The concept of a “hydrogen economy” took a hit in the first years of the new century when hydrogen was touted as the clean transportation fuel of the future. Prolonged failure to deliver on that vision tarnished popular perceptions. The concept didn’t die, but it receded from public attention.
In the past few years, interest has revived. That’s in part because practical fuel cell vehicles may finally have arrived. (I’ve started seeing hydrogen-fueled Toyota Mirais on the roads in the San Francisco Bay area.) But there’s also a lot of interest — notably in Europe — in the use of hydrogen for buffering the irregular energy output of wind and solar PV resources. Hydrogen would be produced by electrolysis when the supply of electricity exceeded demand, and then used to meet demand during shortfalls.
That particular use of hydrogen is controversial, due to the low 40% round-trip efficiency of P2G2P (power to gas to power) when electrolytic H2 is used for energy storage. Advocates don’t dispute the low efficiency, but contend that renewable energy will soon be so cheap that low efficiency won’t matter.
However there’s no avoiding the fact that 2.5 kWh of electricity would need to be generated for every kWh delivered from storage. In a 100% renewable energy economy, a large fraction of energy must be delivered from storage: 50 – 75%, depending on geographical region and on the state of transmission systems and demand side management. That means a rough doubling of the RE capacity that would have to be installed. The average cost of electricity would more than double.
There is an alternative that could be quicker and more economical. That alternative is hydrogen made using the chemical potential energy of fossil fuels, but with no CO2 emissions.
Avoiding CO2 emissions
There are two general ways the chemical energy of fossil fuels can be tapped without emitting CO2 to the atmosphere:
- Thermal cracking. H-C bonds are broken, yielding hydrogen and carbon. Solid carbon has commercial value, and is easy to store;
- Steam reforming. High temperature steam converts hydrocarbon feedstocks to hydrogen and CO2.
The pure carbon side product makes cracking an attractive alternative — in principle. Quite a lot of R&D has been devoted to improving the energy efficiency and economics of this approach. A recent DOE report provides a good overview. However, almost all of the work seems focused on carbon black as the primary product. Carbon black has significant value as an industrial product, but the market would be quickly saturated if cracking were employed at the level needed to address energy markets. In most cases, that wouldn’t be possible anyway; the overall energy efficiency of most cracking processes is too low. They consume more energy than the resulting hydrogen can supply.
An exception is found in some of the more recent work to crack light hydrocarbons — methane in particular — using molten metals as both a catalyst for the cracking reaction and a medium for separating the resulting carbon and hydrogen. The process is energy efficient, but so far the reactors are too expensive for the modest production rates they’ve been able to demonstrate. Commercial feasibility may yet be realized, but it doesn’t seem immanent. Hence, I’ll set aside further consideration of cracking and focus on reforming.
Steam reforming of hydrocarbons to yield hydrogen and CO2 is well established. It’s widely used to produce hydrogen on an industrial scale for oil refinery and fertilizer plant operations. But as the foundation for a prospective hydrogen economy it has two drawbacks. One is that, for economic efficiency, current processes need the economies of scale that come only with large chemical plants. They’d work in an all-out national program to implement a hydrogen economy, but are complex and not practical for small seed projects in local communities.
The second drawback is that steam reforming produces CO2. It thus depends on some form of CCS or CCU to avoid releasing CO2 to the atmosphere. The “carbon capture” part or CCS / CCU is no problem; the SMR process can be — and often is — engineered to deliver a nearly pure CO2 side stream. Nor would the “storage” part of CCS be a technical problem; there are proven methods. However it would face political hurdles. Greenpeace and some influential “climate hawks” are firmly against it.
CCU (“U” for “utilization”) has the inverse problem. There are no political hurdles; everybody likes the concept. But all of the uses implemented or proposed to date are limited. They’re niche applications that couldn’t begin to handle the amount of CO2 that large scale H2 production by reforming would create.
Before we can adopt national energy policies supporting CCS, Greenpeace and like minded organizations must be willing to relax their opposition. They’ll need to drop their insistence on “keep it in the ground” as the only acceptable policy for slashing carbon emissions. If they do so, they can — in my opinion — expect to see adoption of robust pricing on carbon emissions, an end to industry-sponsored campaigns to undermine climate science, and much more rapid progress overall toward an emissions-free energy economy.
Those will strike some as wild assertions. I’ll explain shortly below why I think they’re justified, but first I need to say something about recent technical developments. They have a lot to do with my optimism.
Better approach to steam reforming?
Research sponsored by CoorsTek has led to the development of a new type of hydrogen permeable ceramic membrane. Although hydrogen permeable ceramic membranes are nothing new, this one is different. Other membranes with selective permeability to hydrogen have been passive. They rely on diffusion of hydrogen across the membrane from a region of higher H2 partial pressure to a region of lower partial pressure. This one is active. It uses an applied electric potential to actively transport hydrogen across the membrane. That enables it to scavenge hydrogen present in a reaction chamber at low concentration and compress it for delivery as a pressurized H2 output stream.
The beauty of that is something that perhaps only a physical chemist can appreciate. I’ll try to explain anyway. It has to do with the thermodynamics of equilibrium reactions. Readers with no interest in chemistry may skip ahead. In the next section, I’ll take up the practical implications.
In a chemical reaction, chemical equilibrium is the state in which both reactants and products are present in concentrations which have no further tendency to change with time. That’s quoted from Wikipedia. It’s a fair definition for chemical equilibrium, but doesn’t capture what is meant by an “equilibrium reaction”. The term refers to a reaction that takes place at or near equilibrium conditions. When equilibrium is approached, the concentrations of reactants and products approach stability, but not because the reactions themselves actually stop. Rather, reactions and their corresponding counter reactions balance. The distinction between reactants and products blurs, as the reactants for one reaction are the products of the counterreaction.
If the concentration of one of the species in a chemical equilibrium is disturbed — say by adding more of that reactant species from an external source — the rate of reactions that consume that species increases. It continues until a new equilibrium is achieved. Conversely, if one of the species is systematically withdrawn, the rate of reactions consuming that species slows. The reactions that produce it will continue at the same rate, with the net result being an excess of production over consumption. The surplus production of the species in question offsets what was withdrawn.
That’s just what happens in the case of withdrawal of hydrogen from a reaction chamber — or tube, more likely — fed with high temperature steam and methane. At the high temperature in the tube, the steam partially dissociates into oxygen and hydrogen. As the initial mix of steam and methane entering the tube flows, dissociated hydrogen is steadily withdrawn through the membrane running the length of the tube. That leaves oxygen to react with the other gases in the tube.
The flowing mix becomes increasingly depleted in methane, enriched in CO2. Carbon monoxide (CO) is also generated, but the concentration never gets high. The CO gets oxidized to CO2. At the end of the tube, what exits is a stream of nearly pure CO2 and steam, with only traces of CO, hydrogen, and methane. The steam is easily condensed and the CO2 can be compressed to a liquid. Any residual hydrogen, methane, and CO are then recirculated.
A key point is that for this type of progressive quasi-equilibrium reaction process, there is very little increase in system entropy and loss of exergy. Being endothermic, the overall reforming reaction:
CH4 + 2H2O + heat ⇒ CO2 + 4H2
requires an input of thermal energy to drive it. That thermal energy converts to increased potential energy in the produced hydrogen over and above that of the input CH4. Thus, the electrical energy expended to pump hydrogen across the membrane does not end up as waste heat; it supplies the high temperature thermal energy needed for the endothermic reaction.
To quantify, one mole of CH4 (methane) has a combustion energy of 890 kilojoules (kJ); four moles of H2 yield 1,144 kJ. The difference, 254 kJ, is what must be supplied to drive the reaction.
Another point worth noting is that the electrical energy supplied to the membrane also serves to maintain a low concentration of hydrogen within the reaction gases. That speeds reforming and eliminates the need for the complex multi-stage reaction process that conventional SMR requires. So the electricity serves triple duty — thermal energy, H2 separation, and acceleration of the reforming reaction. That triple duty contributes to the low capital cost and to the energy and economic efficiencies that the process promises for even small implementations.
In terms of hydrogen out to electricity in, the new process is about six times more productive than conventional electrolysis. If the hydrogen produced is subsequently converted to electricity in a fuel cell or a combined cycle gas turbine, it will give roughly four times more electricity out than was consumed in making it — an effective 400% return on electrical energy invested. Compare that to the 40% return for the P2G2P scheme often proposed for long term storage of renewable energy. Leveraging the chemical potential energy in methane makes the process literally ten times more productive.
The process should also be quite productive in terms of capital. Its simplicity means that the equipment should be cheap enough for economical operation at low capacity factors. It could be run from cheap surplus power on an as-available basis. Moreover, the process can be throttled over a wide range under real time control to provide regulation service to the grid. It would thus make an ideal discretionary load.
As a consequence of these factors, a hydrogen-fueled power system using these reformers could operate equally from H2 made using cheap surplus power and stored temporarily, or from natural gas reformed in real time to meet demand. The latter requires a portion of generated power to be diverted back to the reformer; it’s less desirable than using stored H2 made from surplus power, but it’s only needed when H2 stores have been depleted by an extended period of sub-normal power production from other sources. The fallback option means H2 stores could be optimized for typical rather than worst cases. Dunkelflaute weather or seasonal variability would not be problems.
Of course, all this is speculative. To date, the new reforming process has been demonstrated only in the lab and in small prototype. But CoorsTek appears to have big hopes for it. They project its use in private refueling stations for hydrogen vehicles in homes supplied with natural gas. That may sound radical, given the internal 800 ℃ operating temperature and the safety concerns surrounding hydrogen in general. But the reformer would be stationary and well insulated, while small hydrogen tanks can be isolated below ground. So it could work.
Regardless of whether this advanced SMR takes off at the home level, there is little doubt that it would be practical at a neighborhood level. It would provide an ideal backup supply for local microgrids. Hydrogen fuel cells of the type developed for the auto industry have been projected by DOE to cost $53 per net kilowatt in mass production. That’s incredibly cheap; $1000 per kilowatt is usually the low end of capital cost for the least expensive class of commercial power plants.
The $53 figure is admittedly a projection for production levels that have not yet been reached. But current costs can’t be very much more than that. The fuel cell system in the Toyota Mirai is spec’d at 113 kW, and Toyota is selling or leasing a few thousand cars per month at a pre-subsidy price point of roughly $60,000. Even if the fuel cell system were to account for as much as 25% of the vehicle’s MSRP, it couldn’t be costing Toyota more than $130 per kW.
The low specific capital cost of these hydrogen fuel cells means that it wouldn’t be a problem to maintain a large reserve capacity idled much of the time. The ramp rate of automotive style fuel cells is better than the most responsive type of gas turbine power plants. They’re not as fast as batteries, but are in the same class. Hence they can easily be paired with variable renewables or with baseload generation to follow the load curve.
Use of hydrogen in neighborhood micro-grids would be self-contained. Making H2 by advanced SMR avoids the need for hydrogen pipelines. Other gas users need not switch from natural gas to hydrogen. The waste heat from producing power could be used to supply district hot water and boost overall energy efficiency. A nice perk is that the process is insensitive to the specific hydrocarbon being reformed.
Operating from natural gas might be the normal mode, but the process should work equally well for bottled propane or natural gas liquids, or from tanks of methanol, ethanol, or even gasoline. The only proviso is that the fuel be free of sulfur or halides that might degrade the membrane. Hence the system could operate from local stored fuel reserves if cut off from its NG supply.
That just leaves the question of how to dispose of the CO2 from the reformers.
Disposition of CO2
Different solutions for disposition of CO2 from the reformers are available depending on the scale of operations. The first couple aim to minimize the capital expenditure for demonstration projects:
- For small neighborhood demonstration projects, CO2 from the reformer could simply be pumped into pressure bottles and sold into the merchant gas market. Small delivery trucks would visit the reforming station nightly, pick up a load of filled bottles, and leave off a load of empties.
- For somewhat larger district level demonstration projects, CO2 from the reformers could still be handled by pumping into pressure bottles, but the bottles would need to be of the larger type permanently mounted on semi trailers.
- For deployment at the level of small cities or suburban counties, the merchant gas market is inadequate to accommodate the amount of CO2 that would be created. In states with oilfields, the EOR market should be large enough. Oil prices permitting, the CO2 price operators would be willing to pay could be sufficient to fund collection and distribution even without credits from carbon pricing. But to keep operational costs low, it might be necessary to build pipeline links from CO2 collection centers to the oilfields.
For large scale H2 deployment or in states with no oil fields, even the EOR market would be inadequate for the amounts of CO2 created. Long term geological storage in depleted oil and gas fields or in deep saline aquifers would be needed. An extensive network of pipelines for delivery of CO2 to injection wells would have to be built, with a robust price on carbon emissions to pay for construction and operation.
The good news is that with hydrogen from reformed fossil fuels used only to provide flexible backing power for renewables or peaking supply for base load generation, the amount of CO2 created would be an order of magnitude less than it would be for handling CO2 from the current fleet of coal-fired power plants.
The political landscape
I said earlier I feel that if Greenpeace and like-minded environmental organizations were to drop their insistence on “keep it in the ground” as the only acceptable way to cut carbon emissions, robust pricing on carbon emissions would be adopted fairly quickly, and the net rate of CO2 emission reductions accelerated. What aspects of the political scene support that conclusion?
I believe that much of the popular opposition to CCS stems from its usual association with the concept of “clean coal”. The surface mining methods employed to minimize the cost of coal are devastating to the environment; there’s good reason to oppose coal apart from CO2 emissions. But at least within the US, coal is losing out to natural gas for economic reasons, even without pricing carbon emissions. Its competitive position would be weakened, not helped, by carbon emissions pricing with credits for carbon sequestration.
CO2 emissions per MWh of electricity from coal are double what they are from natural gas to start with, and post-combustion carbon capture adds far more to the energy and financial cost of coal than reforming adds to natural gas.
“Keep it in the ground” is a rationale for opposition to CCS, but it’s counterproductive for ending coal use while being a loaded gun aimed at other parts of the fossil fuel industry. It threatens the economies of oil exporting nations. The oil and gas leases that would have to be abandoned and the reserves that would have to be left in the ground have present values measuring in the tens of trillions of dollars. No wonder we see well-funded think tanks that challenge climate science and project dire consequences for the economy from carbon pricing. Given what’s at stake and their financial firepower, is it any surprise that fossil fuel interests have been able to defeat policies that threaten them?
Some 3500 years ago, the legendary Chinese military strategist Sun Tsu, in his treatise on “The Art of War”, wrote that one should always leave one’s enemy an avenue of retreat. Not out of benign concern for the enemy’s welfare; it was simply that cornered enemies fighting for their lives and fortunes fight very hard. It may still be possible to defeat them, but it will be costly. In general, the less an enemy stands to lose, the easier the victory over them.
CCS is an avenue of retreat for non-coal fossil fuel interests. It removes the gun to their heads that “leave it in the ground” represents, and offers a prospect of profitable redirection of business. They are, after all, uniquely positioned with the skills and equipment that large scale CCS would employ. But a robust price on carbon emissions, coupled with a corresponding credit for sequestration, is absolutely essential to enable that redirection.
Opponents of CCS need to figure out whether they are more interested in slaying enemies or in achieving their war’s core objectives. I know what Sun Tsu would advise.