-scale water electrolysis hydrogen production can meet the demand for vehicle fuel. The development of hydrogen energy industry is mainly centered on provinces with abundant hydrogen sources (abundant fossil fuels, industrial by-product gas or renewable energy) or rapid development of fuel cell technology, exploring the form of hydrogen energy development, and forming a pattern of self-production and self-consumption of hydrogen energy in the province.
The medium term is the transitional stage of the development of China's hydrogen energy industry. In 2030, China's demand for hydrogen energy will reach 35 million tons, accounting for 5% of the energy system. In 2035, the number of hydrogen fuel cell vehicles in China will reach 1.3 million, and the demand for hydrogen energy in the field of transportation will increase significantly. With the reduction of CCS cost and the industrialization of coal supercritical water gasification process, clean coal hydrogen production will replace the traditional coal hydrogen production process. The amount of industrial by-product hydrogen in China is large, but the utilization rate is low at present. It is estimated that the industrial by-product hydrogen discharged in China in 2022 will be 135,000 tons. Assuming that a fuel cell vehicle runs 100 km per day and the hydrogen consumption is 1 ~ 1.25 kg/100 km, it can be used by 330,000 fuel cell vehicles
[196]. The full utilization of by-product hydrogen can provide a large amount of high-quality hydrogen for the medium-term development of hydrogen energy industry in China. The cost of hydrogen production from water electrolysis by renewable energy has met the commercial requirements, and it has gradually become an effective hydrogen source donor. The three low-carbon hydrogen production routes are the main donors of hydrogen. The development of hydrogen energy industry will also expand from the initial core provinces to the surrounding provinces, and build the industrial chain of multi-province coordinated development.For example, the western Guangdong line in the "one ring and three lines" hydrogen energy development plan of Guangdong Province is to use Yunnan's abundant water resources to develop renewable energy to electrolyze water to produce hydrogen, and to provide hydrogen sources for Guangdong Province through long-distance transportation
[5]. This form of inter-provincial coordinated development, on the one hand, can promote the development of hydrogen energy industry in both places at the same time, on the other hand, it can also accumulate the experience of long-distance transportation of hydrogen.
After 2050, China's hydrogen energy industry will be fully developed, and the primary energy structure will be transformed from fossil energy to renewable energy. Hydrogen production from water electrolysis by renewable energy will become the main hydrogen production route, and hydrogen production from solar energy, biomass and nuclear energy is expected to become an effective means of hydrogen supplement. Hydrogen energy industry will be laid out nationwide and become an indispensable part of the energy system.
Hydrogen production from fossil energy is a realistic resource choice in the early stage of the development of China's hydrogen energy industry, and its high carbon emissions are not in line with the concept of "carbon neutral" development in the future. The clean hydrogen production process from fossil energy can realize the low-carbon utilization of high-carbon energy, alleviate the problem of carbon emissions to a certain extent, and is an important transitional form to promote the development of energy system from high carbon emissions to zero carbon emissions in China. Clean energy water electrolysis and other clean energy hydrogen production processes can improve the utilization rate of clean energy, increase its total proportion in the energy system, and promote the low-carbon transformation of the energy system, which is an inevitable choice to achieve the goal of "carbon peak" and "carbon neutrality".
7 Application of Artificial Intelligence in Hydrogen Production System
As mentioned above, there are many ways to produce hydrogen, and the various hydrogen production technologies are very different, and the energy sources are also different. How to select the appropriate hydrogen production path according to local conditions is an important research topic for the future development of hydrogen energy
[199]. While promoting the development of "hardware" system of hydrogen production process, the development of "software" system such as prediction, evaluation and optimization of hydrogen production system is also very important. China is rich in renewable energy, and hydrogen production from renewable energy is expected to become the mainstream hydrogen production route in the future. However, China has a vast territory and great differences in ecological environment, and the same hydrogen production method may have great differences in technical implementation paths in different environments, which increases the complexity of hydrogen production system construction in different regions of China. At the same time, the complete hydrogen production system itself is very complex, as shown in Figure 19. Taking the wind power hydrogen production system as an example, the whole system includes not only the electrolytic hydrogen production device itself, but also the front-end energy supply system and the terminal hydrogen application system. Accordingly, there are many parameters involved in the process of hydrogen production, such as the output voltage of the front-end energy supply system, the pressure, liquid level and temperature of the electrolytic cell, the output and purity of hydrogen at the end, and the purity of oxygen, all of which need to be reasonably controlled, and many of these parameters are interrelated, which further increases the difficulty of controlling the parameters of the hydrogen production system
[200]. The problem of evaluating the feasibility and applicability of a method in a particular environment can be solved by using traditional mathematical modeling methods. However, for the complex system such as hydrogen production system, the establishment of mathematical model inevitably requires complex mathematical equations, boundary conditions and reasonable assumptions to achieve, the workload is very huge, and the small differences in different environments may make the model complex changes, which makes the prediction, evaluation and optimization of hydrogen production system very difficult.
The rapid development of artificial intelligence and machine learning technology provides a new solution to this problem. Computational intelligence, an important form of artificial intelligence, has obvious advantages in solving such problems. The most important feature of computational intelligence is that it does not need to establish a precise mathematical or logical model of the problem itself, but has excellent learning ability for a large number of data sets embedded in the input target, and can model the target variables by learning and identifying complex behaviors hidden in the data
[201,202]. At present, the modeling algorithm of computational intelligence has been developed into many forms. The commonly used methods in hydrogen production include artificial neural network (ANN), fuzzy Delphi method (F-D), fuzzy analytic hierarchy process (FAHP), multi-layer perception method (MLP), adaptive neuro-fuzzy inference system (ANFIS), genetic algorithm (GA) and so on
[202]. Based on the actual situation of Taiwan's energy resources, Chang and others used F-D method to evaluate and analyze 14 indicators of SMR, CCS + SMR, conventional grid water electrolysis hydrogen production, photovoltaic power generation water electrolysis hydrogen production, wind power water electrolysis H yrogen production, biomass gasification hydrogen production and biomass fermentation hydrogen production from four aspects of technology, environmental protection, economy and society.The results show that water electrolysis by photovoltaic power generation and water electrolysis by wind power are two hydrogen production technologies suitable for future development in Taiwan
[199]. Yilmaz used genetic algorithm to carry out thermal economic optimization and economic analysis of hydrogen production system by alkaline water electrolysis with geothermal energy
[203]. The results show that the electrolysis device, turbine and heat exchanger are the main cost sources in the system, and the cost of hydrogen production can be reduced by reducing the number of equipment. Developing multiple intelligent systems based on computational intelligence that can accurately simulate and predict multiple hydrogen production pathways, and designing accurate and powerful intelligent platforms for unsupervised autonomous learning of hydrogen production data are important development directions for the application of artificial intelligence in hydrogen production technology in the future
[202].
The richness of China's energy structure, the diversity of hydrogen production routes and the complexity of hydrogen production systems make it very difficult to select and optimize hydrogen production routes. Some foreign researches have shown that computational intelligence has unique advantages in product prediction, path selection and evaluation, and performance optimization of hydrogen production system. With the breakthrough of 5G technology and the rapid development of commercialization, the response speed of artificial intelligence will be faster, the application mode will be more intelligent, and the application of hydrogen energy system will be more promising. However, there are few reports on the research of hydrogen production system using computational intelligence in China. China's hydrogen energy industry is entering a critical period of development, and a complete industrial chain is taking shape. At this time, it is very important to evaluate and optimize the hydrogen production system and even the hydrogen energy industry chain. While developing hydrogen production technology, we should recognize the importance of artificial intelligence in the application of hydrogen production technology, invest in corresponding research, make up for shortcomings, and develop software and hardware together, so as to better promote the development of hydrogen energy industry in China.
8 Conclusion and prospect
China's coal-based energy endowment will not change in the short term.In order to achieve the goal of "carbon peak" in 2030, the clean utilization of coal is an inevitable way for future development, and the coal supercritical water gasification process and coal-to-hydrogen + CCS technology are the important development directions of clean utilization of coal in the future, which are expected to provide sufficient low-carbon hydrogen sources for China's hydrogen energy industry in the near and medium term. The amount of industrial by-product hydrogen in China is huge, and the utilization rate is insufficient, which is expected to provide a relatively clean hydrogen source for China in the near and medium term. Due to the lack of natural gas resources in China, large-scale hydrogen production from natural gas nationwide does not have advantages, and some areas with abundant natural gas resources are more likely to achieve small-scale, regional SMR + CCS clean hydrogen production in the near and medium term. In addition, hydrogen production from methane cracking is a zero-carbon-emission natural gas hydrogen production process that has attracted much attention, and the development of low-cost, high-activity and long-life catalysts is the key to accelerate the industrialization of this technology and promote "carbon neutralization".
In the long run, it is an inevitable requirement to achieve the goal of "carbon neutrality" by transforming China's primary energy donors into clean energy based on renewable energy. Large-scale renewable energy hydrogen storage by water electrolysis is an important way to solve the problem of "three abandonment" of renewable energy and promote "carbon neutralization". PEMEC has fast start-up, fast fluctuation response, high SOEC energy conversion rate, and obvious advantages in small-scale distributed energy storage and large-scale centralized energy storage, both of which are ideal electrolytic cells for hydrogen production coupled with renewable energy generation systems.
In order to achieve the goal of large-scale, low-cost and environmentally friendly hydrogen production, technological breakthroughs must be made in the field of PEMEC and SOEC. For PEMEC, reducing the catalyst loading or developing non-noble metal catalysts with good performance, increasing the research and development of battery materials, and realizing the localization of key components as soon as possible are effective ways to reduce its cost and realize large-scale hydrogen production applications. For SOEC, by improving the material composition and preparation process, the development of electrolytic cells with excellent electrolytic performance and high temperature stability can increase the service life. At the same time, the development of other key equipment (such as compressors, heaters, etc.) With good stability for SOEC stack is also the key to realize its coupling with the power generation system and long-term operation.
In addition, other new hydrogen production technologies, such as solar water photolysis, I-S cycle and biomass hydrogen production, provide diversified options for clean energy hydrogen production. R & D investment in new technologies and new directions can make technical reserves for hydrogen production in China, and improve the freedom of route selection and fault tolerance rate in China.
While increasing the breakthrough of "hardware" technology in hydrogen production process, attention should also be paid to the development of "software" system for prediction, evaluation and optimization of hydrogen production system.Recognizing the importance of artificial intelligence in the development of hydrogen production system under the rapid development of "5G", developing artificial intelligence algorithms for hydrogen production system adapted to China's national conditions will inject new strength into the development of hydrogen energy industry in China.
the National Natural Science Foundation of China(91645126)
the National Natural Science Foundation of China(21273128)
the Tsinghua University Initiative Scientific Research Program(2018Z05JZY010)
the Tsinghua-MIT-Cambridge Low Carbon Energy University Alliance Seed Fund Program(201LC004)
PDF
