Fusion Plasma Rockets

Propulsion using nuclear fusion

Mirrorfusion Ship
Image from Steve Bowers
A Gasdynamic Mirror Fusion ship; note the long fusion tube and the extensive cooling radiators

Fusion Plasma Rockets- Data Panel

Summary:Developed in the Early Interplanetary Age, fusion plasma rockets were the workhorses of the old Solar System, especially favoured by the Genetekker and Space Adapted tweak cultures at that time. A few very early interstellar missions used fusion propulsion, but more advanced propulsion systems are almost always used for interstellar transport in the modern era.

Even today fusion rockets are used for interplanetary transport in many middletech polities or for recreational use
See alsoExternal Fission-Fusion pulse drives
Basic Propulsion:Reaction
Specific Impulsevarious depending on fuel and design (see below)
FuelsD-T, D-D, D-He3, He3-He3, D-D, p-B11
Reaction Mass:usually hydrogen; sometimes water or other fluids
Minimum Technology Required:Middle to High Tech (Information/Interplanetary Age Equivalent)
Matter Manipulation:microscale/nanoscale (precision materials)
First Introduced:Information Age
Used by:Many polities, clades, individuals and groups away from the Nexus, in low resource solar systems and among members of the Deeper Covenant, out on the periphery, and among some anti-ai luddite groups
Used in:medium speed interplanetary transport, some slow interstellar missions
Construction Costs:Autofac: high (bulky, precision materials); Hylonano: reasonably cheap assuming presence of component materials
Running cost:depends on availability of fuel
Advantages:reasonably reliable, reasonably good performance, can be made with materials in most solar systems; doesn't require amat

Disadvantages: Interplanetary or very slow interstellar use only

Fusion Plasma Rockets

Basic Principles

A fusion plasma rocket heats or otherwise energizes a diffuse plasma of exotic isotopes of light elements until a sustained reaction occurs in which these isotopes combine into the nuclei of heavier atoms. This releases energy in the form of charged particles and neutrons. The neutrons are un-wanted by products and typically end up heating the surrounding structure and machinery or escaping into space. The charged particles, however, add their heat to the plasma to maintain its temperature. A leak is introduced into the magnetic confinement for the plasma, allowing the hot exhaust to vent into space at high velocities. This venting exhaust provides the rocket propulsion.

The very hot, diffuse plasma from a fusion rocket often provides a thrust that is too feeble to be practical, with a high exhaust velocity but minimal reaction mass for propulsion. Consequently, most fusion plasma rockets incorporate a second plasma chamber, sometimes called an "afterburner" in analogy with atmospheric jet turbine engines. Hot fusion plasma is vented into the secondary chamber, where it encounters a denser gas or plasma. The mixing of the hot fusion gas with the secondary fluid results in a plasma that is denser and colder. This plasma is allowed to escape into space, providing more thrust at a lower exhaust velocity for the same power. The heated afterburner plasma behaves like any thermal plasma rocket with an input energy equal to the energy in the fusion rocket plasma exhaust and a power efficiency of not more than 85%. In particular, for a given rate of propellant mass flow m'(in kg/s) and power P(in watts), the force (in newtons) is F = √ (1.7 P m') for 85% nozzle efficiency. The propellant used in the afterburners does not need to be the same as that of the fusion plasma - bulk afterburner propellant is usually whatever easily vaporizable material is most readily available and which can be easily stored. These afterburners are often needed for rapid interplanetary travel, but for interstellar missions the price in ΔV is too high and afterburners are commonly not employed for journeys between the stars.

Fusion plasma rockets are similar in many ways to electrically driven plasma rockets. The main difference is that the fusion plasma rocket generates its energy internally, whereas conventional plasma rockets rely on an external source of power to energize the plasma.

Fusion plasma rockets are distinguished from fusion pulse rockets in that the fusion reaction occurs inside an enclosed reactor, rather than detonating small fusion explosions exterior to the spacecraft within the magnetic nozzle.

Most fusion reactions produce radiation that escapes the thin fusion plasma and interacts with the walls of the reactor, either in the form of neutrons or high energy bremsstrahlung photons. This is wasted energy as far as propulsion goes, since these radiations do not heat the fusion plasma used for the rocket. Further, it heats the reactor walls, which increases the heat load on the spacecraft and thus increases the area and mass of the radiators required to shed the heat. On the positive side, a heat engine can be run between the reactor walls and the radiators (usually a high temperature gas turbine) that can be used to generate power. To minimize the radiator area, the heat engine will typically have a maximum efficiency of 1/4 or less, so that at most 1/4 of the radiated fusion power can be extracted as electricity or mechanical power for practical on-board use. It also requires the addition of a turbine and electric generator, which together add mass to the spacecraft. Energy in the fusion plasma that is not being used for propulsion can be directly captured with a magnetohydrodynamic generator at high efficiency (generally between 80% to 95%) to produce electrical power.

Occasionally, a dual-mode fusion rocket are seen which use the neutron or bremsstrahlung radiation from fusion to heat hydrogen propellant. The hot gaseous hydrogen is allowed to escape at exhaust velocities on the order of 10 km/s. This produces a fusion thermal rocket that can provide significantly higher accelerations than the fusion plasma rocket at the cost of much more profligate propellant use. When high acceleration is not desirable, the spacecraft transitions to a pure fusion plasma rocket. In this way, much of the energy that would otherwise be wasted or even add to the heat load can be used for propulsion instead. With a 70% efficient solid nozzle, a thermal rocket using hydrogen propellant at 10 km/s can achieve thrusts of 1 newton for every 7 kW of neutron and bremsstrahlung power, and will use 0.1 gram of propellant per newton of thrust per second.

Thermal Fusors

Thermal fusors confine a hot plasma to sustain fusion. There are three primary thermonuclear fusion reactions that thermal fusors can use, and most thermal fusion reactors can make use of any of them, although some may be optimized for only one (such as the addition of a lithium blanket for breeding tritium for a D-T reactor, or not making the engineering trade-offs to handle large neutron fluxes for a D-He3 reactor).

Thermal fusion rockets are largely obsolete, having been superceeded first by athermal reactors and then by conversion reactors.


The simplest fusion reaction to produce is between two isotopes of hydrogen: deuterium (often denoted D) and tritium (denoted T). This D-T fusion occurs at an optimum temperature of 13.6 keV, far lower than that of any other fusion reaction, which makes it the easiest to ignite. Further, for a reactor capable of confining a plasma at a given pressure, the D-T reaction will produce more power by two or more orders of magnitude than any other fusion reaction.

There are, however, two serious drawbacks to D-T fusion. The first is that 80% of the energy of the reaction goes into neutrons, which cannot be used for propulsion, are a radiation hazard, and gradually transmute and embrittle the reactor. The other is that tritium is itself radioactive, with a half-life of 12.32 years. In addition to posing a radiological hazard in the event of a leak, this means that tritium cannot be stored for long periods of time as it will eventually decay. In many cases, the tritium is generated on site from a blanket of lithium that surrounds the reactor. The neutrons react with the lithium to produce tritium, which is then scavenged from the molten lithium metal. The reaction of a neutron with Li-6 to produce a tritium atom and a helium atom releases even more energy that must be disposed of as heat, whereas the reaction of a neutron with Li-7 produces tritium, a helium nucleus, and an additional neutron. Often the reactor is simply made thin-walled without a lithium regeneratior so that as many neutrons as possible can escape into space, reducing the spacecraft's thermal load.

With a nominal 85% nozzle efficiency, the exhaust velocity of a D-T fusion rocket will be 0.00385 c or 1150 km/s if run with an even mix of deuterium and tritium. This will require a fusion power of 3.4 MW per newton of thrust, of which 2.7 MW will be in the form of neutrons and 0.7 MW is present in the kinetic and thermal energy of the exhaust. The fuel required is 0.35 milligrams of D and 0.52 milligrams of T per second per newton of thrust.

The first fusion plasma rockets were D-T thrusters with a specific power of close to 1 kW/kg, neglecting the lithium breeder blanket. By the time of the incorporation of the Jovian League, specific powers had risen to 10 kW/kg. The difficulty of storing tritium, however, usually meant that these designs either used a lithium blanket with its added mass, or burned D-He3 instead of D-T with the resultant 80-fold power loss. In theory, Technocalypse-era fusors runing on D-T fuel could approach 1 MW/kg, but in practice these reactors were run on D-He3 and the resulting neutron flux from a full power D-T burn would have melted the reactor in short order.


The deuterium-deuterium reaction is primarily valued because its fuel is abundant nearly everywhere. It is not radioactive and it can be found anywhere there is normal hydrogen at about 1 part D in 6,500 parts H. However, the D-D fusion reaction proceeds much slower than for D-T. Unfortunately, a major energy loss mechansim, the emission of x-rays due to electrons colliding with nuclei known as bremsstrahlung, remains the same for both D-D and D-T. These x-rays easily penetrate the thin plasma, moving energy from the fusion plasma to the walls of the reactor where it does no good. In order to have the fusion power exceed the bremsstrahlung power, the reactor must be run at a much higher temperature, which means a more diffuse plasma for the same pressure in the reactor and consequently less power. D-D reactors are generally run at between 100 keV and 500 keV of temperature to minimize bremsstrahlung for a given amount of fusion power, where they produce between 10,000 and 25,000 times less power than if the reactor was burning D-T.

There are two possible fusion reactions for deuterium - one produces a proton and a tritium nucleus, the other a neutron and a helium-3 nucleus. Both occur at equal rates in the fusion plasma. At the temperatures typical of a D-D reactor, the tritium or helium-3 nucleus produced immediately fuses with a deuterium, producing more energy and, in the case of tritium, another neutron. Overall, 38% of the energy released by deuterium fusion is in the form of neutrons. As with D-T, these neutrons are a nuisance and useless for propulsion. In addition, at 100 keV about half the energy is lost to bremsstrahlung, wereas at 500 keV this decreases to a third of the total power. Approximately the same fraction of fusion power is available for propulsion as with D-T, the difference is that the bremsstrahlung portion of the wasted power always interacts with the reactor walls and does not cause radiological hazards nor degrade the structure of the reactor. This can lead to larger thermal loads since the bremsstrahlung x-rays cannot simply pass through the reactor walls into space.

At 85% nozzle efficiency, a D-D fusion plasma at 100 keV will deliver an exhaust velocity of 0.0116 c, or 3,500 km/s. This will produce 1 newton of thrust for every 17.1 MW of fusion power, of which 6.5 MW will be lost as neutrons, 8.6 MW will be lost as bremsstrahlung, and 2.0 MW goes into the exhaust plume, while consuming 0.29 mg of deuterium per second. At 500 keV, the exhaust velocity is 0.026 c, or 7,800 km/s, and requires a power of 15.8 MW per newton, of which 6.0 MW is lost as neutrons, 5.2 MW is lost as bremsstrahlung, and 4.6 MW is dumped into the exhaust, while using 0.13 mg of deuterium per second per newton.


The quest for a reaction that does not generate a significant amount of neutrons leads to a reactor that can fuse deuterium and an exotic isotope of helium: helium-3. The fusion of a deuteron with a helium-3 nucleus results in a proton and a nucleus of the common helium isotope, helium-4. Both reaction products are charged, and quickly dump their energy into the fusion plasma. The D-He3 fusion reaction is optimized at a temperature of 50 keV to 60 keV, both in terms of optimium power and minimizing bremsstrahlung. In this range of temperatures, a reactor running on D-He3 will generate 80 times less power than if it were fusing D-T.

A D-He3 reactor is not entirely aneutronic, however - a small amount of D-D fusion takes place among the deuterium in the plasma. In a well designed reactor, approximately 1/20 of the total energy is lost as neutrons, so neutron activation, transmutation, and embrittlement will need to be taken into account when designing a D-He3 reactor, as well as shielding against neutron radiation for rad-sensitive equipment, crew, and passengers. Running the reactor helium-3 rich will decrease the amount of neutrons produced relative to the total power, but will also decrease the total power available. Further, approximately 1/5 of the fusion power is lost to bremsstrahlung. The remaining 3/4 of the fusion power remains in the plasma, and can be used for propulsion.

Helium-3 is available wherever primordial helium can be found - typically in stars and gas giants. It is very rare on rocky or icy planets, planetoids, moons, and asteroids. A small amount of helium-3 is embedded by the solar wind into the surface regolith of airless rocky bodies, and an even smaller amount can be extracted from wells of rocky planets that results from radioactive decay. Helium-3 can also be created artificially by producing tritium from lithium with a strong neutron source (such as a D-T fusion reactor) and waiting for the tritium to decay into helium-3. Of these sources, only gas giants and industrial manufacture produce enough helium-3 for large scale economic use of fusion power and fusion rockets (by the time stars can be economically mined for helium-3, there are better alternatives for power and propulsion).

A reactor running on an even mix of D and He3 with 85% nozzle efficiency will produce an exhaust velocity of 0.0077 c, or 2,300 km/s. It requires 1.82 MW per newton of thrust, of which 0.36 MW is lost as bremsstrahlung and 0.09 MW is lost as neutrons. The remaining 1.37 MW is put into the exhaust stream. Each newton of thrust requires 0.17 mg/s of D and 0.26 mg/s of He3.

Athermal fusors

An athermal reactor does not keep the fusion plasma at a well defined temperature, in order to limit the collisions between the atomic nuclei and electrons that produce bremsstrahlung. Reactors using athermal designs were instrumental in the early history of the First Federation, although they were later replaced by conversion reactors when monopoles become available.


The fusion reaction between a proton and the boron-11 isotope results in no neutrons, and lacks side-chains that lead to neutron producing fusion reactions. In addition, both reactants are commonly available from rocky planets, leading to cheaper fuel since it does not have to be hauled up from the deep gravitational potential wells of gas giants. However, the p-B11 fusion reaction has one significant drawback - a hot fusion plasma of protons and boron-11 will always lose more energy to bremsstrahlung than it gains from fusion. Consequently, a hot p-B11 plasma can never ignite in a self-sustaining fusion burn.

To get around this limitation, p-B11 reactors must work with athermal plasmas. Since bremsstrahlung is a result of collsions between electrons and nuclei, a p-B11 fusion plasma needs to have energetic protons and boron-11 nuclei that nevertheless do not run into electrons at high speed. This is a challenging design constraint and the reason the early fusion reactors used either D-T, D-D, or D-He3 reactions.

When p-B11 reactors can be made, they can directly harvest the reaction products, which are much more energetic than the particles undergoing fusion and can fly out of the containment for the diffuse plasma. This has two significant effects. First, the exhaust stream of a p-B11 fusion plasma rocket consists entirely of already burnt fuel, rather than venting mostly un-burnt fusion plasma into space. Second, the exhaust velocity is particularly high - 0.044 c or 13,300 km/s. A p-B11 fusion plasma rocket requires 7.86 MW per newton of thrust, all of which is imparted to the jet of exhaust. Each newton of thrust also consumes 6.3 micrograms of hydrogen and 68.9 micrograms of boron-11 each second.

During the great expulsion, GAIA provided a number of p-B11 fusion plasma rockets to the fleeing ark-ships. Reverse engineering of these reactors allowed modosophonts to build p-B11 reactors and rockets with a specific power of nearly 10 kW/kg by the time of the First Federation. These same reactors can also be run on D-He3 at 9 times the power output.


By reducing the bremsstrahlung of a hot fusion plasma, the D-He3 reaction can proceed more efficiently and the reaction products can be harvested to better effect. The exhaust from an athermal D-He3 rocket is two separate streams - hydrogen with an exhaust velocity of 0.177 c (53,100 km/s) and helium with an exhaust velocity of 0.044 c (13,200 km/s). Every newton of thrust consumes 18.9 micrograms of deuterium and 28.3 micrograms of helium-3, and requires 85.8 MW of fusion power of which 4.3 MW is lost as neutrons and 81.5 MW of which goes into the exhaust jet.

D-T and D-D

Athermal fusors can also use D-T and D-D fusion, but the resulting neutron flux is generally prohibitive. Occasionally, an athermal reactor is designed to produce very high power outputs using D-T fusion, at 700 times the power of the nominal p-B11 fusion reaction, but the extraordinary neutron flux generally limits this to pulsed operation or results in very short reactor lives.

If the exhaust stream of an athermal fusion rocket is directed into an afterburner, the afterburner plasma is a thermal plasma which acts just like any other thermal plasma rocket.

The Almucantar, a magnetic confinement fusion drive freight class

Some common designs

Magnetic confinement: The Magnetic confinement fusion rocket uses bucking coils to extract a magnetic flux tube from a toroidal magnetic fusion reactor and exhaust the thrust. There were many technical difficulties to be overcome, especially involving magnetic field strength and the size and weight of the coils, and this engine only became practical with the invention of lightweight supercompact fusion reactors during the mid 2nd century. They very quickly became standard in the Interplanetary Age, until being supplanted in the inner solar system by the anti-matter fusion hybrid drive that became economical with increased output by the big antimatter cartels.

Gasdynamic Mirror Fusion: A form of magnetic confinement where the fusion does not occur in a torus, but instead linearly, with mirror magnets at each end. This cylindrical arrangement is necessarily very long, but the long reaction chamber allows a wide array of heat radiators to be conveniently located along the length of the chamber. (see image at head of page)

Inertial Confinement: The fuel is targeted by a halo of beams from lasers or particle accelerators. The fuel is held in place long enough to undergo fusion by its own inertia, and by the pressure of the beams themselves. A typical inertial confinement motor has a central reaction chamber entirely surrounded on all sides with inwardly directed beam generators, forming an outer concentric shell with only one opening - the rocket nozzle itself.

Inertial Electrostatic Confinement: These rockets use electrostatic fields to contain and direct the fusion reaction. The configuration can be spherical or cylindrical; one useful design, the Polywell fusor, uses a polyhedral arrangement, giving these rockets a distinctive geometry. Only advanced IEC fusion motors are capable of creating the conditions required to contain p-B11 fusion.

Some design considerations

Radiators: In all fusion motors, the intense heat of operation requires efficient cooling systems and large radiators to prevent a melt-down. Radiators are most efficient in space if they do not radiate towards each other, so either a single flat sheet of radiating surface or two sheets arranged at right angles (or combinations thereof) are most efficient.

Shielding: Fusion drives produce levels of neutrons and high energy photons, depending on the reaction. The best shielding for fast neutrons is hydrogen-containing materials, such as water or polymers, while slower neutrons can be stopped by boron. High energy photons are best stopped by a later of dense metal.

Interstellar Use: Internal fusion drives can have an Isp between 400,000 and 3,000,000 seconds, a thrust to weight ratio of 30 to 1 or better, and a maximum delta-v of 30,000 km/sec or maybe more, if the extreme mass fraction required can be tolerated. As a note, 30,000 km/sec is roughly equal to 0.1c, and thus internal fusion drives are the first drives to even potentially have use as interstellar drives. However the amount of payload a fusion drive would be able to propel to this speed is so small that there would be no capacity for deceleration. Interstellar craft powered by ordinary fusion are vanishingly rare.

See also Antimatter-catalysed Fusion, Conversion Drive

Related Articles
  • Antimatter Catalysed Fusion
  • Antimatter Propulsion
  • Chemical Rockets
  • Fission Drives
  • Fusion Reactor
  • Nuclear Fusion
  • Nuclear Pulse Drives
  • Propulsion Performance Statistics
  • Proton - Text by M. Alan Kazlev
    One of the two basic elementary particles found the atomic nucleus, the other being the neutron. It has a positive charge equal and opposite that of the electron, and a mass similar to the neutron. Protons have a mass of 1.007276 daltons, or 1.6726 = 10-27 kg.
  • Proton-Proton Chain - Text by M. Alan Kazlev
    A series of three thermonuclear reactions that convert hydrogen nuclei to helium nuclei, converting a tiny amount of mass into energy. The process is efficient at temperatures above 10,000,000 K. In most less massive stars, this chain is the primary source of heat and radiation. The proton-proton chain converts hydrogen into helium releasing energy in the form of particles and gamma-rays. Hydrogen is converted into helium in a chain of reactions. The first reaction takes an average of 1 billion years to occur while the others are much shorter. Generally there are so many hydrogen nuclei that the 1 billion year waiting period does not stop it from producing tremendous radiation.
  • QED Fusion Reactor - Text by M. Alan Kazlev, Inspired by an entry in the Lunar Institute of Technology
    The electrostatic fusion reactor first suggested by Robert Bussard of 20th century Old Earth (C.E.), uses fusion fuel confined by a spherical voltage potential well of around 100,000 volts. It makes an excellent, and fairly low tech (late Information Age / early Interplanetary Age equivalent) light power plant, that it can be used as part of a basic high thrust to weight ratio propulsion system for spacecraft and aircraft. Basic propulsion versions have a specific impulse of between 1500 and 6000 sec.
  • Rocket Performance
  • Superconductors
  • Torch Drives
Appears in Topics
Development Notes
Text by Luke Campbellsome material by M. Alan Kazlev, Mark Mcamuk, Chris Shaeffer, Richard Baker and David Dye
Initially published on 04 June 2010.