Regrettably, a relatively small number of skeptics have been given undue publicity, presenting inaccuracies, distortions, errors and outright falsehoods, even in respectable peer reviewed journals. Misleading papers have appeared this year in Science, The Scientific American, and Solar Today to name just a few. It is important to try to present a more balanced view, especially as the future of the electrical industry will be shaped to no small degree by the evolution of the "hydrogen economy".

Skeptics
Anti-hydrogen published papers all suffer from some or several of the following thought/analysis weaknesses:

• A failure to see the nature and evolution of the HE in practical terms and in the context of the total energy picture, including obvious alternatives.
• Postulating a near 100% switch to hydrogen, implying renewable electricity generation, conversion to hydrogen, transport of hydrogen to region of energy need, and then reconversion into electricity.
• Incorrect figures for efficiency of hydrogen generation and compression.
• Wrong infrastructure assumptions, primarily relative to hydrogen transport.
• Storage density assumptions that are much worse than is already being achieved.
• Analyses based on old technology, with the implicit assumption of no technological progress.
• Consideration of hydrogen production only by electrolysis.
• Cost comparisons done per unit of energy contained instead of per unit of service provided.
• Failure to consider whole system implications when estimating energy efficiencies and costs.
• Grossly exaggerated safety issues, and failure to consider the natural safety features of hydrogen.

One of the most frequently quoted references is a paper^1,2 by Bossel and Eliasson which has the appearance of being quite scientific, but is in fact largely wrong, due to both direct errors and wrong thinking.

Quick research
Any of the authors that have misrepresented the practicality of the HE could have discovered the following facts with an hour or so of googling:

• Stuart Energy sells prepackaged alkaline electrolysis systems with 85% cell efficiency LHV, and 73% system efficiency compressed to 5000 psi, using unheated feedwater.
• Fully absorbed cost of hydrogen at 10% after tax IRR, with electricity at average industrial cost of 4.83cents/kwh, would be $5.60/kg for such a system as a forecourt supply, and at today’s capital cost. Electricity is 88% of this cost.
• Such cells can reach near 90% efficiency with 100+ deg. C feedwater
• Successful experiments have been conducted with ionic activators to raise the efficiency of alkaline electrolysis by 10%.
• PEM electrolysers operate at 80 - 85% efficiency, and have a theoretical upper limit of 94% if we can develop high temperature membranes. Progress on high temperature membranes is promising.
• Solid oxide steam electrolysers have reached 90 -95% efficiency in the lab, but are not yet commercially available.
• Hydrogen can be compressed to 5000 psi at >90% efficiency and 95% efficiency with electrolyser precompression.
• The oxygen "byproduct" of electrolysing water can be used in many processes to raise the efficiency and lower the NOX of those processes. Oxygen enhanced combustion raises combustion efficiency by 25-30%.
• PEM air fuel cells are already at 45-50% efficiency. PEM direct oxygen FCs are at 59% efficiency at rated load and 69% at 20% load. Up to 80% is expected in the future.
• Through 2002 almost all fuel cell design/improvement was empirical. Only in 2003 did advanced computer models become available to accelerate progress.
• Microturbine/SOFC hybrid FC systems are at 53% electrical efficiency installed, and modeling during 2004 has shown the path to 67%. The goal is 70% by2010.
• Hydrogen can be generated during off peak hours to operate coal and nuclear generators at max. efficiency 24/24. Waste heat from the generators will improve electrolysis efficiency. Nighttime electricity can be profitably provided at <4 cents/kWh.
• Targeted worst case storage density for automotive fuel tanks is >5.5% of hydrogen by weight, and 11% at 5000 psi has already been achieved in conformable tanks that exceed requirements for overpressure and cycling, and 11.4% in lithium nitride.

When these facts are factored into anti-hydrogen papers, the papers become nonsensical. New equipment that uses all of the efficiency advances listed above, can reach >84 efficiency for hydrogen compressed to 5000 psi, and given energy credit for the oxygen generated, total system efficiency can reach at least 90%. Hydrogen cost would then drop to $4.60/kg. Electricity to hydrogen to electricity efficiency (e.g. for wind turbine buffering) will reach >60% efficiency, double what is assumed now.

What is the Hydrogen Economy?
The primary uses of hydrogen will be replacement of dwindling supplies of petroleum derived transport fuel, and buffering of intermittent renewable electricity sources. Hydrogen appears to be the best choice for the transport task, but critics feel that batteries or battery/FC stores like zinc/air fuel cells will prove to be viable alternatives. Quite possibly all 3 will prove feasible.

With NG also in decline, we have to look at how it is used. 8% is used today as a source of hydrogen, and almost 90% of the hydrogen is used to produce gasoline, and ammonia for fertilizer. As NG prices rise in the USA ammonia production is being moved offshore. As petroleum availability declines, the need for hydrogen for gasoline will decline. NG cost is already at a level to make hydrolysis competitive with NG reforming. This is the likely path to large-scale commercialization of hydrolysis.

More important is the 22% of NG used for residential purposes, mainly home heating, and 23% used for electricity generation. Some of the electricity generation is CCGT with efficiencies near 50%, but the bulk is older SCGTs with <25% efficiency. Globally the efficiency is probably <30% today. If the residential gas was consumed in an efficient SOFC with waste heat captured for heating and cooling, the NG could be used about 70- 80% efficiently (instead of the centralized <30%). The fuel cells exist, but are still expensive. However as NG prices rise, and carbon penalties enter the scene, gas suppliers or electric utilities will probably get into the home fuel cell business, instead of building more centralized generating plants. This is already happening in Japan. RMI sees the secondary aspect of the HE as building cogeneration, with NG being a bridge to the hydrogen economy. Certainly such a step can at least double the efficiency of NG consumption, and is therefore desirable, but the benefit with hydrogen is not clear.

How much hydrogen?
The main purpose of the HE is to use hydrogen to replace petroleum as a portable fuel for vehicles. For cars the key is to have lightweight, efficient FC vehicles, e.g. "hypercars" with fuel cells. The present car fleet averages 24 mpg. The Prius HEV already gets better than 40 mpg, and is not lightweight. A hypercar version of the Prius would get 80-90 mpg, and a full size family car would get at least 70 mpg. If we then replace the ICE with a FC we will more than double tank to wheel efficiency, giving us >140 mpg equivalent. In trucks and buses there is not much room to improve rolling efficiency, but better FC saving is possible, probably close to factor 3. Overall factor 5 is probably achievable. (Note-PEM fuel cells are already being used to power submarines). The approximately 28 quads of petroleum used in the USA for road transportation can be replaced in the next 25 to 30 years by efficiency plus < 6 quads of hydrogen. Buffering 15 quads of wind at 2010 expected efficiencies can be done with less than 3 quads of hydrogen. A mere 10 quads of hydrogen will go a long way to powering the so-called hydrogen economy. If batteries and/or metal/air fuel cells prove out, even less hydrogen will be needed. Another alternative to hydrogen for transport is likely to be oil generated from the thermal depolymerisation of agricultural waste, scrapped tires and municipal sewage. If this technology is widely deployed during the next 2-3 decades it could provide 1-2 Gb of diesel oil equivalent, or perhaps 20% of our current annual oil consumption, replacing at least 1 quad of the above estimated hydrogen.

Where do we get the hydrogen?
It has been estimated that conversion of current USA waste biomass could supply 2.7 quads of hydrogen. The 15 quads of dispatchable wind described in “The Energy Challenge 2004 – Wind”, would produce enough excess energy to generate 4 quads of hydrogen above that needed for buffering. Clean coal electricity generation byproduct would supply another significant increment. There are several other potential sources, including photolysis from waste water treatment plants or algae, solar thermochemical water splitting, and industrial process byproduct. We probably need additional electricity for no more than 1-2 quads of hydrogen. This amount can be supplied from the present electrical infrastructure on less than 6 hours per night.

Cost considerations
The 2005 cost of hydrogen by electrolysis given above is $5.60/kg with electricity at 4.83 cents/kWh and 73% efficiency. At 3 cents/kWh and 80% efficiency this would translate to $3.45/kg. Hydrogen from NG reforming, with NG at $6.00/Mbtu and electricity at 3Cents/kWh has been estimated by Amory Lovins as $2.50/kg. Hydrogen from waste biomass is estimated at $2.80/kg. It seems safe to count on a hydrogen cost =< $5.00/kg, with some probability of reaching $3.00/kg average. As 1 kg of hydrogen is comparable to 1 gal. of gasoline on an energy basis this cost still appears uncompetitive. However, recall that it is the cost per unit of service, not per unit of energy that counts.

At a before tax cost of $1.50/gal. and 25 mpg, gasoline costs 6 cents/mile. At a cost of $5.00/kg and 140 mpg equivalent, hydrogen costs only 3.6 cents/mi. At $3.00/kg the hydrogen powered car would only have to deliver 50 mi/gal equivalent to equal gasoline at today’s price. This is much less than the FreedomCAR goal, and the FreedomCAR does not consider low rolling resistance design. Hydrogen is not just competitive, it is sufficiently attractive relative to rising gasoline prices, to drive a free market economy switchover during the coming years. However an intelligent energy policy could accelerate the transition, and make it smoother rather then letting it be crisis driven.

Interestingly, given credit for the byproduct oxygen produced, hydrolysis would be competitive with reforming NG at a NG price of <$7.00/Mbtu, and we are already there. Critics have suggested that the oxygen credit is insignificant because oxygen is priced at about 10 cents/kg today, or $.80 of oxygen per kg of hydrogen. This criticism misses the point that oxygen today is a byproduct of producing nitrogen, and is not produced in enough quantity to support a use like oxygen enhanced combustion on a useful scale. Produced in sufficient quantity, e.g. as a byproduct of hydrogen production, oxygen would have a considerably higher value because of its contribution to combustion efficiency, making hydrolysis even more attractive.

The other key issue is fuel cell cost. In 1990 the estimated high volume production cost of PEM fuel cells was $3000/kW, dominated by platinum at 20 grams/kW. In 2004 platinum is down to 0.8 grams/kW and the estimated high volume FC cost is $225/kW. A FC car is estimated to need 50-80 Kw and the target price for FCs is $30.00/kW by 2015, but this does not take into a count a super efficient hypercar. The hypercar needs closer to 20 kW, so $100.00/Kw would be good, and that may be achieved before 2010.

What about infrastructure?
For filling station forecourt generation of hydrogen, energy will be transferred to point of use as electricity, using the existing infrastructure. For locally sited building cogeneration NG will be transported by the existing infrastructure. The major need for new infrastructure will be the forecourt hydrolysers. The technology already exists but large refueling stations would need scaled up units. The cost of converting the needed filling stations over a 30 year period is estimated to be 100-200 billion dollars, much less than the petroleum industry puts into E&D and infrastructure over a similar period (or less than the war in Iraq in a much shorter period).

Of course there is always the chicken and egg problem of which comes first, the vehicles or the infrastructure, and without some volume how do we get the infrastructure cost down. The cost problem is likely to be solved by Japan and China, both of whom are pushing ahead very rapidly. Japan has hydrogen FC cars on the road and China has ordered 10,000 hydrogen fueled buses to be delivered in time for the 2008 Olympics. This problem will get solved, whether by us or not.

As NG becomes scarcer it can be supplemented with hydrogen to make town gas for many uses, and finally, much of the NG pipeline network can be lined with a polymer barrier and used for hydrogen transmission. Salt caverns have already been used successfully for large-scale storage. A large share of the hydrogen needed for wind buffering can be stored under pressure in the wind turbine towers.

Hydrogen safety^3,4
RMI has produced a few foils nicely addressing this issue. The main points deal with the inherent characteristics of hydrogen that make it safer than gasoline. Hydrogen is 8 times lighter and 4 times more diffusive than methane and 12 times more diffusive than gasoline fumes. It dissipates quickly and cannot accumulate unless trapped by e.g. a ceiling. In air, exposed to a flame or spark, it burns long before it reaches an explosive concentration, and is 22 times less explosive than gasoline. It burns upwards in a narrow plume and emits very little radiant heat. You almost have to be in the flame to be burned. Tanks designed to contain hydrogen have survived 50 mph crash tests without rupturing.

Transition to the HE
NG production in North America is likely to be halved before 2015. It is likely that onshore production plus LNG imports can then maintain 10 Tcf or so of supply for a couple of decades after that. Much of the early NG shortfall will be compensated by efficiency improvements and demand destruction. Large-scale hydrogen replacement will be 1-2 decades away. If petroleum availability declines at 5%/yr, it will take 26 years to get to ¼ of the present supply. Very early on, rising gasoline prices will stimulate the shift to more efficient cars, and finally fuel cell cars. One leading market research firm has estimated that HEVs will be 20%+ of North American auto production by 2010, and that without incentives. Hydrogen FC cars will likely appear shortly thereafter, but demand will grow gradually. Without energy policy incentives car fleet replacement is unlikely to be complete before 2040. A government “feebate” policy to encourage replacement of gas guzzlers with efficient vehicles could accelerate this process. Wind energy growth is also likely to be a 30-year proposition.

Conclusions

• The driver of the HE will be declining NG and petroleum supplies.
• The primary need is efficiency, especially more efficient FCs and cars.
• The primary problem is still cost, mainly of FCs and storage tanks, but the major cost issues will be resolved by a combination of research and volume production.
• Supporters, in addition to skeptics, are suffering from important misconceptions, e.g.:
  • The government is spending major sums to drive down FC cost below what would be necessary with more efficient cars. Focusing on more efficient cars would be more productive.
  • RMI sees NG as a bridge, rather than a problem and driver.
• The electric utility and gas industries need to start looking at distributed FC cogeneration, urgently.
• Large-scale electrolysis, raising the return on off-hour generation, is a major profit opportunity for electric utilities.
• The transition to the HE will be a rapid evolution, not a revolution, and it is already well under way.

References:
1) http://www.efcf.com/reports/E02_Hydrogen_Economy_Report.pdf

2) http://www.hyweb.de/News/LBST_Comments-on-Eliasson-Bossel- Papers_July2003_protected.pdf

3) hgovttp://www.rmi.org/images/other/Energy/E03-15_H2FutureOfEnergy.pdf

4) http://www.hq.nasa.gov/office/pao/History/SP-4404/ch8-6.htm H2  

Other Sources:
http://www.eere.energy./hydrogenandfuelcells/pdfs/35948.pdf

http://www.eere.energy.gov/hydrogenandfuelcells/annual_report03.html

http://www.eere.energy.gov/hydrogenandfuelcells/pdfs/iid1_milliken.pdf

http://www.eere.energy.gov/hydrogenandfuelcells/pdfs/separ_02_intro_anderson.pdf

http://www.fuelcellseminar.com/pdf/2003/Davis.pdf

http://www.protonenergy.com/index.php/html/gasproducts/datacharts/index.html

http://www.sciencenews.org/articles/20041030/bob10.asp