A comparison between the Polymer Electrolyte Membrane (PEM) Fuel Cells and the Alkaline Fuel Cells (AFC) with liquid electrolyte

by K. Kordesch, Technical University Graz, A-8010, Graz, Austria

 

Abstract:

This comparison will discuss the electrode and system design differences between fuel cells with an immobilized electrolyte (e.g. contained in a matrix, in a gelled layer or in a membrane) and a free flowing alkaline or acidic electrolyte. The inherent electrochemical and technological advantages and disadvantages are pointed out. The likely reasons for the historically often changed selection of certain fuel cell types for various applications are presented, relating the choices to opportunities and technical circumstances. These choices are by no means final and still open, as will be shown. However, they depend not so much anymore on technical or electrochemical problems and solutions, but on available materials, manufacturing feasibilities and especially on the cost of the total system to be used for a specific purpose. A modern example is the planned use of PEM Fuel Cells for Electric Vehicles. These membrane fuel cells are supposed to replace a combustion engine which costs $ 50 - $ 100 per kW power output, but do require at the present time an investment of thousands of Dollars per kW. Why did so many proponents chose this system and / or this application? Can other systems do better? This comparison suggests alternatives.

 

The Technical-historical Background

Early fuel cells were built with liquid electrolytes, solutions of potassium- or sodium hydroxide or diluted acids. The classical example is the alkaline fuel cell by Francis Bacon, who built his models around 1945 in a bathtub. The advantages of Hydrogen - Oxygen fuel cells for space application became clear and it was also the only application which could afford them. It led to the elimination of any mechanical pumps, which were not reliable enough. The use of matrices (for instance microporous asbestos) soaked with KOH became standard for NASA space fuel cells and it is still the present method, in spite of the fact that better matrices were found. Membrane fuel cells (by General Electric Co) existed already 30 years ago, but the possibility of pin holes and gas cross leakage was too high at that time. The Nafion membranes changed that.

The fact that liquid circulating electrolytes offered great advantages for heat management and water removal requirements was believed to be negated by the disadvantage of creating parasitic shunt currents in series connected cell assemblies. Delayed start-up procedures often led to the reversal of cells and irreversible cell failures. The use of combining fuel cells with rechargeable batteries in a hybrid system and the advantages in this respect (see reference) were not recognized.

Phosphoric acid fuel cells, were then considered the best fuel cells for power plants. They tolerated the high percentage of CO2 in any not completely shifted reformer gas. The phosphoric acid crystallized or gelled in the electrolyte space when heated to the operating temperature of about 210 oC thereby creating a fixed electrolyte which gave the desired performance. This system could not exchange or circulate the electrolyte. The use of the heat for water warming, etc, was done in separate heat exchangers, which, of course did use pumps.

As a matter of convenience and for time saving, all the testing of fuel cells was (and is) done in a continuous mode operation. The need to operate fuel cells in an interrupted fashion, with sometimes long idle periods did not seem to be important. However it is ! If one considers the operation of an electric vehicle in normal use. Kordesch demonstrated this fact the first time with his Hydrogen- Air City car, driving it daily, for over 3 years on public roads in the early 1970's (see reference).

Consider the use of a gasoline engine: it operates only when the car is in use. 4000 operating hours (about 4 months when counted continuously) correspond to about 200.000 miles of operation, may be within two years or 10 years of calendar time. Only large, special gas motors can run without failure for thousands of hours continously.

A similar situation is obvious in fuel cell operation of an electric vehicle. It must be able to shut down completely for longer time periods, hours, days, weeks, etc., it must be safe in the garage, with turned off gas supply, at ambient and even low temperatures.

The alkaline fuel cell with circulating and removable electrolyte can do it. It should also be mentioned that on activated stand, without load, fuel cell electrodes and catalysts degrade more than under load. The high voltage on open circuit is the reason for carbon oxidation processes, catalyst changes, etc. Unfortunately, the alkaline matrix fuel cells with immobilized KOH electrolyte combined all possible disadvantages: the electrolyte had to stay in the cells, residual carbonate (from any uncomplete air cleaner) accumulated, separators (matrices) deteriorated, gas cross leakage started during drying out or crystallisation periods during storage times without careful maintanance. Therefore it is not surprising that the AFC's got into miscredit (with help from the competition).

Also the phosphoric fuel cell manufacturers found out soon, that prolonged storage of a shut down system changed the crystallized phosphoric acid and it could not be changed. The saving point was (and is) that power station usually stay activated and are supposed to operate continuously (for 40.000 hours or about 4 years at least). This was achieved by close to commercial units in the 200.000 kW range.

No endurance reports were really published about the Megawatt size units which were tested for a few thousands of hours (a few months) and then shut down for inspection. With molten carbonate and solid state systems shut down procedures are avoided if possible.

PEM fuel cell operate perfectly under 45 oC. They can be shut down. They are ideal for small specific applications which can also afford the hig cost. Above 55 oC the water balance and membrane dry-out difficulties start and require carefully controlled accessories. More about that later, when costs of fuel cells and accessories are compared.

 

System design Comparison between Alkaline and Membrane Fuel Cells

A frequently mentioned point is the high current capability of the PEM fuel cells compared to other systems. In this connection it is important to look at the efficiencies achieved at different current densities. The cell voltages allow a rough estimate. Alkaline Air-Hydrogen fuel cells have an operating voltage of about 0.8 V at 300 mA / cm2. Assuming that PEM fuel cells deliver 600 mA/cm2 at 0.7 V, this is a loss of 10-15 % in efficiency.

Engineers calculate that doubling the current density corresponds to half the weight or size of the fuel cell. Yes, but only the stack is reduced. The required system accessories and controls may more than offset this advantages, not even considering life expectancy questions which are not investigated easily and catalyst cost relations.

The high current multiplies the membrane water balance difficulties, especially increases the air flow requirements. High air flow means a careful control of water production (load and temperature depending) versus membrane drying out. Air supplied by a controlled compressor can only solve such problems at a considerable over-all system efficiency loss and a high cost increase.

The alkaline system produces the reaction water at the hydrogen anode, the acidic membrane system at the air-cathode. That makes the difference. Hydrogen can be efficiently circulated in a closed system with a simple condenser arrangement, Air must be blown through (80 % is nitrogen!) at a high stoichiometric rate (or pressurized) if the current is high.

High current densities and high air flow are also modeling parameters for the alkaline cells, especially if the size of the soda-lime CO2-Air cleaner is considered , but it is not a serious system problem because a suitable cleaner (which lasts for 5000 hours) and the KOH can easily be exchanged (like an oil change in an automobile) or refilled.

System considerations also determine the cost of the system. Prices in the range of more than a few 100 Dollars per kW eliminate the chances for general electric vehicle applications. Catalyst questions come up. Due to the sensitivity of PEM cells against CO, the level of noble metal catalysts is kept high. Alkaline fuel cells using large surface carbon based low level Pt-metal catalysts are not much sensitive to CO in the hydrogen, a lower grade of purity hydrogen or converter gas can be used. The use of non-noble metal catalysts in PEM cells is not likely because of the acidic pH. Alkaline fuel cells can at least at the air electrodes use conventional low-cost perovskites or spinells.

 

The Fuel Question:

Liquified or compressed hydrogen is not likely to be used in every-day vehicles. Perhaps metal hydrides. However, the most frequent suggestions propose methanol converted to a hydrogen rich gas. The production of a pure fuel is possible, but not cheap.

The efficiency of the total system is not so good anymore, especially if compared with improved engines. We propose to use liquified ammonia, which is available in low pressure cylinders. Ammonia can be converted in a cracker to 75 % hydrogen. The efficiency of conversion is high. The capacity per weight and volume of ammonia as fuel surpasses methanol (see reference). Ammonia is produced worldwide in amounts of 100 million tons per year and is one of the mostly used chemicals, produced from natural gas. Ammonia is distributed and available in farm operations. For decades the household refrigerators worked with ammonia as cooling medium. The smell is strong, but that is an advantage because it will indicate any leakage in the cracker or fuel cell system.

Ammonia poisoning - if it should happen accidentally, is medically completely reversible, which is not the case with methanol. Small amounts of ammonia are even proposed as gasoline additive, it makes the engine exhaust gas absolutely NOx-free.

Acidic PEM fuel cells can use hydrogen produced in ammonia crackers, but the remaining traces of ammonia must be removed. Alkaline fuel cells are less sensitve, the electrolyte simply rejects ammonia and the residual ammonia in the hydrogen can be recirculated through the cracker catalytic heating unit.

 

SUMMARY

PEM-Fuel cells presently dominate the low temperature fuel cell literature. The expectations are very high and supported by large companies. The competition in the automobile and oil industry is high and only the cost factor will finally be deciding the use of fuel cells. If the expectations are not fulfilled, fuel cells will revert back into the already repeatedly experienced role of the promising power source, available in 15 years. The choice of applications will change with time and experiences.

Literature:

See the paper presented by K. Kordesch at the International Power Sources Symposium, May 12, 1999, in Brighton, England.