**Radioisotope Power Systems for Space Applications**

#### Antonio Sanchez-Torres

*Universidad Politécnica de Madrid, Escuela Técnica Superior de Ingenieros Aeronáutcos, Departamento de Física Aplicada, Spain* 

#### **1. Introduction**

456 Radioisotopes – Applications in Physical Sciences

Engler, R.E., "Atomic Power in Space – A History," U.S. Department of Energy Report DE-

Furlong, R.R. and Wahlquist, E.J., "U.S. Space Missions Using Radioisotope Power

Hammel, T.E., and Osmeyer, W.E., "The Selendie Isotope Generators," AIAA-1997-498, Conference on the Future of Aerospace Power Systems, March 1977. Hammel, T.E., Bennett, R., Otting, W., and Fanale, S., "Multi-Mission Radioisotope

Lange, R.G. and Mastal, E.F., "A Tutorial Review of Radioisotope Power Systems," in *A* 

National Research Council, *New Frontiers in the Solar System – An Integrated Exploration* 

National Research Council, *The Sun to the Earth – and Beyond: Panel Reports*, National

Schmidt, G., Wiley, R., Richardson, R. and Furlong, R., "NASA's Program for Radioisotope

Schmidt, G., Abelson, R., and Wiley, R., "Benefit of Small Radioisotope Power Systems for

Surampudi, R., Carpenter, R., El-Genk, M., Herrera, L., Mason, L., Mondt, J., Nesmith, B.,

Wiley, R. and Carpenter, R., "Small Radioisotope Power Source Concepts," in proceedings

U.S. Department of Energy, *Atomic Power in Space*, Prepared by Planning & Human Systems,

El-Genk, American Institute of Physics, Melville, New York, 2004.

American Institute of Physics, Melville, New York, 1994, pp. 1-20.

*Strategy*, National Academies Press, Washington, DC, 2003a.

Thermoelectric Generator (MMRTG) and Performance Prediction Model," AIAA-2009-4576, 7th International Energy Conversion Engineering Conference, August 2-

*Critical Review of Space Nuclear Power and Propulsion*, edited by M.S. El-Genk,

Power System Research and Development," in proceedings of *Space Technology and Applications International Forum (STAIF-2005)*, edited by M.S. El-Genk, American

NASA Exploration Missions," in proceedings of *Space Technology and Applications International Forum (STAIF-2005)*, edited by M.S. El-Genk, American Institute of

Rapp, D. and Wiley, R., *Advanced Radioisotope Power Systems Report*, Jet Propulsion

of *Space Technology and Applications International Forum (STAIF-2004)*, edited by M.S.

Inc. under Contract DE-AC01-NE32117, 2nd ed., Springer-Verlag, New York, 1983,

AC01-NE32117, March 1987.

5, 2009.

Systems," *Nuclear News*, April 1999, pp. 26-34.

Academies Press, Washington, DC, 2003b.

Institute of Physics, Melville, New York, 2005.

Physics, Melville, New York, 2005.

Chaps. 7, 14.

Laboratory, Pasadena, CA, March 2001.

At the beginning of the Space Age, both propulsion and power generation in the spacecraft has been the main issue for consideration. Considerable research has been carried out on technologies by several Space Agencies to reach outer planets and generate electric power for the systems and subsystems in the spacecraft (SC). Various types of power source such as solar photovoltaic, Radioisotope power systems (RPS) have been used by Space Agencies. New technology such as reactor based, electric solar sail and electrodynamic bare tethers might be used in the future for both propulsion and power generation. Mainly, both NASA and Russian Agency worked separately using nuclear technology to obtain more efficiency in their systems for deep space exploration.

Radioisotope Power Systems (RPS), is a nuclear-powered system to generate electric power to feed communication and scientific systems on a spacecraft. Radioisotope Thermoelectric Generators (RTGs), a type of Radioisotope Power System, were used in the past as electric power supplies for some navigational and meteorological missions, and most outer-planet missions. Radioisotope power systems use the natural decay of radionuclides produced by a nuclear reactor. The expensive, man-made Plutonium-238 (238Pu) is the appropriate source of energy used in RPS fueling; its long half-life (~87 years) guarantees long time missions. The limited avability of Plutonium-238 is inadequate to support scheduled NASA mission beyond 2018. After the Cold War, throughout the Non-Proliferation of Nuclear Weapons Treaty, the production and processing of these resources have been severally reduced. There is a high-priority recommendation to reestablish production to solve the severe 238Pu demand problem (National Reseach Council, 2009).

The isotope initially selected for terrestrial and space power applications was Cerium-144 because it is one of the most useful fission products available from nuclear reactor (Furlog, 1999; Lange, 2008). Its short half-life (about 290 days) made Cerium-144 compatible with a possible short-time mission. However, the high radiation associated with a powerful beta/gamma emission produces several problems with the payload interaction and safety in the case of reentry orbit. The development of RTGs was assigned to The Atomic Energy Commission in 1955. The first system developed for space situation was the System for Nuclear Auxiliary Power (SNAP). The Cerium-144 fueled SNAP-1 power system was never used in space. The first flight with a RTG was SNAP-3 in 1961 delivering 11.6 kW over a 280 days period, using as fueling Polonium-210 (Po-210) isotope. Po-210 is an alpha emitter with

Radioisotope Power Systems for Space Applications 459

would also generate low energy bremstrhlung x rays that is easy to shield against. This suggests isotopes such as T3, Ni63, Sr90, Tc99, Pw147, Curium-242 and Curium-244 are other

When solar panels cannot be used efficiently for planetary missions, RPS becomes the best available alternative. Typical RTG structure consists basically on a couple of metallic conductor, with hot and cold end-connectors. The system operates under thermoelectric generation principle, the so-called Seebeck effect. Heating one end from the natural decay of a radioactive isotope and the other end keeping cold, the gradient of tempreature between two ends will produce a voltage drop. Connecting the terminals through a resistive load causes an amount of current flowing in the electric external circuit, and then generating

Considerable research has been carried out to develop new technologies to improve RTG efficiency using more efficient thermoelectric materials with low thermal conductivity. The dynamic conversion systems, which convert partially the thermal energy in the fluid into mechanical work to drive an alternator to produce electricity, would provide higher electric

Power source (number) Spacecraft Mission type Launch

RHU Heater (1-4) Titan Micro-Rover rover 2015 GPHS (1) Europa Lander Planetary 2015 GPHS (1) Titan Moon Lander Planetary 2015 GPHS (1) Ganymede Lander Planetary 2015 GPHS (1) Callisto Lander Planetary 2015 GPHS (1) Titan Rough Lander Planetary 2015 GPHS (1) Europa Rough Lander Planetary 2015 GPHS (1) Callisto Orbiter Subsatellite Planetary 2015

GPHS (1) Europa Orbiter Subsatellite Planetary 2015

GPHS (2-4) Titan Rover Rover 2015

MMRTG-ASRG Jupiter Europa Orbiter Planetary 2020

GPHS (1) Titan Amphibius Rover Amphibius

Lander Planetary 2015

Subsatellite Planetary 2015

Subsatellite Planetary 2015

Array Mini-Lander Planetary 2020

Micro-Sat Planetary 2020-2030

Array Micro-Sat Planetary 2020-2030

Rover <sup>2015</sup>

power per unit mass, reducing the amount of Plutonium-238 required.

RHU Heater (1-4) Europa Impactor Micro-

GPHS (1) Ganymede Orbiter

GPHS (1) Outer Planets Magnetosphere

GPHS (1-3) Lander Amorphor. Rover

RHU Heater (7-9) Prospecting Asteroid Mission

RHU Heater (7-9) Saturn Autonomous Ring

Table 2. Several US future missions with RPS

possibilities.

electric power.

a very high power density and low radiation emissions. Since Po-210 has short half-life (138 days), space missions are highly limited. The early RTGs developed a specific power slightly larger than 1 W/kg. SNAP-9A system reached 20 W/kg whereas later systems such as Galileo developed 5.4 W/kg (Brown, 2001; Griffin, 2004). Several past missions have used RPS as it shown in Table 1. Table 2 shows several future missions that will use RPS as main power system.


Table 1. US spacecraft with RPS

The RTG fuel must be produced in adequate quantities with appropriate nuclear safety requirements for space missions. There are only a limited number of radioisotopes available for space power system applications. Using isotopes with pure low-energy beta emission would eliminate the requirements to shield against gamma radiation. Low energy particles

a very high power density and low radiation emissions. Since Po-210 has short half-life (138 days), space missions are highly limited. The early RTGs developed a specific power slightly larger than 1 W/kg. SNAP-9A system reached 20 W/kg whereas later systems such as Galileo developed 5.4 W/kg (Brown, 2001; Griffin, 2004). Several past missions have used RPS as it shown in Table 1. Table 2 shows several future missions that will use RPS as main

Power source (number) Spacecraft Mission type Launch SNAP-3 RTG (1) Transit 4A Navigational 1961 SNAP-3 RTG (1) Transit 4B Navigational 1961 SNAP-9A RTG (1) Transit 5BN-1 Navigational 1963 SNAP-9A RTG (1) Transit 5BN-2 Navigational 1963 SNAP-9A RTG (1) Transit 5BN-3 Navigational 1964 SNAP-10A Reactor Snapshot Experimental 1965 SNAP-19B RTG (2) Nimbus B-1 Meteorological 1968 SNAP-19B RTG (2) Nimbus III Meteorological 1969 ALRH Heater Apollo 11 Lunar 1969 SNAP-27 RTG (1) Apollo 12 Lunar 1969 SNAP-27 RTG (1) Apollo 13 Lunar 1970 SNAP-27 RTG (1) Apollo 14 Lunar 1971 SNAP-27 RTG (1) Apollo 15 Lunar 1971 SNAP-19 RTG (4) Pioneer 10 Planetary 1972 SNAP-27 RTG (1) Apollo 16 Lunar 1972 Transit-RTG (1) Triad-01-1X Navigational 1972 SNAP-27 RTG (1) Apollo 17 Lunar 1972 SNAP-19 RTG (4) Pioneer 11 Planetary 1973 SNAP-19 RTG (2) Viking 1 Planetary 1975 SNAP-19 RTG (2) Viking 2 Planetary 1975 MHW-RTG (4) LES 8, LES 9 Communication 1976 MHW-RTG (3) Voyager 2 Planetary 1977 MHW-RTG (3) Voyager 1 Planetary 1977 GPHS-RTG (2) RHU Heater Galileo Planetary 1989 GPHS-RTG (1) Ulysses Planetary 1990 RHU Heater (3) Mars Pathfinder Planetary 1996 GPHS-RTG (2) RHU Heater Cassini Planetary 1997 RHU Heater (8) Mars MER Spirit Mars rover 2003

Opportunity Mars rover 2003

GPHS-RTG (1) New Horizons Planetary 2006

The RTG fuel must be produced in adequate quantities with appropriate nuclear safety requirements for space missions. There are only a limited number of radioisotopes available for space power system applications. Using isotopes with pure low-energy beta emission would eliminate the requirements to shield against gamma radiation. Low energy particles

RHU Heater (8) Mars MER

Table 1. US spacecraft with RPS

power system.

would also generate low energy bremstrhlung x rays that is easy to shield against. This suggests isotopes such as T3, Ni63, Sr90, Tc99, Pw147, Curium-242 and Curium-244 are other possibilities.

When solar panels cannot be used efficiently for planetary missions, RPS becomes the best available alternative. Typical RTG structure consists basically on a couple of metallic conductor, with hot and cold end-connectors. The system operates under thermoelectric generation principle, the so-called Seebeck effect. Heating one end from the natural decay of a radioactive isotope and the other end keeping cold, the gradient of tempreature between two ends will produce a voltage drop. Connecting the terminals through a resistive load causes an amount of current flowing in the electric external circuit, and then generating electric power.

Considerable research has been carried out to develop new technologies to improve RTG efficiency using more efficient thermoelectric materials with low thermal conductivity. The dynamic conversion systems, which convert partially the thermal energy in the fluid into mechanical work to drive an alternator to produce electricity, would provide higher electric power per unit mass, reducing the amount of Plutonium-238 required.


Table 2. Several US future missions with RPS

Radioisotope Power Systems for Space Applications 461

The radioisotope fuel must be not be very expensive. Additionally, the radionuclide proposed has to be easily shielded against deep penetration radiation, as gamma radiation, avoiding the destruction of the electronic components on the spacecraft onboard. The fuel capsule must withstand impact against the ground at high velocity in case of a rocket launch

High *P0* and *t1/2* half-life values are required for space applications, reducing the valuable radioisotopes. Isotopes without powerful radiation such as gamma or beta is also required. The negative beta emitters can be recovered abundantly from fission fuel reprocessing plants. The alpha emitters with weak gammas are easier to shield. However, they are more

The radionuclide most used in RPS, Plutonium-238, is produced by the isotope Np-237.

1 238 1 237 <sup>92</sup> <sup>92</sup> *n U nU* + →+ <sup>2</sup>

<sup>237</sup> <sup>237</sup> *<sup>U</sup>*<sup>92</sup> <sup>93</sup> → + *e Np* .

Separating Np 237 from reactor fuel and further irradiated in a neutron flux, the plutonium

1 237 <sup>238</sup> *n Np*<sup>93</sup> *Np*<sup>93</sup> + →+ γ

<sup>238</sup> <sup>238</sup> *N* <sup>93</sup> <sup>94</sup> *p* → + *e Pu* .

The Plutonium-238 is selected for both high *t1/2* and specific power, producing heat by emitting alpha particles. The fuel is prepared in the form of pure plutonium oxide (*PuO2*) with 0.7 ppm Plutonium-238 and less than 0.5 percent Thorium-238 and Uranium-232.

The appropriate isotope combined with other components create a heat source that efficiently transfer the isotope heat to electrical power. The most used system for space missions is the general purpose heat source (GPHS). Fig. 1 shows the GPHS structure used in missions such as Galileo, Ulysses, and Cassini. Each module is designed to produce about 250 W at the beginning of mission. Its weight is about 1.43 kg, and its size and shape are selected to survive orbital reentry and post-impact into the ground at high terminal

Each GPHS module contains four pressed *PuO2* fuel pellets. Both diameter and length of the cylindrical fuel pellet is about 2.75 cm. An iridium alloy containment shell and clad made of 0.05 cm aluminum thickness encapsulate the fuel pellet. The iridium alloy is made to resist oxidation in a post-impact environment scenario. The fueled clad is the combination of fuel

Two of these clads are confined in a Graphite Impact Shell (GIS) made of carbon material. The GIS structure is designed to decrease the damage to the iridium clads during a possible free-fall accident. Two GISs are inserted into an aeroshell that is composed in graphite material. A thermal insulation layer of carbon-fiber cover each GIS decreasing the high temperature supported to the clads during atmospheric reentry heating. The aeroshell

failure, and an Earth-reentry situation. These accidents will be described in section 6.

Using U 238 in the nuclear reactor, the isotope Np 238 is produced by decay reaction

expensive than the beta emitters.

required is generated by

pellet and cladding.

**3. General purpose heat source** 

velocity. Typical dimensions are 9.72 cm × 9.32 cm × 5.31 cm.

In section 2 we review the fuel requirements for an optimal RPS. The main part of the RPS, the well-known General Purpose Heat Source, is described in section 3. In section 4 we study the static conversion energy (without movable parts), analyzing thermoelectric effects in the conductors. Additionally, we describe both RTG and Multi-Mission-RTG structures and principle characteristics. The dynamic conversion energy is reviewed in detail in section 5, focusing on Stirling and Brayton power systems. Due to Planetary Protections Requirements, some tentative outer-planets missions like Jovian moons exploration in Europa Jupiter System or re-entry missions which use RPS have to be safety enough. In section 6, we review the safety models for possible RPS accidents. In section 7, RTG will be compared with solar arrays. Conclusions are written in section 8.

#### **2. Radioisotopes for power generation**

At least 1300 radioisotopes, both natural and man-made, are available for terrestrial and space applications. Many are generated in both nuclear reactors and particle accelerators. The initial activity of the isotope is

$$A\_0 = \mathcal{Z}N\_0 \text{ [Bq]} \tag{1}$$

where *N0* is the initial isotope amount and ( ) 1 2 2 λ = *Ln t* / / is the decay constant of the isotope for a *t1/2* half-life. Table 3 shows several characteristics of useful radioisotopes for RPS. The specific electrical power generated by the heat of the source is given by

$$P\_0 = 1.6 \cdot 10^{-13} \eta \times \frac{E[\text{MeV}] \,\lambda \Big[\text{s}^{-1}\Big] N\_A \,\text{[nuclei/mol]}}{M \,\text{[amu]}} \tag{2}$$


where η is the conversion efficiency from thermal energy to electricity, *E* is the energy release per decay, *NA* is the Avogadro's number and *M* the atomic mass.

Table 3. Characteristics of isotopes useful for RPS. Notice both high *t1/2* and specific power of the Plutonium-238

In section 2 we review the fuel requirements for an optimal RPS. The main part of the RPS, the well-known General Purpose Heat Source, is described in section 3. In section 4 we study the static conversion energy (without movable parts), analyzing thermoelectric effects in the conductors. Additionally, we describe both RTG and Multi-Mission-RTG structures and principle characteristics. The dynamic conversion energy is reviewed in detail in section 5, focusing on Stirling and Brayton power systems. Due to Planetary Protections Requirements, some tentative outer-planets missions like Jovian moons exploration in Europa Jupiter System or re-entry missions which use RPS have to be safety enough. In section 6, we review the safety models for possible RPS accidents. In section 7, RTG will be

At least 1300 radioisotopes, both natural and man-made, are available for terrestrial and space applications. Many are generated in both nuclear reactors and particle accelerators.

λ

[] [ ] [ ]

=⋅ × (2)

*M*

is the conversion efficiency from thermal energy to electricity, *E* is the energy

Isotope Radiation emission *t1/2* Specific Power (W/g)

Tritium-3 β- ,no ϒ 12.3 years 0.26 Cobalt-60 β- , ϒ 83.8 days 17.70 Nickel-63 β- ,no ϒ 100.1 years 0.002 Krypton-85 β- , ϒ 10.7 years 0.62 Stronium-90 β- ,no ϒ 29.0 years 0.93 Ruthenium-108 β- ,no ϒ 1.0 years 33.10 Cesium-137 β- , ϒ 30.1 years 0.42 Cerium-144 β- , ϒ 284.4 days 25.60 Promethium-147 β- , ϒ 2.6 years 0.33 Polonium-210 α, ϒ 136.4 days 141.00 Plutonium-238 α, ϒ 87.7 years 0.56 Americium-241 α, ϒ 432 years 0.11 Curium-242 α, ϒ 162.8 days 120.00 Curium-244 α, ϒ 18.1 years 2.84 Table 3. Characteristics of isotopes useful for RPS. Notice both high *t1/2* and specific power of

MeV s nuclei/mol -1

isotope for a *t1/2* half-life. Table 3 shows several characteristics of useful radioisotopes for

λ

[Bq] (1)

= *Ln t* / / is the decay constant of the

*A N* 0 0 =λ

RPS. The specific electrical power generated by the heat of the source is given by

. amu *E NA <sup>P</sup>*

<sup>−</sup>

compared with solar arrays. Conclusions are written in section 8.

where *N0* is the initial isotope amount and ( ) 1 2 2

<sup>0</sup> 1 6 10

13

η

release per decay, *NA* is the Avogadro's number and *M* the atomic mass.

**2. Radioisotopes for power generation** 

The initial activity of the isotope is

where

η

the Plutonium-238

The radioisotope fuel must be not be very expensive. Additionally, the radionuclide proposed has to be easily shielded against deep penetration radiation, as gamma radiation, avoiding the destruction of the electronic components on the spacecraft onboard. The fuel capsule must withstand impact against the ground at high velocity in case of a rocket launch failure, and an Earth-reentry situation. These accidents will be described in section 6.

High *P0* and *t1/2* half-life values are required for space applications, reducing the valuable radioisotopes. Isotopes without powerful radiation such as gamma or beta is also required. The negative beta emitters can be recovered abundantly from fission fuel reprocessing plants. The alpha emitters with weak gammas are easier to shield. However, they are more expensive than the beta emitters.

The radionuclide most used in RPS, Plutonium-238, is produced by the isotope Np-237. Using U 238 in the nuclear reactor, the isotope Np 238 is produced by decay reaction

$$\begin{array}{rcl} n^1 & + \cdot \mathcal{U}\_{92}^{238} \rightarrow \ 2n^1 + \cdot \mathcal{U}\_{92}^{237} \\\\ \mathcal{U}\_{92}^{237} & \rightarrow & e & + \; N p\_{93}^{237} \end{array}$$

Separating Np 237 from reactor fuel and further irradiated in a neutron flux, the plutonium required is generated by

$$\begin{array}{rcl} \text{Np}^1 + \text{Np}\_{93}^{237} \rightarrow \text{ } \text{ } & + & \text{Np}\_{93}^{238} \\\\ \text{Np}\_{93}^{238} \rightarrow \text{ } & e \rightarrow \text{ } & \text{Pu}\_{94}^{238} \end{array}$$

The Plutonium-238 is selected for both high *t1/2* and specific power, producing heat by emitting alpha particles. The fuel is prepared in the form of pure plutonium oxide (*PuO2*) with 0.7 ppm Plutonium-238 and less than 0.5 percent Thorium-238 and Uranium-232.

#### **3. General purpose heat source**

The appropriate isotope combined with other components create a heat source that efficiently transfer the isotope heat to electrical power. The most used system for space missions is the general purpose heat source (GPHS). Fig. 1 shows the GPHS structure used in missions such as Galileo, Ulysses, and Cassini. Each module is designed to produce about 250 W at the beginning of mission. Its weight is about 1.43 kg, and its size and shape are selected to survive orbital reentry and post-impact into the ground at high terminal velocity. Typical dimensions are 9.72 cm × 9.32 cm × 5.31 cm.

Each GPHS module contains four pressed *PuO2* fuel pellets. Both diameter and length of the cylindrical fuel pellet is about 2.75 cm. An iridium alloy containment shell and clad made of 0.05 cm aluminum thickness encapsulate the fuel pellet. The iridium alloy is made to resist oxidation in a post-impact environment scenario. The fueled clad is the combination of fuel pellet and cladding.

Two of these clads are confined in a Graphite Impact Shell (GIS) made of carbon material. The GIS structure is designed to decrease the damage to the iridium clads during a possible free-fall accident. Two GISs are inserted into an aeroshell that is composed in graphite material. A thermal insulation layer of carbon-fiber cover each GIS decreasing the high temperature supported to the clads during atmospheric reentry heating. The aeroshell

Radioisotope Power Systems for Space Applications 463

where Δ*V* is the potential difference across a piece of metal due to a temperature difference Δ*T* . The sign of the Seebeck coefficient represents the potential of the cold side with respect to the hot side. For electrons diffusing from hot to cold end, the cold side is negative with respect to the hot side, making *S < 0*. Since the Seebeck coefficient depends on temperature,

*i*

Using the Fermi-Dirac distribution, the average energy *Eav* per electron in a metal is given by

π

= +

where *EF0* is the Fermi energy at 0 K. The average energy in the hot end is greater, and energetic electrons in the hot end diffuse toward the cold region until the potential prevents further diffusion. Notice that the average energy in Eq. (5) also depends on the material

0

<sup>E</sup> <sup>E</sup>

Ef Ef


Conductor

( ) (). *av av* −= + − *eV E T T E T*

δ

ߜ

<sup>0</sup> <sup>1</sup> <sup>0</sup> <sup>1</sup>

V

Considering a small temperature difference *δT* produces a voltage *δV* between the accumulated electrons and exposed positive metal ions as it is shown in Fig. 2. For electrons diffusing from the hot region to the cold part, the system would work against the potential difference *δV* , i.e. -e*δV* , decreasing the average energy of the electron by *δEav*,yielding

> δ

*T*, neglecting

*k T e V E* π − ≈ δ

+ -

Hot Cold

T

T+ T T

3 5 <sup>1</sup> 5 12 *av F*

*kT E E*

2 2

0

f(E) f(E)

ߜ

*T2* term we obtain,

*F*

*E*

Δ = *V S dT* . (4)

, (5)

(6)

(7)

*T*

*T*

the voltage between two hot/cold regions is

through *EF0.* 

Fig. 2. Seebeck effect diagram.

Using Eq. (5) in (6), and expanding *T+*

the Seebeck coefficient reads

provides protection against surfaces. Step 1 GPHS module, which is used on the New Horizons eexploration mission to Pluto, improves the initial GPHS device including an aeroshell between the two GISs. A second aeroshell improvement, known as Step 2 GPHS module, gives additional protection in the clads for hipervelocity reentry into the atmosphere (Benett, 2006; Brown, 2001; Griffin, 2004; Hastings, 2004).

Fig. 1. General purpose heat source (GPHS) structure. (Source NASA/DOE/JPL)

#### **4. Static conversion energy**

The static conversion energy use the well-known thermoelectric or Seebeck effect. The thermoelectric effects in metals depend on the electronic structure of the materials. A temperature difference between two points in a conductor or semiconductor results in a voltage difference between these two regions. The Seebeck coefficient gives the magnitude of this effect. The thermoelectric voltage generated per unit temperature difference in a conductor is called the Seebeck coefficient.

Consider a metallic rod that is heated at one end and cooled at the other end as represented in Fig. 2. Since the electrons in the hot region are more energetic with greater velocities than those in the cold region, the electrons from the hot end diffuses toward the cold part. This situation prevails until the electric field developed between the positive ions in the hot region and the excess electrons in the cold region prevents further electron motion from the hot to cold end. A voltage is therefore gathered between the hot and cold ends with hot end at positive potential. The Seebeck coefficient *S* is given by the potential-to-temperature difference ratio

$$S = \frac{\Delta V}{\Delta T} \,\tag{3}$$

provides protection against surfaces. Step 1 GPHS module, which is used on the New Horizons eexploration mission to Pluto, improves the initial GPHS device including an aeroshell between the two GISs. A second aeroshell improvement, known as Step 2 GPHS module, gives additional protection in the clads for hipervelocity reentry into the

atmosphere (Benett, 2006; Brown, 2001; Griffin, 2004; Hastings, 2004).

Fig. 1. General purpose heat source (GPHS) structure. (Source NASA/DOE/JPL)

The Seebeck coefficient *S* is given by the potential-to-temperature difference ratio

The static conversion energy use the well-known thermoelectric or Seebeck effect. The thermoelectric effects in metals depend on the electronic structure of the materials. A temperature difference between two points in a conductor or semiconductor results in a voltage difference between these two regions. The Seebeck coefficient gives the magnitude of this effect. The thermoelectric voltage generated per unit temperature difference in a

Consider a metallic rod that is heated at one end and cooled at the other end as represented in Fig. 2. Since the electrons in the hot region are more energetic with greater velocities than those in the cold region, the electrons from the hot end diffuses toward the cold part. This situation prevails until the electric field developed between the positive ions in the hot region and the excess electrons in the cold region prevents further electron motion from the hot to cold end. A voltage is therefore gathered between the hot and cold ends with hot end at positive potential.

> , *<sup>V</sup> <sup>S</sup> T*

<sup>Δ</sup> <sup>=</sup> Δ (3)

**4. Static conversion energy** 

conductor is called the Seebeck coefficient.

where Δ*V* is the potential difference across a piece of metal due to a temperature difference Δ*T* . The sign of the Seebeck coefficient represents the potential of the cold side with respect to the hot side. For electrons diffusing from hot to cold end, the cold side is negative with respect to the hot side, making *S < 0*. Since the Seebeck coefficient depends on temperature, the voltage between two hot/cold regions is

$$
\Delta V = \bigcap\_{T\_i}^{T} S \, dT \, . \tag{4}
$$

Using the Fermi-Dirac distribution, the average energy *Eav* per electron in a metal is given by

$$E\_{av} = \frac{3}{5} E\_{F0} \left[ 1 + \frac{5\pi^2}{12} \left( \frac{kT}{E\_{F0}} \right)^2 \right] \tag{5}$$

where *EF0* is the Fermi energy at 0 K. The average energy in the hot end is greater, and energetic electrons in the hot end diffuse toward the cold region until the potential prevents further diffusion. Notice that the average energy in Eq. (5) also depends on the material through *EF0.* 

Fig. 2. Seebeck effect diagram.

Considering a small temperature difference *δT* produces a voltage *δV* between the accumulated electrons and exposed positive metal ions as it is shown in Fig. 2. For electrons diffusing from the hot region to the cold part, the system would work against the potential difference *δV* , i.e. -e*δV* , decreasing the average energy of the electron by *δEav*,yielding

$$-e\delta V = E\_{av}\left(T + \delta T\right) - E\_{av}\left(T\right). \tag{6}$$

Using Eq. (5) in (6), and expanding *T+*ߜ*T*, neglecting ߜ*T2* term we obtain,

$$-e\mathcal{S}V = \frac{\pi^2 k^2 T}{2E\_{F0}}.\tag{7}$$

the Seebeck coefficient reads

Radioisotope Power Systems for Space Applications 465

Fig. 3. Radioisotope Thermoelectric generator (RTG). (Source NASA/DOE/JPL)

Fig. 4. Multi-mission Radioisotope Thermoelectric Generator (MMRTG). (Source

The multi-mission radioisotope power generation (MMRTG) is the next generation of space RTGs (see Fig. 4). MMRTG is being developed by The Department of Energy (DOE) for

**4.2 Multi-mission radioisotope thermoelectric generator** 

planetary missions.

NASA/DOE/JPL)

$$S = -\frac{\pi^2 k^2 T}{2eE\_{F0}}.\tag{8}$$

Table 1 shows typical experimental values for the Seebeck coefficient for several metals. Notice that some metals have positive *S* such as copper. The sign means that the electrons moves from cold to hot end of a copper rod.

Considering an aluminum rod heated at one end and cooled at the other end, the voltage difference reads

$$V\_{AB} = \int\_{T\_0}^{T} (S\_A - S\_B) dT \,\prime \tag{9}$$

where *SA-SB* is the thermoelectric power for the thermoelectric couple given by both rods joined in a closed circuit. The voltage produced by the thermocouple pair depends on the metal used. Some conductor doped by the addition of impurities can produce deficiencies or an excess of electrons providing greater efficiency. The power extracted of the thermoelectric material is a function of its operating temperature. Elements with high enough thermal conductivity produce energy looses. Heat entering into the hot end would escape without much conversion to electricity. For a thermoelectric generator the thermoelectric rating, *Z =S2/RK*, depends on the characteristic of the material, i.e. the voltage produced for the difference of temperature. Both *R* and *K* are electrical resistivity and thermal conductivity of the material, respectively. The thermoelectric generator will be more efficient with high *Z* values, i.e. high *S,* 1*/R* and 1*/K.* Ordinary metals like cooper are very good heat conductors.


Table 4. Seebeck coefficients for several metals.

#### **4.1 Radioisotope thermoelectric generator**

The typical static conversion system used in all outer planet mission is the well-known RTG (see Fig. 3), which is composed by a stack of 18 GPHS modules. The joined module GPHS-RTG, operates at normal voltage output of 28 V-dc. Both diameter and length of the RTG are 0.42 and 1.14 meters, respectively, and its weigh is about 55.9 kg.

The heat source assembly is surrounded by 572 silicon germanium (SiGe) thermocouples, known as unicouples. The unicouples are connected in two series-parallel electric wiring circuits providing the full output voltage. The induced magnetic field by the wires in the RTG is minimized, rearranging the electrical wiring (Abelson, 2004; Lange, 2008). The most recent use of a GPHS–RTG module was built for the New Horizons mission, launched in January 2006 to reach Pluto in 2015.

*k T <sup>S</sup> eE* π

Table 1 shows typical experimental values for the Seebeck coefficient for several metals. Notice that some metals have positive *S* such as copper. The sign means that the electrons

Considering an aluminum rod heated at one end and cooled at the other end, the voltage

( )

where *SA-SB* is the thermoelectric power for the thermoelectric couple given by both rods joined in a closed circuit. The voltage produced by the thermocouple pair depends on the metal used. Some conductor doped by the addition of impurities can produce deficiencies or an excess of electrons providing greater efficiency. The power extracted of the thermoelectric material is a function of its operating temperature. Elements with high enough thermal conductivity produce energy looses. Heat entering into the hot end would escape without much conversion to electricity. For a thermoelectric generator the thermoelectric rating, *Z =S2/RK*, depends on the characteristic of the material, i.e. the voltage produced for the difference of temperature. Both *R* and *K* are electrical resistivity and thermal conductivity of the material, respectively. The thermoelectric generator will be more efficient with high *Z* values, i.e. high *S,* 1*/R* and 1*/K.* Ordinary metals like cooper

Metal S at 0º C (*µ*V K-1) S at 27º C (*µ*V K-1) *EF0* (eV) Al -1.60 -1.80 11.6 Cu 1.70 1.84 7.0 Ag 1.38 1.51 5.5 Au 1.79 1.94 5.5

The typical static conversion system used in all outer planet mission is the well-known RTG (see Fig. 3), which is composed by a stack of 18 GPHS modules. The joined module GPHS-RTG, operates at normal voltage output of 28 V-dc. Both diameter and length of the RTG

The heat source assembly is surrounded by 572 silicon germanium (SiGe) thermocouples, known as unicouples. The unicouples are connected in two series-parallel electric wiring circuits providing the full output voltage. The induced magnetic field by the wires in the RTG is minimized, rearranging the electrical wiring (Abelson, 2004; Lange, 2008). The most recent use of a GPHS–RTG module was built for the New Horizons mission, launched in

are 0.42 and 1.14 meters, respectively, and its weigh is about 55.9 kg.

,

0

*T AB A B T*

≈ − (8)

*V S S dT* = − (9)

difference reads

moves from cold to hot end of a copper rod.

are very good heat conductors.

Table 4. Seebeck coefficients for several metals.

**4.1 Radioisotope thermoelectric generator** 

January 2006 to reach Pluto in 2015.

Fig. 3. Radioisotope Thermoelectric generator (RTG). (Source NASA/DOE/JPL)

#### **4.2 Multi-mission radioisotope thermoelectric generator**

The multi-mission radioisotope power generation (MMRTG) is the next generation of space RTGs (see Fig. 4). MMRTG is being developed by The Department of Energy (DOE) for planetary missions.

Fig. 4. Multi-mission Radioisotope Thermoelectric Generator (MMRTG). (Source NASA/DOE/JPL)

Radioisotope Power Systems for Space Applications 467

High efficiency in the RPS would both reduce system mass and fuel requirement, decreasing the total cost. Normally, RPS are tested at the ground in a vacuum chambers for a long time (>1000 hours). This would demonstrate that the design work also in the space with high

Since the Brayton cycle is useless for power generation under 0.5 kW, missions with lower

The Stirling power system is based on a kinematics engine driving a three phase alternator. The initial design only worked during six months. It would not be appropriate for long-term missions. using a free piston combined with a linear alternator would be a promising technology for space power applications. The Stirling system, at difference of Brayton type, might provide low power. Further, Stirling engines operating in reverse have already used in space to provide cryogenic cooling for imaging sensors. The device has reciprocating pistons and displacers as it shown in Fig. 6. The motion of the components depends on physical springs or gas and on the cycle pressure swing of the engine. Typically, the engine

Displacer Piston Gas Displacer

Cocler Regenerator Heater

Two types of free pistons, linear alternator, Stirling machines are under development for mission in the range between 10 and 100 watts. The first type, under development by NASA. The power conversion efficiency for the Stirling producing 100 W is about 30%. The second type uses flexural spring to support the moving component to prevent friction and to provide enough axial springing for the free piston movement. The engine relies on the gap between the cylinder and the displacer to serve as regenerator for the system. As with

Spring

Head

power requirement need an auxiliary electrical power generator.

Fig. 6. Schematic diagram of the Stirling Isotope Power System

power conversion efficiency.

contains only two moving parts.

Displacer Drive Rod Power Piston

The MMRTG will generate 120 W of power at launch from a Pu-238 heat source assembly containing a stack of 8 Step 2 GPHS modules, which are described in section 3. The MMRTG operates at a normal output voltage of 28 V-dc. Both diameter and length of MMRTG are 64 cm diameter and 66 cm, respectively. The central heat source cavity is separated from the thermoelectric converter by a helium isolation liner. The helium generated by the Pu-238 is dumped to the environment by diffusion through an elastomeric gasket seal. The thermoelectric converter cavity can operate in both atmospheric environment or space vacuum (Ritz, 2004; Lange, 2008).

The thermocouples are connected in a series/parallel electrical circuit to improve the efficiency up 6.8%. Waste heat is radiated from the eight radial fins. These fins are made of aluminum alloys coated with a high-emissivity to disintegrate and release the GPHS modules in the case of reentry into the Earth's atmosphere. The MMRTG is both lighter and smaller than RTG system.
