Chapter 8: Satellite Propulsion Systems¶
Key Concepts: Tsiolkovsky rocket equation, specific impulse, chemical propulsion (monopropellant, bipropellant, cold gas), electric propulsion (gridded ion, Hall effect, electrospray, PPT), propellant budgeting, all-electric GEO satellites, advanced and nuclear propulsion.
Propulsion is a critical enabling technology for satellite missions. Whether correcting orbital injection errors, executing station-keeping maneuvers over a 15-year operational lifetime, or raising a spacecraft from low Earth orbit to geostationary orbit, the propulsion system determines mission feasibility, lifetime, and mass budget. This chapter provides a comprehensive treatment of the physical principles, hardware implementations, performance trade-offs, and mission applications of satellite propulsion systems, from classical chemical thrusters to modern electric propulsion that is transforming commercial satellite operations.
8.1 Propulsion Fundamentals¶
8.1.1 The Tsiolkovsky Rocket Equation¶
The foundational relationship governing all rocket propulsion is the Tsiolkovsky rocket equation, derived in 1903. It relates the velocity change achievable by a spacecraft to the mass of propellant consumed:
where \(\Delta v\) is the total velocity change (m/s), \(v_e\) is the effective exhaust velocity (m/s), \(m_0\) is the initial (wet) mass, \(m_f\) is the final (dry) mass, \(I_{sp}\) is the specific impulse (s), and \(g_0 = 9.80665\) m/s\(^2\) is standard gravitational acceleration. The ratio \(m_0/m_f\) is the mass ratio, and the propellant mass fraction is \(\zeta = (m_0 - m_f)/m_0\).
For a spacecraft requiring multiple \(\Delta v\) maneuvers, the total propellant mass required for each maneuver \(i\) is:
This exponential dependence on \(\Delta v / (I_{sp} g_0)\) is crucial: doubling \(I_{sp}\) for the same \(\Delta v\) reduces propellant fraction dramatically, which is why electric propulsion with \(I_{sp} > 1500\) s is so attractive for large \(\Delta v\) missions.
8.1.2 Thrust and Specific Impulse¶
The thrust force produced by a rocket engine is:
where \(\dot{m}\) is the propellant mass flow rate (kg/s), \(v_e\) is the exhaust velocity at the nozzle exit (m/s), \(P_e\) is the nozzle exit pressure (Pa), \(P_a\) is the ambient pressure (Pa, effectively zero in space), and \(A_e\) is the nozzle exit area (m\(^2\)). The first term is the momentum thrust; the second is the pressure thrust. In the vacuum of space, \(P_a = 0\), and for a well-designed nozzle with exit pressure matched to near-vacuum conditions, the two terms combine into an effective exhaust velocity:
Specific impulse is defined as thrust per unit weight flow of propellant:
\(I_{sp}\) has units of seconds and is the primary figure of merit for propellant efficiency: higher \(I_{sp}\) means less propellant is consumed for the same impulse delivered. The total impulse delivered by a thruster is \(J = F \cdot t_{burn} = \dot{m} \cdot I_{sp} \cdot g_0 \cdot t_{burn}\).
8.1.3 Nozzle Theory and the Ideal Rocket¶
For an ideal rocket with isentropic flow through a converging-diverging (de Laval) nozzle, the theoretical exhaust velocity is:
where \(\gamma\) is the ratio of specific heats, \(R_u = 8.314\) J/(mol\(\cdot\)K) is the universal gas constant, \(T_c\) is the combustion chamber temperature (K), \(M_w\) is the molecular weight of the exhaust products (kg/mol), and \(P_c\), \(P_e\) are chamber and exit pressures. This equation reveals the key drivers of high \(I_{sp}\): high combustion temperature \(T_c\) and low exhaust molecular weight \(M_w\). This is why hydrogen (\(M_w = 2\)) is such a valuable rocket propellant (Sutton and Biblarz 2017).
8.2 Chemical Propulsion Systems¶
Chemical propulsion exploits exothermic chemical reactions to heat propellant and expel it through a nozzle. It offers high thrust and heritage reliability at the cost of modest \(I_{sp}\) compared to electric alternatives (Wertz et al. 2011).
8.2.1 Cold Gas Propulsion¶
Cold gas thrusters are the simplest propulsion devices: a pressurized inert gas (typically nitrogen N\(_2\), or occasionally helium or R-134a refrigerant) is stored in a tank and expelled through a small nozzle. No combustion or heating occurs.
Performance is limited by the speed of sound in the stored gas. For nitrogen at 20\(^\circ\)C, the theoretical maximum \(I_{sp}\) is approximately 80 s, with practical values of 50–75 s. Thrust levels are typically 0.05–10 N, and system simplicity (no pyrotechnic valves, no propellant compatibility issues, no ignition system) makes cold gas ideal for:
-
CubeSat attitude control where volume and cost constraints dominate.
-
Proximity operations and formation flying where contamination of sensitive surfaces by chemical exhaust products must be avoided.
-
First-generation micro-satellite propulsion serving as a technology baseline.
The PRISMA formation-flying mission (2010) flew, among other systems, a MEMS cold gas micro-propulsion demonstration. Many 6U CubeSats now fly MEMS-fabricated cold gas micro-thrusters with nozzle throats of 100–500 \(\mu\)m diameter, produced by companies such as Vacco Industries and Bradford ECAPS. The key limitations are low \(I_{sp}\) (limiting \(\Delta v\) budget severely) and the large propellant tank volume required to store gas at useful pressures (typically 100–300 bar).
8.2.2 Monopropellant Hydrazine Systems¶
Hydrazine (N\(_2\)H\(_4\)) monopropellant thrusters have been the workhorse of satellite propulsion since the 1960s. The propellant decomposes exothermically when passed over a catalytic bed:
followed at high temperatures by:
The net reaction produces a gas mixture (N\(_2\), H\(_2\), NH\(_3\)) at temperatures of 600–900\(^\circ\)C, yielding an \(I_{sp}\) of 220–235 s in vacuum (Pisacane 2005). The Shell 405 catalyst (granular iridium on alumina support) is the industry standard, developed at Shell Development Company and qualified for space use in the 1960s. Alternative catalysts have been developed to improve cold-start behaviour and ignition reliability at low temperatures (\(-10^\circ\)C to \(+60^\circ\)C operational range).
Hydrazine thrusters range from 0.5 N (attitude control, \(< 1\) g/s flow rate) to 22 N (orbit adjust, \(\sim 10\) g/s). The system architecture uses one of two pressurization schemes:
-
Regulated: A pressurant gas (N\(_2\) or He) maintains constant feed pressure through a pressure regulator, providing constant thrust and \(I_{sp}\) throughout the mission but requiring a separate pressurant tank.
-
Blowdown: The pressurant and hydrazine share the same tank; as propellant is consumed, pressure drops (typically by a factor of 4:1 from 28 to 7 bar), reducing thrust substantially and \(I_{sp}\) by several percent over the mission. Simpler and lighter, this is the preferred choice for small satellites.
Hydrazine is classified as a toxic, carcinogenic substance (REACH SVHC list in Europe), requiring full SCAPE (Self-Contained Atmospheric Protective Ensemble) suit handling and specialized ground facilities. This has driven significant investment in green propellant development.
8.2.3 Green Monopropellants¶
Two ionic liquid-based green propellants have reached flight maturity as hydrazine replacements:
AF-M315E (ASCENT – Advanced Spacecraft Energetic Non-Toxic): Developed by AFRL (Air Force Research Laboratory) and commercialized by Aerojet Rocketdyne, AF-M315E is a hydroxyl ammonium nitrate (HAN)-based ionic liquid blend. Key properties:
-
\(I_{sp} \approx 257\) s (approximately 12% improvement over hydrazine at same chamber conditions).
-
Specific gravity \(\approx 1.47\) (50% denser than hydrazine at 1.004, reducing tank volume for the same propellant mass).
-
Non-toxic handling eliminates SCAPE suit requirements; classified as a Class 1.3C explosive only.
-
Flash point \(>\)107\(^\circ\)C (much safer than hydrazine flash point of \(\sim\)52\(^\circ\)C).
-
Higher catalyst preheat required (several hundred \(^\circ\)C versus roughly 100–150\(^\circ\)C for hydrazine), adding preheat time and power draw.
AF-M315E was first demonstrated in orbit on the GPIM (Green Propellant Infusion Mission, 2019) aboard a Ball Aerospace BCP-100 spacecraft, validating the 1 N-class ASCENT thrusters in orbit.
LMP-103S (HPGP – High Performance Green Propellant): Developed by ECAPS (a Bradford ECAPS / Bradford Space company) as an ammonium dinitramide (ADN) and water/fuel blend. Demonstrated on the PRISMA spacecraft (2010) at the 1 N thrust level, it achieves \(I_{sp} \approx 252\) s and has since flown on several commercial small satellites.
8.2.4 Bipropellant Systems¶
Bipropellant engines use separate fuel and oxidizer streams that combust upon mixing. This enables higher combustion temperatures and thus higher \(I_{sp}\) than monopropellants, at the cost of two independent propellant feed systems (valves, lines, tanks, and mixture ratio control).
The most common satellite bipropellant is MMH/MON-3 (monomethylhydrazine fuel / mixed oxides of nitrogen with 3% NO oxidizer), which is hypergolic – spontaneously ignites on contact – eliminating the need for an ignition system:
MMH/MON-3 achieves \(I_{sp} \approx 315\) s at a mixture ratio (O/F by mass) of approximately 1.65:1. The Marquardt R-4D thruster (490 N, \(I_{sp} = 311\) s) and its variants are used on hundreds of GEO spacecraft for liquid apogee engine (LAE) burns and attitude control. For apogee burns with thrust of 400–1000 N, bipropellant engines deliver GTO-to-GEO transfer in 1–3 firing sequences spanning 24–72 hours, compared to 6–12 months for all-electric orbit raising.
The N\(_2\)O\(_4\)/UDMH (unsymmetrical dimethylhydrazine) combination has been widely used by European Ariane upper stages and Russian/Chinese spacecraft, with \(I_{sp}\) of 310–316 s. The Russian S5.92 engine (about 20 kN, \(I_{sp} \approx 327\) s, used on the Fregat upper stage) exemplifies mature UDMH engine technology.
For future missions requiring higher performance without the toxicity of storable propellants, LOX/LCH\(_4\) (liquid oxygen/liquid methane) offers \(I_{sp} \approx 363\) s in vacuum. SpaceX Raptor achieves \(I_{sp} = 380\) s (vacuum) and is non-toxic, with potential for in-situ resource utilization (ISRU) at Mars where CH\(_4\) can be synthesized from atmospheric CO\(_2\) via the Sabatier reaction.
| Propellant | Type | \(I_{sp}\) (s) | Thrust Range | Notes |
|---|---|---|---|---|
| Cold gas (N\(_2\)) | Monoprop | 50–75 | 0.05–10 N | CubeSat ACS; no combustion |
| Hydrazine (N\(_2\)H\(_4\)) | Monoprop | 220–235 | 0.5–445 N | Heritage workhorse; toxic |
| AF-M315E (ASCENT) | Monoprop | 252–257 | 0.5–25 N | Green; preheat required |
| LMP-103S (HPGP) | Monoprop | 248–254 | 0.5–25 N | ADN-based; green |
| MMH / MON-3 | Bipropellant | 310–316 | 5–1000 N | GEO apogee standard |
| N\(_2\)O\(_4\) / UDMH | Bipropellant | 310–315 | 50–1000 N | European/Russian heritage |
| N\(_2\)O\(_4\) / MMH | Bipropellant | 318–325 | 100–1000 N | High-performance GEO |
| LOX / LCH\(_4\) | Bipropellant | 355–380 | 100 N–MN | Future interplanetary, ISRU |
| LOX / LH\(_2\) | Bipropellant | 440–455 | 50 N–MN | Upper stages; not storable |
8.3 Electric Propulsion¶
Electric propulsion (EP) uses electrical power from solar arrays or a radioisotope/fission source to accelerate propellant to exhaust velocities far exceeding those achievable by chemical combustion. The \(I_{sp}\) is not thermodynamically limited by combustion temperature but by the available electrical energy per unit propellant mass. The key trade is power: EP thrusters require substantial electrical power (100 W to 100 kW) but consume far less propellant for the same \(\Delta v\) (Wertz et al. 2011).
The jet power efficiency of an EP thruster is:
Typical EP thruster efficiencies are 40–70%, depending on technology. For a given power \(P_{in}\) and efficiency \(\eta\), there is a fundamental trade between thrust and \(I_{sp}\): for fixed power, increasing \(I_{sp}\) (exhaust velocity) reduces propellant flow rate and thus thrust.
8.3.1 Gridded Ion Thrusters¶
Gridded ion thrusters (GITs), also known as electrostatic ion thrusters, use a plasma discharge to ionize propellant (typically xenon Xe, occasionally krypton Kr or iodine I\(_2\)) and then electrostatically accelerate ions through precision-drilled metal grids.
The Kaufman-type thruster (named for Harold Kaufman, NASA Lewis Research Center, 1960s) operates in three stages:
-
Ionization: Xenon gas is fed into a cylindrical discharge chamber where electron bombardment from a hollow cathode creates a dense plasma (\(> 90\)% ionization efficiency). The discharge chamber operates at 25–35 V.
-
Extraction: A multi-aperture screen grid (at plasma potential, \(+\)1000 to \(+\)5000 V) and accel grid (at negative high voltage, typically \(-200\) V, to prevent back-streaming electrons) extract and collimate positive Xe\(^+\) ions from the plasma boundary.
-
Acceleration: The net beam voltage \(V_b\) (sum of screen and accel potentials, 1000–5000 V) accelerates ions to exit velocity:
where \(q = 1.602 \times 10^{-19}\) C is the elementary charge and \(m_i\) is the ion mass. For Xe\(^+\): \(m_i = 2.18 \times 10^{-25}\) kg (atomic mass 131.3 amu). For \(V_b = 1100\) V (approximately the NSTAR full-power beam voltage), the ideal exit velocity is \(v_{exit} \approx 40\) km/s, corresponding to an ideal specific impulse of:
The general relationship between \(I_{sp}\) and beam voltage for a gridded ion thruster is:
where \(\eta_b < 1\) is an efficiency factor that accounts for beam divergence, doubly charged ions and incomplete propellant utilization (it reduces the ideal 4100 s above to the roughly 3100 s actually measured for NSTAR). Total thruster efficiency (beam power / input DC power) ranges from 60% to 80%.
Key flight examples of gridded ion thrusters:
-
NSTAR (NASA Solar Electric Propulsion Technology Application Readiness): 30 cm diameter thruster, 2.3 kW input power, \(I_{sp} = 3100\) s at full throttle, 93 mN thrust, throttle range 0.5–2.3 kW (\(I_{sp} = 1979\)–3100 s). Flew on Deep Space 1 (1998–2001, first use of ion propulsion as primary propulsion on a deep-space mission) and Dawn (2007–2018, Vesta and Ceres orbit insertion).
-
T6 (QinetiQ, now the Moog T6 for ESA): 22 cm thruster, 4.26 kW input, 145 mN thrust, \(I_{sp} = 4120\) s at design point. Four T6 thrusters power the BepiColombo Mercury Transfer Module (MTM) for the ESA/JAXA Mercury mission (launched 2018, Mercury orbit insertion planned for late 2026 after a revised trajectory), contributing to the deep-space \(\Delta v\) needed for solar system braking.
-
NEXT-C (NASA Evolutionary Xenon Thruster - Commercial): about 7 kW input, 236 mN, \(I_{sp} \approx 4190\) s. A single unit flew on NASA’s DART mission (2021). A long-duration life test exceeded 50,000 hours.
-
RIT-22 (Airbus Defence and Space): Radio-frequency ion thruster using RF inductive discharge (eliminating hollow cathode life limits), 22 cm beam, \(I_{sp} = 4600\) s at 5 kW, developed for GEO orbit raising and station keeping.
A critical component is the neutralizer cathode, which emits electrons into the ion beam downstream to prevent spacecraft charging to multi-kV potentials that would attract the beam back, and to prevent repulsion of subsequent ion beam pulses.
Grid erosion via charge-exchange (CEX) ion sputtering limits thruster life. NSTAR completed a 30,472-hour extended life test (ELT) with acceptable grid erosion, demonstrating the feasibility of multi-year deep space missions.
8.3.2 Hall Effect Thrusters¶
Hall Effect Thrusters (HETs), also called Hall thrusters, Closed Drift Thrusters (CDTs), or Stationary Plasma Thrusters (SPT, the Soviet/Russian designation), are the dominant electric propulsion technology for commercial GEO satellite station keeping. They have accumulated hundreds of thousands of hours of flight operation since the Soviet SPT-50 flew in 1971 on Meteor weather satellites.
The operating principle exploits the \(\mathbf{E} \times \mathbf{B}\) drift of electrons in crossed electric and magnetic fields:
-
A radial magnetic field \(\mathbf{B}\) (100–300 G) is applied across the annular discharge channel by electromagnets or permanent magnets.
-
An axial electric field \(\mathbf{E}\) is applied between the anode (upstream, at \(+V_d\)) and the external hollow cathode (at ground or negative potential).
-
Electrons emitted by the hollow cathode drift into the channel but are trapped by the crossed fields in an azimuthal Hall current (the \(\mathbf{E} \times \mathbf{B}\) drift), forming a rotating electron cloud that acts as a virtual cathode for ionizing neutral xenon.
-
Positive Xe\(^+\) ions, being far too massive to be significantly magnetized (ion Larmor radius \(\gg\) channel length), experience the full axial electric field and are accelerated out of the thruster.
The discharge voltage \(V_d\) (200–800 V for typical operation) determines exhaust velocity and \(I_{sp}\):
where \(\alpha \approx 0.85\)–0.95 accounts for the effective voltage fraction used for ion acceleration (the remainder drives ionization and electron heating). At \(V_d = 300\) V with xenon the ideal value is about 2000 s, and measured values are lower (typically 1500–1800 s) because of incomplete propellant utilization and plume divergence; high-voltage Hall thrusters operating near 1000 V reach roughly 2500–3000 s.
HETs have significantly lower \(I_{sp}\) than gridded ion thrusters but substantially higher thrust density (thrust per unit channel area) and simpler construction with no fragile high-voltage grids subject to ion sputtering. Practical discharge voltages for commercial GEO station keeping are 300–500 V, giving 1500–2000 s \(I_{sp}\).
Key flight Hall thruster examples:
-
SPT-100 (Fakel Enterprise, Russia / licensed by Safran, France): 100 mm mean channel diameter, 1.35 kW, about 83 mN, \(I_{sp} \approx 1600\) s. It has flown on many Russian and Western commercial spacecraft since the mid-1990s and is among the most widely flown Hall thrusters.
-
PPS-1350 (Safran Snecma, France): 1.5 kW, 90 mN, \(I_{sp} = 1650\) s. Used on the ESA SMART-1 lunar mission (2003–2006), which demonstrated electric propulsion for lunar capture, and on several commercial GEO satellites.
-
BHT-200 (Busek, USA): 200 W small HET, 12.8 mN, \(I_{sp} = 1390\) s. Flew on AFRL TacSat-2 (2006) and on several other small satellites.
-
PPS-5000 (Safran, France): 5 kW class, roughly 300 mN, \(I_{sp} \approx 1800\) s. Used on Airbus and Thales Alenia Space GEO platforms.
-
X3 (University of Michigan / NASA Glenn): Multi-channel (nested) HET, demonstrated operation at about 100 kW (about 5.4 N of thrust) in ground tests. Intended for future crewed Mars missions requiring megawatt-class SEP.
Krypton propellant for LEO constellations: SpaceX adopted krypton (Kr, atomic mass 83.8 amu, versus Xe at 131.3 amu) for the Hall thrusters of its first-generation Starlink satellites, and has used argon in later (V2 Mini) satellites. At the same discharge voltage, krypton produces higher exhaust velocity (lower mass) for a higher \(I_{sp}\) at the same voltage, but with lower efficiency than xenon because krypton ionizes less readily. The dominant motivation is economics: krypton and argon are far cheaper and more abundant than xenon, which matters when thousands of satellites each need propellant. Starlink is the largest fleet of Hall-thruster-propelled spacecraft in history.
8.3.3 Electrospray Thrusters¶
Electrospray thrusters (also known as field-emission electric propulsion, FEEP, when using liquid metals, or colloid thrusters when using ionic liquids) extract and accelerate charged species from a liquid surface at a sharp emitter tip subject to a strong electric field.
When the applied voltage exceeds a threshold, the balance between electrostatic pressure and surface tension forms a Taylor cone at the tip apex, from which a fine jet of ions or charged droplets is emitted. For ionic liquid propellants (e.g., 1-ethyl-3-methylimidazolium tetrafluoroborate, [EMIm][BF\(_4\)]), both positive and negative ions are emitted alternately (or from dual emitter arrays), avoiding spacecraft charging issues.
Key performance characteristics:
-
Thrust range: 1–1000 \(\mu\)N (micro-Newton class), making them uniquely suited to precision drag-free and attitude control applications.
-
\(I_{sp}\): 500–5000 s depending on whether droplet mode (lower, 500–1000 s) or ion evaporation mode (higher, 3000–5000 s) dominates.
-
Power: 1–50 W per thruster head; very power-efficient at micro-Newton thrust.
-
No moving parts; fast electrical response (millisecond time constant, limited by propellant wetting dynamics).
-
Propellant stored as liquid at room temperature in simple reservoirs; no pressurization required.
The LISA Pathfinder mission (ESA, with a NASA contribution; launched December 2015) demonstrated drag-free flight with test-mass acceleration noise better than required. ESA’s micro-propulsion system used Gaia-heritage cold-gas micro-newton thrusters, while NASA’s ST7 Disturbance Reduction System flew Busek colloid (electrospray) micro-newton thrusters and also achieved drag-free control. These results support the micro-propulsion needs of the LISA gravitational-wave observatory (adopted by ESA in 2024, launch planned for the mid-2030s).
Commercial electrospray systems include Accion Systems TILE (Tiled Ionic Liquid Electrospray) modules, which integrate arrays of silicon micro-fabricated emitter tips for small-satellite propulsion, and which have been flown in orbit on small satellites.
8.3.4 Pulsed Plasma Thrusters¶
Pulsed Plasma Thrusters (PPTs) ablate a solid Teflon (PTFE, polytetrafluoroethylene) propellant bar using a pulsed capacitor discharge, producing a plasma that is electromagnetically accelerated by Lorentz forces. Each pulse operates as follows:
-
A high-voltage capacitor (charged to 1–2 kV, energy \(E_{cap} = \frac{1}{2}CV^2\), typically 2–20 J) discharges across a spark-plug triggered electrode gap adjacent to the Teflon face.
-
The spark ionizes a thin surface layer of Teflon, forming a transient plasma. The plasma current (\(10^4\)–\(10^5\) A flowing for \(\sim\)5 \(\mu\)s) interacts with the self-generated magnetic field.
-
Lorentz force \(\mathbf{J} \times \mathbf{B}\) accelerates the plasma axially, while the ablated neutral gas is also entrained and heated, contributing to thrust.
-
The Teflon bar is spring-fed (or gravity-fed in testing) to advance fresh surface into the discharge gap.
Each pulse delivers an impulse bit (\(I_{bit}\)) of typically 100–500 \(\mu\)N\(\cdot\)s, with \(I_{sp}\) of 500–1500 s and pulse repetition rates of 1–100 Hz. Total thruster efficiency (jet power / input power) is low (5–20%) due to late-time ablation of neutral gas that does not participate in electromagnetic acceleration.
PPT advantages: solid propellant eliminates all feed system components (valves, lines, tanks, regulators); extremely simple and low-cost; propellant is non-toxic; capable of very small impulse bits useful for attitude control. NASA’s EO-1 (2000) flew a PPT as an experimental attitude-control actuator, and PPTs have since flown on various small satellites and CubeSats.
8.4 All-Electric GEO Satellites¶
Traditional GEO satellites use a chemical bipropellant liquid apogee engine (LAE, \(\Delta v \approx 1800\) m/s, firing duration 1–3 hours) to transfer from geostationary transfer orbit (GTO, \(e \approx 0.73\), \(a \approx 24{,}500\) km) to GEO, then chemical or electric thrusters for station keeping. All-electric satellites eliminate the LAE entirely, using electric thrusters for the complete GTO-to-GEO transfer and all subsequent station-keeping (Wertz et al. 2011).
The Boeing 702SP (small platform) was the first operational all-electric commercial GEO satellite platform. ABS-3A and Eutelsat 115 West B were launched together in March 2015 aboard a Falcon 9. The 702SP uses XIPS-25 gridded ion thrusters for orbit raising and station keeping. Orbit raising from GTO to GEO takes several months of continuous thrusting (against days for chemical).
The mass advantage of all-electric propulsion is derived from the Tsiolkovsky equation. For NSSK over 15 years with total \(\Delta v \approx 800\) m/s:
For a 2000 kg dry satellite, this represents 840 kg (chemical) versus 94 kg (EP) for station-keeping propellant – a saving of 746 kg that can be applied to payload capacity or reduced launch cost. Eliminating the LAE propellant (of the order of 1200 kg for a 3000 kg wet satellite) adds to this saving, which is why all-electric satellites can be launched at a much lower mass than chemical equivalents (at the cost of a longer orbit-raising time).
The all-electric orbit-raising trajectory is a low-thrust spiral from GTO to GEO. The spacecraft fires thrusters near apogee (to raise perigee) and near perigee (to raise apogee and reduce inclination). The spiral passes through the Van Allen radiation belts for several months, requiring:
-
Electronics designed for a substantially higher total ionizing dose than for chemical orbit raising.
-
Radiation-resistant solar arrays (for example thicker cover glass) to limit degradation during the belt crossings.
-
Careful thermal management during long-duration thrusting.
Subsequent all-electric or hybrid GEO platforms include Airbus’s Eurostar E3000 EOR (for example Eutelsat 172B, launched 2017), the Eurostar Neo, and Thales Alenia Space’s Spacebus Neo, using Hall or ion thrusters.
8.5 Propellant Budget Analysis¶
A systematic propellant budget quantifies all \(\Delta v\) demands across the mission lifetime and derives total propellant mass. The sequence accounts for each maneuver separately (Eq. (8.2)) since spacecraft mass decreases with each consumption event (Wertz et al. 2011).
For a GEO communications satellite with 15-year design life, the major \(\Delta v\) requirements are:
| Maneuver | \(\Delta v\) (m/s) | Prop. Mass (kg) | \(I_{sp}\) (s) | System |
|---|---|---|---|---|
| GTO apogee kick (LAE) | 1,820 | 1,772 | 315 | Bipropellant |
| Inclination trim at GEO | 50 | 36 | 315 | Bipropellant |
| Injection error correction | 30 | 21 | 315 | Bipropellant |
| N–S station keeping (15 yr) | \(15 \times 50 = 750\) | 598 | 235 | Hydrazine |
| E–W station keeping (15 yr) | \(15 \times 2 = 30\) | 20 | 235 | Hydrazine |
| Attitude control (momentum dump) | 20 | 14 | 230 | Hydrazine |
| Repositioning maneuvers | 20 | 13 | 235 | Hydrazine |
| End-of-life disposal to graveyard | 11 | 7 | 230 | Hydrazine |
| Uncertainties and residuals (5% of propellant) | — | 124 | — | — |
| Total wet mass | \(\sim\)4105 | \(\sim\)2605 |
The north–south station-keeping (NSSK) \(\Delta v\) of \(\sim\)50 m/s/year is driven by solar and lunar gravitational perturbations that cause orbital inclination to increase at approximately 0.75/year without correction. Maintaining the inclination within about \(\pm\)0.05 of the equatorial plane requires frequent NSSK maneuvers (typically every one to two weeks for chemical systems, or near-continuous thrusting with electric propulsion). East–west station keeping (EWSK) corrects longitudinal drift caused by Earth’s equatorial ellipticity (triaxiality), requiring only \(\sim\)2 m/s/year.
End-of-life disposal to a graveyard orbit at least 300 km above GEO requires \(\Delta v = 11\) m/s, recommended by ITU-R S.1003 and the IADC guidelines and required for U.S.-licensed satellites by FCC rules. The graveyard orbit lies at \(\sim\)36,086 km altitude where the satellite will remain for hundreds of years (there is no atmospheric drag at GEO altitudes).
8.6 Advanced Propulsion Concepts¶
8.6.1 Nuclear Thermal Propulsion¶
Nuclear Thermal Propulsion (NTP) heats hydrogen propellant using a fission reactor, achieving exhaust temperatures of roughly 2500–3000 K. This is lower than chemical flame temperatures (about 3500 K), but the very low molecular weight of hydrogen gives a much higher exhaust velocity. The theoretical \(I_{sp}\) for hydrogen heated to 3000 K:
where \(M_w = 2\) kg/kmol for molecular hydrogen and \(T_c = 2700\) K is a representative reactor exit temperature. The U.S. Rover/NERVA (Nuclear Engine for Rocket Vehicle Application) programs (1950s to 1972) ground-tested engines (for example Phoebus-2A and XE-Prime) with specific impulses of the order of 800 s and thrust levels up to roughly 900 kN, demonstrating the concept before the programs were cancelled.
The DARPA–NASA DRACO (Demonstration Rocket for Agile Cislunar Operations) program aimed to demonstrate NTP in Earth orbit (BWX Technologies for the reactor, Lockheed Martin for the vehicle) but was reported to have been cancelled in 2025. NTP’s principal appeal is shortening crewed Mars transit times compared with chemical propulsion, which would reduce crew exposure to galactic cosmic rays.
8.6.2 Solar Electric Propulsion at Scale¶
SEP performance scales directly with available solar power. The BepiColombo mission (ESA/JAXA, launched 2018, Mercury orbit insertion planned for late 2026) uses T6 ion thrusters (two operating at a time, roughly 4.5 kW each) together with planetary gravity assists to reach Mercury.
NASA’s Dawn mission (2007–2018) accumulated more than 50,000 hours of ion thruster operation on NSTAR thrusters powered by solar arrays of about 10 kW at 1 AU, consuming most of its roughly 425 kg of xenon over the 11-year mission. Dawn is, as of 2025, the only spacecraft to have orbited two extraterrestrial bodies (Vesta and Ceres).
For future high-power applications, the 100 kW class Hall thruster X3 has been tested at NASA Glenn Research Center, while the Advanced Electric Propulsion System (AEPS), a Hall thruster of about 12–13 kW, was developed by Aerojet Rocketdyne for Gateway’s Power and Propulsion Element (the future of the Gateway program is uncertain as of 2025). Megawatt-class nuclear electric propulsion (NEP), combining fission reactors with ion thruster clusters, has been studied for crewed Mars missions (Isp \(>\) 5000 s, continuous thrust for months) and rapid cargo delivery to the outer solar system.
| System | \(I_{sp}\) (s) | Thrust | Power | TRL | Best Application |
|---|---|---|---|---|---|
| Cold gas (N\(_2\)) | 50–75 | 0.05–10 N | None | 9 | CubeSat ACS |
| Hydrazine monoprop | 220–235 | 0.5–22 N | Heater only | 9 | Small sat \(\Delta v\), ACS |
| Green prop. ASCENT | 252–257 | 0.5–25 N | Heater only | 8 | Hydrazine replacement |
| Biprop. MMH/MON-3 | 310–316 | 5–1000 N | None | 9 | GEO apogee, maneuvers |
| Gridded ion (Xe) | 2000–10000 | 10–250 mN | 0.5–7 kW | 9 | Deep space, GEO-all-elec |
| Hall effect (Xe) | 1500–3500 | 10–600 mN | 0.2–50 kW | 9 | GEO SK, orbit raising |
| Hall effect (Kr, Ar) | 1500–2200 | 10–100 mN | 0.25–5 kW | 9 | LEO constellations |
| Electrospray | 500–5000 | 1–1000 \(\mu\)N | 1–50 W | 8–9 | Drag-free control, fine ACS |
| PPT (Teflon) | 500–1500 | 0.05–2 mN | 5–200 W | 9 | CubeSat ACS |
| NTP (H\(_2\)) | 800–1000 | \(10^4\)–\(10^6\) N | Nuclear (GW\(_{th}\)) | 5 | Crewed interplanetary |
| NEP (gridded ion) | 5000–10000 | 10–1000 N | Nuclear kW–MW | 4 | Deep space cargo |
8.7 Thruster Selection and System Design¶
Selecting a propulsion system for a given mission requires balancing multiple constraints simultaneously. The primary decision variables are:
\(\Delta v\) requirement and time budget: High \(\Delta v\) (\(>\)1 km/s) strongly favors EP if sufficient power is available and if schedule permits slow orbit raising. For rapid maneuvering (\(<\)days), chemical high-thrust systems are necessary. Mixed architectures (chemical for apogee and emergency, EP for station keeping) are common in traditional large GEO satellites.
Power availability: EP requires electrical power proportional to thrust level. At 60% total efficiency, generating 200 mN at 1800 s \(I_{sp}\) requires \(\sim\)3 kW. Large GEO satellite solar arrays provide 10–25 kW, easily supporting one to four HETs in parallel. CubeSats generating 5–30 W can only support micro-thrusters (electrospray, cold gas, micro-PPT).
Contamination: Ion thruster charge-exchange (CEX) products deposit on solar cell cover glass (sputtered molybdenum, tungsten), reducing transmission. Hall thruster backflux contains sputtered boron nitride channel material. Electrospray ionic liquid droplets can deposit on optical surfaces. Thruster placement geometry and solar array orientation during thrusting are designed to minimize plume impingement.
End-of-life compliance: ITU, FCC, and ESA guidelines require LEO satellites to deorbit within 5 years and GEO satellites to achieve graveyard orbit. Propulsion systems must retain sufficient propellant for deorbit (LEO) or raise to graveyard (GEO). Passivation (venting remaining propellant) is required to prevent on-orbit breakup from residual energy sources.
8.8 Summary¶
This chapter has surveyed the full breadth of satellite propulsion technologies from first principles. The Tsiolkovsky rocket equation (Eq. (8.1)) remains the central governing relationship: the exponential dependence of propellant mass on \(\Delta v / (I_{sp} g_0)\) drives the quest for higher specific impulse and makes electric propulsion with \(I_{sp} > 1500\) s enormously attractive for high-\(\Delta v\) missions.
Chemical propulsion (hydrazine monopropellant, MMH/MON-3 bipropellant) dominates where high thrust, rapid maneuvering, or power-limited environments prevail. Green propellants (AF-M315E, LMP-103S) are progressively replacing toxic hydrazine as regulatory pressure increases. Electric propulsion, particularly Hall effect thrusters, has transformed commercial GEO satellite architecture: the all-electric 702SP and Eurostar 3000 Neo platforms eliminate the heavy liquid apogee engine, saving 400–600 kg of launch mass. The Starlink constellation’s krypton HET fleet represents the largest deployment of electric propulsion in history.
Looking forward, nuclear thermal propulsion (NASA DRACO program) offers a path to crewed Mars missions with substantially reduced transit times, while high-power SEP (BepiColombo T6, NEXT-C, X3) is enabling increasingly ambitious robotic science missions to the inner and outer solar system (Sutton and Biblarz 2017; Wertz et al. 2011; Pisacane 2005).
Chapter Summary – Key Equations:
-
Tsiolkovsky: \(\Delta v = I_{sp} g_0 \ln(m_0/m_f)\)
-
Thrust equation: \(F = \dot{m} v_e + (P_e - P_a)A_e\)
-
Ion exit velocity: \(v_{exit} = \sqrt{2qV_b/m_i}\)
-
GEO NSSK budget: \(\approx 50\) m/s/year; EWSK: \(\approx 2\) m/s/year
-
GEO disposal: minimum \(\Delta v = 11\) m/s to graveyard (\(+\)300 km above GEO)