Nuclear Hydrogen Production

In the next decade, the hydrogen demand is expected to rise with a factor of four. This rise is independent of any future hydrogen economy that people expect to come, but is due to an increase in demand from the chemical industry, the fertilizer industry and the petrochemical or the oil-refinery industry. Hydrogen is added to crude oils to produce lighter fuels like gasoline (the so-called "cracking" process), to remove impurities such as sulfur, and to reduce its toxicity. Oils contain a number of carcinogenic compounds like benzene (C₆H₆), which can be removed by adding hydrogen. Whereas high-quality crude oil has a hydrogen-to-carbon ratio similar to that of gasoline, heavy oils have a ratio one half of that. Therefore large quantities of hydrogen are needed to make gasoline out of heavy crude oils. The use of economic non-fossil hydrogen reduces substantially the carbon dioxide emission and protects the chemical and refinery industry from high natural gas prices. Assuming a thermochemical hydrogen production process with efficiency of 50%, the energy required for the hydrogen production in the US would exceed the thermal output of all nuclear plants in the US. At present, over 99% of all hydrogen is being produced by steam reforming of natural gas and from lower value refinery streams.

Since the mid-19^th century, the world has been slowly shifting from one form of energy to another; this means from wood to coal to liquids to gases. The latter includes hydrogen. The driving forces behind these gradual shifts are twofold: increase of energy density at the one hand, and easiness of transport at the other. Oil is easier to transport then coal, and natural gas is easier to transport through pipelines then oil.

The move from solid to liquid to gas fuels involves another sort of transition: the less visible process of "decarbonization". By moving from wood to coal to oil to natural gas, the ratio of hydrogen to carbon in the molecule of each successive source has increased. Roughly speaking this ratio is 1-5 for wood, 1-2 for coal, 2-1 for oil, and 4-1 for natural gas. For hydrogen this ratio would be infinite.

The hydrogen, which is an energy carrier and not a primary energy source, has to be produced from other resources like natural gas (by means of steam reforming, for example), nuclear energy (by means of electrolysis or thermochemical processes) or others. Nuclear energy is expected to play a prominent role in the next coming century.

The electricity production from nuclear energy is steadily increasing with a 10-years annual average of 2.4%. For the last couple of years, nuclear energy has been the fastest growing primary energy resource worldwide. So nuclear energy is not a declining business at all. This strengthens our opinion that nuclear energy can continue to play a dominant role this century and that hydrogen production could become a major nuclear energy application with a share in the world’s energy consumption of several tens of percents!

Hydrogen can be produced in many different ways, using a wide range of technologies. Some of these involve mature industrial processes, like the steam-reforming process to convert natural gas into hydrogen and carbon dioxide; others are still in the laboratory. Electrolysis, which is splitting of water by using electricity, is a commercially available and well-proven technique. It produces high-purity hydrogen, but at a price that strongly depends on the electricity price. Steam reforming, which is splitting of hydrocarbon gases using heat and steam, is also well understood at large scale. Barriers are the large scales of the production process and the CO₂ emission. Gasification, which is splitting heavy hydrocarbons and biomass (but this is still under research) into hydrogen and gases for reforming, is also a well-understood process at large scale, but needs extensive cleaning to purify the hydrogen. Thermochemical cycles using cheap high-temperature heat from nuclear power or solar energy is potentially a massive production process at low cost, with high efficiency (>50%) and with no emission of greenhouse gases. Barrier, however, is the research needed to commercialize the processes. Think of corrosion resistant materials research and chemistry technology. Biological production of hydrogen by algae and bacteria is potentially a large resource. Barriers are the large area needed and the slow hydrogen production rates. Much research is needed in this field, and it is far from commercialization. At present, about 96% of all the hydrogen consumed nowadays is being produced from fossil fuels (reforming processes), and only 4% from electrolysis.

The processes for hydrogen production discussed in the remainder of this presentation are those for which water is the only material input, and hydrogen and oxygen are the only material outputs. Furthermore the required energy inputs are meant to be only heat (rather than useful work, such as electricity). These processes may be visualized as a set of chemical reactions that sum the decomposition of water. The ultimate aim, of course, is the discovery of a viable two-reaction process, which offers the potential for high-efficiency and low costs.

Hydrogen production via thermochemical processes is not a new development. By General Atomics, a literature survey was designed to locate all thermochemical water-splitting cycles that have been reported in the open literature. Interest in thermochemical water splitting has varied greatly with time. Already in the 1960's a project was carried out by the military, called the "Energy Depot", which had as an objective the production of a fuel from indigenous materials. The primary energy source was thought to be high-temperature heat from a portable nuclear reactor. Quickly after the military project, interest switched to civilian use. Interest boomed in the 1970's at the time of the oil crises, but faded with the onset of cheap oil and plentiful of natural gas. At present, most of the continuing work takes place in Japan, where dependence upon foreign energy sources continues to be of national concern.

The literature search by General Atomics turned up a number of cycles (about 115) far too large to analyze in depth. Therefore the researchers established some objective screening criteria to reduce the number of cycles to a manageable number.

The preliminary screening process consisted of applying the metrics to each thermochemical splitting process and summing the scores with equal weighting to get an overall score for each process. Of course not all metrics could be easily calculated and expert judgements were needed to fill the gaps. It turned out that the totals did not really cluster into a group of favorable and unfavorable processes.

Subsequently a second screening process was applied to reduce the number of cycles further. Three experts based on the most recent papers published made detailed investigations. Preliminary block-diagrams were designed to get a better understanding of each process, and thermodynamical calculations were made. Finally the three experts independently rated the viability of each cycle by assigning a score to it ranging from favorable (+1), acceptable (0), or unfavorable (-1). Two cycles stood out from all the others with a score of +3, namely the adiabatic version of the UT-3 process based on bromium and the sulfur-iodine process. The UT-3 process consists of four chemical reactions at temperatures varying from 200 till 800 °C. Essential work on this cycle has been performed in Japan and first publications date from the late 1970's. Over time the flowsheet has undergone several revisions and the most recent one based on the adiabatic implementation of the cycle was published in 1996. For a stand-alone plant, the efficiency of hydrogen generation is estimated to be 36 to 40%, while in case of cogeneration electricity and hydrogen efficiencies between 45 and 49% can be reached.

The sulfur-iodine process was first described in the mid-seventies. It is based on a sulfur (sulfuric acid) circulation and an iodine circulation. Both cycles are highly endothermic and need heat at temperatures of 900 and 400 °C, respectively. In the iodine circulation there is some uncertainty about the best way to accomplish the hydrogen-iodide decomposition step. Using this cycle, hydrogen can be produced with efficiency of 52%, which is significantly higher then the efficiencies reported for the UT-3 cycle. Furthermore, for the UT-3 cycle, a 10% efficiency increase was observed in case of cogeneration electricity and hydrogen. If this number would also be applicable to the S-I cycle, overall efficiencies could be as high as 60%. This is the reason that the S-I cycle attracts interest from research groups all over the world and that even the Japanese are considering the S-I cycle as a backup for the UT-3 cycle. The S-I process has been demonstrated in Japan, where a pilot plant has been built to confirm the flowsheet of the chemical process and the calculated thermal efficiency.

As mentioned before, the heat needed for the thermochemical water splitting processes is generated by nuclear power. At present around 440 nuclear power plants are operating worldwide, all but a few generating electricity (a few reactors are used for desalination purposes and others). Most reactors are located in the US (105), Europe, and the Far East. As I have mentioned before, if all the hydrogen needed in the US in ten years from now would be generated by nuclear, all reactors would have to generate hydrogen with 50% efficiency. So hydrogen production could give an enormous push to nuclear energy in the US and worldwide.

Light Water Reactors generate around 85% of all nuclear energy produced today. The group of light water reactors can be subdivided into Boiling Water Reactors and Pressurized Water Reactor. In the first-mentioned reactor type, water flows through the reactor core from the bottom to the top and is allowed to boil. The pressure is about 75 bars and the outlet temperature around 290 °C. Current reactors provide steam at saturation conditions. The steam generated is used directly in a set of high-pressure and low-pressure turbines to generate electricity. Current design BWRs are incapable of providing steam at the required high temperature. If the reactor is used to superheat the steam, however, it may be possible to provide thermal energy at 900 °C. BWRs employing nuclear superheat have been operated in the 1950's and 1960's with reactor steam outlet temperature of around 500 °C, however, at the expense of severe corrosion of fuel elements. Steam at a temperature of 900 °C would be highly corrosive, which practically eliminates the BWR from hydrogen production applications.

In the other type of light water reactors, so-called Pressurized Water Reactors, the pressure in the core is twice that in a boiling water reactor (150 bar). The water in the primary loop does not boil and a separate heat exchanger or steam generator is needed to transform the heat generated in the reactor core to usable steam. The reactor core outlet temperature is around 320 °C. In order to provide thermal energy at 900 °C, the water coolant must be in the supercritical state. This means the temperature of the water must be above 374 °C. Supercritical Pressurized Water-cooled Reactors have been proposed, however, only at a temperature of 500 °C and an operating pressure of 250 bars. The high pressure needed to achieve an outlet temperature of 900 °C eliminates in practice the PWR from hydrogen production applications.

Both these reactor types belong to the so-called generation II and III reactor types. The first generation nuclear power plants were prototypes of reactors. The reactors from the second generation are the workhorses of nuclear-based electricity production today. Third generation nuclear power plants have been successful in several countries. Advanced plants based on US technology have been build and are being constructed in Japan, South Korea, and Taiwan, and are expected to be selected by other countries in the coming years. Small but important enhancements to these plants lead to the so-called generation III+ designs that could help to make them state-of-the art and deployable in 2010. Except from these "near-term deployment" concepts, complete new reactor designs are being developed that could enter the market in 2030 and beyond. These plants, which are said to belong to the fourth generation of nuclear energy, should be highly economical (this means they will have a clear life cycle cost advantage over other energy sources). Furthermore, they will excel in safety, reliability and sustainability, they will minimize their waste production and they will be proliferation resistant.

The key parameter to economic and sustainable nuclear energy is the coolant outlet temperature and thereby the thermal efficiency of the reactor. Both the power costs, the waste production and other environmental parameters benefit from a rise in coolant temperature. Present-day nuclear reactors have an efficiency of 33%, while future reactors with high coolant outlet temperature could reach 45 to 50% efficiency.

Many applications need higher temperatures than can be produced by the current Light Water Reactors, and new concepts like the High Temperature Reactor or the Very High Temperature Reactor are needed. I will come later to these new reactor types.

Especially for the S-I hydrogen production process there has been an assessment of reactor types with the goal to identify the most promising ones. In this assessment, the reactor heat source is conceptually decoupled from the hydrogen production plant with a heat exchanger providing the interface between the two systems. This intermediate helium loop assures that any leakage from the reactor coolant loop will not contaminate the hydrogen production system, and that any corrosive process chemicals cannot enter the nuclear reactor core. The temperature drop over the intermediate loop is estimated to be 50 °C. The coolant type generally classifies the reactor. The reactors were graded based on 5 requirements and 5 criteria. As I mentioned before, the light water reactors cannot be used for the S-I hydrogen production process. The best reactor type turns out to be a gas-cooled HTR, while the second best is a salt-cooled HTR. The latter reactor type possesses both the HTR safety advantages as well as the molten salt advantages, like the excellent heat transfer capability that reduces the maximum fuel temperatures relative to gas-cooled reactors and improves passive heat removal. Also the metal-cooled reactors like the lead, bismuth and lead-bismuth eutectic coolants give a good score.

High temperature reactors are not a new development. Actually several have already been built, put into operation and shutdown. The AVR, which is an experimental pebble-bed HTR that started in 1967 in Germany and operated for 21 years. The THTR (Thorium HTR) started in 1985, but operated only for 5 years due to political and economic reasons. In the US, Fort St Vrain started in 1974, the same year that the prototype in Peach Bottom was shutdown. At present two new HTRs are in operation: the HTTR in Japan and the HTR-10 in China. The first one uses so-called block-type fuel, while the second one uses pebble fuel.

^* Conceptual generation IV designs (see later on).

A present-day pebble-bed High Temperature Gas-cooled Reactor (HTGR) belongs to the generation III+ of nuclear reactors and consists of a reactor core filled with fuel pebbles and surrounded with a graphite reflector. The reactor core itself would be several meters in diameter and similar height. Each fuel pebble with a typical diameter of 6 cm contains about 10 to 20 thousand TRISO coated fuel particles, which constitute one of the major barriers against fission gas release. Each TRISO coated fuel particle contains a fuel kernel with diameter of about 0.5-mm (500 micrometer) and four protective layers. One porous carbon layer to absorb all the gaseous fission products that are produced in the fuel kernel during irradiation, one pyrocarbon layer, one SiC layer and again a pyrocarbon layer. These three protective layers form a dense barrier that prevents the fission products from escaping the particle. The pyrocarbon layers will contract under irradiation creating a compressive strain in the rigid SiC layer that compensates for the expulsive strain from the gaseous fission products.

The key point to the safety of each HTR is that the fission products will be kept inside the protective layers of the TRISO coated particle as long as the fuel temperature does not exceed 1600 °C. Each HTR has to be designed such that during all incidents that one can imagine, the fuel temperature stays well below this limit, which assures the integrity of the fuel and which prevents any fission product release from the fuel. After a loss of coolant incident, the maximum fuel temperature is reached after 1 to 2 days, but due to the relative low power density of the reactor and the passive heat transfer mechanisms that transport the decay heat of the fuel to the environment, the maximum temperature stays well below the limit.

Instead of helium, a molten salt coolant at ambient pressure can be used. For these purposes one uses a mixture of LiF and BeF, which is highly stable under irradiation, and which has a high heat transfer capability and a high atmospheric boiling temperature near 1400 °C. As mentioned before, this reactor possesses both the high temperature gas-cooled reactor safety advantages as well as the molten salt advantages. Because of the excellent heat transfer capability similar to that of water, the temperature between the fuel pebble and the coolant could be reduced, leading to a higher coolant outlet temperature from which the efficiency of the hydrogen production process can benefit.

The VHTR is the next step in the evolutionary development of high temperature reactors. It is a graphite moderated helium cooled reactor with a power of 600 MWth that can supply heat with core outlet temperatures of 1000 °C. Of course, the increase of the core outlet temperature results in an increase of the fuel temperature and reduced margins in case of heat up accidents. To permit the higher temperatures, ZrC coated fuel particles are considered, while the reduced margins should be supported by more accurate calculations. Except for the change in materials applied, the fuel would not look much different from that for the current design HTRs.

To meet the generation IV goals of sustainability, a helium-cooled fast reactor is being developed that makes better use of the uranium resources. It delivers heat at a temperature of 850 °C and at a pressure of 90 bars. With regard to fuel consumption this reactor is aimed to be at least self-sufficient, which means that it breeds as much plutonium from U-238 as is being consumed. Several fuel designs are being investigated, among which the pebble-bed concept in use for thermal spectrum HTRs, and the prismatic block type. Because of the fast neutron spectrum, more neutrons per fission are released, which can be used to breed new plutonium from uranium. In this way the usable energy contents of natural uranium increase a hundred times, simultaneously reducing the nuclear waste arising.

A more futuristic type of nuclear reactor is the one that makes use of fluid fuel based on molten salt, which could again be a mixture of LiF and BeF or a mixture of NaF and ZrF₄

In conclusion, nuclear energy is a viable option to produce hydrogen with no emission of greenhouse gases. In principle, a nuclear reactor can be coupled via an intermediate heat exchanger to a hydrogen production plant. To start with, an HTR of generation III+ can be used in combination with a steam reforming plant to produce hydrogen from natural gas . If the emission of carbon dioxide would be a serious obstruction, one could sacrifice some of the hydrogen production yield to produce methanol instead of carbon dioxide. Methanol is a valuable resource for the chemical industry worldwide. At a later stage the generation III+ reactors can be gradually replaced by generation IV reactors like the AHTR and the VHTR, while the steam-reforming hydrogen production plants can be replaced by thermochemical hydrogen production plants based on the thermochemical processes.

The slides can be downloaded here.

For more information, please contact  j.l.kloosterman (at) tudelft.nl