I'm going to make a prediction today: you will never drive a hydrogen fueled car. Although hydrogen does indeed have some benefits in certain applications, it's my task today to separate the reality of useful fuel cells from the hydrogen hype. That may seem like a bold statement to you now, but by the end of this article, you'll understand why.

Much has been made of the concept of a "hydrogen economy," because it offers the possibility of a portable fuel that can be generated from any number of sources and consumed without greenhouse gas emissions.

That's a major win-win against the twin devils of peak oil and global warming, and as such it has attracted the support of an unlikely alliance including environmentalists, technologists, politicians and automakers.

It's important to realize that hydrogen is not a fuel source; it's an energy carrier. Hydrogen does not exist freely in the universe; it's always bound to something else. So it takes an investment of energy to free hydrogen from its existing arrangement and make it available as a stored fuel.

The hydrogen fuel cycle goes like this: hydrogen is liberated from some source, compressed or liquefied for storage and transport, then "burned" in a device called a fuel cell, in which energy is captured from the hydrogen as it combines with oxygen from the air to form water. The captured energy can be used to power electric motors and generators, and the only emissions are pure water.

It's an elegant vision, and has captured the imagination of such luminaries as Stan Ovshinsky, wunderkind founder of the advanced energy company Ovonics (Energy Conversion Devices, ticker symbol ENER). Proponents imagine a future wherein the original hydrogen is generated by the electrolysis of water, using electricity generated from renewable sources. Thus the hydrogen fuel cycle would begin and end with plain water, and would still offer portability, as well as a basis for a distributed clean, green energy cycle.

They envision homeowners generating their own renewable power (using solar, geothermal, micro-hydro, or whatever they've got) and turning it into hydrogen that they can store on-site, then consume in their hydrogen-powered cars or in the fuel cell stack that powers their home.

Unfortunately, the vision breaks down when we analyze the energy return on investment (EROI) of the process. According to the second law of thermodynamics, when energy is converted from one form into another, a little energy is lost in the process, usually as heat. Essentially, every time you convert energy, you pay a tax.

ROI: The Hydrogen Buzzkill
Calculating the EROI of a hydrogen fuel cycle requires a good many assumptions about how it will be generated, transported, stored and consumed. So different sets of assumptions can produce quite different results. In the aforementioned example of home-based hydrogen generation, where the hydrogen is generated and consumed in a single site, losses along the way are low. But when it is used in a vehicle, losses are much higher.

Let's explore a typical calculation of the EROI of the hydrogen fuel cycle for cars:

1. Suppose we generate the hydrogen by the electrolysis of water. First we must "rectify" the grid's AC electricity into DC, at a cost of about 2% to 3% of the energy contained in the hydrogen.

2. Now we can electrolyze the water, but that process is only about 70% efficient, so we lose another 30% there.

3. Now we have hydrogen gas, but it takes up a lot of space. We could compress it to around 10,000 psi to make it fit in a reasonably sized tank, which costs another 15%. But even then, it would only have about one-fifth of the energy density of gasoline, and the pressurized tank needed to store it is very heavy, large and expensive. So if we wanted to use it in a vehicle, we would have to liquefy the hydrogen by cooling it down to about -253°C and keep it in a pressurized, insulated container instead. This process would cost another 30% to 40% of the energy in the hydrogen.

4. We lose some more during storage because hydrogen boils off above -253°C, so it's very difficult to keep it from escaping its container. In vehicles, about 3% to 4% of the hydrogen boils off every day. And at least 10% of the hydrogen will boil off during delivery and storage.

5. Then we burn the hydrogen in a vehicle's fuel cell at an efficiency of about 50% (for a proton membrane fuel cell stack).

6. And finally, we lose another 10% of the energy that makes it to the electric motors driving the wheels, because they are only about 90% efficient.

7. In the end, about 80% of the original energy generated in order to produce the hydrogen is lost, for an EROI of 0.25. Since it doesn't pay to have an energy regime with an EROI of less than one, hydrogen cars seems a permanent improbability.

Carbon Emissions Persist
There's another dirty little secret about hydrogen that is rarely mentioned by hydrogen hypers: the vast majority of hydrogen manufactured today is not made from the hydrolysis of water, because of the energy inputs needed. Instead, it's made from natural gas, because it's a ready and easily exploited feedstock for hydrogen production that can be transported more easily in liquid form. And that means that the hydrogen production does, in fact, produce carbon dioxide emissions, effectively nullifying the environmental benefits of fuel cells.

When natural gas is the feedstock, as it is today, the hydrogen fuel cycle amounts to going around the block to get to the back door, for nothing.

A final problem with the concept of a "hydrogen economy" is that we'd essentially need a whole new infrastructure for it, from "wells to wheels." Nothing in our current energy infrastructure is compatible with hydrogen.

A major reason for that is that it's the smallest element, so it wants to escape from just about anything you use to contain it. Tanks, pipes, valves, and fittings all along the way leak constantly. For another, it's highly reactive, and makes metal brittle and prone to leakage. The storage and transport losses can be considerably worse than in the above example.

Starting Over
To build a "hydrogen economy," we would need to start over with everything. Hundreds of thousands of miles of pipeline, 90,000 new pumps at service stations, 210 million vehicles—everything.

Given what we know about the peak oil situation, one wonders just how much of the remaining fossil fuel energy would be needed to replace all that stuff. Let's just say it would be a sizable chunk, a chunk we'd probably be better off using for food and shelter, and making solar panels and wind turbines.

And then there is the old chicken-and-egg problem: who's going to pony up the hundreds of billions (actually probably closer to the low trillions) of dollars to build all that infrastructure until the cars are in the showroom, and who's going to put hydrogen cars into a sufficient number of showrooms until the customer has easy access to a refueling station?

There are a few other alternative hydrogen infrastructures, but each has daunting challenges associated with it:

• Hydrogen could theoretically be produced on-board a vehicle from liquid methanol or gasoline, but it's going to be difficult, inefficient, and expensive. Big R&D money needed for that direction.

• Hydrogen could be produced at local centers, but then we're back to the aforementioned problems of storage, transfer, and the lack of infrastructure.

• It could also be produced right at the fueling station, from methane gas or from water via electrolysis, but the cost of building such stations will be enormous and the infrastructure needs would be great (either to ship natural gas to the stations or to upgrade the grid to handle all that extra electricity). And again, who's going to make that investment before the cars are there?

Now, although it doesn't make sense as a transportation solution, in the right applications hydrogen can be a useful storage system. For example, a large commercial building equipped with a solar system and a fuel cell stack could generate, store, and use much of its own power with minimal losses along the way and no emissions. In such applications, hydrogen is smart. Consequently, I believe the future is bright for companies that focus on that market segment.

But you will never drive a hydrogen car.

Chris Nelder is a solar designer in Marin County, California and a contributing editor for EnergyandCapital.com as well as GreenChipStocks.com; an investment advisory service that focuses solely on renewable energy and organic food markets.