Researchers at Ford Scientific Laboratories, in Dearborn, Mich., first conceived AMTEC technology in 1968 when they identified and patented a converter known as the sodium heat engine. This heat engine was based on the unique properties of b-alumina solid electrolyte (BASE), a ceramic material that is an excellent sodium ion conductor but a poor electronic conductor. BASE was used to form a structural barrier across which a sodium concentration gradient could be produced from thermal energy. The engine provided a way to isothermally expand sodium through the BASE concentration gradient without moving mechanical components. Measured power density and calculated peak efficiencies were impressive, which led to funding from the Department of Energy for important material technology development.

In the early 1980s, the technology was brought to the Jet Propulsion Laboratory (JPL), in Pasadena, Calif., for possible use as a conversion device in a spacecraft power system. Converters on board spacecraft such as Voyager and Galileo operated at approximately 5 percent efficiency, while this new conversion technology was predicted to provide more than 20 percent efficiency when fully developed.

Work at Ford and JPL continued through the 1980s as materials for the electrodes were developed and bench-top devices were tested. Efficiencies of up to 19 percent were measured, and cells were operated for up to 14,000 hours, although the devices were not yet capable of operating outside the laboratory environment.

The AMTEC cell, now the size and configuration of a battery, is no longer a bench-top device with sodium plumbing, filters and pumps. Instead, heat is added to one side of the cell at 600° to 850°C and rejected at the other side at 150° to 450°C, producing dc-electric power from the terminals at up to 40 percent thermal-to-electric conversion efficiency. The process is simple because as long as there is heat, there is efficient power.

Much of the recent AMTEC development has been directly related to space power applications: Small radio-isotope-powered AMTEC systems are being considered by JPL for planetary exploration; solar thermal systems are being developed by the Air Force Phillips Laboratory, in Albuquerque, N.M., for earth-orbiting spacecraft; and the Robotics Institute at Carnegie Mellon University, in Pittsburgh, is considering an AMTEC power system for lunar surface rovers. Predicted cell efficiencies for these systems range from 20 to 30 percent, with cell power densities near 80 watts per kilogram. The efficiency, power density, and high heat-rejection temperature, as well as the reliability of direct thermal power, make AMTEC technology attractive for such applications.

The Pluto Express Mission is one of the proposed NASA endeavors that may benefit from AMTEC. This mission, currently in the definition and spacecraft design phase, is proposing two spacecraft that will fly past Pluto between 2010 and 2015 and take a number of scientific measurements and photographs. Each spacecraft will weigh approximately 100 kg and need 75 to 100 W of electric power. The light level at Pluto's orbital distance—approximately 30 times the earth-sun orbital radius at flyby—has been describe as "deep twilight." A solar-powered spacecraft would need a solar panel with an area of over 250 square meters to provide 75 W of power using current technology, which is not a practical solution.

The converter technology traditionally used on such missions as Pioneer, Voyager, and Galileo has been thermoelectrics. This technology has proven to be highly reliable. System efficiency, however, is generally only 5 to 6 percent. A thermoelectric power system designed for the Pluto Express spacecraft would weigh approximately 18 kg. As an alternative, a radioisotope heat source coupled to a device to convert the heat directly to electric power is much more compatible in size with this small spacecraft.

An AMTEC system designed for the Pluto Express would be about four times more efficient than a thermoelectric system. It would also use less heat-source material and reduce the power system's mass to only 6 kg, 10 kg less than a thermoelectric unit, a significant reduction on a 100-kg spacecraft. Design studies have shown that the high efficiency and heat rejection temperature of the AMTEC cells result in a compact power system that fits neatly on the spacecraft.

AMTEC cells are supported by their mounting studs from a housing that cools the cell condenser and radiates the waste heat to space. The other end of the cell faces the heat source and receives the thermal energy to be used in the conversion process.

The system's high-efficiency potential and ability to mate with a range of heat sources have led to work at the Air Force Phillips Laboratory on a solar-thermal AMTEC power system for earth-orbiting spacecraft. Solar energy is concentrated by parabolic mirrors or Fresnel lenses on an AMTEC receiver, which has both AMTEC cells and thermal-energy storage material. During the insolation period of the orbit, a portion of the solar energy is converted to electric power by the AMTEC cells, and another portion is stored in the solid-liquid phase change of the thermal-energy storage material. During the eclipse period of the orbit, the thermal-energy storage material releases its stored energy to the AMTEC cells, keeping the spacecraft fully powered during this phase of the orbit. The system that has been designed will provide 1 kW of power to the spacecraft.

This power system design offers several features important to spacecraft operation in low and middle earth orbits (LEO and MEO). The thermal energy storage can be designed for a very high number of thermal cycles and can therefore have long LEO lifetimes. The AMTEC cells are radiation hard, which will enable them to operate without degrading in the high radiation zones found in MEOs. AMTEC cells can also be low cost and could therefore reduce spacecraft power system cost. These features would be useful to both military and commercial communications spacecraft.

Development is proceeding on the cell and individual AMTEC components to improve efficiency. Cell durability and compatibility with a zero-g environment must then be demonstrated with a flight test. Life models for the converter that would predict power levels after 10 years of operation must also be completed. Once these primary issues are resolved, the technology can be made available for communications, earth observation, and space exploration.

The effort to develop AMTEC technology for space has produced converters with useful features for terrestrial power systems. Some of the applications include self-powered furnaces, recreational vehicle components, highway hazard lights, and portable/remote sites. The system designs and specific development, however, are still in an early stage.

To date, self-powered gas furnaces appear to be the largest and most likely ground power application of AMTEC. There are over 35 million installed gas furnaces in the United States and 2.5 million new or replacement installations each year. This represents a very substantial market for a device that could provide either emergency replacement power or dedicated (electric-grid independent) power for the furnace. A survey by the Arthur D. Little Co. found that at an installed cost of $400, 21 percent of homeowners in the United States would be likely to purchase a backup electric power unit for their furnaces.

There are several ways to extend the advantages of a natural-gas-powered electric generator. First, the same electric generator that provides power to the furnace blower could also provide power for the air conditioning system during the summer. This system would be independent of the electric grid problems common in many locations during peak cooling periods or when thunderstorms strike. Second, waste heat from a dedicated electric power system on the furnace or air conditioner could be used to heat water. These methods would put the dedicated power system to use year round, conveniently and economically.

In applications where the waste heat can be fully used, the consumer cost to generate electric power from gas can be based on the gas-flow rate required to produce thermal energy equal to the electrical energy generated. The thermal energy that would normally be considered waste heat in the thermal-to-electric conversion process can be used for space heating, thereby reducing the gas heater thermal load.

A typical electric power equivalent for natural gas is $0.012/kW-hour. By replacing electric power at this figure with a 500-W gas-powered AMTEC unit, a consumer would save approximately $40 in a year that required 800 hours of furnace operation. The replacement would require that the AMTEC-unit exhaust be routed through the furnace heat exchanger. The resulting annual cost savings, however, would significantly offset the initial cost of unit purchase and installation.

Dedicated electric power systems for these applications must be reliable and quiet and have a long operational lifetime. A 500-W AMTEC system has been designed to provide these features. The converter would initially be installed beside the furnace to minimize the impact on present furnace designs and allow after-market installation options in those regions with frequent electric power outages. A natural gas burner in the unit would heat a bank of AMTEC cells. Power from the cell bank would be run through an inverter to power the original blower motor(s). Efficient dc motors would be considered for furnaces where the AMTEC power unit is offered as part of the original equipment package, thereby eliminating the need for the inverter. The heat from the power system would be rejected into the furnace heat exchanger or routed to the water heater, thus recovering the thermal energy for household use.

The basic unit described for the furnace application is readily adaptable to a number of fuels and power levels, so the AMTEC device can serve a range of other needs. AMTEC units are lightweight, quiet, conserve fuel through higher efficiency, and minimize exhaust emissions by using external combustion. These features make the unit useful in commercial and military remote sites, recreational vehicles, and hybrid electric vehicles. Some of the small 10-W cell packs that have been the basis for laboratory development could be used for small power applications where a combustion source is available. Other potential applications include charging batteries for mobile electronics operations, supplying power for telemetry to transmit data from fixed remote locations, and operating small electronic appliances.

Many remote sites require 10 to 500 watts of power. The power systems for these sites may provide environmental, process or seismic monitoring, data transmission, cathodic pipeline protection, or low-level facility power. Many of these systems must operate reliably for many years on limited maintenance and refueling.

Thermoelectric generators are typically used where long life and high reliability are required. These generators operate at 3 to 7 percent efficiency, compared to the 18 percent efficiency produce by current AMTEC cells. Switching from thermoelectric to AMTEC could therefore lower substantial fuel costs and lessen the impact on the environment at these sites. The cost savings would be especially important where fuel must be transported to the sites by helicopter. At these sites, the transport cost is significantly higher than the fuel cost.

AMTEC systems have many other design options. The hot-side temperature range of the technology makes it compatible, as a bottoming cycle, with other conversion systems. It can, for example, be coupled with thermionic converters. The operating temperatures are well matched so that the heat rejection side of the thermionic cell can be directly coupled to the hot side of the AMTEC cell. The output currents of the devices are also well matched, so that the units can be connected in electrical series. Tests are currently under way at the Air Force Phillips Laboratory to investigate this coupling. The hot side of the AMTEC cell can be coupled to solid oxide fuel cells (SOFC) operating at 1000°C. The AMTEC cells would produce power, as well as control the SOFC temperature by modulating the AMTEC cell current. AMTEC cells can also use the waste heat from the high-temperature emitters of the thermophoto-voltaic converter.

Cold-side temperatures for AMTEC are compatible with other converter technologies. Both thermoelectric and the Rankine cycle could use the AMTEC waste heat. Process heat, as with any of the possible combined cycles mentioned here, could further extend the efficiency of these combined technologies.

Depending on market size, AMTEC systems could be produced cheaply. Cells are constructed out of abundant, inexpensive basic materials. Stainless steel or nickel alloys are used for most of the components. Small amounts of niobium and molybdenum are used to join the BASE to the cell structure and provide current collection from the BASE. The BASE itself is formed from aluminum oxide and sodium oxide, with small amounts of a lithium or magnesium oxide stabilizer. BASE production will still be the key cost element. Production cost estimates range from $0.05 to $1 per W, depending on production volume. Small, high-end markets, such as spacecraft and remote site power, will have higher production costs because of low volume and initially high BASE-tube costs. These markets, however, are already paying $10 to $20 per W for remote terrestrial systems and more than $5000 per W for complete spacecraft power systems.

All of the systems described above are built on the small AMTEC cells being developed for space power. The fully self-contained cells can be operated in a variety of configurations. Cells have been operated as single units and as cell bundles on flame heating. Single cells have been operated in electrically heated systems that provide controlled thermal boundary conditions. Power and efficiency have been measured under these controlled conditions and found to agree with the analytic models used to predict the performance of the various AMTEC systems.

Recent AMTEC development has focused on increasing the cell's thermal-to-electric conversion efficiency, now rated at 18 percent. The key to this efficiency has been the improved thermal isolation between the heat source and sink. Heat shields and materials with low thermal conductivity have provided most of this improved isolation. Heat shields, placed between the hot BASE tubes and the condenser, significantly reduce the radiation heat loss. Cell walls made from low-conductivity materials, such as stainless steel or Inconel, reduce the parasitic heat loss conducted between the hot and cold ends of the cell.

Features that provide additional gains in efficiency through modest improvements in thermal isolation and increased power density are now in the design and testing phase. Near-term goals have established 25 to 30 percent as a target that could be achieved in about a year. Further improvements are possible in later years through advanced materials and higher temperature operation. Numerical models, validated against test data, predict that efficiencies as high as 40 percent (with power densities near 500 W/kg) could be achieved with continued development. This efficiency and power density would open up additional applications, such as hybrid electric vehicles.

Increasing the power density in AMTEC systems is an important development because it reduces the size and mass of a system and improves efficiency. Volumetric power density depends on the power produced at the surface of the BASE tube and the amount of BASE tube surface area that can be located in a given volume. Current work has established a baseline for a BASE and titanium-nitride electrode combination that provides a known and repeatable power density. However, electrodes that will offer higher surface-power density have been identified. These electrodes will be evaluated as the cell design matures.

BASE-tube surface area located in a given cell volume can be increased by reducing the diameter and increasing the number of the BASE tubes. This design approach is used with the multi-tube cell. The single 1.5-centimeter diameter BASE tube is replaced by a shorter 0.7-cm diameter tube, which results in a more compact cell with higher power density. The efficiency also increases because the parasitic heat losses remain constant while power increases.

Efficiency and power density must remain essentially constant over many years of operation. Systems that provide spacecraft power must operate continuously for over 10 years, and systems for the self-powered furnace must be capable of many startup cycles over 20 or more years of service and two years of cumulative operation. Cell operating life depends on the behavior and stability of the cell materials during startup cycles and at normal operating temperatures. Materials in components such as the electrode and BASE must have similar thermal expansion coefficients and be well bonded to withstand many thermal cycles without increasing the charge transport impedance at the interface. These same materials must not undergo chemical or physical changes caused by diffusion processes that take place at high temperatures. Key components include the BASE, electrode, BASE-to-metal braze, feedthrough, and cell wall. The effects of temperature and thermal cycles, in conjunction with sodium exposure, are being examined to fully understand and control cell lifetime.

Material life models—and the supporting life test data for the cell and key components—are also important aspects of AMTEC development. Models that have been developed for constant-temperature electrode operation and validated against test data have shown that one or more of the electrode design options can meet system lifetime requirements. Models for other components are still needed. Early cell prototypes have been tested under continuous load at 600°C for over a year without performance degradation. Mechanical tests on BASE tubes operated under these conditions have shown that the mechanical strength of the tube is not affected.

Higher-temperature continuous- and thermal-cycling tests are still needed. Development of wicks with small pores (5 microns) and high permeability has advanced enough that cells can operate at temperatures up to 827°C. Tests at these high temperatures are being performed at the Air Force Phillips Laboratory in the vacuum environment typical of space power systems. Burner tests typical of terrestrial power systems and thermal cycling tests have also begun at Advanced Modular Power Systems Inc., in Ann Arbor, Mich. Current test systems are cycling the BASE, electrode, and BASE-to-metal seal between 300° and 800°C every 30 minutes. Cell operation under higher-frequency thermal cycling, and rapid cell-startup rates will also be evaluated. Preliminary cell testing with a 5-minute room-temperature-to-775°C startup has been successful.

To demonstrate the practicality of AMTEC for field and space applications, shock and vibration tests on a mock-up of an individual cell were also performed. A cell assembled with the BASE tube (but without sodium) was subjected to vibration test protocols for typical spacecraft launch vehicles. When shaken and shocked along both longitudinal and transverse axes, the cell withstood 19 g_rms vibration and 3800 g_rms at 1000 hertz shock loading without damage. Maximum loads of this magnitude could be encountered in launch accelerations and stage separation for some spacecraft operations. Generally, these loads exceed those experienced in handling, shipping, and operation.

A series of tests were performed to understand the effects of cell wall penetration during handling and field operation. In one set of tests, mixtures of air and water were suddenly introduced to the hot operating cell interior through a remotely controlled solenoid valve. These tests simulated accidental cell wall ruptures or corrosion that could occur in either air or water during power-producing operation. In another set of tests, an operating cell was penetrated by a high-velocity bullet to simulate a military application.

To produce maximum reaction rates, all tests were performed while the cell was near the maximum operating temperature. In every case, the test results were remarkably unspectacular. In the air and water injection tests, the cell wall temperature briefly rose by approximately 100°C. In the projectile penetration, a small puff of smoke was observed, but the temperature rise was small.

These tests suggest that the sodium-water reaction rates under these simulated failure conditions are controlled well enough to safely operate the cell in proximity to human activity. This reaction rate control is attributed to the small amount of sodium, generally less than 2 grams, and to containing sodium in wicks to minimize the exposed sodium surface area. Additional design and testing will be required to evaluate the implications of the sodium hydroxide that forms as a result of cell penetration. This material, often found in drain cleaners, is caustic; but in more than 40 cells built and operated over the last two years, no cell has posed a health hazard to a test operator.

The ability of AMTEC systems to meet the needs of several applications is made possible by a number of important engineering advances. Replacing the electromagnetic pump with wicks reduces the cell design complexity, eliminates parasitic heat losses through pump current leads, and makes the cell more compact. Introducing remote condensing increases efficiency by isolating the hot zone from the cold zone of the cell. It also improves the cell's reliability by reducing the thermal stresses on the BASE tube. Multi-tube cells will further increase cell power density. These advances are the reason the technology has emerged from the laboratory.

AMTEC technology will continue to move into the marketplace. Currently the converters are produced and sold in limited quantities for prototype testing and demonstration. Space power programs will continue for several years, spurring improvement of the cells. AMTEC cells will need additional modifications before they can be assembled into fully operational systems and placed in the field for testing commercial applications. Fabrication facilities will also need to be developed for both the cells and the BASE tubes. But based on the potential AMTEC systems have shown to date, their future looks promising.

Robert K. Sievers, Joseph F. Ivanenok III, and Thomas K. Hunt are on the technical staff of Advanced Modular Power Systems Inc., in Ann Arbor, Mich.

"A Thermoelectric Device Based on Beta-Alumina Solid Electrolyte," N. Weber, Energy Conversion, 14, 1-8 (1974).
"Thermoelectric Energy Conversion with Solid Electrolytes," T. Cole, Science, 221, 915 (1983).
"Reversible Thermodynamic Cycle for AMTEC Power Conversion," C.B. Vining, R.M. Williams, M.L. Underwood, M.A. Ryan, and J.W. Suitor, Proceedings of the 27th Intersociety Energy Conversion Engineering Conference, Vol. 3, p. 3.123, (1992).
"Performance of the Terrestrial Power Generation Plant Using the Alkali Metal Thermo-Electric Conversion (AMTEC)," E. Sasakawa, M. Kanzaka, A. Yamada, and H. Tsukuda, Proceedings of the 27th Intersociety Energy Conversion Engineering Conference, Vol. 3, p. 143, (Aug. 1992).
"AMTEC Power Systems for Remote Site Applications," T. K. Hunt, J.F. Ivanenok III, and R.K. Sievers, American Institute of Physics Conference Proceedings, 324, Vol. 2, p. 761, (Jan. 1995).
"Low Emission AMTEC Automotive Power Supply," T.K. Hunt, R.K. Sievers, and J.F. Ivanenok III, Proceedings of the 30th Intersociety Energy Conversion and Engineering Conference, Vol. 3, p. 145, (Aug. 1995).
"AMTEC Radioisotope Power System for the Pluto Express Mission," J.F. Ivanenok III and R.K. Sievers, Proceedings of the 30th Intersociety Energy Conversion and Engineering Conference, Vol. 1, p. 661, (Aug. 1995).

home | features | weekly news | marketplace | departments | about ME | back issues | ASME | site search

© 1995 by The American Society of Mechanical Engineers