In a series of diary entries here I have discussed the wonder fuel DME, dimethyl ether, detailing the international effort, centered in Asia, but also under review around the world, to bring this fuel into broad use both as a substitute or a means to transport natural gas, as a replacement for LPG, and as a motor fuel for diesel engines. The industrial infrastructure is already under construction. Since diesel engines are, of course, currently used both in cars and trucks and because DME is a demonstrated option for fueling them, DME potentially is a real option to replace oil in transportation. Given that personal transport systems and trucks are likely to be with us for a long time, we must attempt to find ways to reduce their onerous impact. In my opinion DME represents the best option for doing this.
To start, let me say, I'm not particularly enamored with the automobile. I’d rather we use some other option to get around. To me, the automobile represents the real worst case potential of so called "distributed energy" schemes. Every automobile, with its output of exhaust, its load of partially oxidized heavy metals, it’s sometimes leaky contingents of toxic and sometimes water miscible fluids - not limited to fuel - is a dangerous source of point source pollutants. Vast waste heaps of various types derive all around the planet from the automobile, and they cannot be controlled in any systematic way simply because there are too many of them, so many that they cannot all be recorded, never mind ameliorated. I own an automobile and guiltily, and sometimes not so guiltily, drive one as a part of my daily life, but I feel like an animal rights activist living on hamburgers when I do this.
I’m not crazy about trucks either. To my mind railroads, preferably electrified railroads are a better option. But no one is listening to me.
The fact is, wise or not, the world as a whole is not quite ready to abandon the automobile or trucks. Of course, even if it did so, say in favor of mass transit strategies that characterized the 19th century and early 20th century – railroads, ferries, canals and boats – sources of energy would still be required. We could, by returning to the past, maximize our efficiency but unless we are willing to accept a vast human die-off (and we may be compelled to accept one). Otherwise we will still need industrial energy, not only to manufacture our familiar household goods, but also to provide our food and warm our shelters.
I have discussed why DME is far superior to the chimerical hydrogen. Everyone from George W. Bush to Arnold Schwarzenegger to Greenpeace postures about hydrogen to no real result, but except for having an exhaust product that is pure water, there is little else to recommend it. I have given a brief overview of the wide variety of sources from which DME contained and described the how the most unacceptable source of DME coal is already being industrialized in China with Japanese assistance. I have also noted that DME can be readily accommodated to a variety of renewable energy strategies for the production of primary energy and in that context, discussed how it is even conceivable – should the Olah reversible methanol fuel cell prove industrially viable – that wind energy could be involved in the production of DME.
I have not yet touched, though, on what I believe is probably the best option, how DME can be manufactured through the use of nuclear energy. When one says "nuclear energy" as often as I do, one is prepared to hear all sorts of negative commentary. My view is that this commentary is irrational and is predicated on a kind of dangerous selective attention wherein a segment of the public refuses to do what should be simple comparisons. Whether anyone wants to face it or not, the critical task before all of humanity is to displace fossil fuels. This must be done or we face unacceptable risks not only to humanity but to the entire biosphere.
The 60 year history of nuclear energy has not been marked by perfection. As is the case with any new technology, there have been disasters and set backs. However all advanced technology is subject to disaster and the greatest disaster in human history is not likely to occur from a nuclear accident, but from the normal operations of the fossil fuel infrastructure. To the extent that nuclear power is merely imperfect, fossil fuels are disastrous by way of comparison. Nuclear power may, albeit rarely, be responsible for fatalities in accidents, but fossil fuels kill continuously in normal operations.
This does not mean that nuclear energy is perfect and cannot be improved. Although it is already very safe, I have discussed some of this efforts to make it even better through the use of advanced fuel cycles for instance. That said, nuclear energy is already superior to all of its alternatives, the chief alternatives being coal and biomass. (The use of biomass, however, much superior to coal, and I do not mean to say that biomass is terrible but only to point out that it is not as safe as nuclear.) In fact, nuclear energy need not be perfect, need not be wholly free of risk to be vastly superior to the continued use of fossil fuels and other energy options. It is precisely because nuclear energy has faced demanding standards that it is so safe - no such effort has been extended to any other form of energy and, like it or not, no other form of energy is globally capable of meeting such standards.
Now, after that long introduction, I am going to talk about producing DME with nuclear energy.
Let me say this on the relationship of hydrogen and DME: In fact, DME can – and probably in some cases should – be manufactured from hydrogen by the catalytic hydrogenation of carbon dioxide (or carbon monoxide. The reason, again, that hydrogen – while it is clearly a useful industrial material, is not really suitable as a consumer product. Its storage is too dangerous and energetically wasteful and its use would require the construction of an entirely new infrastructure. The difficulties with hydrogen contrast with DME which can use existing natural gas and LPG infrastructure and can – with some modifications – use existing diesel infrastructure. DME can be liquefied at temperatures higher than the boiling point of water. Hydrogen can only be liquefied at temperatures lower than the boiling point of liquid nitrogen. So the argument is not about whether we should make hydrogen, but is instead about how we use it. I contend that hydrogen is only suitable for use as it is currently used – in captive manufacturing programs. It is simply not a good idea to shoving hydrogen into homes and automobiles because its combustion product is water.
There are many ways that carbon dioxide can be recovered from the air or sea water or from other sources like burning biomass or – as we all know – from burning fossil fuels. The latter option is my least favorite, but that said, hydrogenation of fossil fuel carbon dioxide offers the possibility of rapidly slowing the growth of carbon dioxide in the air – allowing us to use the carbon contained in fossil fuels 1 or 2 or more times before releasing it into the atmosphere. (Of course as things stand now, we use the carbon just once.) If the carbon dioxide is obtained by degassing seawater or by removal from air, we will be in the happy place of having industrially modeled our energy production on that used by living things. (Living things let us know it is possible, for schemes like this to work.)
So when I discuss nuclear energy options for the manufacture of DME, I will be referring to programs that produce hydrogen, not DME and considering that this hydrogen will be used to hydrogenate carbon dioxide. There is one exception, and that would be the Olah reversible methanol fuel cell. (Methanol is easily dehydrated to give DME – DME being superior to methanol because DME, unlike methanol, is not toxic.) I have discussed this fuel cell in the renewable energy based DME discussion. (Let me first explain that this fuel cell is an R&D project, and not a commercially available product – a critical caveat intended to limit the magical thinking that is so prevalent in energy discussions.). That said, it is very promising since it can address drawbacks not only to wind and solar, but to nuclear, as well. One can imagine an Olah fuel cell that was constructed to remove carbon dioxide from the air by equilibrium shifting, the same kind of process that photosynthetic plants use.
Both wind and nuclear (and even solar) could be used with a methanol fuel cell by simply taking electricity that is generated but not immediately needed and diverting it to methanol synthesis in the fuel cell. Wind and solar, as is well known, suffer from not being continuously available – limiting their utility and raising their cost. This is less of a drawback with solar, since solar is generally available when electricity demand is highest, peak of noon on a hot sunny day. So if solar were cheaper and affordable – and it still isn’t – it would meet a major niche, peak demand. A major drawback for nuclear energy is that to produce low costs, nuclear power plants must run nearly at full capacity. Even if nuclear power plants were not so capital intensive, turning a nuclear power plant on and off is not quite the same as turning a switch. Reactor physics limits the speed at which reactors can power up and power down. Thus nuclear power plants cannot do what some gas and renewable capacity can do – meet peak demands. Many types of nuclear reactors take many hours, even days, to adjust their power levels up although all modern reactors can shut their power down. For this reason currently nuclear generally only displaces coal among the fossil fuels. Even if this is highly desirable, as coal is unacceptably dangerous, nuclear cannot still do everything.
If however, methanol fuel cells were connected to the grid, nuclear power plants could be diverted to producing motor fuels in off peak hours while serving an important function in the grid for which they are now not well suited, as "spinning reserve," power that is generated but not actually consumed. "Spinning reserve" must be present in power grids to address unanticipated changes in demand. Most of the time this is defined as power than can be fed into the grid in less than 30 minutes. Generally this function is served at present by natural gas power plants – but natural gas is an unacceptably dangerous fuel because of its climate change implications. If nuclear power plants were powering Olah fuel cells though, their power could easily be diverted in short order when needed. When the power is not needed, it would be making motor fuels. Under these circumstances nuclear could do everything, including making motor fuels.
Nuclear energy can, of course, be used to run electrolysis units as well, where hydrogen is produced by electrically splitting water. Electrolysis is not a major player in the world of hydrogen production today. The steam reforming of methane is still cheaper, as is the steam reforming of coal. This is strictly true of course, only in the case where the external costs are excluded, and not in the case where external costs are included through mechanisms like carbon taxes. Were their carbon taxes, hydrogen would be produced only with wind and nuclear power (as well as some hydroelectric as is the case in places like Norway) and electrolysis would become an economically important way to make hydrogen. Regrettably there are no carbon taxes. There should be such taxes but there aren’t. Interestingly, however, nuclear could in theory be slightly better at it than either wind or solar, were the generating costs equal. (Actually nuclear is cheaper than either wind or solar, although wind is getting better all the time.) This is because the energy efficiency of electrolysis can be raised by increasing the temperature at which the electrolysis is performed. Nuclear energy, in contrast to wind or PV solar, involves high temperatures. Appeal to this fact is a bit of a dodge though, since nuclear must always reject heat to the environment and is itself thermodynamically limited in efficiency. Recapturing some of this heat would increase the thermodynamic efficiency of both nuclear and electrolysis from their existing values, but there still would be a considerable loss of primary energy. Electrolysis is not, in my view, a real winner though. It has some utility but it is still expensive.
The use of the heat side product of nuclear energy to increase the efficiency of electrolysis does raise an important point though, the point of using nuclear energy to provide process heat. As we turn on our lights, drive our cars, and heat our homes, we do not think to ourselves that much of the world’s energy demand is used for industrial processes. It happens that the utility of nuclear heat has been recognized for some time. In fact, the first commercial nuclear reactors in the world not only generated electricity but also provided heat for district heating and for industrial processes. Some of the processes amenable to the use of nuclear energy are discussed here/ Therein we read:
Some of the first civilian reactors in the world were used to supply heat, e.g., Calder Hall in UK (1956) and Agesta in Sweden (1963). Calder Hall provided electricity to the grid and heat to a fuel reprocessing plant, and Agesta provided hot water for district heating of a suburb of Stockholm. The first nuclear power station in Russia (1954) was also a multi-purpose facility providing electricity and heat to the closed city of Obninsk in Kaluja region, near Moscow. Currently less than 1% of the heat generated in nuclear reactors is used for non-electric applications1. Direct use of heat energy is more desirable from an energy efficiency point of view and nuclear energy is an enormous source of greenhouse-gas-free energy. However, nuclear power has remained primarily a source for electricity generation. Presently about 30% of the world’s primary energy is used for electricity production, and approximately 2/3 of this energy is thrown away as waste heat. Yet despite past and current use models, it is possible to optimise the use of nuclear heat for both electric and nonelectric applications, thereby making more efficient use of nuclear energy. Experience in co-generation of nuclear electricity and heat has been gained in Bulgaria, Canada, China, Hungary, Kazakhstan, Russia, Slovakia and Ukraine...
The capture of heat that is the side product of electricity generation is happily, becoming increasingly important and is the core of what is involved both in combined cycle power plants and in "cogeneration" plants. Currently use of both of these techniques are dominated by fossil fuels, but the idea of using nuclear energy in exactly the same way is well known and has been industrially characterized since the dawn of the nuclear era. (The new Romanian CANDU reactors will all be co-generation plants providing district heating – so it’s not like the idea of using nuclear energy in this way was ever abandoned.) The paper lists some 18 processes that could use nuclear energy – several of them regrettably involving fossil fuel processing like coal gasification (the thing that all environmentalists must fight) and petroleum refining (which environmentalists must get banned.) Nuclear plants around the world also desalinate water, mostly for plant use, although the relatively small 150MWe Kazakh breeder reactor, which operated from 1973 until 1999, produced 80,000 cubic meters of water for city supplies from brackish water. (This is about 0.01% of the water demand of Los Angeles.)
There is one interesting process though that I would like to discuss further. This is the thermochemical splitting of water.
If you heat water to high enough temperatures – an energetically demanding process – at temperatures of several thousands of degrees Celsius, water will be split into hydrogen and oxygen. Because of the high temperatures involved, temperatures higher than industrial materials can accommodate, the process is of no practical utility. Even if it were, the hydrogen and oxygen would be mixed together and would therefore be highly explosive. It happens though that there are a multitude of ways to reduce the temperatures at which this process can be accomplished while also generating the hydrogen and oxygen in separate places. Collectively these processes are known as thermochemical hydrogen cycles. There are many such processes but the one that has generated the most attention is the "sulfur-iodine" cycle.
The sulfur iodine cycle produces hydrogen in a series of chemical reactions designed so that the starting material for each is the product of another. Two products are removed from the cycle, hydrogen and oxygen, and one is added water. It works like this: Sulfuric acid (battery acid) is heated to 850C where two molecules decompose into two molecules of Sulfur Dioxide, two molecule of water, and one molecule of oxygen. The oxygen is separated for either use or discharge into the atmosphere. Two molecules of iodine and four molecules of water are added to the two molecules of sulfur dioxide to give two molecules of sulfuric acid – which is sent back to the reactor where the oxygen is made – and four molecules of hydrogen iodide. This is accomplished at between room temperature and 120C. At around 400C, the hydrogen iodide is heated to decompose into iodine – sent back to the reactor where the sulfuric acid is made - and hydrogen, which is separated for use such as the hydrogenation of carbon dioxide to make DME.
This prose description of the chemistry is somewhat belabored. It can be elucidated more clearly, for those with a knowledge of chemical symbols on the website of the Franco American research team that is working to commercialize this process. In this link we see some photographs of actual lab scale process equipment, as well as a description of the process issues and challenges that still need to be addressed, and a discussion of the time lines that have been planned for the completion of the laboratory scale work, with all of the work being completed by 2008. A more detailed description of the portion of the work being conducted in France can be found using this link.
The Japanese Nuclear Hydrogen Program shown in this link is far more advanced than the Franco-American program. Lab scale work was complete 4 years before the Americans and French intend to complete theirs, and thermochemical hydrogen has been produced using the full cycle. In this link one can see the mock-ups of the pilot equipment.
The Test High Temperature Gas Cooled Reactor that will be used to run the pilot has been constructed, brought to criticality more than 8 years ago and is currently undergoing operational testing.
The Japanese program is highly ambitious and, as shown, highly advanced. Typically modern nuclear reactors run about 33% thermal efficiency where – in the absence of cogeneration technologies – the remainder of the energy is rejected to the environment as waste heat. (Most of the world’s power plants behave similarly, although new combined cycle natural gas and coal plants have achieved efficiencies higher than 50% for electrical generation.) In contrast, the Japanese intend to capture much of the energy of the waste heat to perform useful tasks: They intend to make hydrogen, from which they can manufacture motor fuel, they intend to generate electricity that they can sell to the grid and then intend to desalinate water. The completion simultaneously of these three missions will mean that the overall thermodynamic efficiency of the process will approach 80% . If the program is industrialized this means that the Japanese will get a very big bank for their buck for every kg of uranium they fission – meaning less spent fuel to handle for recycling and very low energy costs.
The hydrogen plant for the commercial reactor – which will be only about 20% as large in terms of thermal output as a standard modern electrical generation reactor – will be about 53 MT/day. Because of hydrogen’s very low molecular weight, following through on a study of hydrogen by the US National Academy of Sciences, this would be enough energy to fuel about 100,000 automobiles. (Please keep in mind my caveat that I don’t think this energy should be used as hydrogen itself but that it should first be transformed into the safer fuel, DME by hydrogenation of carbon dioxide.) At the same time the reactor is producing this fuel it will be producing 200 MW of electricity continuously, and producing each day about 90,000 tons of desalinated fresh water.
These reactors, though small, are designed to be modular and easy to build
The Japanese plan to complete their pilot demonstration of full system integration by the end of this decade. The first commercial plant to achieve all of these goals will begin operating around 2020, with design and construction being under taken in the 2010s.
Note that at the same time that Japan is developing this technology they are constructing a DME based infrastructure. Thus Japan has real hope of an energy future.
It is worthwhile I think to note the status of other national thermochemical hydrogen cycles programs. The brief discussion of these programs will come from some reports on a recent meeting international meeting on nuclear hydrogen that was held in October of 2005 in Oarai, Japan.
Canada: Canada intends to produce hydrogen using advanced electrolytic schemes, known as PEM integrating the hydrogen from its electrical grid using high efficiency supercritical water high temperature CANDU type reactors. Wind power, which has been problematically integrated into the Canadian grid because of its off peak integration, will play a roll. Although the economics of this program may be suspect compared to sulfur-iodine cycles, the program will have the considerable advantage of integrating two greenhouse gas free sources, nuclear and wind, to produce hydrogen.
China: China has been operating its HTGR-10, a "pebble bed" type reactor for several years now. This reactor has completed safety testing and is now intended to be used for the testing of the Sulfur-Iodine (SI) cycle. Process lab scale testing is only expected to be complete by 2008, so they are roughly about as advanced as the Americans are in the SI process, but they are far more advanced in reactor technology and have a firmer commitment to their timeline. They plan to pilot the (SI) cycle by running reactor connected SI systems during the first part of the next decade. A larger pilot will operate from 2015-2020 and the first commercial plant will begin operation around 2020, around the same time as the Japanese commercial effort ensues. Also like the Japanese, the Chinese have already begun to construct a DME infrastructure that is, regrettably for now, coal based.
Korea: South Korea is planning on using a type of reactor it has designed known as a VHTGC reactor (Very High Temperature Gas Cooled Reactor). Construction of a small test version this reactor is not expected to begin until 2010, so they are behind the Japanese and Chinese in terms of reactor development, and matched with the US. They are, like the US, developing the SI process independent of reactors, and will pilot their system by non-nuclear means. They are developing some interesting ceramic materials in their program to address some separation and corrosion issues. They plan to construct a pilot facility that will be able to produce between 25 and 75 tons of hydrogen per year using the SI cycle by 2013. By 2019 they intend to have an SI unit coupled to a reactor that will be able to produce about 22 tons of hydrogen per day, enough to fuel around 40,000 cars. In the 2020s they plan to commercialize the process.
I do recognize, by the way, that many people around the world, including many of my fellow democrats and members of DKos (but surely not the majority) have a visceral negative reaction to nuclear power. I hope to alleviate that reaction by appeal to reason and the presentation of information – fear can often be displaced by knowledge.
In sum the SI process for the production of hydrogen is being explored in significant detail around the world. It seems that it will become a standard tool for hydrogen production in Asia in the next several decades, fueled by nuclear energy. The processes offer a real approach to eliminating the chief role played by oil in the international economy, the need for transportation fuels. This will particularly be true if the world elects to use its hydrogen not as a motor fuel itself, but as a precursor to the manufacture of dimethyl ether, DME, the remarkable broadly employable fuel that can displace natural gas, diesel fuel, and LPG gas.