Energy Storage
 
 
 
 
Rogerkb [at] theworldisfinite [dot] com 
 
Essays     Essays Archive     Book Reviews     Links 
 
Essays Archive  > Energy Storage
 
 
 
 
 
Anyone who believes that exponential growth can go on forever in a finite world is either a madman or an economist.    
 
Kenneth E. Boulding 
eSolar power tower technology licensed by CO2 splitting company Sundrop Fuels 
 
A while back I read about a group Los Alamos scientists who had developed a method for using solar energy to split carbon dioxide (CO2) into carbon monoxide (CO)  and oxygen. They achieved this result by heating CO2 gas to very high temperature in concentrated sunlight. At these high temperatures the CO2 molecules are directly ripped apart by thermal energy. I call this method of chemical decomposition the sledgehammer approach to converting sunlight into chemical energy. Once the CO2 molecules have been split into CO and oxygen the two gases have to be separated to prevent them from recombining when the gas cools down. The Los Alamos group who called themselves LARE (Los Alamos Renewable Energy) did not give any information about how they achieved this separation. One possible method would be a semi-permeable ceramic membrane. 
 
You might be asking yourselves: What does producing carbon monoxide from carbon dioxide have to do with renewable energy? The answer to this question is that carbon monoxide is a combustible gas with a high energy content per mole (or per molecule), higher, in fact, than hydrogen gas. It is not considered to be a good fuel gas because of its low energy content per kilogram (not to mention the fact that it is a deadly poison). However, carbon monoxide can be transformed into hydrogen via the water gas shift reaction: 
 
CO+H2O=H2+CO2 
 
The H2 thus produced has about 92% of the energy content (HHV) of the original CO (assuming 100% conversion). Therefore this process can be thought of as an indirect method of splitting water to  produce hydrogen. At the end of the process the CO2 is recovered so that it can be recycled to produce more hydrogen. 
 
I should point out here that I am not a promoter of the so called "hydrogen economy". For a number of reasons I believe that hydrogen is a poor choice for an energy carrier. However, hydrogen can be further processed into more convenient chemical fuels, so that if a source of renewable hydrogen became available we could probably find a way to utilize it. At the end of this post I will consider briefly possible transformations of hydrogen into more useful energy carriers. 
 
The question I wish to address right now is whether their is any reason to believe that the method of production devised by the LARE team can produce hydrogen at a reasonable cost.  Considering the fact that the LARE group has licensed their process to a private startup company called Sundrop Fuels, it appears that they believe that real commercial potential exists. Considering that the landscape is littered with the corpses of renewable energy startups that did not deliver on their promises the mere existence of such a company should not cause any great excitement Still, in the world of renewable hydrogen startup companies are few and far between. Most of the time when you read about the results of people doing research on renewable hydrogen production you hear language like: "Yes we have split water in the laboratory and produced some hydrogen, but a lot more research and some major technological breakthroughs are needed before this idea can be converted into an economical method for producing hydrogen". So the fact that a group of Los Alamos scientists claim to have already developed a practical process for producing hydrogen may have some significance. 
 
On the other hand the direct thermal splitting of water using concentrated sunlight has been tried and rejected as an impractical process. Extremely high temperatures are required to obtained a reasonable efficiency, an no one has figured out a practical method for separating the hydrogen from the oxygen at these high temperatures. Is their any reason to believe (apart from the assertions of the LARE team that this feat is any easier to accomplish for carbon dioxide than for water?  
 
In searching for information on this subject on the Internet I found an interesting paper[1] published in 1986. These researchers asked precisely the same question as I did above. That is they asked whether CO2 is a better candidate for direct thermal splitting than H2O which achieves only 4% dissociation at 2400ºC. They concluded that the answer is yes the thermal dissociation of CO2 is more favorable than H2O at temperatures around 2000ºC. Furthermore they build an experimental reactor using calcia-stabilized zirconia as a semi-permeable membrane to separate the oxygen and carbon monoxide and succeeded in converting up to 20% of the input CO2 into CO at 1960ºC. They also said that they believed substantially higher dissociation rates could be achieved if the ionic conductivity of the membrane was improved. Of course "easier to split than water" is not the same thing as "economically practical", but this result gives some reason to hope that the Los Alamos scientists are not just blowing smoke. 
 
One reason that I am skeptical about the economics of this process is that the original implementation described by the LARE team (This description is no longer available on the web.) involves the use of high solar concentration parabolic mirrors with small processing furnaces held at the focal point by long carbon fiber booms. This kind of configuration seems likely to have high capital costs and probably high maintenance costs as well. 
 
Therefore I was interested to learn recently that Sundrop Fuels has signed an agreement with eSolar to license their power tower solar technology. Power tower technology uses an array of mirrors to focus solar energy on a relatively large central tower. Thus the carbon fiber booms and the miniature processing furnaces on each parabolic dish are eliminated. The array of small flat mirrors used by the eSolar power tower may also offer a cost advantage compared to the large parabolic mirrors  of the original LARE design. I am not holding my breath waiting for  cheap hydrogen to come rolling out of the desert, but this licensing decision by Sundrop Fuels is interesting. 
 
Now I wish to briefly discuss possible uses of this hydrogen, assuming that it can be produced at a reasonable cost. Neither compressing it into tanks and shipping it to end users nor transporting it in pipelines is likely to be economical. The LARE team's original proposal was to store the hydrogen locally and use it as a fuel to power steam turbines. The idea is that the same solar field would produce both steam and hydrogen, and the hydrogen would be used as a backup fuel to power the steam turbines when insufficient sunlight is available (i.e. at night or during cloudy weather). I am not convinced that this scheme would be any cheaper than proposals for thermal storage using nitrate salts, and it would not address the need for long term storage to even out seasonal changes in insolation. Hydrogen storage is expensive and storing months worth of supply is probably not practical. 
 
Another storage option is to chemically transform the hydrogen into some form that can be stored more conveniently. One possibility is to combine hydrogen with nitrogen to produce ammonia (NH4). The Haber-Bosch process is a well established for accomplishing this synthesis. The chemical equation for ammonia synthesis is: 
 
4H2+N2==>2NH4 
 
The energy content (HHV) of the resulting ammonia is 91% of the input hydrogen (assuming 100% efficiency of synthesis). Ammonia has some disadvantages as a fuel which I have discussed elsewhere. It cannot be used as a drop in replacement for hydrocarbon fuels in internal combustion engines or in gas turbines. However, I see no reason why it could not be used to heat water for a steam turbine; The great advantage of external combustion engines is that they are fuel flexible. 
 
I have never heard of anyone attempting to run a steam turbine using ammonia as a fuel, but, naturally, there has not been much motivation to do so. At present ammonia is most cheaply synthesized using hydrogen derived from natural gas via steam reforming and the water/gas shift reaction. Since natural gas is itself a high quality fuel and nearly one third of its energy content is thrown away in the process of turning it into ammonia, no good motivation exists for trying to use ammonia as  fuel. Only if an economical source of renewable hydrogen became available would it make sense to consider ammonia as a fuel. 
 
If the performance of nitrogen as a fuel for running steam turbines was unacceptable one could synthesize methanol via the reaction: 
 
CO2+3H2==>CH3OH+H2O 
 
The resulting methanol would have 88% of the energy content of the original hydrogen. In order for such a fuel cycle to be carbon neutral one would have to capture the CO2 generated during methanol combustion so that it can be recycled into more methanol. Obviously capturing and storing CO2 would add to the complexity and cost of the fuel cycle. Still, the cost of carbon capture from methanol might be cheaper than carbon capture from coal since the carbon to hydrogen ratio methanol is 0.25 compared to typical values of 1.0 for coal. The use of ammonia as an energy storage medium eliminates the problem of recycling a secondary element since nitrogen can be extracted from and returned to the atmosphere. 
 
By no means do I regard renewable hydrogen as a holy grail that will save our civilization, particularly so if 'civilization' is taken to mean growth based private finance capitalism. Previous readers of this website (if there are any such creatures) will know that I do not believe that this beast can or should be saved. The last thing that any person who desires the long term welfare of humanity should be hoping for is a dirt cheap, carbon free alternative to fossil fuels. We should hope that renewable energy sources are expensive enough to bring to an abrupt end to our dreams of endless increases in material wealth, but cheap enough to support modest but comfortable life styles. 
 
 
[1] Yutaka Nigara and Bernard Cales, Production of Carbon Monoxide by Direct Thermal Splitting of Carbon Dioxide at High Temperature, Bulletin of the Chemical Society of Japan, Vol.59 , No.6(1986)pp.1997-2002 
 
January 1, 2008 
 
Some More Energy Storage Options for Wind Turbines 
 
One possibility for storing the energy from wind turbines is to convert it into heat and store it in one of the heat storage forms proposed for solar thermal energy (e.g. Nitrate salts, or dissociated ammonia). Electricity can be converted to heat in resistive elements with near to 100% efficiency. The stored energy would be converted back into electricity via a steam turbine. The problem with this idea is that the steam turbine efficiency is unlikely to be much greater than 1/3 so that the electricity cost will increase by a factor of three based on efficiency considerations alone, and in addition one has to add on the capital costs of heat storage and the steam turbine. Before dismissing this idea as absurd you should not forget that quite a number of people with advanced scientific and engineering education have proposed with a straight face that water electrolyzers and hydrogen fuel cells (both very expensive pieces of capital equipment) could be used as an electricity storage system. Since the efficiency of water electrolysis is approximately 0.7 and the efficiency of fuel cells is a little under 0.5, this scheme has about the same round trip efficiency as the heat storage scheme just described. Electric resistors are cheap and steam turbines are at least an order magnitude less expensive than fuel cells. I am not sure about the cost of thermal storage, but I think it is significantly cheaper than electrochemical batteries. I am not claiming that this idea is a route to dirt cheap dispatachable electricity from wind, but it is achievable with current technology and does not require hydrocarbon fuel as does a compressed air / gas turbine storage scheme. 
 
Another energy storage possibility would be to combine compressed air and thermal storage. That is part of the time the wind farm would be storing energy as compressed air and part of the time it would be storing energy as heat. These two forms of energy could be recombined and used to run an expansion turbine. This idea has been used on a small scale by the company Active Power as parts of an energy storage system for uninterruptible power supplies (UPS). I believe that the storage efficiency of compressed air is higher than that of thermal storage (70% compared to 33%), plus, if an appropriate geological formation is present for storing the air, the capital cost will probably be lower than for thermal storage. Thus the combination of compressed air and thermal storage should be cheaper than thermal storage alone. 
 
Given the relatively low energy density of compressed air and thermal storage, I think it is unlikely that such storage schemes would allow the storage of energy for time periods of weeks or months.  Still they would add significant utility to electricity generated by wind farms, albeit at a price substantially higher than that of fossil fuels in their prime. However, the era of dirt cheap energy is passing away, and in the new era that emerges many technologies than now appear ‘impractical’ in a world obsessed with constant productivity improvements will prove useful in maintaining modest but adequate standards of living in a sustainable future.   
 
July 6, 2007 
 
 
Methanol as an Energy Storage Medium 
 
 
As I have discussed in previous posts, the intermittent and highly variable nature of wind generated electricity will prevent it from becoming a fossil fuel replacement without some effective means of energy storage.  The most desirable energy storage medium would have the same properties as fossil fuels: high energy density, ease of storage and transport within the current energy infrastructure, and low volatility so that storage over periods of many months are possible. Since hydrocarbons have proved to be such a great energy storage medium why not continue to use them? This idea has been promoted by Nobel prize winning chemist George A. Olah in his book Beyond Oil and Gas: The Methanol Economy (co-authored by Alain Geppert and G.K. Surya Prakash). From the title of this book it is clear that Olah is promoting methanol (CH3OH) as such an energy carrier. The hydrogen and oxygen necessary for synthesizing methanol could come from electrolyzed H2O, but where will the carbon come from? 
 
The most obvious sources of carbon are biomass, fossil or living. If one uses fossil sources of carbon then, when the methanol is burned as a fuel, this fossil carbon is released into the atmosphere as CO2 and contributes to increasing the percentage of this gas in the atmosphere.  If one uses living biomass (or at least recently deceased) the process is at least carbon neutral, but if dedicated energy crops are used then the usual economic competition with other valuable uses of land and fresh water takes place. 
 
Olah has a short term proposal for the source of carbon and a long term proposal. The short term proposal is to recycle CO2 from fossil fuel power plants (notably coal fired plants). This proposal does not prevent the fossil carbon from ending up in the atmosphere as CO2, but it makes it do double economic duty. That is the carbon first provides electricity when it is burned in the generating plant, and then it is removed from the exhaust and combined with H2 to produce methanol which can provide the same services as petroleum. Thus the total CO2 emissions per unit of economic output is reduced. Clearly such a strategy mitigates CO2 emission and extends the useful lifetime of fossil fuels, but does not address the goal of a long term CO2 neutral energy system or of a long term energy system which is completely independent of fossil fuels. Olah makes no serious attempt to analyze the economics of such a fuel production scheme. That is how much will methanol synthesized from electrolyzed hydrogen and CO2 captured from coal plant emission cost per unit of energy compared to the cost per unit of energy of conventionally derived gasoline and diesel fuel? Part of the answer is given in my discussion of methanol economics below. 
  
In the long term Olah proposes to extract CO2 directly from the atmosphere by means of exposing large areas of chemical absorbents to the atmosphere and then later desorbing the CO2 for use in methanol synthesis. Olah admits that this procedure for collecting CO2 is not at present ‘economical’ but asserts that it will become so in the future. Not a single shred of evidence is provided in support of this assertion. 
 
The key to this whole scenario is economics, not technology. The path to methanol production described by Olah can, in a purely technological sense, be carried out today. We can absorb CO2 from the atmosphere. We can electrolyze water to produce hydrogen. Well established processes for the catalytic synthesis of methanol out of H2 and CO2 exist. The question is whether or not such synthesis can be done at a cost which would make economic  large per capita consumption of liquid hydrocarbon fuels derived from such a source. Since Olah does not make a serious attempt to answer this question, I do not feel like spending time and energy trying to debunk his assertion. My intuition, (For what it is worth. My training is in physics, not chemistry) is that the answer is a decided negative. 
 
However, other possibilities for the use of methanol as an energy carrier exist. Toshiba Corporation researchers have published an interesting proposal for the use of methanol as an energy carrier. They also envision combining electrolyzed H2 with CO2. However, their proposed source of CO2 is methanol fired gas turbines. That is methanol (which is a high energy density liquid at ambient temperatures) would be transported from the synthesis site to large gas turbines. The gas turbines would be designed in such a way that the CO2 in the exhaust to could be cooled and separated from H2O (the other combustion product). The CO2 could then be transported back to the methanol synthesis site to be combined with hydrogen to produce methanol again. Compressed CO2 is a liquid at ambient temperatures and well established procedures for transporting large quantities of it already exist. Thus the CO2 would be shuttled back and forth between methanol synthesis sites and power generation sites. Some loss of CO2 would take place, so that some amount of CO2 replacement from other sources would be required over time, but  if the loss rate was small enough, then biomass could be used as a source of CO2 without placing huge additional demands on the biosphere. Of course such a CO2 recycling scheme could produce transportation fuel only in the case of an electrified transportation system. 
 
What are the likely excess costs of such an electricity generation scheme relative to today’s natural gas fired turbines? We can make a preliminary attempt to answer this question by assuming that the methanol turbine has the same cost per kW of power output as a conventional natural gas turbine. This assumption is probably false since provisions for collecting and transporting CO2 back to the methanol synthesis site would add to the cost. Nevertheless this assumption can give a lower limit on the excess costs. If we make this assumption, then the methanol synthesis plant can be regarded as a ‘methanol well’ and we can compare the price per unit energy of methanol at the ‘well head’ to the cost of natural gas at the well head. The current cost of natural gas at the well head is about US $0.02 per kWh. Of cource natural gas prices are not usually given in these units. $/MBTU are more typical (1 MBTU = 1 million BTU = 2930kWh. MBTU is also sometimes called MMBTU. Don’t you just love engineering units?) Translated to these traditional units we find the cost of natural gas to be  $0.02/kWh = $6/MBTU. The reason I use kWh as my energy unit will become apparent in a moment. 
 
The NREL’s latest report on wind energy called Wind Powering America quotes wind electricity prices a $0.04 per kWh. As a sanity check on this figure, the currently quoted priced of wind generation is  $1000/kW. If we assume a wind turbine capacity factor of 30%, a discount rate of 8%, a loan and turbine lifetime of 15 years, and ignore  maintenance costs then the calculated cost/kWh over a 15 year time period is $0.044/kWh which is reasonably close to the NREL estimate.  
 
The second piece of the cost puzzle is the electrolytic production of hydrogen. Water electrolyzers are expensive pieces of capital equipment. Typical electrolyzers on the market today cost $3000/kW, three times the cost of wind turbines/kW. Furthermore the lifetime of an electrolyzer is less than that of a wind turbine. Norsk Hydro, a major manufacturer of water electrolyzers, claims on their website that the service interval of their electrolyzer’s is seven to ten years. They do not say exactly what ‘service’ means, but I assume that the electrodes have to be replaced. Of course the cost of this service would be discounted since it happens years after the initial capital outlay. For simplicity let us say that the total cost of the ectrolyzer over the lifetime of the wind turbine is $4000. The efficiency of water electrolysis is about 70% so that adding electrolysis to the system bring the cost of 1kWh of hydrogen up to (5/.7)*0.04=$0.28/kWh or 14 times the price of natural gas at the well head. 
 
The final piece of the cost puzzle is methanol synthesis. Currently almost all methanol is synthesized from methane (CH4, the main component of natural gas). Methanol synthesis from methane is a two stage process. In the first step methane and steam are reacted in the presence of a nickel based catalyst to produce a gaseous mixture of CO, CO2, and H2. The reaction of steam and CH4 is highly endothermic (i.e. It absorbs a lot of energy.), and this energy is provided by combusting methane (natural gas). Steam reforming is the major energy consuming part of the methanol synthesis process. In the second stage of methanol synthesis H2 is reacted with CO and CO2 in the presence of copper-zinc based catalyst to produce methanol (CH3OH) and water. The reaction of CO2 with H2 to form methanol is also endothermic but has a much lower heat of formation than the reaction of CH4 and steam. It is only the second part of the methanol synthesis process that is relevant to our current discussion, since we are assuming that H2 is supplied from electrolysis and CO2 is supplied from methanol turbines. The energy efficiency of methanol from natural gas synthesis is a little over 70%. Presumably synthesis starting from H2 and CO2 would be more efficient since the steam reforming step would not be necessary. I have been unable to find any discussion of the energy efficiency of the reaction of H2 and CO2 to from methanol. I will assume 90% for purposes of the present discussion. 
 
The current price of methanol synthesized from natural gas is $1.00/gallon. If this number is translated into cost per kWh of energy content (methanol energy density = 4.6kWh/liter and 1 gallon=3.785 liters) we find that methanol cost is 0.06/kWh. If methanol synthesis is 72% efficient then the cost of the natural gas feedstock required to produce 1kWh of methanol is about $0.03. The remaining cost of $0.03/kWh is presumably manufacturing cost plus the producer profit. For simplicity I will assume that the cost of methanol synthesis from H2 and CO2 is $0.03/kWh. Although the step of steam reforming methane is not required, thus lowering costs, the lower capital utilization factor enforced by the low capacity factor of the wind plant will increase costs. In any event, for the purposes of the present argument this is the number I assume. 
 
The total cost for 1kWh worth of methanol via the wind/hydrogen route then becomes: 
 
1kWh Methanol Cost = (1kWh hydrogen cost/.9) + $0.03 = $0.34/kWh 
 
This cost should be compared to a cost of $0.02/kWh for natural gas at the well head. kWh here refers to the energy content of the fuels in question and not to kWh of electricity delivered when the fuel are used to power a gas turbine generator. Advance combined cycle gas turbines can convert chemical fuel to electricity with about 70% efficiency so that the contribution of fuel purchases to electricity costs for these two cases are $0.03/kWh and $0.48kWh respectively. This estimate for methanol cost is low because there will be extra cost associated with collecting the CO2 from the turbine output and transporting it back to the methanol synthesis site. The total energy efficiency of this process (ignoring the energy embedded in capital equipment) is 0.7*0.9*0.7=0.44. That is 44% of the energy contained in the output of the wind generators is recovered as electricity at the end of the process. 
 
Clearly using methanol as an energy storage medium is not a route to cheap electricity with current technology. The biggest cost contributor is water electrolysis. The efficiency of electrolysis is unlikely to much exceed 70% but a number of researchers are attempting to develop lower capital cost electrolyzers. Although one can find optimistic stories on the internet about companies seeking to develop such electrolyzers, I am not holding my breath waiting for them to deliver the goods. Progress in electrochemistry is notoriously slow. The lead acid batter was invented 147 years ago, and it still reigns supreme in the field high capacity electricity storage in spite of the high desirability of developing alternatives. The prospect of low cost electrolyzers feeds the fantasy of the ‘hydrogen economy’, but just because people are willing it into existence does not mean that it will appear any time soon. 
 
At a cost of over $.50/kWh such an electricity source could not support current levels of economic production in the highly developed countries, let alone power the economic growth required for the ‘healthy’ functioning of private finance capitalism. However, even such expensive electricity might have marginal economic utility. That is if you were faced with a choice of zero electricity or modest amounts of electricity supplied by this relatively expensive method you might find a net benefit from purchasing the expensive electricity. 
 
April 4, 2007 
 
 
What About Carbon as an Energy Storage Medium? 
 
Intermittency is the Achilles heel of renewable energy sources. The NREL stimates that the cost of wind generated electricity as it comes out of the generator on the tower from high wind sites are currently $0.03/kWh, and projects that they will drop to $0.02 by 2020[1]. This cost was undoubtedly calculated using discount economics in which it is assumed that if our money is not invested in a wind turbine it can be invested in mutual funds which will rise 10% a year forever. The cost of the lost income from these stock funds is counted against the cost of generating electricity from wind. When the madness of discount economics finally comes to an end the economics of wind energy will look even better than they do today. 
 
However, when one considers trying to turn the highly variable output from wind arms into a dispatchable base load power system, this rosy cost picture disappears in a puff of smoke (See my previous post: Is Wind Energy an Effective Substitute for Coal and Natural Gas? ). To get dispatchable base load power from wind one must add an energy storage andre-conversion system. If the costs per kWh of delivered energy of the storage/re-conversion system are expensive, the costs of the whole system are expensive.  I emphasize the re-conversion (i.e. turning the stored energy back into AC electric current) aspect of the storage system, since in some cases this is the most expensive aspect of storage (think of hydrogen fuel cells).  One must not only consider the capital and maintenance costs of energy storage system, but one must also consider the energy losses involved in such storage systems. The efficiency of an electricity storage system is the fraction of the original electric energy input to the system which is recovered after the storage and re-conversion processes. Costs are automatically multiplied by a factor of 1/efficiency.  
 
Unfortunately, low cost energy storage systems have proven very difficult to develop. The energy storage medium which has received the most public attention is electrolyzed hydrogen. Hydrogen does have the virtue of high gravimetric energy density, but aside from this property it has very serious drawbacks as an energy storage medium: High capital costs for production and re-conversion, high storage costs, and low round trip energy efficiency. 
 
Other proposed storage media tend to have very low energy densities (order of magnitude lower than fossil fuels). Among the low energy density media, compressed air and flow batteries appear to have the most reasonable costs. I do not think that combination of wind with either of these technologies is yet competitive with fossil fuels, but higher fuel prices and/or carbon taxes may allow these technologies to find a niche in the market. However, because of the low energy densities, no one is suggesting storing energy for a time period much beyond 24 hours. The relatively low costs of compressed air and flow batteries are achieved by high having high duty cycles.  What happens if the wind howls for two weeks and then disappears for two weeks? What happens if the blows in the winter, but you need to run your factory in the summer? The capital cost associated with storing very large quantities of energy in these low density media would be prohibitive. 
 
In addition to low capital costs for production/re-conversion and high round trip energy efficiency, the ideal storage media would have he same virtues as fossil fuel; i.e. high energy density and low volatility allowing cheap storage for long periods of time. The other day a thought came to me  about a possible storage medium; Instead of hydrogen, what about the other member of the hydrocarbon pair? What about using carbon as an energy storage medium? This idea might seem crazy at first sight, but bear with me for a moment. 
 
This idea came to me when I was reading about a new technology called direct carbon fuel cells. In a direct carbon fuel cell carbon micro particles combine with oxygen to directly produce CO2 and electricity. These fuel cells have been developed by researcher at Lawrence Livermore Labs[2]. These fuel cells are a very interesting technology.  The entropy change between the beginning and final states of this electrochemical process is near zero so that the theoretical maximum efficiency is near 100%. Furthermore the Lawrence Livermore researchers claim to have already achieved 80% efficiency in bench top prototypes of these fuel cells, and they believe that 85% efficiency is achievable in a practical real world realization of such fuel cells. This situation should be contrasted to hydrogen fuels cells in which the maximum theoretical  efficiency is 73%, while real life hydrogen cells achieve only a little over 40% efficiency. The Livermore researches also claim that they foresee a reasonable path to low cost manufacturing of this type of fuel cell ($200/kW). Again this claim should be contrasted to hydrogen fuel cells where cost remain stubbornly high even after decades of research and development. 
I should point out, by the way, that no one is proposing to use direct carbon fuel cells to run automobiles. Like solid oxide fuel cells they are too heavy for such mobile applications. 
 
The Livermore researchers are proposing direct carbon fuel cells as a potential ‘clean coal’ technology. The idea is that the carbon fuel for these cells would be processed from coal. The fuel cells would then produce a pure output stream of CO2 which would be captured and sequestered (i.e. injected into underground geological formations). My purpose in this post is not to discuss the virtues/disadvantages of clean coal technology. But whatever else can be about such technology, it is clear that its utilization is only a stop-gap, since coal is a finite resource just as much as oil and natural gas.  
 
However, as I was thinking about the stream of CO2 emerging from the direct carbon fuel cell, an idea came to me: What if we could reverse the process and electrolytically  decompose CO2 into carbon and oxygen with the same high efficiency? 85% times 85% is 72% which is a respectable efficiency. It is certainly a lot better than the water electrolysis/hydrogen fuel cell combination than many people  have proposed with a straight face as a possible electrical energy storage system. If this reduction of CO2 could be done with reasonable cost, we would have the basis for an energy storage system. When the wind was blowing we could reduce CO2 to carbon and oxygen. Carbon has a reasonable energy density and can easily be stored. When we need electricity we combine the carbon with oxygen in a fuel cell and collect the CO2 to be reduced to carbon and oxygen again when the wind is blowing. CO2 can be stored as liquid at ambient temperatures so that storage costs should be reasonable. Some amount carbon loss will happen over time, but as long as the rate of loss is low, we can replenish it from biomass without putting a large demand on the ecosystem. 
 
Of course this idea is just a speculation in the mind of a person largely ignorant of chemistry. Serious practical barriers might exist to its implementation. The first question to be asked is: Can CO2 be electrolytically decomposed in C and O2 (cheaply or otherwise)? Searching on the internet did not reveal a lot of research activity in this field. When you think about it, such a lack of research is not surprising. When vast stores of carbon are available in the form of coal, what fool would spend time and energy trying to extract it from CO2? However, it turns out that NASA has done some research on the electrolysis of CO2 in a effort to find a method of producing oxygen and fuel from CO2 in the Martian atmosphere. They found that it was possible to electrolytically decompose CO2 into carbon monoxide and oxygen via the reaction: 
 
     2CO2 ==> 2CO + O2 
 
Their orginal experiment using a zirconia based catalyst obtained an efficiency of only 50%[3]. However they are also developing a molten carbonate electroysis system which they believe will have suprior perfomance to the zirconia based system[4]. Of course, this reaction only gets you half-way to the goal. One oxygen atom has been removed, but one still remains. Searching further on the internet, I found that CO can react with itself in something called a disproportionation reaction to produce carbon and CO2: 
 
     2CO ==> C + CO2 
 
This reaction is not electrolytic, but it will proceed spontaneously in the presence of a suitable catalyst (Fe, Ni, and Co all work as catalysts for this reaction.). This latter reaction has received considerable recent study as a  method of producing carbon nanotubes. If CO2 from this second reaction is fed back into the first reaction the following CO2 reduction system is obtained: 
 
   CO2 (from fuel cell) 
      | 
 2CO2==> 2CO + O2 
      |__________ 
                             | 
 2CO ==> C + CO2 
 
This system of reactions has the desired result; When one mole of CO2 is fed in, one mole of C and one mole of O2 emerge. 
I am not claiming that this is a practical system for reducing CO2 to C and O2. I do not know what energy efficiency and capital cost could be achieved in a practical, continuous flow, high throughput system. I am just pointing out  that the reduction of CO2 is technically feasible, and it might be worthwhile to invest some effort into trying to find an opmtimal process.  
Of course, no matter how efficiently and cheaply CO2 reduction could be achieved, such as system would add significant cost to a wind generation system. CO2 reduction and direct carbon fuel cells would add capital cost, and all costs must be multiplied by a factor of 1/efficiency. However, if wind costs really do drop to $0.02/kWh, then substantial costs could be added to the system without making it economically useless. This is probably just a crazy idea, but occasionally, an apparently crazy idea turns into a useful invention.  
 
[1] NREL Wind Powering America Update, 01/22/2007: http://www.eere.energy.gov/windandhydro/windpoweringamerica/pdfs/wpa/wpa_update.pdf 
 
[2] Lawrence Livermore Nation Laboratory Carbon Conversion Fuel Cell on line report, 2001: http://www-cms.llnl.gov/s-t/carbon_con.html 
 
[3] NASA Tech Report, 2002, Mars In Situ Resource Utilization (ISRU) Oxygen Production: http://rtreport.ksc.nasa.gov/techreports/2002report/600%20Fluid%20Systems/609.html 
 
[4] NASA Tech Report, 2002, Space Habitat Carbon Dioxide Electrolysis to Oxygen
 
Feb 22, 2007 
 
 
Thermochemical Production of Hydrogen 
 
I am not a fan of the hydrogen economy. Hydrogen is a poor choice of energy carrier. Hydrogen is expensive to transport and store, and cost effective means do not exist for converting its chemical potential energy into electricity or mechanical motion (When you have a fuel cell available at less than $5000/kw with a five year warranty drop me a line.). On the other hand the idea of converting intermittent renewable energy sources into a high energy density chemical form for convenient transport and long term storage is obviously very attractive. Renewable hydrogen could be the starting point for the synthesis of methanol (CH3OH) or ammonia (NH3) both of which are liquid at ambient temperatures and have energy densities of 6.4kWh/Kg and 5.1kWh/Kg compared to 12.2kWh/Kg for gasoline and 0.25kWh/Kg for advanced Lithium ion batteries.  
 
Even if cheap hydrogen from renewable resources was available, I am not claiming that conversion of that hydrogen to methanol or ammonia is a miracle solution to our energy problems. Each choice has special problems, and I doubt that either one could compete economically with fossil fuels in their prime. However, in this post I want address the issue of hydrogen production costs alone.  
 
The only well established commercial technology for the production of hydrogen from water is alkaline electrolysis. In a report on an LAX project to build a hydrogen filling station published here, the developers state that the capital cost of  a Stuart IMET 1000/15/10 electrolyzer capable of producing 1kg of hydrogen/hour was $155,000. The high heating valued of 1kg of H2 is 39kWh. If we assume an electrolyzer efficiency of 0.7 then power rating of this electrolyzer must have been  39/0.7=55.7kW and the cost/kW = $155,000/55.7kW = $2782/kW. According the formula given in my post on renewable energy economics the cost of hydrogen produced from this electrolyzer is given by 
 
Hc =(RE/0.7) + M×2872/(0.7×CF×L×365×24) + OM 
 
where RE is the cost of the renewable electricity per kWh,  L is the electrolyzer lifetime, CF is the capacity factor of the renewable generator, OM is the operating and maintenance costs, and M is a capital cost multiplication factor reflecting the effect of interest paid on the capita loan (For example a 15 year loan at 8% interest results in M=1.75). So if RE is $0.04/kWh, CF=0.3, L=15 years, M=1.75, and OM=0 then the cost of hydrogen production is 
 
Hc= ($0.057 + $0.176)/kWh = $0.23/kWh 
 
For reference $0.11/kWh corresponds approximately to $4/gallon of gasoline equivalent. This is not cheap fuel, and this is not fuel ‘at the pump’; Additional energy loss and cost will be added in further processing/distribution. Just to give you an idea of how this calculation compares to hydrogen costs quoted elsewhere here is a link to an article on GE efforts to develop a lost cost electrolyzer in which the current cost of electrolytic hydrogen is quoted at $6 to 8$ per kg. Since a kg of H2 contains the energy equivalent of 1.06 gallons of gasoline my calculation is not too far away from these quoted costs.  
 
So is there any other route to cheaper hydrogen other than cheaper electrolyzers? One method of hydrogen production from water that has been researched fairly extensively (but not commercialized) is thermochemical production of hydrogen. In this method of production water is reacted with chemical reagents at high temperature in a series of reaction that liberates H2 and O2 and regenerates the original reagents (That is the chemicals effectively act as a catalyst driving the breakup of water into it elemental constituents without themselves being consumed.). One example of such a process is called the S-I cycle and is given by the following reactions 
 
Prime or Bunsen reaction 
(2H2O + SO2 + I2  H2SO4 + 2HI) 
 
Sulfuric acid decomposition 
(2H2SO4  2SO2 + 2H2O + O2) 
 
Hydrogen iodide decomposition 
(2HI  I2 + H2) 
 
The prime reaction of the cycle runs at a temperature of 900C and the estimated efficiency is 50%. That is 50% of the energy used to heat the chemical baths is converted into hydrogen chemical potential energy. 
 
Conceivably such a reaction could run from the heat of concentrated solar energy. However, most people who are promoting this method of hydrogen production are advocating using the waste heat from advanced thermonuclear fission reactors as the source of heat. Part of the reason for such advocation is that these people are smart enough to understand that transporting hydrogen over long distance is a difficult, costly task. If solar thermochemical power plants are used the lowest production cost will be in high insolation desert areas which are likely to be far from the point of end use. Nuclear reactors, on the other hand can be built anywhere (assuming that people are willing to live close to them) so that hydrogen production facilities could be located near to end users. 
 
However, if we consider hydrogen only as an intermediate product to be converted into some other chemical substance which is easier to transport then desert locations are not necessarily a problem. 
 
What is the likely cost of hydrogen produced by a solar thermochemical process? I have come to believe that attempting to estimate the production costs of non-existent technologies is somewhat of a fools game, but I will take a stab at it anyway. For the basis of my speculation on costs I will use the known costs of solar thermal trough plants which are used to generate electricity from sunlight using single axis tracking parabolic mirrors. Solel, an Israeli company which operate more than 100MW of such solar thermal generation in the Mohave desert in California, has recently announced an agreement with the Spanish company Sacyr to build 150MW of solar thermal generation in Spain. The announced cost of this plant is U.S. $890 million. If you divide the cost by the amount of generation you find a cost of approximately $6000/kW. Elsewhere on the internet I have seen several analysts claiming capital costs of $3000/kW or less for such plants, but I have to accept the Solel announcement as reality since it relates to a real world project and not to somebody’s speculation. 
 
My key assumption (which has no solid justification) is that the cost per unit of solar energy collected will be the same for a solar thermochemical H2 facility as for a solar trough electric generation facility. If the cost of the collectors dominates the overall costs then this assumption is equivalent to the assumption that collector cost peer unit of concentrated light energy is the same for the two types of plant. A thermochemical facility will have a different design than a solar trough plant. The most likely design is a power tower design in which a field of mirrors will focus light on a central processing facility. Such a design will require two axis tracking rather than the single axis tracking and thus more complex motor. Such a design will also avoid the need for a heat transfer fluid and an extensive plumbing system to transport the hot fluid. Whether the assumption if equal costs for the two types of plant is good or bad I will make it. 
 
A second piece of information need is the efficiency with which the collected solar energy is converted to the desired end product. For a solar thermal plant I will assume that the steam plant efficiency is 35%. As I have already mentioned researchers claim that the efficiency if the I-S hydrogen production cycle operating at 900C is 50%. Therefore the cost of the thermochemical plant will reduced relative to the solar trough plant by a ratio of 35/50 so that $6000/kW becomes $4200/kW for the hydrogen producing plant. 
 
Other assumptions are: 
 
N = plant lifetime = 20 years 
 
f = capacity factor = 0.3 
 
M = interest multiplication factor = 1.75 (6% interest for 20 years) 
 
OM =  operating and maintenance cost per kWh of H2 produced = ? 
 
The cost of hydrogen then becomes (see my post on renewable energy economics for a further discussion of this formula): 
 
     H2 Cost = [(1.75×4200)/(0.3×20×365×24)] + OM = $0.14/kWh + OM 
 
$0.14/kWh (excluding OM) is certainly cheaper than $0.23/kWh I calculated for electrolysis although it is far from being inexpensive (Again recall that $0.11/kWh is the same as $4/gallon of gasoline equivalent.). 
 
Wind turbine capital costs are usually quoted at U.S. $1000/kWh which is much lower than the costs quoted above for concentrating solar power. Wouldn’t it be nice if wind energy could be used to drive thermochemical H2 production? Well, actually, you could use wind power in that way. Electric can be converted to heat with near 100% efficiency, so there is nothing preventing the use of wind turbines as a source of heat  for a thermochemical reaction. If the efficiency of hydrogen production is 50% then the wind cost contribution to hydrogen production is double the cost of wind electricity cost. So if electricity from a wind turbine cost $0.04/kWh then the wind contribution to hydrogen cost via thermochemical production is $0.08/kW. The question then becomes what are the capital costs of the chemical processing facilities? I have no means of addressing this question directly. However, by way of example we can examine the costs of another large scale chemical production process: The production of ammonia (NH3) from natural gas. I found a web posting which claimed than in 2005 natural gas costs of $9.38/MBTU constituted 90% of the cost of ammonia production. I have also read that the energy efficiency of the production of ammonia from natural gas is 65%. That is 65% of the energy content of the natural gas feedstock ends up in the form of ammonia potential energy. These two numbers along with conversion factor 1MBTU = 293kWh can be used to calculate the production cost of ammonia per kWh of energy content:  
 
Production Cost =(1/9)×[$9.38/(0.65×293kWh)] = $.0055/kWh 
 
That is, excluding the cost of the natural gas feedstock, the cost of ammonia production is less than ½ cent per kWh. Of course this number applies to ammonia production and not to H2 production, but it illustrates the fact that large scale chemical production processes can be relatively cheap. So the idea of turning wind generated electricity into heat in order to drive a thermochemical H2 production process is not as absurd as it might appear at first sight. 
 
As I mentioned at the beginning this post, I believe that in order for renewable hydrogen to be economically useful it has to be transformed into a more convenient chemical form. I will discuss the economic prospects for achieving such a transformation in a later post. 
 
June 6, 2007 
 
 
Carbon Monoxide As an Energy Storage Intermediate 
 
I have written before about the desirability of converting intermittent renewable energy sources into liquid or solid chemical fuel. Being able to deliver energy from intermittent renewable when and where we want it requires either some form of long term energy storage or a global supergrid capable of moving energy around to where it is needed. The idea of a global supergrid and the necessary spirit of international cooperation required to construct and manage it appeals to me, but I am not holding my breath waiting for such a system of planetwide cooperation to emmerge. For long term (i.e. months, which is to say the time frame of seasonal variations in wind and sunlight) energy storage as liquid or solid chemical fuel is the most obvious choice. Water splitting to produce hydrogen has been the most vigorously pursued option of this kind. Hydrogen itself is not a great energy storage medium because of its low volumetric energy density. However hydrogen can be combined with other common elements to form liquid fuels which would be good energy storage media. Two often mentioned candidates are ammonia (NH3) and methanol (CH3OH). Methanol contains carbon, so that if one desires an energy producton process which largely independent of both fossil carbon and atmoshpheric carbon then the CO2 which is formed when the stored energy is extracted from methanol must be captured and recycled back into methanol via the reaction 3H2 + CO2 ==> CH3OH + H2O. Such CO2 capture is acheivable for stationary generation of electricity, but is impractical for mobile transportation applications. 
 
The problem with this idea is that a low cost method of splitting water to produce hydrogen has not been found. An alternative to water splitting is CO2 splitting. If CO2 can be split to produce carbon monoxide (CO) then hydrogen can be produced via the water gas shift reaction CO + H2O ==> H2 + CO2.  This reaction is slightly exothermic (it give off energy) and the resulting H2 has 93% of the energy content of the orginal CO. Thus CO2 splitting is an indirect route to producing H2 and H2 derived fuels. CO2, like H2O, is a very stable molecule and splitting it is not particularly easy. Nevertheless efforts are being made to find efficient methods to produce CO from CO2. The same array of optinons is available for CO2 as for H2O 
 
Electrolysis: 
 
In the presesence of a charged conductor (platinum seems to be the preferred material) CO2 can split into CO and into charged oxyen atoms. In an electrolysis cell the charged oxygen atoms (or ions) then migrate to the other electrode where the oxygen is converted to normal neutral O2 molecules. This process is expensive and inefficient. Nasa has pursued it as mean of producing oxygen and fuel from CO2 on space missions, but the likelihood of large scale fuel production by such mean seems small. 
 
Photocatalysis: 
 
This process is very similar to electrolysis. In the presence of electric charges and an appropriate catalytic material CO2 splits up into CO and into a charged oxygen atom (or ion). The oxygen ions then migrate to another location where they give up their charge and recombine into netural O2 molecules. In photocatalysis, however, the elecrons are not provided by an external current source. The electrons are excited into the conduction band of an semiconductor by incident sunlight. This same method has been pursued as a means of water spltting for several decades without any notable practical success. One problem with attempting to split water by this method is that water is a very corrosive substance and finding semiconductor materials and catalysts that can survive for long periods of time immersed in water is difficult. CO2 may be less restrictive in this regards. Recently Chemists Clifford Kubiak and Aaron Sathrum at the university of San Diego have succeeded in using sunlight to split CO2 using a gallium phosphide semiconductor coated with a nickel based catalyst. The fact that electricity does not have to be separately generated and fed into the CO2 splitting system could give photocatalytic methods an economic advantage relative to tradtional electrolysis. Still, I would not hold my breath waiting for cheap carbon monoxide (and by implication cheap hydrogen) to emmerge from hectares of semiconductor material sitting in the desert. 
 
Thermochemical Splitting: 
 
In thermochemical splitting a series of chemical reactions occurs at high temperature which results in an oxygen atom being removed from a CO2 molecule while the other chemical particpants in the reaction return to their initial state. The high temperature necessary to drive this reaction can be supplied by concentrated sunlight. This same approach has been pursued for water splitting without any real practical success. A group and Sandia National Laboratory is working on a CO2 splitting device which uses a ferrite (iron oxide) material doped with cobalt. The ferrite is deposited on a large disk one half of which is exposed to concentrated sunlight. At high temperature oxygen atoms in the ferrite are released. The material is then rotated into a second half chamber which is filled with CO2 gas. The CO2 molecues donate one oxygen atom to the ferrite and become CO. The disk then rotates back to the heated half chamber and the process is repeated. This same device could also be used to split H2O, but the Sandia group seems to feel that CO2 splitting will be an easier process to optimize. They plan on having a prototype system up and runnng some time in 2008, but claim that even if the prototpe works, commercially viable systems might be more than a decade away. 
 
Direct Thermal Splitting: 
 
I call this the sledge hammer approach.  Concentrated sunlight is directed into a thermal chamber filled with CO2. At a sufficiently high temperature (>2000ºC) the CO2 molecules disintegrate into CO and O. This distintegration is partly due the the CO2 molecules banging into each other at high speed and partly due to the action of the solar photons themselves. The key is separating the carbon monoxide from the oxygen at high temperature. If you allow the gas to cool down the ingredients will recombine and your effort has gone for naught. The same approach has been tried for direct thermal splitting of water, but has largely been abandoned because separation of hydrogen and oxygen at high temperature has very low efficiency. In the case of CO2 splitting a group of scientists at Los Alamos Nation Lab claim that they have developed a process which can achieve an efficiency and cost which makes it commercially viable. They use parabolic dish concentrators which direct sunlight into a ceramic thermal chamber which is held at the end of a carbon fiber boom attached to the center of the reflecting dish. The temperature of the reaction chamber is 2400ºC. The plan of the Los Alamos group is to produce both electricity (via a steam turbine) and hydrogen (derived from CO via the water gas shift reaction) with the same set of solar dishes. The hydrogen will be stored locally and used to power the steam turbine when insufficient sunlight is available.  
 
The claim (unproven of course) that this process is commercially viable is surprising to me. All other schemes for producing chemical fuel from sunlight that I have read about are research projects which people are hoping will some day lead to useful applications, but no one claims that such application are anywhere close to being realized. This particular project sounds expensive because of the carbon fiber booms and the high temperature cermamic chambers. Conceiveably the carbon fiber boom could be eliminated if back focus reflective lenses such as this one invented by Viktor Vasylyev were used instead of parabolic reflectors. The Los Alomos group has recently received $20 million dollars from a silicon valley venture capitalist, so somebody thinks that they have a chance of delivering the goods.   
 
March 15, 2008 
The real science of political economy, which has yet to be distinguished from the bastard science, as medicine from witchcraft, and astronomy from astrology, is that which teaches nations to desire and labor for the things that lead to life; and which teaches them to scorn and destroy the things which lead to destruction. 
 
John Ruskin
 
 
 
 
Roger K. Brown 
Rogerkb [at] theworldisfinite [dot] com