[September 11, 2002]

Regenerative Fuel Cells: Eliminating The Infrastructure Requirement

BY DANIEL SWEENEY

Both the strength and weakness of fuel cell technology sui generis is the matter of fuel. By utilizing an external, high energy fuel source for the chemical reactants that generate electricity, a fuel cell is able to achieve a considerably higher energy density per mass than does a battery using a closed cycle, relatively low energy internal chemical reaction for the same purpose. At the same time, by disposing of the end products of the reaction outside the fuel cell proper, the fuel cell avoids the buildup of internal residues that eventually degrade the performance of batteries. Nevertheless, it can be argued that the fuel itself represents a significant obstacle to the establishment of fuel cell technology in anything beyond a few high value applications.

The fuel in most types of fuel cells is ultimately hydrogen because a chemical reaction involving hydrogen and oxygen actually generates the electrical charge; however, the hydrogen itself may or may not be stored in free form. The fuel actually supplied to the hydrogen-based fuel cell by the user may be gaseous or liquid hydrogen in pure elemental form stored in tanks or metal hydrides, or, conversely it may be methanol, ethanol, natural gas, or even gasoline. In cases where fossil fuels are used, additional chemical reactions are required to extract the hydrogen, and those additional reactions normally occur in a separate device called a reformer. On the other hand, in the methanol and ethanol types, both reactions occur right at the membrane with no external apparatus required, and for this reason these types are known direct methanol or direct ethanol fuel cells. Metal hydrides represent a special case. A reformer as such is not used, but the need for at least one additional processing stage results in increased structural complexity.

Other types of fuel cells exist which are not based upon reactions combining hydrogen and oxygen, and one of these, the zinc air fuel cell developed by Metallic Power (Carlsbad, CA) is a regenerative type. Assessing the competitiveness of these purely metallic types with more traditional hydrogen fuel cells is difficult at the current minuscule level of commercial penetration for either basic type.

In any case, where alcohol, zinc, or pure hydrogen is used, the lack of a comprehensive distribution infrastructure unquestionably hinders mass acceptance of the technology. If the fuel is hard to get, the potential user is going to weigh the inconvenience of refueling against the high power output, low pollutant emissions, and relatively small form factor of the fuel cell. And where some kind of fossil fuel is the primary fuel for the fuel cell, one of the chief rationales for using fossil fuels in the first place, i.e., reducing dependence on fossil fuel, is thereby negated.

But what if an external fuel source were not required? What if a sort of fuel-less fuel cell were possible that dispensed with the need for infrastructure? That is in fact the promise of the regenerative fuel cell, in many ways the most interesting fuel cell technology on the horizon today.

The Versatile Regenerative
A regenerative or reversible fuel cell is one that does not require either fossil fuel or some poorly distributed chemical to sustain its chemical reactions. Instead this kind of fuel cell has the capability of reconstituting its fuel from the waste products generated in the final reaction that produces the positive ion and free electrons which give the fuel cell its charge.

The actual reconstitution involves an input of electrical current, whether from the grid or from an independent power source such as a wind turbine, solar array, or microturbine. In the case of those regeneratives operating on hydrogen (the majority) what occurs is electrolysis. Waste water produced by the fuel cell is broken down into hydrogen and oxygen, which are then recombined to produce water, as well as free electrons. A different chemical reaction occurs in the regenerative zinc fuel cell, whose sole representative is made by the aforementioned Metallic Power.

A regenerative fuel cell mimics, to a considerable extent, the operation of a rechargeable battery, but in theory it holds a number of significant advantages over any secondary battery yet developed.

"The biggest advantage so far as we're concerned is energy density," says Tim Conver, president of AeroVironment (Monrovia, CA) a diversified alternative energy company. "Our device gets about 600 watt hours per kilogram, much better than any battery we've found."

Furthermore, the regenerative need not share the vulnerability of the rechargeable battery to overcharging, or to the effects of full discharge, or for that matter, to the progressive breakdown effects and performance degradation of the secondary battery. No one's predicting an unlimited lifespan for regeneratives, but most researchers in the field believe that operating lives of up to a decade may be possible.

One could of course argue that a regenerative fuel cell actually is a sort of battery in as much as the reactants no longer come from an external source but cycle indefinitely within a closed system. Ultimately, it's a matter of semantics, but in fact most regeneratives are derived from ordinary PEM (polymer electrolyte membrane) fuel cells; in other words, they don't resemble any form of secondary battery in their basic chemistry. What distinguishes them from conventional PEMs is the inclusion of a specialized hydrogen storage system, which, as it happens, is also what distinguishes them from secondary batteries, where a separate storage system as such does not exist. In regeneratives the reactants are ultimately segregated in structurally sealed chambers rather than being merely separated by an electrolyte as in all true secondary batteries.

The Regenerative Advantage
On the surface the regenerative fuel cell appears extremely attractive as an energy source. It shares the very high energy density of conventional PEM systems, and thus lends itself to a range of applications where high power in a small form factor is desired. These would include backup power in residential and industrial settings, energy storage for off-grid electrical systems, automotive and marine propulsion systems, industrial self-propelled robots, power tools, portable computers, and perhaps even cellular phones. Beyond these few generalizations, further characterizing the potential of the regenerative fuel becomes difficult because several different design approaches are currently extant.

As indicated, most regeneratives are PEM types, and these include fuel cells under development by Proton Energy Systems (Wallingford, CT), Hydrogenics (Toronto), and AeroVironment. Two other companies active in the field, Metallic Power and Ovonic Batteries (Troy, MI), employ non-PEM designs utilizing metallic fuel sources and electrolyte membranes that do not require noble metal catalysts. Another company, Regenesys of Cambridge, England makes a regenerative based on a totally proprietary operating principle of which little has been disclosed.

PEM Regeneratives
PEM type regeneratives, based as they are on a reasonably mature fuel cell technology, are fairly straightforward in terms of their electrical generation functions. Like all current generation PEMs, they remain expensive to manufacture, but their general feasibility has been established. They do, however, present the user with some problems in respect to storage of the hydrogen gas. The AeroVironment and Hydrogenics products both require pumping systems to pressurize the gas subsequent to its release, and that increases the weight and electrical inefficiency of the overall system. Proton Energy claims to be able to generate hydrogen gas at pressure, eliminating the need for an external pump, but has not disclosed the technique by which this is accomplished.

Storage of high pressure gas is always a concern, both for reasons of safety and of footprint and weight. The carbon fiber and Kevlar composite storage tanks that have recently become available promise to solve the weight problem, and greatly reduce the danger of rupture due to their much superior tensile strength by weight as compared to metal containers, but such composite tanks remain expensive, adding to the cost of what is already a prohibitively expensive technology in most applications. And they still take up considerable space. Storage of the hydrogen in metal hydrides is the reverse; hydride storage is space efficient but carries with it a significant weight penalty.

Another problem with PEM type regeneratives, and presumably with other types as well, is that an electrolyte that can reverse the chemical reaction cannot be optimized for both halves of the reaction, and so separate electrolytes are usually employed for electrolysis and energy generation (a regenerative that uses only one membrane is properly called a reversible). This is true of the products of AeroVironment and Hydrogenics, both of which are actually combining electrolyzers and fuel cells in one housing.

Of the metallic fuel type manufacturers, Ovonic Batteries will not disclose if their design is truly reversible, but Metallic Power indicates that its zinc air fuel cell uses an entirely separate device for regeneration.

Obviously, a true reversible would be simpler, more compact, and arguably cheaper to manufacture than a two electrolyte regenerative, but the latter appears to be further along the road to productization. The fact that AeroVironment has chosen a two membrane design, in spite of the stringent weight requirements of its application, argues strongly that the design problems in perfecting reversibles may be particularly intractable.

A final disadvantage of PEM regeneratives, and all PEM fuel cells for that matter, is the requirement for a precious metal catalyst in the membrane ¾ platinum in most current designs. Much design effort has gone to reducing the amount of platinum needed to sustain the reaction, and current PEM fuel cells use a fraction of the metal required in the designs of a decade ago. Still, some have argued that if PEM fuel cells were to be truly mass produced, they would place a strain on the world's supply of the precious metal unless further significant reductions could be achieved. The issue is far from settled at present.

Metal Fuel Regeneratives
Metallic Power's and Ovonic Batteries' designs are so different from one another that drawing generalizations regarding metallic compound types from them is rather difficult. Both do share one important advantage, however: namely, that neither gold nor platinum are required in the electrolyte.

The Metallic Power design, which is currently in beta testing and is aimed at providing backup power in cellular telephone base stations, uses a supply of zinc pellets and air to fuel the reaction. Zinc oxide, the principal waste product, is then broken down into metallic zinc in the reconversion process.

In the Ovonic design, which incidentally is not the only fuel cell to utilize metal hydrides as a fuel source, the hydrides function as particularly efficient means of storing hydrogen -- the density of hydrogen by volume far exceeding that of the best high pressure tanks, and, at the same time, posing no safety or handling problems. While Ovonic would disclose few details of operation, the basic cycle consists of extracting the hydrogen from the hydrides according to an already established fuel cell technology, and then reconstituting the hydrides according to a proprietary process for which the company is seeking a patent.

As a matter of interest, Ovonic Batteries' parent company, Energy Conversion Devices, is a diversified manufacturing and research firm with key patents relating to recordable optical discs, nickel metal hydride batteries, and solid state memories, and a total patent portfolio numbering in the hundreds. Energy Conversion Devices is promoting the new fuel cell technology in partnership with ChevronTexaco, which like many traditional energy companies, has begun to invest heavily in alternative power.

Toward Commercialization
As of this writing, only one regenerative fuel cell is commercially available, that manufactured by Regenesys, a multi-kilowatt model intended to provide massive energy storage at power generation facilities and backup power for large industrial settings. The AeroVironment regenerative, which has been demonstrated in the world record setting drone aircraft, Helios, is a completed product, but is not on the market as yet. As AeroVironment president Conver explains, "We haven't been able to identify markets for it other than our own drone aircraft. The technology is just too expensive at present."

Ovonic, Proton Energy, Metallic Power, and Hydrogenics all see the cellular telephone carriers as the likeliest early market for the devices. Explains Proton's P. Tombaugh, "the smaller footprint of the regenerative as compared to rechargeable batteries is the big selling point."

Susan Connell, a spokesperson for Metallic Power, adds that "the cathode life is still an issue at this point, which makes backup power the most promising early application. Eventually we may see the use of the batteries in vehicles or power tools."

Atakan Ozbeck, an analyst with Allied Business Intelligence (Oyster Bay, NY) who tracks the fuel cell industry, believes that micro regenerative fuel cells are not near to being perfected though. "The utility of such a device is so great, that if anyone were close to succeeding, we'd be hearing a lot more announcements."

Proton's Tombaugh cautions that in certain applications currently dominated by secondary batteries, the latter are unlikely ever to be superseded by regeneratives. "The fundamental problem with regeneratives is that they are inefficient energy storage devices in spite of their high energy density. You may be able to store energy as hydrogen at a level exceeding 90%, but a fuel cell is less than 50% efficient at converting that energy into electricity, so the round-trip efficiency is less. With secondary batteries the basic overall efficiency is in the 90% range, best case. In a public utility or distributed power situation where you're trying to store large amounts of energy, this is unacceptable."

Relatively low power density, that is low peak to average power output, as compared to batteries is another inherent limitation that will surely limit the applications in which these devices are used. Where brief spurts of high power are required, as in starter systems or residential power systems where motors and transformers draw high in-rush currents, fuel cells must be supplemented with batteries or ultra-capacitors.

Nevertheless, in view of the highly accelerated progress in fuel cell design over the course of the last two or three years, and the predictions of many manufacturers and analysts of numerous product introductions in the mid term, the peculiar advantages of the regenerative are apt to receive more and more attention.

Daniel Sweeney is a freelance writer on telephony and power systems. He lives in Burbank, California.