≡ Menu

Spacecoaches and Beamed Power

If you’re planning to make it to the International Space Development Conference in San Juan, Puerto Rico next month, be advised that Brian McConnell will be there with thoughts on a subject we’ve discussed in several earlier posts: A ‘spacecoach’ that uses water as a propellant and offers a practical way to move large payloads (and crews) around the Solar System. Based in San Francisco, Brian is a technology entrepreneur who doubles as a software/electrical engineer. In the essay below, he looks at the spacecoach in relation to the Breakthrough Starshot initiative, where synergies come into play that may benefit both concepts.

by Brian McConnell


The spacecoach is a design pattern for a reusable solar electric spacecraft, previously featured on Centauri Dreams here and developed in A Design for a Reusable Water-Based Spacecraft Known as the Spacecoach (Springer Verlag), which I wrote with Alex Tolley. It primarily uses water as its propellant. This design has numerous benefits, chief among them the ability to turn consumables, ordinarily deadweight, into working mass.

The recent announcement of the Breakthrough Starshot project, which aims to use beamed power to drive ultra lightweight lightsail probes on interstellar trajectories, is of note. This same infrastructure could be used to augment the capabilities and range of spacecoaches (or any solar electric spacecraft), while providing a near-term use for beamed power infrastructure as it is developed and scaled up.

The spacecoach design pattern combines a medium sized solar array (sized to generate between 500 kilowatts and 2 megawatts of peak power at 1AU) with electric propulsion units that use water as propellant (and possibly also waste streams such as carbon dioxide, ammonia, etc). We found that, even when constrained to these power levels, they could fly approximately Hohmann trajectories to and from destinations in the inner solar system. Because consumables are converted into propellant, this reduces mass budgets by an order of magnitude, and effectively eliminates the need for an external interplanetary stage, all while greatly simplifying the logistics of supporting a sizeable crew for long duration missions (more consumables = more propellant).


The primary constraint for space coaches, especially if you want to travel to the outer solar system, is available power. This is an issue for two reasons. First, solar flux drops off by 1/r2, so at Jupiter, a solar array will generate roughly 1/25th the power as it does at Earth distance. Second, trips to more distant locations will typically require a greater delta V (and thus higher exhaust velocity to achieve this with a given amount of propellant). The amount of energy required to generate a unit of impulse scales linearly with exhaust velocity, so the net result is the ship’s power requirements are increased, all while the powerplant’s power density (watts per kilogram of solar array) is decreased.

Testing Beamed Power

Beamed power infrastructure would enable space coaches and solar electric spacecraft in general to operate at higher power levels for a given array size, which would enable them to operate at higher thrust levels, and to utilize higher exhaust velocities to maximize delta V and propellant efficiency. This means they would be able to accelerate faster, achieve higher delta-v, while using less propellant. In effect beamed power to SEP spacecraft will give their operators the equivalent of a nuclear electric power plant (without the nukes).

A spacecoach built for solar only operation would be able to serve as a testbed for beamed power. For example, a space coach departing Earth orbit could be illuminated with a beam that increases its power output by a small amount, say 10% (large enough to make a measurable difference in performance, yet small enough that major modifications are not required to the ship as it just experiences slightly brighter illumination while in beam). At higher light levels, this technique could also be used to simulate lighting and heat loading conditions expected at the inner planets while remaining in near Earth space. Note also that lasers can be tuned to the absorption wavelength(s) of the photovoltaic material, greatly improving conversion efficiency (and reducing heat gain per unit of power delivered). An even cheaper way to build out and test power beaming infrastructure will be with satellites and probes that utilize solar electric propulsion.

The pathway to a system based primarily on beamed power then becomes one based on incremental improvements, both for the ground based facilities and for the ships. This would result in near term applications for the beamed power facilities while the much more technically challenging aspects of the starshot project are sorted out. Meanwhile, satellite and space coach operators could test ships with ever higher levels of beamed power until they hit a limit (heat rejection is probably the main limit to how much power can be concentrated per unit of sail area, as this is similar to concentrated photovoltaics).

The chart below illustrates the power/performance curve by showing the amount of impulse that can theoretically be generated per megawatt hour using electric propulsion, as a function of exhaust velocity. Real world performance will be somewhat lower due to efficiency losses, but this shows the relationship between thrust, ve and power. We see that impulse per MWh varies from 72,000 kg-m/s (ion drive, ve ~ 100,000 m/s) to 1,400,000 kg-m/s (RF arcjet, ve ~ 5000 m/s). A Hall Effect thruster, a flight proven technology, would yield about 300,000 kg-m/s per MWh. Compare this to pure photonic propulsion, which would yield only 12 to 24 kg-m/s per MWh. Clearly photonic propulsion will be necessary to achieve a delta v of 0.2c, but for more pedestrian applications such as satellite orbit raising, launching interplanetary probes or cargo ships from LEO to BEO (beyond earth orbit), electric propulsion will work well at power levels many orders of magnitude lower than what’s required for a starshot.


Driver for an Interplanetary Infrastructure?

Closer to home there could be lots of opportunities to sell beamed power to space operators. It’s costly to launch large payloads beyond low earth orbit (which isn’t cheap in the first place). Meanwhile, payload fairings limit the size of self-deploying solar arrays, which limits the use of electric propulsion for satellites and probes. If one could launch spacecraft with small solar arrays to LEO, and then use beamed power to amplify their power budget they could use electric propulsion to boost themselves to their desired orbits or interplanetary trajectories within a reasonable time frame. The beamed power infrastructure can also be built up incrementally. Early systems would beam 100 kilowatts to 10 megawatts of power to targets measuring meters to tens of meters in diameter. This should be readily achievable, and can be scaled up from there in terms of power output, beam precision, etc. The result: lower costs per kilogram to deliver a payload to its destination or desired orbit compared to all chemical propulsion.

This could make electric propulsion for transit from LEO to GEO and beyond an attractive option. Meanwhile, the power beaming operator would accrue lots of operational experience with beam shaping, tracking objects in orbit, etc, all things that will need to be mastered for the starshot project, while providing an economic foundation for the power beaming facilities during the buildup to their intended purpose.

In fact, one can imagine the starshot project becoming a profitable LEO to BEO (beyond earth orbit) launch operator in its own right. The terrestrial power beaming infrastructure is one component. A standardized “power sail” that can be fitted to many different payloads, from geostationary satellites to interplanetary probes, is another. The power sail would consist of a self-deploying solar array that is sized to work well with beamed power, heat rejection gear, and electric propulsion units. It would use beamed power during its boost phase to rapidly accrue velocity for its planned trajectory, and then as it leaves near Earth space, would transition to use ambient light as its power source from there. Meanwhile these power sails would provide an evolutionary path from conventional spacecraft to solar electric propulsion to the nanocraft envisioned for purely photonic propulsion.

As a starting point, it would be interesting to conduct ground based vacuum chamber tests to see how a variety of PV materials respond to being illuminated with concentrated laser light tuned to their peak absorption wavelengths. What do the conversion efficiencies look like? How much waste heat is generated? How do the materials perform at high temperatures in simulated in-beam conditions? Building on that one can imagine experiments involving cubesats to validate the data from those experiments in real world conditions, and if that all works out, one could scale up from there to build out beamed power infrastructure for use by many types of solar electric vehicles.

Ambitious R&D projects have a way of generating unintended side benefits. It’s possible that the starshot initiative, in addition to being our first step toward the stars, will also make great contributions to travel and exploration within the solar system.


Comments on this entry are closed.

  • Dani Eder April 29, 2016, 10:36

    Current space solar arrays (such as http://www.spectrolab.com/DataSheets/Panel/panels.pdf ) have total thicknesses of 290 microns (11.5 mils), while the Planetary Society’s experimental solar sails are 4.5 microns, and with further development can probably be brought down to 1 micron. Therefore it makes sense to dispense with the lasers, and use sail-like reflectors to concentrate sunlight onto solar cells.

    Assuming 4.5 micron reflectors, the can increase the sunlight collected by 64 times in area at a cost of doubling the mass. Allowing for 80% reflector efficiency, we then get 25.6 times the net sunlight on the solar panels, or 1 AU effective power at Jupiter.

    Locating all the power collection on the spacecraft avoids the problem of the Earth’s rotation making the ground lasers point the wrong way a lot of the time, and having to split their time between multiple targets.

  • Robert April 29, 2016, 12:23

    What kind of total mass for water as propellant is envisioned for say, a trip to Mars?

  • Michael April 29, 2016, 13:31

    It is much more efficient for solar cells to have a matched wavelength to the band energy, stray heating would be at a minimum as almost all of the energy will be used to make electrons move. It is just not that efficent to make laser light. Prehaps a mixture of reflectors and laser light would be better, the laser light would allow a greater acceleration over a shorter distance.

    I have always wanted to use particle accelerators over lasers, not only can they generate enormous amounts of kinetic energy to move things they can also transmit fuels, say using oxygen ions first and then using hydrogen ions which can then be recombined to allow a more aggressive deep gravity well maneuvoir, they are also a lot easier to make.

  • Jens Knudsen April 29, 2016, 14:26

    Try both, and see what works the best? I see the reasoning behind what you present, Dani, but I also have a sensation we have a powerful case for beamed power on our hands.

    • Dani Eder May 1, 2016, 14:11

      I am by no means against beamed power in general, but it’s more a matter of timing and use case. The article above talks about transport in the inner solar system. For that purpose, solar-electric with some level of concentration of sunlight would be perfectly adequate.

      For climbing out of Low Earth Orbit, where you are in shadow 40% of the time, or simply to speed up the climb, beamed power may be very useful. The range is much shorter, so aim and focus are easier to maintain.

      The more general point is in engineering you can’t categorically say A is better than B. Depending on the mission and circumstances, either one may turn out to be the best answer. So I agree with your “try both” statement – it’s better to have more options to choose from when planning a particular project.

  • Brian McConnell April 29, 2016, 15:58

    A point for commenters to bear in mind. Alex and I are thinking about applications for the starshot power beaming infrastructure during the long period of development and testing.

    Solar electric spacecraft, especially unmanned craft, are constrained by solar array area (payload fairings limit the size of the craft, and any self-deploying array packaged within). That limits you to tens of kilowatts of electric output, which limits how much thrust a SEP engine can generate.

    If you can light the craft up at 10-100x suns, with light matched to the PV material’s band gap to minimize heat gain, you’ll have something approaching a megawatt to work with. A cluster of Hall Effect thrusters at this power level would generate tens of Newtons, enough to build up substantial delta-v within a few days of operation.

    So apart from a spacecoach, imagine something like the DAWN spacecraft launched to a low-ish orbit on a cheap vehicle like Falcon 9, which then boosts itself via SEP to high delta-v, then transitions to ambient light (and much lower thrust) once away from Earth. There should be lots of interesting scenarios like this.

    • Michael April 30, 2016, 12:50

      We would need to chose a wavelength that is not absorbed by the atmosphere for the laser. The higher you go up the less absorption by mass, water vapour or ozone so you can chose a better wavelength. If we used balloon platforms we could move well above most of the absorption bands and also give paying guests a lift. Balloon platforms offer a reasonable way of managing laser systems and quite large turbine generator sets can be used for electrical generation.

    • Robert April 30, 2016, 13:39

      Interesting that you mention concentrated solar up to 100 suns. The (I’ll
      say hypothetical for the moment to avoid debates) hydrino reaction can provide light at multi thousand sun concentration for CPV conversion. Actually, the reaction emits light in the range of the sun’s blackbody intensity at it’s surface and has to be reduced to match the commercially available CPV of around 2000 suns. Also, the spectrum can be engineered to best match the CPV for efficiency and minimize heating. The cost of water mass is about 3.3 liters per hour for a constant 1MW electric at 30% efficiency. But that mass can also be used as propellant mass consisting of oxygen and di-hydrino gasses.

      This is why I was asking what the water mass requirements are for missions such as a mission to Mars for the Spacecoach as currently envisioned?

    • Dani Eder May 1, 2016, 14:31

      > payload fairings limit the size of the craft, and any self-deploying array packaged within

      The ISS demonstrated you can assemble large objects in orbit from smaller pieces, and DARPA is working on robotic servicing ( http://www.darpa.mil/news-events/2016-03-25 ), so in the not too distant future, orbital assembly may not need humans on site. They likely will still be involved, but remotely from the ground.

      Even without assembly, one of the ISS’s eight solar array wings fits in a box 0.5 x 1.0 x 4.6 meters when folded, which certainly can fit in a conventional payload fairing. With modern cells (The ISS’s are quite old), it could produce 140 kW. A Falcon 9 payload fairing should be able to fit 4 such wings around an electric thruster core, giving 560 kW available power.

    • Michael May 6, 2016, 16:07


      If the orbiting craft were pumped with intense laser light on an increasing eccentric orbit on close approach to the Earth it maybe worth it. Been near to the Earth during intense pulses it would get maximum laser light.

      You could ask Hop David http://hopsblog-hop.blogspot.co.uk/ for any advice or pointers.

      • Jens Knudsen May 7, 2016, 3:44

        This kind of technology could be put to good use long before it needs to accelerate a ‘starchip’ to 20% of the speed of light!

        • Michael May 7, 2016, 14:13

          We need an integrated laser system where one purpose serves another i.e. say an interstellar craft is propel out of the solar system and it energises satellites in orbit and it powers Brain/Alex concept and it powers relay deep space satellites and…

  • Andrew Palfreyman April 29, 2016, 16:41

    For stagecoach mission costing, how sensitive is the overall mission cost to that $1,700/Kg launch cost? Does it make sense to build StarTram v1, with capital cost $19B and launches at $43/Kg?

  • Erin S April 30, 2016, 7:56

    Somewhat tangential, but whatever happened to spacecoach.org? I’ve been trying to look it up as I’ve grown interested in the spacecoach concept for a science fiction project, but it seems like the domain was taken over by a squatter and I haven’t found much relevant information on Google.

  • Jens Knudsen April 30, 2016, 15:25

    A weighty reason we should put a beaming facility on Earth is, it’s here. It looks like a big, heavy thing to me.

  • Eniac April 30, 2016, 20:41

    It is surprisingly difficult to outshine the sun at a distance, with lasers. See this: https://what-if.xkcd.com/13/ for the lighter side. Then remember that the Earth is much closer to the moon than the sun, adding another few orders of magnitude if we want to outshine the sun at long (i.e. multiple AU) distances.

    I wonder if people have thought this laser thing all the way through. Can anyone show me a laser that has even remotely the radiant intensity (W/sr) of the sun? Because that is what it would take before light beams could become useful around the solar system, for anything.

    Such lasers would also be needed to address optical METI: A laser that would make the sun seem brighter when looked at from a star that the beam is pointing at.

    • Eniac May 8, 2016, 15:36

      I did some calculations, and here, for what it’s worth:

      The relevant quantity for “shining” is radiant intensity. It is the power that goes into a solid angle, and it is measured in Watts per steradian, i.e. W/sr.

      1) Radiant intensity of sunlight
      We know that the sun shines at 1.4 kW/m^2 at 1AU distance. A square
      meter at 1AU spans a solid angle of (1m/1AU)^2. Therefore:

      I_sun ~ 3 * 10^25 W/sr

      2) Radiant intensity of a diffraction limited laser
      The diffraction limit gives a divergence of theta = lamda/(pi*w),
      where w is the radius of the aperture, and lambda the wavelength.
      The solid angle of the beam is theta^2, so the radiant intensity becomes

      I_laser ~ (4 * 10^13 m^-2) * w^2 * P

      In order, then, for a laser with a 1 m^2 aperture to shine as brightly as the sun at the same distance, it would have to have a power of roughly 1 TW.

      At 100 m^2 aperture, the best telescopes we can currently build, it would be 10 GW. The 1 km^2 optical phased array that is envisioned for Starshot would only need a power of 1 MW. A square kilometer optical phased array is not much more than fantasy at this point in time, but it could probably be built with enough funds and perseverance.

      One thing is for sure, though: A laser that is weaker or smaller than this will be useless at interplanetary distances. It can still be useful when it is much closer to the receiver than the sun, i.e. for powering satellites from the ground or for propulsion in the Earth/moon system.

  • Andrew Palfreyman May 1, 2016, 6:25

    It seems possible that the kinetic energy from a particle beam could compete with the force exerted by a laser, at the same input power levels. Has a comparitive analysis been done?

    • Michael May 1, 2016, 14:02

      I personally would like to see a comprehensive analysis of the two systems, lasers versus particle beams. My money would be on particle beams, they have so many advantages over lasers, but combining the two may be the best way forwards. A particle beam packs more punch per region than a laser will ever do and they are very, very versatile i.e. tunable to the requirement.

      • Eniac May 1, 2016, 14:41

        I always though lasers are the most versatile and tunable beams we had. They are also much better collimated, and come at higher power levels. I am not sure what your “many advantages” are, but neither high power (punch?) nor collimation are amongst them. Judging from how many hand-help proton beam pointers I have seen, versatility isn’t one, either.

        Given that collimation is so very essential here, particle beams need not apply.

      • Alex Tolley May 1, 2016, 16:19

        How would you keep the particle beam collimated? To accelerate it the beam must be charged, which will result in divergence. If the beam is neutralized after acceleration, e.g. injecting electrons into a proton beam, how will the target extract the energy of a neutral beam? Physically stopping it, like Dyson’s Medusa sail to capture neutrons from nuclear explosions?

        • Michael May 2, 2016, 1:21

          @Eniac May 1, 2016 at 14:41

          ‘I always though lasers are the most versatile and tunable beams we had.’

          Accelerators can be attenuated over a huge range, lasers offer less power control unless they are used in a phased array. Lasers limit the sail fabrics construction where as magnetic fields offer a stand off non touching design having less temperature sensitivity.

          ‘They are also much better collimated, and come at higher power levels. I am not sure what your “many advantages” are,’

          Particle accelerators come in very high power levels, they can smash atoms (punch), lasers in or near the optical wavelength can’t unless they are used to accelerate atoms to do there bidding.

          ‘Given that collimation is so very essential here, particle beams need not apply.’

          Particle beams can be collimated by spreading the particles out over a larger initial area reducing the repulsive force to the sqr, increasing the particles velocity at the same power also increases the distance between them and again reducing the repulsive force to the sqr. Increasing the mass of the ionised atoms reduces the spread as well.

          ‘but neither high power (punch?) nor collimation are amongst them. Judging from how many hand-help proton beam pointers I have seen, versatility isn’t one, either.’

          We would not use them in air, that is a disadvantage, unless very high up in the atmosphere where a piece of the atmosphere could be accelerated by impact to interact with the magnetic field coil for first send off.

          • Mark Zambelli May 5, 2016, 6:00

            The Starshot laser system will be a phased array.

        • Michael May 2, 2016, 1:27

          @Alex Tolley May 1, 2016 at 16:19

          ‘To accelerate it the beam must be charged, which will result in divergence. If the beam is neutralized after acceleration, e.g. injecting electrons into a proton beam, how will the target extract the energy of a neutral beam? Physically stopping it, like Dyson’s Medusa sail to capture neutrons from nuclear explosions?’

          If the beam is neutralised the velocity of it interacting with the magnetic field will ionise it again. There is also the possibility of using a physical layer of material like a sail together with a magnetic field to take the impacts and convert it to momentum, if the interacting velocity is low enough erosion should be quite low.

          • Jens Knudsen May 2, 2016, 16:45

            The particle accelerator would still have to be placed in space somewhere?

            • Michael May 3, 2016, 8:25

              yip, but putting all the inefficiencies of a laser system together and it may be better to have a particle accelerator in space.

              • Adam May 7, 2016, 22:29

                Michael, as discussed on this very web-site, Jim Benford has demonstrated that particle beams diverge into uselessness within thousands of kilometres. They’re pointless for beaming across significant fractions of an AU. Landis, and Dana Andrews before him, don’t have the particle physics background that Jim (and his brother Greg) do, so they missed that particular limitation of particle beams when they advocated them as pushers for interplanetary and interstellar vehicles. The problem is that the internal temperature of a neutralised particle beam is too high for it to not significantly spread over the range desired.

                • Michael May 8, 2016, 10:43

                  If we spread out the ‘beam’ the divergence is a lot less. If I had say 100 ionised particles in 1 mm^3 the forces would be huge and throw them apart. Now say I have 100 particle beams and spread the ions out over 1 m^3 the forces have dropped enormously. Spreading the initial beam area and increasing the space between each emission particle reduces the forces and therefore the repulsive spread.

                • Michael May 8, 2016, 12:25

                  ‘The problem is that the internal temperature of a neutralised particle beam is too high for it to not significantly spread over the range desired.’

                  At larger fractions of c time dilation aids us by reducing the time it can spread out and heavy atoms resist thermal divergence well. I believe once we have mastered one atom manipulation on a large scale particle accelerators will come to the fore for interstellar propulsion. Particle accelerators can have energies thousands of times that of anti-matter energy releases.

                  Here are some interesting articles on ultra cold plasmas.


                  • Eniac May 10, 2016, 23:18

                    Michael, all these ways of improving the focus of particle beams are purely hypothetical, technology readiness level zilch. Yes, in theory there is no diffraction limit and so, theoretically, a neutral particle beam could be focused better than a laser. But the engineering obstacles are many, enormously difficult, and mostly unexplored.

                    Looking at particle and laser beams that we can create now, today, particle beams are lagging by many orders of magnitude. Lasers are commonplace and everywhere, in all sizes and powers, and they are now even reaching good conversion efficiencies.

                    • Michael May 11, 2016, 9:39

                      ‘Michael, all these ways of improving the focus of particle beams are purely hypothetical, technology readiness level zilch’

                      Cooling the particles to reduce transverse velocity has already been researched and proven.

                      And accelerating a piece of material thinner than a human hair at 60000 g’s has no technological hurdles…miracles^x required.

                  • Jens Knudsen May 11, 2016, 14:22

                    There are many things about the ‘starchip’ that are far fetched. But to me, they look attainable. And there aren’t many possibilities when we talk about instellar flight. One of the things that make it hard, is that we need to make choices, and even the most conservative ones entail considerable risk, and (many) tons of money. We should go for the best solution if we can find it in advance, but sometimes it is not possible to guess what it is, and there will be surprises coming along the way. At the end of the day, what we need is a solution that works. The group of people that have started this program have been trying to find one.

                    This forum is the first place I have heard about particle beams as a means of beamed propulsion. Why not try it at the ISS? Sooner or later somebody will send an ion engine up there, and the Planetary Society have shown us how to build a deployable sail based on a cubesat. There must be ground facilities also, used for ion engine testing, that could be used to see if a decent sail could survive the blast. But the ‘sending part’ of the particle beam still looks big and heavy to me, if we want it to propel a small sail to 0.2c. I realize the SLS is coming on stage, but…

                    • Michael May 12, 2016, 3:21

                      ‘This forum is the first place I have heard about particle beams as a means of beamed propulsion.’

                      Particles beams for propulsion go back a long way.

                      ‘There must be ground facilities also, used for ion engine testing, that could be used to see if a decent sail could survive the blast.’

                      Particle beam powered craft we most likely use a magnetic field as opposed to direct impact to prevent damaging the sail. I prefer a combination sail/magnetic field setup.

                      ‘But the ‘sending part’ of the particle beam still looks big and heavy to me, if we want it to propel a small sail to 0.2c. I realize the SLS is coming on stage, but…’

                      Particle beams will be heavy but in space the contruction can be a lot lighter because there is no need to keep the air pressure out. Nanotubes can also be used in particle beam accelerators as cooling and guides.

  • Paul D. May 1, 2016, 13:53

    I think one should also consider the possibility of laser cooling, since it could increase the achievable power density at the spacecraft (both in the collector and in the engine.)

    Perovskite PV materials have recently been shown to be good laser cooling candidates. Perhaps one could make a self-cooled laser PV collector with very high power density.