Thermionics Quo Vadis?: An Assessment of the DTRA's Advanced Thermionics Research and Development Program (2001)

As part of the mandate from the Defense Threat Reduction Agency, the committee looked carefully at missions that might use thermionic power conversion in space. In short, there are no current or near-term missions for space that require a thermionic energy conversion system. Furthermore, thermionic technology has not been developed to the point where it is ready for an end user. However, the committee believes that one thermionic energy conversion subsystem using concentrated solar energy may be suitable for use in the far term.

Development of thermionic converters for space has been directed predominantly at applications that use a nuclear heat source. However, a few programs have been undertaken using a solar heat source that concentrates solar energy into a central receiver.

The first major program to consider a solar thermionic heat source was the Solar Energy Technology (SET) program at the Jet Propulsion Laboratory (JPL) in the 1960s. More recently, the U.S. Air Force (USAF) and NASA have considered in a limited way both solar thermionic power systems and hybrid propulsion and power systems. The hybrid systems would first use solar energy and a solar concentrator for thermal propulsion to raise a satellite from low Earth orbit to final orbit, such as to a geosynchronous orbit. Once the vehicle is on station, the same solar concentrator heat receiver would be used as a heat source for a thermionic power system.

In this chapter, the committee discusses potential missions that could use concentrated solar energy as the heat source for a thermionic power conversion system in both standard and hybrid systems.

Solar thermionic converters do not involve many of the problematic technical issues associated with nuclear thermionic conversion, nor does solar conversion technology have to contend with the social controversy surrounding the use of nuclear power systems. Most notably, in the technical regime, solar heated converter lifetimes can be very long compared with in-core nuclear thermionic systems, which must overcome the effects of a harsh radiation environment and fuel swelling. Historically, the problems associated with thermionic power systems have been the size and mass of the solar concentrator, in the case of solar thermionics, as well as the complexity and cost of developing an entirely new spacecraft power system in both cases, solar and nuclear. However, recent advances in solar concentrators have mitigated those problems to a large extent.

The primary benefit envisioned with a solar thermionic space power system, such as the high-power, advanced, low-mass (HPALM) system discussed in the next section, is that such a system could deliver high electric power, from 20 to 100 kilowatts, in a low stowed volume, lightweight package. At the projected performance parameters, a 50 kilowatt solar thermionic system would take up less room on the launch vehicle and weigh about the same as a solar array of current design for a 20 kilowatt communications satellite. It is in this high power range that solar thermionic systems have the potential to perform better than other power conversion systems. Below 20 kilowatts, however, solar thermionics would most likely not be com-

petitive when compared to the industry standard, photovoltaic battery systems. Of course, for missions where there is little or no available sunlight, such as outer planet exploration, solar powered conversion systems are not feasible.

Based on current preliminary studies, the committee believes that a solar thermionic power system could possibly facilitate the launch of non-nuclear spacecraft with up to 100 kilowatts of onboard power in a single launch with a standard U.S. launch vehicle. This achievement would more than double the power available from current photovoltaic systems, although future photovoltaic systems may also offer competitive performance. The key is to develop and demonstrate a long life solar thermionic power system that meets or exceeds the performance goals set for the HPALM system described in the next section.

The committee believes that the following types of missions fit the criteria for using solar thermionic systems if the thermionic technology has been developed to a point where it is demonstrated and readily available:

• Space transportation using electric propulsion,

• Advanced high power communications, and

• Space-based military applications.

Space transportation using electric propulsion is perhaps the most interesting and compelling of these potential solar thermionic missions. In typical chemical propulsion systems, combustion provides propulsion energy to accelerate the propellant. In electrical propulsion systems for space, electrical energy provides the electromagnetic and electrostatic fields as well as the heat to accelerate the propellant. The solar thermionic power conversion system could be used in a dual mode power and electric propulsion configuration. The power conversion system is first used for electric propulsion to maneuver a spacecraft into the proper orbit. Once the vehicle is on station, the system could then be used for the mission power requirements. Therefore, the power conversion system has two modes of operation, propulsion and power supply, hence the term “dual mode.” Electric propulsion should not be confused with traditional rocket engines or rocket launches from Earth to space. Electric propulsion is not used to place the spacecraft in orbit but, rather, to maneuver the vehicle once it is in space.

One unique benefit for electric propulsion systems is that nuclear powered and some solar powered thermionic systems designed to operate over normal capacity at the beginning of life, can operate in a surge power mode (also referred to as peak power mode). For example, in Figure 3.7 in Chapter 3, if the emitter temperature is increased from 1800 K to 2100 K, the output power can be doubled for a relatively short time, 30 to 90 days. To configure a system to operate in a surge mode requires that all components be designed to achieve equilibrium at a higher temperature. This type of operation could require spacecraft designers to accept a somewhat shorter vehicle operating lifetime. Such surge mode operation can be used to decrease the time required for orbit positioning or transfer. In this dual mode operation, electric propulsion can be combined with other mission requirements to achieve a synergistic benefit.

A wide range of missions could potentially take advantage of recent advances in electric propulsion systems. Electric propulsion uses fuel 4 to 10 times more efficiently than chemical propulsion and 2 to 4 times more efficiently than solar thermal propulsion.

With the smaller volume and mass afforded by a solar thermionic electric propulsion system, a number of configuration scenarios can be imagined. For instance, smaller, less expensive booster rockets might be used for launch, or more satellites might be launched per boost vehicle. Or, a larger on-orbit, station-keeping fuel payload might be allowed, which would extend the lifetime of a spacecraft on orbit.

The committee believes that spacecraft designers may require that many new high altitude spacecraft, such as satellites in geosynchronous orbits, be inserted above the Van Allen belt by their launch vehicles in order to avoid payload exposure to radiation. From that point, electric propulsion could be used to raise the vehicle to a final orbit. Currently, spacecraft are inserted directly into geosynchronous orbit, but this practice significantly limits the cargo space and mass available for payload.

The committee believes that the spacecraft design community may make use of high power, low mass applications that solar thermionics has the potential to offer. Higher available power levels would enable shorter insertion times when using electric propulsion in the manner described above.

Appendix C contains details of spacecraft mission considerations and describes a concept for a commercial system coupling high power to electric propulsion

for orbit transfer in order to reduce operational costs. A 100 kilowatt electric propulsion system made up of four 25 kilowatt thrusters is compared with the total impulse of the Thiokol Star 75 motor, a state-of-the-art solid propellant space motor. Based on cost estimates for a thermionic power system, cost savings on the order of $100 million could be realized by satellite operators. The savings associated with dual mode operation could also be substantial. This cost analysis, of course, assumes a fully operational and tested system that would require a substantial investment to develop.

Finding: Solar thermionic power systems may enable dual mode operation: electric propulsion for orbit raising and orbit maintenance and electric power for satellite payloads while on station.

Recent commercial space activities indicate a trend of placing larger and larger satellites into geosynchronous orbits. The geosynchronous market continues to grow, with larger satellites with higher power levels being built every year. Twenty years ago, aerospace companies were launching geosynchronous satellites that required 1 to 3 kilowatts of electric power. Presently, the standard power requirement for geosynchronous communications spacecraft is approximately 10 kilowatts. Satellites requiring 20 to 25 kilowatts are now being designed for possible use within 3 years.¹

A solar thermionic system could meet the need for potential future growth in the industry, while allowing the industry to explore even higher power applications that could benefit the consumer. For instance, as satellite broadcast power increases, consumers can use smaller and smaller receiver dishes on the ground.

Space-based radar systems have been studied extensively by the military over the past few decades. Most of these studies have recommended a combination of low orbits and low duty cycles. This combination is probably recommended based on the prohibitively large size and mass of traditional power systems required by a more advanced space-based radar system capable of broad coverage, detection of small targets, and frequent contact updates of multiple targets. Such missions require on the order of 100 kilowatts of available electric power. A solar thermionic system could drastically change the feasibility of such a mission. The higher power levels a solar thermionic system could allow for an affordable space-based radar system that improves on the mission performance now provided by airborne radar systems, such as the Airborne Warning and Control System (AWACS) and optical and Long Wave Infrared (LWIR) systems.

For a solar power conversion system using thermionics, solar energy is concentrated into a heat receiver using either reflective or refractive optics. This heat receiver then serves as the heat source for a group of thermionic converters that provide power to the spacecraft.

As presented to the committee by General Atomics, the HPALM concept uses a large off-axis, inflatable parabolic reflector to focus solar energy into a 2000 K heat receiver that is radiatively coupled to the thermionic devices (see Figure 4.1). The program is based on a concept developed in part by General Atomics. An AFRL program to study the HPALM design concept started in mid 2000.

In one approach being evaluated, the cylindrical thermionic converters are designed with an inverted multicell. In this design, the emitter is on the outside and the collector is in the center of the cylinder. This is a nontraditional approach to designing a cylindrical thermionic device. In traditional designs, the collector is outside the cylinder and the heat is generated internally, as it is with a nuclear thermionic fuel element (TFE), and then removed by cooling systems attached to the outside surfaces of the TFE.

In the HPALM concept, the thermionic element is inverted. Heat will be applied to the outside of the thermionic element cylinder, and heat pipes will be used to remove the waste heat, at approximately 1100 K, from the collector surface inside the cylinder. General Atomics anticipates reaching efficiencies of 20 to 25 percent for the converters. According to General Atomics, the conceptual system is sized to provide 50 kilowatts of

FIGURE 4.1 Artist’s rendition of the HPALM solar thermionic concept.

SOURCE: L.Begg, General Atomics, presentation to the Committee on Thermionic Research and Technology, August 2000.

electricity and has a calculated performance of 106 watts per kilogram and 80 kilowatts per cubic meter in a stowed configuration ready for space launch.

The HPALM concentrator consists of an off-axis inflatable dish with a thin film reflector for low mass and low stowed volume for launch. The concept is based on inflatable space structure development that has been sponsored by NASA and AFRL over the past several years. The highlight of this work, which is not directly related to the HPALM concept, was the deployment of a large inflatable antenna structure from the space shuttle in 1997. Large area, thin film solar reflectors have been flown a number of times and are now included in the Hughes Corporation’s concentrator solar arrays for some communications satellites. An inflatable concentrator configured specifically for the HPALM system does not, however, exist at this time. If an inflatable concentrator is not successfully developed, then the projected performance capability, as presented under solar thermionic systems in Table 3.4 (Chapter 3), would not be as appealing due to the increase in mass required for a noninflatable concentrator.

Another important component of the HPALM power system is the heat receiver. This unit would be designed to capture the solar energy from the concentrator through a small aperture. The concentrated solar rays would be partially absorbed and internally reflected to different surfaces. The reflected rays bounce between surfaces and are partially absorbed during each reflection process. This process repeats until essentially all of the solar energy is absorbed inside the heat receiver. The aperture would be kept small to minimize both heat loss and the loss of reflected solar energy back out of the aperture. Refractory or carbon materials can be used to absorb the heat at the working temperature of the receiver. An integral heat receiver with thermal storage designed for use with a 2000 K HPALM system does not exist at present, although the technology has been demonstrated at 1100 K by both NASA and the USAF. One final, yet important, aspect of the HPALM design is the mechanism used to transfer heat energy from the solar receiving cavity to the thermionic converter. The heat from the graphite block receiver is radiated to the emitter in General Atomic’s version of the concept. The HPALM concept does not call for any dynamic mechanical systems to transfer the heat energy.

The committee suggests that the eventual sponsoring agency align a thermionics research and development program to coordinate with the HPALM program.

Recommendation 3. The sponsoring agency should concentrate its near-term thermionic development work on a space-based solar thermionic power system, such as the high-power, advanced, low-mass (HPALM) concept.

The committee makes this recommendation based on the understanding that the HPALM system is designed around a thermionic converter power supply; therefore, successful completion of a thermionics research and development program would be a major HPALM program priority.

The committee finds the HPALM concept interesting for a number of reasons. The calculated performance as presented by General Atomics is approximately twice as good as that of current solar array technology in terms of stowed volume and mass, since the system would make use of a lightweight inflatable concentrator. From a power density perspective, the HPALM system is estimated to provide between 23 to 33 watts per kilogram, which is comparable to photovoltaic systems. A solar power thermionic system providing 50 kilowatts of electrical power could have many interesting potential applications. For instance, a 50 kilowatt HPALM system may enable electric propulsion used to raise satellites to middle and high orbits. The committee anticipates that HPALM performance should improve as the system design is refined. However, it should be acknowledged that comparing the performance of a future thermionic concept system to the current capabilities of a photovoltaic system is not entirely valid. Only when a detailed thermionic system has been designed or built can it be compared to a then-current photovoltaic system.

Based on space vehicle market trends over the past 10 years, the 50 kilowatt design proposed for the HPALM concept appears to be a potentially viable option. The solar thermionic system also appears to have a power density advantage when scaling to higher power levels, while a photovoltaic system has the advantage when scaling down. Compared with a photovoltaic space power system, a solar thermal system starts with a higher percentage of fixed mass. Thermionic generators are usually multikilowatt devices, and both the pointing system and the pressurization system for the inflatable structures, as proposed for the HPALM system, have the desirable trait of scaling nonlinearly relative to the increase in power production.² Therefore, the power density of solar thermionic power systems will most likely increase as they are scaled to higher power levels.

Finding: A solar thermionic conversion system can likely be designed, built, and flight demonstrated using known technology. However, since the lifetime, mass, efficiency, and cost of such a system are uncertain, they need to be addressed in future technology programs.

It is important to note that all of the committee’s assessments of solar thermionic power system technology are based on projected capabilities. Until the potential of such a system is proven, these assessments represent what might be possible. Even though there are potential positive benefits from a solar thermionic system, a number of technical challenges would have to be addressed before such a system could be definitively endorsed.

Another aspect of developing a new power system concept, such as a solar thermionic system, is that detailed design information must be available to the satellite designer before the system can be considered for use. Although the technology exists so that the components of a solar thermionic power system can be developed, no one has developed a detailed, peer reviewed design of a complete system. The committee expects that a HPALM program at AFRL may fill this need.

The committee identified several areas where technical issues must still be resolved. The list should not be considered to be inclusive.

• Power conditioning,

• Energy storage,

• Thermionic converter configuration,

• Cavity/receiver heat flux uniformity, and

• Pointing, tracking, and reacquisition.

The last two are not discussed in the text that follows.

An important consideration that has not been addressed in the HPALM concept description is power

conditioning. Thermionic converters tend to produce low voltage and high current while operating at high temperature. This particular configuration presents an especially difficult problem for a high power system on a spacecraft. Current high power spacecraft, 10 to 20 kilowatts, use bus voltages of 50 to 160 volts to minimize the mass of power conductors. The committee expects that as power levels increase from 20 to 100 kilowatts, voltages will continue to increase to as high as 270 or 300 volts.

A thermionic system such as HPALM is likely to have a much smaller voltage than this future specification even with all of the converters connected in series. The need to handle the high currents that would be characteristic of a high power thermionic system would require large diameter conductors. Such conductors can act as a source of heat leaks from a high temperature device.

Also, power conditioning systems would be very heavy, which is not a desirable property for space launch. For example, a recent power conditioning unit developed for a 5 kilowatt Hall-effect propulsion system that converted 28 volts to 300 volts weighed approximately 25 kilograms. These power conditioning issues can thermally impact both the performance of the energy converter and nearby components on the satellite.

Finding: Power conditioning technology issues are inherent to a space-based power conversion system. So far, this issue has not been addressed by the DTRA-sponsored thermionics program.

For any satellite power conversion system, energy storage must be considered to meet peak power demands and to provide power when the primary power system is not operational, such as prior to system startup during launch and deployment. For solar power systems, energy storage must be considered when the satellite is shadowed by Earth. There are two basic options for storing energy: batteries or thermal energy storage.

Battery system storage, the current baseline for the HPALM concept, is the same energy storage scenario used for photovoltaic power systems. When batteries are used, the solar power system is oversized to generate extra power during the sunlit portion of the orbit. The extra power is stored as electrochemical energy in a battery to provide power to the spacecraft during eclipse periods when sunlight is not available.

One technology now being developed is lithium ion batteries, which deliver up to 100 watt hours per kilogram. A portion of a battery’s energy between 40 and 80 percent is normally used during each eclipse. The amount used is driven by the required lifetime for the batteries since they last longer if they use less of their energy during each discharge cycle.

Thermal energy storage is an alternative approach to energy storage, an approach that may be more attractive than batteries for the HPALM program. A thermal energy storage device would be incorporated into the heat receiver, and energy for operating the spacecraft during eclipse would be stored in a phase change material. Such a material melts as heat is added, then refreezes when the heat is used during eclipse. This process allows the energy conversion device to operate continuously, even during eclipse. The thermal energy storage material must be selected for minimum mass, which can be achieved in part through a high heat of fusion. The material must also have a melting point that matches the operating temperature of the heat receiver. The material is enclosed in a heat exchanger containment system that enables the heat to transfer between the heat receiver and the thermionic fuel elements.

One candidate high temperature material that has been tested is silicon, which has a heat of fusion of 1.80 megajoules per kilogram. However, the melting point of silicon is 1673 K, which is compatible with a thermionic energy conversion system but is well below the heat receiver temperature of 2000 K proposed for HPALM. If the HPALM system used the new baseline of 1673 K to match the melting temperature of silicon, there would be a 30 percent decrease in power generation efficiency in the thermionic converters. A higher temperature material therefore needs to be identified if thermal storage is to be considered. The committee suggests that manganese oxide, with a melting point of approximately 2000 K and a heat of fusion of 0.77 megajoules per kilogram (213 watt-hours per kilogram), may be a promising alternative. However, it should be noted that should such a concept be implemented into a full system, the overall performance of the system with energy storage would be significantly lower.

For a thermal energy storage system, the usable heat of fusion of the pure material is reduced by the inefficiency of the electrical energy conversion process. This

reduction in the heat of fusion is typically fivefold. In the case of manganese oxide, the resulting energy storage system performance would be about 20.56 kilograms per kilowatt for the case illustrated in Table 3.4. This performance compares favorably with lithium ion batteries, which have been demonstrated to provide 18.33 kilograms per kilowatt. Since batteries can actually use only a fraction of this amount, thermal storage may add to the attractiveness of a solar thermionic system. However, this is an area that needs more investigation.

An added benefit of using thermal energy storage is that both the heat receiver and the thermionic converters are kept at a constant temperature throughout the life of the spacecraft. Without a heat receiver, the solar energy would be concentrated on different parts of the heat receiver cavity at different points of an orbit, thus creating hot spots in the cavity. If there is no thermal energy storage system to act as a buffer between the concentrated solar energy and the thermionic converters, they may produce different amounts of power depending on the amount of solar energy directed on any one converter. Unevenly heated converters would mean a significantly more complex power source. The use of isothermal heat pipe cavities has also been demonstrated as a method of avoiding hot spots in the cavity.

Another problem with having orbit-induced hot spots in the receiver cavity is thermal-induced mechanical stress. Varying temperature gradients resulting from solar energy concentration could shorten the life of power system components. A thermal energy storage system could eliminate deep thermal cycling for system components and increase the life and reliability of the system. However, such a system may not be practical for satellites placed in geosynchronous orbit due to the very long eclipse times.

Finding: A phase change design for thermal energy storage should be considered as part of solar thermionic system development.

A final consideration that concerns the committee is that the initial HPALM design includes the use of cylindrical thermionic devices. The HPALM concept proposed by General Atomics uses radiative coupling between a cylindrical heat receiver and a ring of several cylindrical converters that stand out from the receiver like spokes from a wheel hub. An alternative configuration could use planar thermionic converters to couple more directly to the outer surface of the receiver. It is not clear why the cylindrical converter design was the only converter configuration considered for evaluation in the HPALM system. The cylindrical TFE would offer more surface area for the emitter and collector and less surface area for the receiver. However, past solar thermionic designs have used a planar configuration that, in the SET program, for instance, has demonstrated very good performance and lifetime.

Finding: It is not yet known if the performance goals of obtaining a high power solar thermionic system based on the HPALM concept are technically feasible.

Finding: For the solar thermionic HPALM concept as presented by General Atomics, there are no obvious advantages to using a cylindrical geometry rather than a planar geometry for the thermionic fuel element.

To summarize, the committee’s assessment is that a space solar power system is the most promising near-term application for thermionic technology. However, while it is the most promising application, success is not certain. The history of spacecraft performance demonstrates that it is difficult to compete with photovoltaic power systems. Although photovoltaics have been used since the late 1950s, no other technology has been able to replace this technology as the power source of choice for Earth orbiting satellites. Significant progress continues to be made in photovoltaic converters. For instance, triple junction solar cells can now deliver up to 29 percent efficiency.

The key to competing with and potentially replacing photovoltaics as the space power source of choice is to optimize solar-to-electric thermionic system design and quickly demonstrate the capability of the thermionic technology. Success requires a near-term focus with an aggressive system engineering approach.

The primary goal of the Solar Orbital Transfer Vehicle (SOTV) program being conducted by the AFRL is to develop an orbit transfer propulsion system using concentrated solar energy to heat hydrogen. A SOTV could be used to raise a client spacecraft’s orbit, that is, to transfer its orbit to a higher elevation. The transfer vehicle could then detach and return to a lower orbit for refueling. Alternatively, the transfer vehicle could

remain attached to the client satellite and use the solar concentrator and a thermionic conversion system as a power supply for the satellite.

The program is in the proof-of-concept phase, and researchers are conducting experiments to verify component performance. Orbit transfer vehicles appear in the USAF Space Command Strategy Master Plan, but no orbital transfer vehicle has been constructed thus far, and no funds have been budgeted to do so (see Figure 4.2).The SOTV program is the only active program other than HPALM that the committee has identified that is considering thermionics as a power conversion mechanism. However, the committee does not advocate aligning a thermionics research and development program with the SOTV program. The primary focus of the SOTV program is on proving the viability of the hydrogen propulsion system. The committee understands that the limited availability of funding drove this decision and that a critique of the SOTV program in relation to the HPALM concept is beyond the scope of this committee’s responsibility. While not intending for this report to cast any doubt on the performance or capability of the SOTV program, the committee does feel obligated to identify the best potential match for thermionics technology.

The committee was interested in the thermionics testing done to date under the SOTV program. Specifically, the DTRA and AFRL sponsored three tests of thermionic element arrays at the New Mexico Engineering Research Institute (NMERI) of the University of New Mexico to demonstrate the readiness of the technology for the SOTV system. The tests were termed the string thermionic assembly research testbed (START) tests. The committee felt it needed to review the tests in detail since the poor START test results indicated that there was some fundamental failure of the thermionic converter technology.

The SOTV program is a follow-on to the hydrogen fueled Integrated Solar Upper Stage (ISUS) Orbital Vehicle program started in 1994, and the START tests were initiated under that program. Because the ISUS concept vehicle would need very high temperatures for the hydrogen fuel propulsion, thermionics was selected as an appropriate power conversion technology since thermionics also requires a high temperature heat source.

FIGURE 4.2 Artist’s rendition of a solar orbital transfer vehicle.

The START tests were designed to evaluate thermionic power conversion for use with a future SOTV spacecraft. Of special interest to the committee was the conclusion from these tests, namely, that there were too many technical difficulties to make thermionics a viable power conversion technology. In fact, the committee believes that the SOTV program’s shift in focus toward propulsion technology may be due in part to the poor START test results. Thermionic power conversion is still being considered as a possible power source in the SOTV program, but the technology is now in competition with Stirling engines, concentrating solar arrays, and other alternatives.

The committee examined the START tests in detail and found problems with the testing, analysis, design, and test fixture fabrication in the START series that should be carefully accounted for in any future tests. Because these problems bring into question the validity of the data that were collected, the committee determined that the viability of thermionic power conversion technology should not be judged based on these tests.

Finding: The string thermionic assembly research testbed (START) test series should not be used as a basis for evaluating the viability of thermionic power conversion technology.

When the ISUS orbital vehicle program was initiated, a contract was awarded to Babcock and Wilcox to develop a test fixture with a string of 8 series-connected thermionic converters to be used in the first test phase. A string of 16 series-connected converters was ordered to be used in a second test phase. The first test of the 8 series-connected thermionic converters was intended to identify any first-phase problems with the converter test setup prior to initiating the second phase. Other groups involved at various times during the testing were General Atomics, NMERI, and Loral Electro Optical Systems (now a part of Lockheed Martin Corporation).

As a consequence of various technical difficulties with the test equipment and procedures, a total of four tests were conducted under the START test series instead of the originally planned two tests. At the end of the four tests at NMERI, a total of 26 converters had been tested in strings of 8 or 16 converters connected in series. The primary objective of the test series was to demonstrate the performance of series-connected converters in a flight-like environment. However, even with the extra tests, every attempt to operate a string of thermionic converters connected in series ended in failure during the START tests. As a result, there was no opportunity to test the string of diodes in a flight-like environment.

The committee is very concerned about these results. Other tests detailed in this chapter have proven that thermionic converters can successfully operate for very long lifetimes both when operated individually and when connected in series.

One such test was performed by JPL under its SET program in 1967, during which four converters were connected in series (see Box 4.1). A second test was performed by General Atomics in a TRIGA reactor with a six cell TFE and again with a three cell TFE. There have been two Russian TOPAZ reactors flown with thermionic converters wired into 28 volt output configurations. Converters from a TOPAZ reactor were

also electrically tested in the United States during the TOPAZ International Program discussed in Chapter 3. Finally, a planar thermionic converter similar to the converters used in the START tests, and built by the same manufacturer that constructed the START test thermionic devices, demonstrated 24,000 hours of life in a test completed in 1994 (Thayer 1994).

Previous experience in testing thermionic devices indicates that converters can be made to work in series-connected circuits to develop a usable voltage level. The committee therefore recommends that the future sponsoring agency look closely at the START tests in order to identify and make use of the lessons learned and increase the probability of a successful test in the future. To aid in this effort, the committee identified several areas that should be given attention in any future tests:

• As with any difficult test program, additional time and financial resources should be included in any future test plan to accommodate problems that will inevitably arise. Technical difficulties, and the time and resources required to deal with them, should be considered as a part of the standard operating procedure for any high risk, experimental scenario.

• Some members of the thermionic device design and manufacturing team should be involved in the system tests. In this way, their expertise can be used early in the test cycle to minimize errors, help overcome testing obstacles, and avoid previously identified mistakes.

• The test fixture should be tested and characterized to make certain that performance requirements are met before fitting the thermionic devices into the fixture. For instance, temperature stability characteristics and temperature gradients in the heating elements should be clearly identified.

• Any future test setup should account for a high electromagnetic interference environment, because the high temperature test fixtures used in previous thermionic experiments generated a large amount of such interference.

The committee believes that the sponsoring agency should conduct an independent test of the original START test converters. First, the sponsoring agency should determine conclusively if the devices still work or if they are no longer functioning, as shown by the results at NEMERI. If the results of the reevaluation of the converters are different from those of NEMERI, an effort should be made to understand why there are differences. Any discrepancies between the two sets of data must be resolved so that the true test issues and device design issues can be identified.

Once the core issues are identified, the sponsoring agency should gather a group of experts to look closely at the START tests. This group should document proper test methods needed to have a successful test in the future.

Recommendation 7. When working on a system-level solar thermionic design, the sponsoring agency should reexamine the string thermionic assembly research testbed (START) tests to record lessons learned. The reexamination should begin with a retest of the original, individual converters to differentiate between problems due to converter design and generator configuration and those due to the test setup. The sponsoring agency should gather an independent group of experts to devise testing methodologies so as not to repeat past mistakes.

Rouklove, Peter. 1967. “Thermionic Converter and Generator Tests,” Jet Propulsion Laboratory research documents, Pasadena, Calif.

Thayer, Kevin. 1994. “Life Test and Diagnostics of a Planar Out-of-Core Thermionic Converter,” paper presented at the Eleventh Symposium on Space Nuclear Power, Albuquerque, N.Mex., January.