Selecting a Satellite Propulsion System

Criteria for Selecting a Propulsion System

Selecting a propulsion system(s) for your mission is a complex and highly interconnected process. To narrow down the solution space, designers must answer a few key questions.

1. What are my constraints? (mass, power, cost, time)
2. What does my satellite need to do? (orbit raising, station keeping, collision avoidance, de-orbiting, etc.)
3. What is my risk tollerance? (tech demo, flight heritage vs. proven flight heritage)

Based on answers to the above, there likely exist numerous solutions that meet your mission requirements. As systems engineers we do not seek the optimal solution, and in fact, the optimial soluition will likely take a sub-optimal time to find. They key is to narrow down your selection to a few suitble options and then engage with those suppliers to see who can deliver on time and on-budget

Design and Project Constraints

If you are in the early stages of your spacecraft development there may not be exact constraints. It is however important to have a general idea of constraints especially for things like power and mass for your design and cost and schedule for the mission. Developing a spacecraft that needs to launch in a year vs. 3 years will significanly narrow your descision space.

What Does the Satellite Need to Do?

From the high level mission objectives you will have a general idea of what the spacecraft has to do in terms of maneuvering. A SpaceX rideshare will typically drop off your spacecraft at an altitude of around 500-600km. At the lower end of the range, most average spacecraft will deorbit naturally within a few years of mission end (some variation with the solar cycle). If your mission is simple and orbital maneuvering is not required you may not need a propulsion system. However, for things like formation flying, precise station keeping, orbit transfers, and for de-orbiting, a propulsion system is required.

You can quantify the "what does my spacecraft need to do" by creating a deltaV budget. DeltaV is a useful reference for "how much maneuvering" a spacecraft can do. It captures your mobility requirement but does not depend on the mass of your spacecraft or efficiency of your propulsion system.

To craft a deltaV budget a mission should be broken down into the various stages and functions that are expected over the course of the mission duration.

Orbital Maneuvers These can vary significantly from mission-to-mission. Orbital maneuvers include any maneuvers that are intended to cause an intentional change to the spacecraft's orbital elements. They include orbit raising/lowering, changing RAAN, inclination, transfering to other orbits, adjusting the semi-major axis for formation flying and much more.

Station Keeping These are also technically orbital maneuvers but the intention here is a bit different. For example, imagine your spacecraft is injected into a 10:00 LTAN sunsynchronus orbit at an altitude of 500km. Over time, due to various disturbances covered in (REF TO ORBITAL MECHANICS) your LTAN will drift. You can correct for this by performing small adjustments periodically to ensure you do not drift too far from your intended orbit.

Collision Avoidance (ColA) Exactly what it sounds like. With the ever growing population of objects, both operable satellites and derilict debris, the probability of a close encounter over a multi year mission is ever increasing. When operators receive a conjunction alert there can be a very narrow window to act and by acting sooner spacecraft can utilize less fuel to maneuver safely out of the way.

Depending on what orbit you are in, there exist conservative estimates for how much deltaV one must account for when mission planning. As the number of objects in LEO increases so will the frequency of required ColA maneuvers. As such it is recommended to maintain a sufficent amount of margin during the planning phase.

De-orbit The FAA require that satellites deorbit within 5 years from the end of their mission life. For many satellites launched into 500-600 km orbits, this can be achieved naturally due to natural effects of atmospheric drag causing the gradual reducion in semi-major axis. However, for satellites with high ballistic coefficents (high mass with low surface area) and spacecraft launched into higher orbits the effects of atmospheric drag alone are insufficent for meeting this requriement. As such, satellites must utilize a propulsion system to either de-orbit directly or reduce their orbit sufficently that atmospheric drag will cause the subsequent de-orbit within the required time period.

Risk Tolerance

When selecting spacecraft subsystems, effectively managing and understanding risk is critical to mission success. The following key areas should be thoroughly evaluated. Note, these are not design requirements they are programatic risks.

Maneuver Execution Speed

• How quickly must critical maneuvers (orbit insertion, station-keeping, collision avoidance, de-orbiting) be completed?

• Faster execution typically means increased complexity and potentially higher system risks, requiring more robust hardware and precise control systems.

Propellant Storage Duration

• How long will propellant be stored onboard before it is utilized?

• Extended storage periods may lead to increased risks associated with propellant degradation, tank leakage, or pressure management.

Flight Heritage and Reliability

• Has the proposed subsystem previously flown in space?

• Systems with extensive flight heritage generally present lower risks, while unproven technologies or components might introduce unknown reliability issues.

Schedule and Budget Margins

• Is there sufficient margin built into the schedule and budget?

• Lack of adequate margin can force undesirable compromises, limit troubleshooting time, and significantly increase mission risk.

---

Mission Analysis Example

Mission Requirements

Distinguish hard requirements with 'shall' and softer requriements with 'should'. Do not be afraid to only use 'should' if at an early stage.

• The spacecraft shall fit on a SpaceX quarter plate and thus has a maximum wet mass of 50kg
• The propulsion system should use a single switch from power system X with a maximum power draw of 50W.
• The propulsion system should cost less than 250k
• The spacecraft shall launch in 2 years
• The spcaecraft shall have a mission life of 2 years

Some of the constraints like mass and power can be assessed early on to narrow down the search. However, cost and lead-time estimates require engaging with suppliers.

Satellite Maneuvering Requirements

• The satellite shall be dropped off in a 550km SSO orbit
• The satellite shall reduce its altitude to 450km
• The satellite shall maintain the 450km altitude for it's mission life
• The spacecraft shall passively de-orbit within 5 years of the end of it's mission life.

Lets decompose these requirements into an initial deltaV budget.

Orbit Lowering We can approximate orbit lowering as a single Hohmann transfer maneuver for now. In reality the maneuver will not be this ideal as we will need to account for the non impulsive behavior of propulsion systems or even the possibility of a "spiral down" maneuver using an electric system.

The Hohmann transfer is best explained in [REFERENCE HERE]. For brevity we will not expand on the calculations but will approximate the deltaV required to be 24 m/s.

Altitude Maintenance For a spacecraft operating at 450km there will be a significant amount of atmospheric drag to contend with. The bulk of our propulsive capability will be focused on raising the orbit every so often to counteract the effects of drag. The approximate deltaV required to maintian a 450km circular orbit assuming a ballistic coefficent of 50kg/$m^2$ at solar max is 74.6 m/s/yr as shown in column 47 Table I-1 in SMAD ¹. We use a low estimate of ballistic coefficent and take the table value at solar max to be conservative. If the ballistic coefficent is higher or the mission occurs during solar min, less deltaV will be required to maintain the orbit.

Collision Avoidance This is slightly tricky to predict. For now lets use a conservative estimate of 5 m/s. At 450km the amount of debris should be somewhat lower than at 500 km or 600 km as objects should encounter significant drag lowering the orbit lifetime.

De-Orbit A spacecraft operating at an altitude of 450km is unlikely to remain in orbit for 5 years post-mission. We can confirm this by looking at column 54 in Table I-1 in SMAD ¹. We use solar min and a ballistic coefficent of 200kg/$m^2$ to be conservative. If the spacecraft ballistic coefficent is lower or the deorbit phase occurs during solar max, the spacecraft should deorbit faster. We see that the spacecraft will have an orbital life of 741 days or 2 years. This is within the 5 year de-orbit requirement and so no deltaV is required for de-orbiting.

Spacecraft Mass (kg):Mission Life (years):

Delta-V Requirement	Delta-V (m/s)	Total (m/s)
Orbit Lowering		24.00
Altitude Maintenance		149.20
Collision Avoidance		10.00
De-Orbit		0.00
DeltaV Margin (20%)		36.64
Total Mission DeltaV		219.84

Total Impulse Required (N·s): 10992.00

Propulsion System Down-Select

From the above we get a total impulse required of about 11 kNs. This value tells us how much thrust and for how long we need to fire the prop system. ex. 11,000s firing a 1N thruster or 22,000s of a 0.5N and so on.

We also know that we need to impart 74.6 m/s of deltaV per year to maintian our altitude. Lets put this in terms of deltaV/orbit. Lets assume we thrust once every 16 orbits (about once a day). At a 450km altitude we have an orbital period of about 93 mins. This would be about 5650 orbits per year. We therefore would need to impart about 0.2 m/s of deltaV once every 16 orbits.

Next, lets estimate the propellent mass for different propulsion technologies and look at that as a percentage of our total spacecraft mass. Note, propellentless technologies were omitted as their effectiveness would be hampered by the high-drag environment.

Technology	Thrust Range	Specific Impulse Range (s)	Propellant Mass for 11 kN·s (kg)	Propellant Mass (% of 50 kg)	Thrust Duration for ΔV = 0.05 m/s (min)	Duration (% of 93 min orbit)
Hydrazine Monopropellant	0.25 – 28 N	180 – 285	4.82	10 %	$<0.01$	0 %
Alternative Mono‑ & Bi‑propellants	50 mN – 22 N	150 – 310	4.88	10 %	$<0.01$	0 %
Hybrids	8 – 222 N	215 – 300	4.36	9 %	$<0.01$	0 %
Cold Gas	10 µN – 3.6 N	40 – 110	14.96	30 %	0.02	0 %
Solid Motors	37 – 461 N	187 – 269	4.92	10 %	$<0.01$	0 %
Propellant Management Devices	–	–	N/A	N/A	N/A	N/A
Electric Propulsion
Electrothermal	0.1 mN – 1 N	20 – 350	6.06	12 %	0.08	0 %
Electrosprays	20 µN – 20 mN	225 – 3 000	0.70	1 %	4.16	4 %
Gridded Ion	0.1 – 20 mN	500 – 3 000	0.64	1 %	4.15	4 %
Hall‑Effect	0.25 – 55 mN	200 – 1 920	1.06	2 %	1.51	2 %
Pulsed Plasma & Vacuum Arc	4 – 500 µN	87 – 3 200	0.68	1 %	165.34	178 %
Ambipolar	0.5 – 17 mN	400 – 1 100	1.50	3 %	4.76	5 %

Data reproduced from NASA Small Spacecraft Technology State‑of‑the‑Art report, “4.0 In‑Space Propulsion,” Table 4‑1. ²

As expected, chemical propulsion systems require higher propellent mass than electrical systems. However they also have the benefit of higher thrust and therefore a shorter thrust duration. This would mean less time performing altitude maintenace maneuvers and more time for payload operations. Additionally, although not shown here, electrical propulsion systems typically consume more power. You need to ensure that the extended duration of high power consumption was workable in your Power Budget.

It is fairly easy to eliminate some technology options. Anything over ~100% of the orbit duration is not practical. Thrusting for 100% of an orbit would drain your battery and also greatly reduce the efficency of the altitude raising maneuver (less impulsive). This percentage can be reduced by thrusting more than once per day however that can eat into payload operational. Conversely, options with propellent mass fraction that are high leave less mass for other parts of your spacecraft.

This is where some additional mission analysis can be handy. You need to quantify what is important to your mission. Is it payload availability, is it cost, or is it something else? If the payload is small and you have plenty of mass budget to fit a chemical system, it might be a better option as the maneuvers are quick to execute. If however you are tight on mass, an electric system might be a better option at the cost of higher power requred and longer thrusting durations.

The down-select process is iterritive. You may choose to go back, reduce conservatism in some key assumptions, adjust requirements, etc. Gradually you will start to narrow down the solution space. Once you narrow down your options, reach out to suppliers, start looking at datasheets and see how real systems compare. SatSearch has a great database of COTS (Commercial Off The Shelf) propulsion systems and components from various vendors. You can also checkout the NASA State‑of‑the‑Art report for a summary of various propulsion systems, their performance capacilities and if any flight heritage exists.

---

Additional Considerations

Accounting for the Real World

Quoted vs Actual Performance Datasheet values of Isp, thrust and total impulse are typically quoted at some nominal operating point. In reality, actual performance can differ for a number of reasons:

• Short duration thrusts: Start up transients are typically at a lower Isp. These effects are not noticeable for longer thrust durations but for quick thrusts it can be significant.
• Pressure drops in blow-down systems: The Beggining-Of-Life (BOL) and End-Of-Life (EOL) feed pressure may vary if using a blow down system. As such you need to account for gradually declining thrust and specific impulse over time.
• Thruster degredation: Electric thrusters may experience cathode/anode erosion, chemical thrusters may encounter fowling/clogging of the throat. Over time the thruster performance may degrade due to these effects. It is important to ask manufacturers explicitly about any life testing that was conducted.

Trapped Propellant and Leakage

• Leak rate: Propellents and pressurants no mater how well sealed will gradually leak over time. The magnitude is can be incredibly small but can still cut into your deltaV margin.
• Trapped propellent: Ever try getting that last bit of toothpaste out of the tube? Same thing. For liquid propellents, regardless if using a diaphragm, bladder, piston or other Propellent Management Device (PMD), there will allways be some quantitiy of propellent left over that is unusable. Be sure not to count this when computing your total useable impulse.

Valve/System Cycles
A system may be qualified to a million Ns of total impulse, but that does not give us much confidence if it was just turned on and then off once. Cycles are what stress propulsion systems and you should know the cycle life of your system regardless of if it is chemical or electric. Chemical systems are often limited by valve cycles while electric systems are stressed by power cycles.

Contingency Margin
Based on all of the above it is important then to include some margin when building out our deltaV budget. Not just to account for the non-idealities of real propulsion systems but also to account for uncertianties in how we plan on using it. The amount of margin will vary depending on the program, mass budget, types of maneuvers, risk tolerance, etc. A margin of 20% is a good rule of thumb but more is allways better if you can afford it.

Recomendations

Electric propulsion is a good option if:

• Lots of DeltaV is required.
• You can support the high power draw.
• You want to make large orbit transfers efficently but don't mind the lengthy maneuver duration.

Chemical propulsion is more advantageous if:

• Not much DeltaV is needed.
• There is insufficent power available for an electric system.
• Time is money and you need to get into your target orbit quickly.
• Sufficent mass margin is available.
• The satellite can compensate for any thrust misallignment (wheels are sized correctly).

Propellent-less is the way to go if:

• The satellite does not need to maneuver.
• Passive deorbit can be achieved easily with drag appendages.
• You don't have to make money and have ample numbers of brilliant PhD and Masters students for your mission to a distant star or planet.

---

References

Footnotes

1. Wertz, J. R., Everett, D. F., & Puschell, J. J. (2011). Space mission engineering: The new SMAD. Microcosm Press. ↩ ↩²

2. NASA Small Spacecraft Systems Virtual Institute. "In-Space Propulsion." NASA, National Aeronautics and Space Administration, https://www.nasa.gov/smallsat-institute/sst-soa/in-space_propulsion/. ↩