SpaceX Super Heavy Starship 2.0 Will Be 8 Times Bigger Than Super Heavy Starship

The SpaceX Super Heavy Starship will be able to launch 100 tons into orbit in a fully reusable basis, but Elon Plans the follow up to be 4 to 8 times bigger. Elon tweeted that Starship Version 2.0 will be 18 meters in diameter instead of 9 meters. This would mean the area of the cross-section would be 4 times higher. If the height was also doubled then it would have 9 times the volume. The engines would likely be upgraded for the Ultra Heavy Starship 2.0. This means the next rocket might be able to launch over 1000 tons per launch.

This would be about twice the payload of the Sea Dragon. The Sea Dragon was a 1962 conceptualized design study for a two-stage sea-launched orbital super heavy-lift launch vehicle. The project was led by Robert Truax while working at Aerojet. It would have had a payload capacity of 550 tons. It would have been 150 meters tall and 23 meters in diameter.

The SpaceX Super Heavy Starship 2.0 may also be limited on height. Tovera Vashini on twitter notes that rocket height is limted by thrust per area. This could mean that SpaceX Super Heavy Starship 2.0 may also end up very near the Sea Dragon in capacity.

In 2016, SpaceX had a 12-meter diameter ITS rocket design. It was designed to launch 300 tons to orbit and with refueling could launch 450 tons to Mars.

The ITS was about double the volume of the current Starship Super Heavy but had 3 times the payload. If the 18-meter Starship 2.0 had similar scaling then it would have 9 times the payload of the Starship.

A 240-meter tall rocket would be as tall as the Woolworth building in New York and about 75% as tall as the Eiffel Tower and twice as tall as the Statue of Liberty.

The first stage of the Saturn V was 10 meters in diameter and the Soviet Union N1 had 17-meter diameter at its widest point.

Elon Musk also tweeted that we are about 2 to 3 months from orbital readiness for the Starship. SpaceX is shipping its tenth raptor engine.

SOURCES -SpaceX, Elon Musk Twitter, Wikipedia
Written by Brian Wang,

142 thoughts on “SpaceX Super Heavy Starship 2.0 Will Be 8 Times Bigger Than Super Heavy Starship”

  1. Remember when Elon said: “the best process is no process”? Apply that to SBSP! And you get Dr. Lewis Fraas’s Space Solar Mirror Concept, adding 8 hours of solar photon energy per day to existing ground-based solar power plants.

  2. “The engines would likely be upgraded” more like improved upon… it will still be the Raptor (methane full flow) engine but build closer to its the max theoretical thrust limit by then.

  3. Well, the aperture has been changed in the plans to match the change in distance. You can see with a telescope the same angular resolution no matter how far away you are looking, for example.

  4. So, I’m now re-thinking the radar aperture question. I had always assumed “bigger” was thus way more expensive. But consider SPS in L-5. Why is it not almost as good as GEO? The radar *size* is the same power, just spread out more. And are not the losses in power beaming independent of the distance thru space?
    Brought on by the fact that light pollution acts the opposite as radar power, in that closer is much squared brighter for the same power, but spreading out the same power radar may be fairly easy.

  5. I have just realized that we are actually discussing things that ARE happening, now that the next step is not Mars Direct. ISRU is now assumed, not too remote to consider. Projects now can work together, everybody can use ISRU. Reminds me of the mind set in 70’s O’Neill!
    And there is the distinction between pumped and convective fluid movement. In various gs. And, of course, these are *real* radiators, not the heat exchangers on cars, so a hard job.

  6. With regards to thickness of the radiators, there are two related issues. I was talking about heat conduction in a solid. There the conductance is proportional to the cross-section area of the solid.

    Similarly, if we’re talking about convection of a heat-carrying fluid inside a pipe, as I recall, its speed is something like proportional to the cross-section area of the pipe, given the same pump pressure. (Or was it inversely proportional, and you need more pressure for same speed? Anyway – harder to push the fluid through thinner pipes, I think). Given the same heat capacity of the fluid, you want to push enough fluid volume, which dictates a minimum inner cross-section area of the pipes.

    Either way, it limits the minimum diameter for amount of heat that needs to be carried away.

  7. I think the LCROSS found quite a bit of C, but it was under-reported. But very early on, asteroids, scooping from Earth or outer moons, whatever is cheapest. ISRU allows more ISRU very quickly. Seems easy to grab stuff compared to being able to use it in manufacture.
    Yes, there are thin ways to make cells on sats, but Criswell claims that making the sheets movable or controllable quickly sends the cost way up. His lowest level *workers* use whatever lunar stuff is underfoot plus needed exotic supplies to lay the sheet down as it is formed, where it never actually moves. A lot of surface area to cover.
    My brainstorm of direct spraying is pretty unlikely, but who knows? It would be really cheap, so could be very inefficient physically and still work.
    I think the thinness of the radiator skin would help. Unless you mean no moving fluid inside? You may be thinking of a curious difficulty in enlarging the radiator, where the need to move the fluid out into a larger surface area is harder than it would seem at first. At any rate, that is a complex issue given all the possibilities. LSP may be able to recover some of the effort spent to handle nite by using the nite period to release heat stored in dirt during the day, rather than radiated. Just leave the cold pipes open and the heat from below will pump itself to the now cold surface. I don’t see an obvious advantage here either way, but that does not mean there isn’t!,

  8. Actually, 500 Gw is the “break even” projection, where no more investment is needed beyond profits. 20 Tw (I’m going to stop the “e”, as even nuke people are dropping the “t”) is the goal, then on to 200 as needed. Just setting up will start a lot of other Space activity.
    Scale up, light pollution, space junk, station keeping. Starts to add up! Need to advance both SPS and LSP with Earth to Earth balancing power beams. Then the SSP pencils will start on the envelope backs.

  9. To add on the structure question:

    If you consider my solar-sail-like proposal, there is very little structure there. Depending on the material, the active layer might even double as its own support structure. You only need very little centrifugal force to spread it out.

    I have my doubts re your proposal for spraying the solar cells directly on the Lunar surface. The surface is too grainy. You want a continuous electrical circuit, not a collection of fragments. So you’d still need some support layer.

    As I’ve said before, at the limit of strength, the two are probably very similar: either you slightly spin up a very thin sheet in orbit, or you lay a very thin sheet onto the Lunar surface. I don’t expect a big difference in thickness.

    Furthermore, if indeed the radiators are the main component of mass, there’s a limit to how thin you can make them. IIRC, the thinner they are, the less heat they can conduct away from the hot parts. So they’re unlikely to be thinner on the Moon.

    The ideal material for both may be carbon based. Doped or coated graphene or CNTs may produce good PV junctions, and they have excellent mechanical properties (so they can support themselves) and excellent electrical and thermal conductivity. But as far as we currently know, there’s not much carbon on the Moon. Which may add an advantage to SPS: it’s easier to bring carbon asteroid stuff to orbit than land it on the Moon.
    (edit: Silicene looks like an interesting alternative doable from Lunar soil.)

  10. You absolutely will need power on the lunar surface, both for ISRU and mass driver. But that’s on the order of hundreds of MW, not the hundreds of GW to TW to which you’d need to scale to make it an economical source of terrestrial capacity.

    There is indeed some station-keeping for SSP, but GEO is pretty clean from an MMOD standpoint. As for light pollution, I assume you’re talking about a big shiny thing illuminating the night side of Earth? Yeah, that’s an issue, but it’s a pretty small price to pay if you get cheap green baseload power in return.

  11. “I don’t think structure is by any means easier on the lunar surface than it is in space.” In that case, I totally agree! But actually, that is the key question. Perhaps SPS is so easy. All the plans I see have to put a structure there. The components themselves are all that is needed on the Moon. Detailed plans are needed, now that the question is clear. Shooting for 5Tw-e at $.001 per kwh-e ASAP.
    As to ELEO processing/construction, that would probably be for the starter stuff, perhaps mass driver parts from rocket lunar lander wet dirt return cargo, as easier than setting up (people esp) on Moon, until the starter stuff is done. . . Let the market decide on that. Stop that asteriod! All sorts of things will be going on if for nothing else than LSP, and it wil really be fast with ISRU. I suspect little lunar manufacturing in the end. But I’m an O’Neill guy, don’t like planets. Too hard to manufacture stuff.

  12. Yes, the space solar power acronym zoo is very painful, but for purposes of this thread, SSP=space solar power and LSP=lunar (surface) solar power.

    I don’t think structure is by any means easier on the lunar surface than it is in space. Your main marginal cost for ISRU components is still the mass of material you mine and reduce, so mass minimization is important no matter what. And the “stable surface area” is effectively infinite in space, without any unfortunate boulders or mountains.

    Just to be clear: I meant 106% more, i.e., you need double the nameplate capacity on the lunar surface that you need on-orbit.

    Finally, I don’t understand your last point. In all cases, you manufacture components on the lunar surface. We’re only arguing about where to put them once they’re manufactured. If what you meant was you’re fine assembling them in earth orbit (probably GEO, not LEO), then we’re in violent agreement–but then you’re pretty much conceding my point.

  13. One thing to think about: When you design a 2-stage rocket with a specific second stage, retrofitting back to a ganged-up set of first stages is inherently sub-optimal, and becomes more so if your goal is to dramatically increase lift to your target orbit. If the goal of a super-duper-heavy launcher is really to octuple the mass to LEO, when you whip out the ol’ Lagrange multipliers and start figuring optimal staging, you’re leaving a lot of performance on the table when you gang things together without changing the second stage.

    One really interesting area that I hadn’t really thought about when diss’ing this up-thread: Tankers. If you’re really going to be reliant on terrestrial prop (which is questionable…), then mounting a radically stretched Starship on a gang of three SuperHeavies might make a lot of sense, in that the airframe modifications might be fairly modest for a near-optimal staging.

    Delivering 500 tonnes of prop to orbit for a fairly cheap development project might make a lot of sense.

  14. And furthermore!
    Probably the first practice case for LSP will be the cells for the mass driver. We will learn a lot!
    A detail, I’m guessing there is a square root of 2 factor needed for changing sun angle (not seasonal, fortunately) during the slow lunar day. So 2*sqrt(2) more than SPS, about 3.
    Another detail, which applies to all ISRU/SSP, is that the costs go down per unit over time. So, esp if replicators or serious bootstrapping is used, The last half should be much cheaper than the first half, for LSP certainly. For SPS, have you seen the plans for 20-200 Tw-e SPS? Lots of station keeping, Space junk and light pollution to consider, isn’t there. (I forget where I saw those plans, for some reason).

  15. I truly appreciate this level of inquiry!
    For convenience, I consider SSP to cover both SPS and LSP. Looks like you are using SBSP for my SSP and SSP for my SPS? To me, SSP is Space Solar Power, generally.
    “Let’s assume that we have solar power modules (PV, structure, radiator, and transmitter) manufacturable on the Moon.” Implicit in this assumption is that the modules will be the same whether part of SPSs or LSP. “structure” will ideally be easier on the Moon, if we can figure out how to use all that stable surface area!
    Also, thanx for the relative *geometry* requirements of the systems. This is the starting point of the comparison, as it is pretty much set in stone. The 106% is correct, but many will mis-read that as “total”, not “more”.
    I assume the mass driver almost from the beginning. LSP is a huge energy project, not a Space project. Long before much is started scale wise, mass driver will be well up and running. Need 5Tw-e pretty fast to be relevant.
    Also, I’m just fine with building the LSP parts in ELEO, thanx mass driver! Whether the machines or the supplies needed to lay down that veeeeery thin layer of collection area. Use the Moon, don’t live there.

  16. Let’s assume that we have solar power modules (PV, structure, radiator, and transmitter) manufacturable on the Moon. Let’s further assume that we have a mass driver, because the tech to build an SBSP module is a superset of the tech to build a mass driver.

    From there, this seems awfully straightforward: LSP has a capacity factor of 50% and higher transmission losses, and SSP has a capacity factor of >90% and lower transmissions losses.

    Let’s call the transmission losses 30% in space and 40% on the Moon. So for 1000 MW delivered to Earth, you need 3300 MW of lunar nameplate and you need 1600 MW of SSP nameplate, i.e., 106% more production of modules for the LSP case than the SSP case.

    Beyond that, we have to assume some logistical support from Earth (specialty parts not available via ISRU and repair crews), which is more expensive on the lunar surface than in GEO or EML. If we estimate that cost just on straight delta-v, logistics cost 40% more for LSP than SSP.

    On the other hand, you have the mass driver capital and operational costs for SSP, and you don’t for the LSP case. I’m willing to believe that the the lower logistical costs and the mass driver costs of SSP offset one another, so SSP has half the deployment cost per unit capacity and roughly the same operational cost.

    What am I missing?

  17. “I’m interested in this for terrestrial power.” That is always the assumption with any SSP, totally separate from Space uses.
    “terrestrial solar farms” have the intermittency problem, along with wind farms, which is not just the day/nite, but randomness of clouds, a real hard thing to have. Check out the use of the redirector sats, which are cheap and fairly low orbit, to balance the existing Earth farms’ output. This would require no ISRU, as the redirectors can *easily* be launched. The radar to send to the redirectors can be numerous and small, as they are hitting a near, compared to GEO or Moon, target. Only a few, full sized, rectennae would have to be built, as they can take in a lot of energy. Power beaming in particular is accomplished, and very little has yet happened in Space. But the system is pretty much up and running!

  18. If I remember the spacewalk to repair the loose panel on ISS, they probably were chaining and had to watch out for high voltage.
    I’ve heard of an engine with a magnet in the piston, a combustion chamber on each side of the piston. and no crank! The magnet moving supplies the electric power to coils in the cylinder.
    I think most simple electronics rectifies the AC it gets, then forms the EM by chopping or modulating the (hopefully!) steady power supply voltage with transistors or such. I had a radio that was AC all the way thru, then filtered out that part, but I believe that was an unusual design. in transmission AC is to cut the long distance current by upping the voltage, then go back down for use. But new electronics allow that with DC and converters, as you mention. They run at high frequencies to make the conversion efficient. High voltage DC easily converts to AC, the voltage is the problem usually.
    Way beyond my professional expertise!

  19. I was complaining about the continuous point (by many!) that the Moon has a day/nite cycle, as if that is the end of the discussion. The Moon’s day/nite cycle has been included in LSP from day one. How to handle it is a part of the plan. Simply bringing it up *conclusively* indicates one does not understand the plan, or they would bring up a more specific problem.
    Anyway, sorry for being rude. I’ve been hearing that for 30 years!
    Comparing SPS, LSP and Earth solar is not a simple thing. A lot depends on whether ISRU is allowed. This is only quite recently a popularly accepted idea. Understanding O’Neill’s observation is crucial as a background for thought.
    For example, “much lower manufacturing costs,” for Earth solar concludes the decision process, if true.

  20. In space, yes, as long as you’re using solar, you’d probably opt for DC for most systems.

    The caveat is that some (many?) systems need high voltage (higher than a few volts). I suppose you could chain the panels in series, but it could be tricky to get all the voltage levels you want while still using as much of the available power as possible.

    So the other option is DC-to-DC converters. But those go through an AC intermediate step, so you may as well use AC systems for some applications, and skip the rectification back to DC.

    With nuclear it can go either way. Fission would most likely use a steam (or CO2) turbine or Stirling engine (if I’m not confusing the engine type), which would produce AC. An MHD generator (from fusion) might give DC. Not sure.

    Regardless, the EM emitters of SPS/LSP have to be AC, because that’s how you generate the EM (unless there’s some DC way to do that, which I’m not aware of).

  21. Maybe I don’t; I certainly don’t know what you’re driving at. I’m interested in this for terrestrial power. If terrestrial solar farms have a higher collection efficiency after you factor in transmission losses, much lower manufacturing costs, and only a 40% lower capacity factor, LSP sounds like a permanent loser to me.

    On the other hand, SSP has more than 100% higher capacity factor, which will cover a multitude of sins when it comes to collection efficiency.

  22. “The biggest problem with LSP as a source for terrestrial power is that the capacity factor is only about 50%” In other words, the day/nite cycle of the Moon. If you continue to bring this up as if it is somehow conclusive or even infomative on the topic of LSP, readers will think you do not understand the proposal. I might as well say that the need to build satellites makes SPS utterly absurd!

  23. Agree with the uncooled collectors, no matter where. I’ve seen a lunar radiator plan that has a layer of mirrors, to cover the hot dirt, a vertical fin that is aligned to the ecliptic/equator and cells flat on the top. The fin gets pretty good look at the dark, but not perfect.
    The heat pipes would not be seeking cold, but thermal mass, so to speak. Like a Tromb wall, to balance out the variability. No actual long term sink.

  24. I agree with the *current* home and industry usage, but if there is “no” reason to do AC, why do it in Space when the starting energy is low voltage DC? If SPS has a clear advantage over LSP, it would be in putting the source near the sink, *directly*.

  25. It’s a plasma up there, not a vacuum in a bell jar. There can be static charges, potentially thousands of volts which must be contemplated, building in the space environment.

  26. > On the negative side, you can’t make regolith completely cold under your collector

    As suggested to Dan, it’s probably colder a little deeper under the surface – precisely because it’s a bad conductor. So you could run heat pipes as deep as necessary. But it adds to cost.

    But as I’ve mentioned, Lunar surface only gets up to ~120 C at the equator, so you may be able to have uncooled collectors. But the beam emitters could still need cooling – maybe not the metal, but at least the electronics.

  27. Anything connected to mains is AC, unless it takes the extra effort to rectify the current/voltage.

    Radars and radios in particular are AC, because they need to generate EM waves. So the current has to have a wave shape.

    OTOH, something like a LED would usually be DC, since it uses a constant current. An electromagnet (e.g. for getting a Lorentz force) can also be constant current (or voltage), hence DC.

    Digital stuff is sort of a mix. You have portions of DC, but it’s switched, which kind of makes it AC. You can sort of think of switched DC as square-wave AC. Which is why electronics engineers have to consider EM interference between devices. The switching produces EM waves.

    When you combine DC and AC, the DC voltage act as a constant offset. And you can make filters to pass one or the other.

  28. Now we are thinking!
    “remain cool a bit under the surface.” Just use the mass that is accessed by the pipes, but don’t have to move the mass. So still cheap . It can heat up in day and get cold at nite just like a home underground unit does, or even an underground house. Big enuf so the average temp is good enuf to get thru the day, it will not “remain” cool, just buffer. This is just a brainstorm, don’t take as a Criswell idea.
    Remember that the SPS ideas were fairly well developed before Criswell started(edit: 1984). $.001 per Kwh-e is really the *bottom line*. That includes goodly profit. Expandable to 200Tw-e w/o anything but MORE.
    All ideas to solve global heating, such as Criswell’s, deserve consideration. I’m as serious as heart cancer!

  29. Well, all spacecraft radiate; it’s just a question of where the equilibrium temperature of the sensitive bits winds up, and how you manage the heat flow.

    SSP has the advantage of having an unobstructed shaded side, which means that it has a bigger temperature difference between the hot stuff and cold space. The problem you have with LSP is that the shaded side points at the ground, which heats up and radiates back at you, reducing the effective temperature difference.

    On top of that, you have conduction issues on the surface that you don’t have in space. That’s a good news/bad news situation. On the plus side, you may be able to run cheap conductive cooling away from your collector and dump heat into the regolith, which will then be a lousy but very large radiating surface. On the negative side, you can’t make regolith completely cold under your collector, because heat will conduct in from the surrounding regolith that’s still in the sun.

    The biggest problem with LSP as a source for terrestrial power is that the capacity factor is only about 50%, unless you’re finding just the right spots at the poles, and then your scale will be limited. Compare that to >90% in GEO or most EML points and the cost of flinging stuff into orbit may easily be offset by the need to make and deploy more stuff on the surface.

    Beam forming with LSP is more challenging, simply because the Moon is farther away than GEO. But that stuff is all black magic to me.

  30. > Making that surface area with sats is not trivial

    I imagine something similar to a solar sail. Minimum thickness and strength, spin it up just enough to spread out. Collector sail can double as shade for radiator sail. Or stack a few layers like that like the shades of JWST.

    Station keeping is tough, but already have conductive stuff for the radar emitters, so maybe can use same for Lorentz force station keeping. Radar is AC, Lorentz is DC, so can be on same conductor.

    – – –

    For Moon, day/night is too long to use night as radiator. But the ground may remain cool a bit under the surface. So can run heat pipes there (at cost of extra mass).

  31. I agree with the asteroid outlook too. In fact, my anti-Mars stuff is not anti-Musk, as it started back in 1977, vehemently O’Neill. It was MY taxes paying the tune for Mars centric gov spending. Spilt milk, for 40 years. Even my lunar interest is O’Neill, just get the goods, don’t live there. LSP is just a special opp within the O’Neill effort. And the dude does make some fine rockets!

  32. Ah! I am taking the assumption that radar will be twice SSP, and then some, for LSP, as it is larger dia but same power, for the half in lunar day. A definite added cost!
    I’m under the impression that the cells do not need cooling. I remember super hot cells (Israel design?) that can have concentrated sunlight for greater efficiency. The radiators are to cool the radar, har! And LSP needs ~3 times the cell area. A definite added cost!
    If radiators are the problem, perhaps the normally ignored mass of the lunar dirt combined with cold nites can use the same parts being cooled in day to dump heat at nite, with the dirt as thermal storage. Radiators on sats need “nite” surface area to work, as would same style on Moon. I see no clear advantage either way here, but needs details. Moon at worst twice the cost, if only cooling radars.
    But each of these things need surface area! The cells, the radars, the radiators. The Moon stinks with surface area, and the sun is bright when up with no atmos. Making that surface area with sats is not trivial.

  33. BFS alone is still ~3 kiloton when fully fueled. I don’t know, it might be possible to escape if given enough thrust-to-weight on the escape system, but sounds doubtful. It’s still a pretty big boom.

    (For comparison, MOAB is “only” 11 tons TNT, and the huge explosion in Tianjin in 2015 was just over 300 tons. Left a big crater behind.)

  34. That is for the full stack with superheavy, though. I was thinking of a BFS only size (or only slightly bigger) SSTO. IMHO launch escape should work for that just fine. After all, it would work for SLS.

  35. I’m not a “Mars person”. I’m a “He who pays the piper” person. If I had a few billion burning a hole in my pocket, colonizing Mars would not be on the top of my list.

    Mining the asteroids would.

  36. Thanx for info!
    I would guess about 0 for beam forming with LSP. That would be to build the sat, right?. The Moon is the sat!
    Radiators can be the lunar dirt, which will get cold at nite. A small advantage from the big cost of nite.
    So, looks like LSP may be a “rounding error” compared to the cost of SPS, except for the increase in the cell area needed for non-direct 24/7 sunlight you get in GEO.
    And remember, the scale has to be at least 20Tw-e or we need another solution!
    An important issue, needs independent look to update 10 year old Criswell ideas. Much like O’Neill, the point is the Physics, not the current engineering.

  37. Keep in mind that that’s the worst case. Realistically it’d be somewhat smaller. But still a nuclear-scale boom. Even the current BFR would be ~13 kiloton. Rather puts into question the feasibility of any crew escape system (somewhat in response to your other post as well).

  38. > most of the mass wasn’t in PV; it was in radiators. Second in terms of mass was beam-forming stuff. The PV and wiring was almost a rounding error.

    If true, that shifts the discussion on LSP vs SPS, since LSP wouldn’t provide a significant advantage on the total mass if the collector’s mass is negligible. Though I wasn’t convinced it would either way – at the limit of strength, I expect the collectors would be about the same thickness (and mass) for both.

    Do you have any numbers on how hot the sat would get without radiators? Moon surface max temp is 390K at the equator (per Wikipedia), so I expect that’d be about the max for LSP. Probably a little less, since some of the energy would be absorbed by the cells. So ~100 C. I think we could come up with solar (or maybe solar-thermal) cells that could tolerate that.

    As I understood, at large enough power scale, the beam-forming stuff would be the same mass for LSP, just spread out differently. No idea idea if that’s true, but if so, the radiators determine which takes less mass overall.

    ( , for your attention)

  39. Or just have a slightly larger Dragon 2 capsule on top, which then would have abort engines and 20 tons to LEO would more than enough to launch Starlink with it, while gaining lots of operational experience with the vehicle. Once they have that, they could still grow the vehicle. But who knows? A different approach may be more sensible by then.

  40. 127 kt of TNT would be 10 times the size of the Hiroshima bomb. I would not want to be anywhere near that if it was to blow up. At the least it would be significant damage to the launch infrastructure and anything else nearby.
    LNG carrier ships do worry me too…

  41. This would be starship as a ISS component or a free flyer in LEO with 20 MT of docking mechanisms
    The docking mechanisms are the real estate to sell.

  42. I foolishly forgot to bookmark the study that looked at a modular SPS, but I do remember that most of the mass wasn’t in PV; it was in radiators. Second in terms of mass was beam-forming stuff. The PV and wiring was almost a rounding error.

    The good news is that that can all be aluminum, which is easy to get via lunar ISRU. But you still have a lot of pretty high tech fab that you’d have to replicate on the lunar surface to get this up and running–and you’d almost certainly need a mass driver.

    BTW: oxygen radicals in LEO are the major cause of space solar panel efficiency degradation. They’re still a problem even at GEO altitudes.

  43. Beam losses are huge, especially if you want the rectennas on the ground and producing adequate capacity through cloud cover. Beyond that, you have diffraction losses. You’ll do well to get 70% of the capacity generated on-orbit to the grid.

    There have been some proposals to use aerostats in the stratosphere to receive very high density beams. I have an open mind on this, but it has trouble passing the smell test.

    I agree that the offset here is that SBSP can have >90% capacity factors, which was the whole point of my comment.

  44. Be careful to define income in *absolute* terms such as food and housing, rather than in relation to other people. We don’t realize how easy we already have it many places!

  45. So, howse about a fuel cell that *emits* some of its charging energy absorbed as fast moving O2, for propulsion?? Whether or not on that, you would refuel by topping off the H2 and getting loads of O2, a waste product. The energy from the fuel cell would accelerate the stored O2 for reaction, but the product H2O would be recovered for fuel cell and other use. Whatever solar cells or other source of power within the ship would re-use the water, to re-fill the H2 supply. Or, more H2 could just be gotten from a depot.

  46. I could believe that. Maybe just building larger cores IS the way to go.

    Of course, at some point you hit diminishing returns. The height is limited by engine pressure, probably not significantly higher. And as the diameter increases, the rocket becomes less aerodynamic. At some point the increased air resistance is going to make larger rockets too much bother, lower the effective capacity due to reducing the speed at max Q.

    But I doubt the demand for cargoes that can’t be split into 500 ton parcels is going to be very high.

  47. Please consider that ISRU is for everyone! While Space Solar Power is O’Neill’s cash cow, it may not even happen. MIGMA, anyone? And long before very much SSP ISRU is done, many other projects will be using it. Stuff we have not had time to think of yet. ISRU is more important than SSP!

  48. It is not a yes/no question, but a balance between time(how long you have to wait for ISRU to start making more ISRU and get going) and cost(how much more the first, bigger, thus faster, steps cost). Bigger, cheaper rockets certainly save time. (edit: in other words, the argument FOR ISRU is scale independent)

  49. Janov sez ideas can be the most powerful painkillers of all. We have evolved very big brains, relative to our cousins, to *jam* pain that is repressed from the past. Pain supplied by our seven million years evolved rituals, “The System”. So, ideas, whether true, good, untrue, bad or just plain silly, can be addictive!
    Mystic ideas are generally driven by deeper birth pain, but the content of any ideas are not necessarily *like* the pain itself, just being a tool to ignore the pain.
    Dreams of Space tend to be big ideas. Very addictive!

  50. Once critical mass of stuff in space is reached and ISRU is viable, the rockets will switch from metal to meat payloads.

  51. My understanding is that this article is wholy conjectural, built around some remark Musk made about a larger fairing.

  52. BFR full stack is ~1000 tons of methane. This thing is 4-8 times larger, so worst case 10000 tons. LNG is 53.6 MJ/kg, so 10000 tons of it is 536 terajoule. Which is 127 kiloton of TNT.

    (For reference, an LNG carrier ship is about 10 times that.)

  53. But you don’t need 1000 tons per launch to land 1000 tons. A single fully refueled BFS v1.x (we’re not at 1.0 anymore with all the revisions they’ve made so far, are we?) can land several dozen tons. Repeat ~15-20 times, and you’ve got your 1000 tons.

    Now, if we needed to land 100000 tons or more, that’d be a different story.

  54. As I understood, the biggest lesson they learned from FH is that strapping on two more cores is more difficult than they thought, and not worth the effort. They get more bang for the engineering buck by making a larger single-core rocket.

  55. If we get a high level of automation, let alone self-replicating automation, the whole “income” concept goes out the window. Or at least will need to be redefined. But if translated to quality-of-life terms, then I generally agree.

  56. not be touched…

    And where would Darth Musk be located, that he could be untouchable by the major superpowers?

    Or does this plan have to wait until he’s moved his home to Mars?

  57. I would not be the slightest bit surprised if it took a 1000 tonnes of equipment landed on the moon to be manufacturing solar cells and microwave transmitters from lunar rocks. (whether setting up said power station on the moon or launching them back to Earth orbit.)

    Remember that the more tonnes you are allowed to start with, the less tech development is required before you can begin.

  58. I’m intrigued by the idea that this would require any mental bracing.

    But then I see people freaking out about fictional character deaths and so forth, so clearly a lot of the population has a level of mental fragility beyond my understanding.

  59. I think that’s the answer to why we need to go to space. With the huge amount of cheap labor on earth, technologies like self-replication machines, propulsion and energy get neglected for the benefits they can bring. Living in space is going to require much more automation to do useful things. That’s going to create a society where people are making 10x higher per capita income than even the richest countries on earth. A lot of technology that gets developed is going to be useful on earth as well as space, but it probably wouldn’t be developed at all if we didn’t travel and live in space.

  60. From rocket science

    Fp = (F * Ve ) / 2


       Fp = thrust power (watts)
       F = thrust (newtons)
       Ve = exhaust velocity (m/s)

    Fp = 1000000 kg * 9,82 m/s^2 * 3500 m/s / 2 = 17,5E9 W
    or 17 GW of power

    For comparison, the heat capacity of a cubic kilometer of air is 1000^3*1200*1 = 1,2E12 J/K

    So this thing would heat a cubic kilometer of air by 0,875 degrees pr minute.

  61. Each lunch of this thing would deploy up to 4000 satellites. Its about 4 per ton. He estimates 20 billion to deploy the Sarlink system using F9 pricing. It supposed to cost 9-10 billion for the satellites (at roughly 700-800k per satellite) and 10-13 billion to launch them. If he could build one of these for lets say 2 billion for the build, another 2 billion for development at 80-100 million a launch, at 4.5-5 billion total, he could deploy Starlink for less than 1/2 the cost from the current F9 model for deployment ~12 billion in launch costs assuming 50-60 million a launch for 60 sats per launch, 200-220 total launches. With this it would take 3-4 launches.

    The Starship with booster might be lower, with around 700-800 sats per go, likely 2 billion in development and I will guess at build cost of 250-300 million per vehicle with booster, maybe 100 million for just the Starship cargo version and 150-200 million for a heavy booster, probably 10-15 million per launch in costs. With the Starship you need to do about 20 launches. It might only cost 3-4 billion to deploy Starlink just with the Starship and booster 1.0.

    So since they have already spent on the Starship, progressing with this doesn’t make economical sense for Starlink. It is only good if he has a bunch of cash and want’s to go big on colonizing Mars.

  62. Bezos is currently, hence why I said in under 10 years. No one else comes close to having their hand in so many multi-billions dollar businesses as Elon does.

  63. Not really afraid of him. Just thought it was interesting that he has the power of a medium sized nation, and some of his ideas would put him near superpower status from a military potential perspective. Bezos is heading that direction as well, but lags. Bezos is the one to worry about, he has an almost Lex Luther vibe. Musk is more like Tony Stark.

  64. I suggest Mars people brace themselves for the possibility that Musk may suddenly wake up to O’Neill. Perhaps he gets mad at Bezos and decides to see what all that *stuff* is all about!

  65. He is perhaps not against SSP, but lunar ISRU, the obvious way to do SSP. Lunar ISRU is not popular amoungst Mars people. Once it starts, you just seem to forget about Mars!

  66. The cost of not blowing away is probably more than the cost of the whole cell in Space, per surface area let alone per illuminated surface area. You go even further by making the actual cell works lighter. I’ve thought of directly spraying the lunar surface to build up cells. If possible at all, they would probably be inefficient, but cheap!

  67. “efficiency of terrestrial solar cells will always beat the efficiency of space-based cells that have to incur transmission losses to get the power to the ground. And that is of course completely true.”
    But. . . but the reason is not to make more efficient cells, it is to get into the sunlight! The power beaming losses (not that great!) are small compared to the same efficiency cells in the dark on Earth! (edit: even the relative dark of midday sometimes!)

  68. I’ve heard there is a 3-D printer head that can make “itself”. And a lot of other stuff too.
    While you state the best possible bootstrapping technique, the sudden advent of large rockets makes the von Neumann starter kit much more massive for the same cost, thus getting to the goal quicker. Imagine such a deep gravity well that we had to wait for Drexler tech to launch!

  69. If you are afraid of Musk, just keep in mind that he can already put 60Ts of Tungsten beach balls in orbit today and wipe out a town or other localized area.

  70. Falcon 9 is already 500 test
    Falcon Heavy is Test/Deca/Dbol
    Starship 1.0 is Test/Deca/Dbol/EQ/Tren
    Starship 2.0…. Test/Tren/Halo/Dbol/Methyltrienonlone

  71. Internet advertising, “big data”, analytics, commerce, and sales is a good example of a technical advent that created the market that now substantiates its existence.

  72. You could maybe haul the materials up from Earth, but I think a key for orbital photovoltaics is building cells that wouldn’t be feasible on Earth, that only survive due to being in vacuum.

    Not cells that can survive exposure to an oxygen atmophere and being beat around on the flight up.

    If you go with space only rated cells, you can pare away almost everything but the junction, minimizing the current carrying capacity the cells need by stacking them to high voltages. They would then become very high power to weight ratio once you’ve eliminated most of the inert structure.

    That’s the key for zero G vacuum applications, high power to weight ratio, not efficiency.

  73. Musk has expressed disdain for SBSP, not just indifference. His argument is that the efficiency of terrestrial solar cells will always beat the efficiency of space-based cells that have to incur transmission losses to get the power to the ground. And that is of course completely true.

    But Musk also happens to own the largest battery company in the world. He has a vested interest in turning intermittent renewable power into baseload and dispatchable load-following via storage.

    On the other hand, SBSP is pretty much suitable for baseload applications from day one–no storage required. So if somebody can crack the scaling, Musk’s argument becomes pretty flimsy–as does the value of Tesla stock. (NB: Tesla is a solar PV and battery company that happens to build cars as a market-priming application. Eventually, they can lose the car business. But they can’t survive losing the PV and battery businesses.)

    But I just can’t see cracking SBSP by hauling stuff up from Earth. The secret to making it successful is to use ISRU to build modules out of lunar or asteroidal material. You con’t need a kilotonne-class launcher to do that.

  74. I actually figured that out a while back for a space-based solar power system, and you’re right: the emissions are minimal.

    But if you have a market for octupling the payload, it’s because you’ve thought up something really high scale to do with it. At some point, emissions really will be a problem.

    Beyond that, even hydrolox and methalox rockets produce a non-trivial amount of NOx compounds, due to the hot radicals from the exhaust whacking into the N2 in the atmosphere. Ozone degradation is already something that people are a little nervous about with high-cadence launches. If you go to the kind of scale being discussed here, you’ll have a problem.

  75. Quite frankly, I am a bit worried about what would happen if this thing blew up on the launch pad. This would be a huge explosion.
    Personally, if I was Musk, I would rather try to get to a SSTO version of Starship with a 20 tonne payload to LEO. I believe there is a bigger market for that. But Musk always dreams bigger than everyone else.

  76. But that scale requires moving resources around at the top of the gravity well, not boosting them from the bottom of the gravity well.

    The highest-scale resource coming up from Earth will be people, and a 150-tonne payload of people on trips from the surface to something in GEO or cis-lunar space lets you move at least 1000 at a time. I assert that that’s about the largest number you want, if only for insurance reasons. We already know that the A380 has trouble boarding and de-planing 600-ish passengers in a lot of airports. I’d expect worse problems doing the same on a spacecraft.

  77. Compared to emissions from other sources, even a couple of hundred Starship launches a year, would hardly have any impact at all. I think people underestimate the amount of fossil fuels we are currently burning for power, heating, etc.

  78. There is PLENTY of O2 out there. And, yes, even with good mass drivers and such there will still be much need for propellant. Good idea, just need a good power source.

  79. We’ve got the rockets now, what we need is a good push to achieve self-replication in factories. THAT is the key technology to expanding into the solar system; The infrastructure requirements for life outside a preexisting biosphere are sufficiently large that colonization of space will be at best a marginal enterprise until amount of infrastructure we can use is uncoupled from human labor.

    I’m convinced that we’re within spitting distance of self-replication if we put in the effort. Not Drexlerian nanotech, (Though that’s a worthy goal.) but what they call “clunking replicators”, factory complexes that run unattended and manufacture all their own parts.

  80. I agree! We certainly do have plenty of Iron, for example, but it is quite costly to pull out of the core. If the goal is to live in Space, rather than on a planet, starting lunar ISRU research 40 years ago would have been quite easy, tho primitive. After all, we seem able to do stuff pretty well on Mars, a much more difficult case. But that is spilt milk. And, the goal is STILL to live in Space!
    How about O2 instead of H2O?(edit: I was thinking of the hot steam microwave rockets, not combustion exhaust)

  81. Mars would be a good candidate for it, and for solettas; Mars synchronous orbit is much closer to ground level, 17,000 km vs 42,160 km for Earth. Makes everything much easier.

    If I were colonizing Mars I’d prefer a mix of ground based nuclear and beamed power from orbit, rather than ground based solar, which is so vulnerable to those dust storms. And maybe solettas to bring the local illumination around the colony up to Earth level.


    On thinking about it, probably no on the Solettas. They are more feasible for Mars, but SpaceX plans on colonizing areas with shallow sub-surface ice.

    Warming the ground in an area with sub-surface ice and vanishingly low surface air pressure could produce geysers and other unfortunate phenomena. And a Soletta would unavoidably do that.

    Best bet is, as I said, ground level nuclear power, and aerostationary solar power satellites. You wouldn’t bother exposing your greenhouses to Martian sunlight, it’s too dim to be worth bothering. You’d bury the greenhouses under at least a couple feet of soil, and illuminate them with purple LEDs for maximum photosynthetic efficiency. A couple feet of dry sand with vacuum between the particles would be adequate insulation to keep the greenhouses warm from the waste heat.

  82. Just WOW!!! Would love to see this built and launched. Could this even be built to be reusable? I’m not sure of what market need it would meet, but I can think of one that it would. The Space Force. Buying a $15 billion dollar rocket would not be too big of a deal (especially reusable), and may be necessary to launch lunar infrastructure so we could build military capable craft on the moon.
    Space command would love to have this capability, even if they are still Air Force. They would have to see the potential. Navy might partner with them for it too considering what value their ships would have with enough military hardware in orbit able to rain fire and brimstone down on them in any ocean in the world. Aircraft carriers are like bottles floating on a lake being shot with a .22 rifle by kids plinking from shore when looking down from orbit. We already spend $13 billion on aircraft carriers so this is not too large of a leap even for hidebound thinkers training and equipping the fight the last war.

  83. The massive advantages due to low launch cost requires a paradigm shift in payload development that will take a while for industry and space agencies to get their minds around. The JWST is a classic example since it was entirely conceived before SpaceX was even founded. If Starship becomes functional you would not even have to fold it up. It could fit it in the cargo bay fully assembled (except for the heat shields) Its a new world.

  84. Making hardware and machinery that can and requires to be launched to space isn’t straightforward nor cheap. And there is no urgent need for making produced-in-series sats, ships or space habs.

    AGREED, But certainly lower launch cost makes series production of payloads much more feasible, including Hab units. Also, with a much launch lower cost per pound they don’t have to pour nearly as much engineering $$s into making the stupid things light (which also makes them less robust) so more can be put into numbers if units.
    Other payloads that are ripe for serial (mass) production are space telescopes and planetary probes and the like. Think of the Webb Telescope. Great work developing it, but they put all their eggs in one basket, as in one unit, one launch. They are so afraid of launching the thing because they put so much into it that if it fails it is a complete wipe-out.
    They should just put a bunch of instruments on it and launch on Falcon Heavy. If it fails (and it probably will due to its fragility and complexity) then at least they will have a bunch of data. In the meantime, start setting up an assembly line to start producing them and then use the data from the previous launch to make mods as necessary and then crank another one out. Launch and mod as necessary until it works then start cranking them out to get a dozen or so. These won’t cost that much per unit because they know how to build them and the key tech which is staggering is already paid for.

  85. You mean ISRU is important because it is just way too COSTLY to get things out of Earth’s gravity well. We actually have PLENTY on Earth for O’Neill cylinders at least for the next couple of hundred years or so. It will be very hard to find enough water in the inner solar system for O’Neill cylinders, especially if we start blowing precious water out the back end of rockets for a one shot momentum boost, which always struck me as a capitally stupid proposal. If reusing rockets is so key to sustainable, thrivable, human presence in space then reusing water (and other volatiles) is even more important. This boils down to using mass drivers wherever possible and reusing that momentum with some sort of tether scheme.

  86. Arguably, he’s trying to move rocketry into a more conventional engineering domain, where you can throw something together following established rules, and be reasonably confident it will work.

  87. If Musk were interested in space solar he likely wouldn’t be showing those cards just yet. Since it is obvious that every business he’s involved himself in relates directly to what is needed to colonize Mars I would be surprised and disappointed if he didn’t have plans brewing for power generation.

  88. Putin should put some nerve agent in his coffee now, or it will be too late to stop this evil man from achieving world domination.

  89. It’s funny that Musk isn’t into space-based solar power. That’s the only industry I can think of that would have a real use for this level of launch for decades on end. Even if you can launch hundreds of megawatts at a time, it’ll take a lot of launches to reach the terawatts the Earth needs.

  90. With 10 of these you have triple the conventional power of the US Navy, but can hit anyone in minutes and not be touched. after take-off These things would fly above all current AA/ABM systems. If reusable within a week , it would take a month to put enough in orbit with just 10 to take out 50% of the 10 largest cities on the planet. Hit the Aswan Dam kill 100+ million.

  91. The Cape will handle anything up to Saturn V scale, which means it will take the Starship 1.0. Beyond that, I expect they’ll need offshore launch from special barges.

  92. That was my thought: Before they went to a super heavy starship 2.0, they’d gang the first stage on the Starship to boost the payload, just as they did with the Falcon.

    Unless Musk just thinks they’re getting enough experience building new rockets that less surprises can be anticipated, of course. That may be what their data is telling them.

  93. > There can be no market for something non-existent.

    Not true. There are often needs that aren’t met with a suitable product or service. People are willing to pay a certain amount to have those needs met. That’s your market for a non-existing product. And that’s the first thing most new businesses are built around.

  94. More than any conventional explosives… but nuclear explosives exist and have done so for many decades.

    He is still overmatched by several different countries. All of whom would take steps to prevent him doing any “holding the world hostage”.

  95. And yet, he has yet to launch a single human being anywhere, let alone LEO. Wasn’t that supposed to happen this month? It’s easier to build big badass rockets than safely move the eggshell human being, but we’ll need to do the latter sometime if SpaceX is going to get beyond cargo.
    Well, maybe he is planning on send robot constructors to the Moon (not Mars yet) to build space habitats for us fragile beings first. Not a bad idea then…

  96. Lets not kid ourselves, if Elon Musk built that thing, he would be the most powerful man alive. 1000Ts of tungsten beach balls can wipe out a big city or two from LEO with a single launch. This is supervillain or if benevolent, Tony Stark, level of power. One of these things would have the fire power of 5-6 aircraft carriers in the US fleet, but with nearly instant response times. 20 and he could hold the world hostage. Also, come to think of it, 2 Starships have more power than an aircraft carrier and 60-70 F-18s in terms of potential energy (conventional explosives at 400 ton TNT max, 3.7 tons per plane vs 100 tons of mass with kinetic energy at mach 15+). 100 tons at mach 15 (1.25 billion kilojules the same energy as ~300 tons of TNT), two is more than a half kiloton, multiply by 10 for this thing, you have 6 kilotons or 1/3 of Little Boy, enough to level a city center.

  97. Often when there are the means, the market develops around them and not vice versa. In other words, we will find what to do with cheap commercial space flight once we have it, I have no doubt about that.

  98. Making a launcher that big implies that there’s enough payload to launch, fairly regularly, from Earth’s surface. At some point, you start running into externalities that prevent that from being true. A few, off the top of my head:

    1) Ozone destruction.

    2) GHG emissions.

    3) Inability to launch from anywhere that isn’t the center of the Pacific Ocean without making so much noise that everybody hates you.

    I think it’s much more likely that the people who put R&D into off-planet ISRU win at this kind of scale: The company that invests in clever gizmos to do this stuff elsewhere is going to beat the company that invests in launching the kind of huge structures that would be needed to make such a rocket economical.

  99. The balance between bootstrapping and brute force has changed, but only in that the number of steps to full scale iSRU is decreased by having big, cheap rockets. ISRU is still the important thing, as we don’t have enuf on Earth to supply O’Neill scale needs!

  100. It’s obvious that Musk has already thought about it… and he’s going to do it.

    “energized by nuclear power” –> NASA KRUSTY

  101. You seem to have a static view of things.

    Starlink satellites have a short lifespan. The goal is to regularly put new, improved, more powerful, and heavier satellites into orbit.

    “I’m more concerned about the financing…”

    I’m not. Starlink’s revenue will be enough to create all the rockets Musk will want to see flying.

    The design of the Falcon 9 cost about $300 million. Recovery cost more than twice as much to fine-tune. And the Falcon Heavy cost just over half a billion. Starship is expected to cost $1.5 billion, and much the same for Super Heavy.

    Compared to the SLS/Orion (~$40 billion by 2034) what SpaceX does is a miracle.

  102. “And let’s see if there actually is a market for that.”

    There can be no market for something non-existent.
    The rocket will create its own market.

  103. Do you think there’s gonna be strong demand for that beyond NASA and other space agencies?

    I guess it depends on the price tag and how validated the launchers become.

  104. Yeah, that’s why it’s good to take the most outrageous Twitter bluffs from Elon with a grain of salt.

    In the case of Starship, I think SpaceX will be busy at least the next decade using and optimizing the reusable architecture, getting some extra Isp from Raptor and a few extra tons to LEO out of it.

  105. WOW, that is crazy!
    But…that’s Elon. Always looking at what’s next, regardless of how impressive things are now. He will be the richest most powerful man on earth in under 10 years. No one else matches his drive & passion for innovation.

  106. 100 tons to LEO (and to Moon/Mars with refueling) is plenty big enough to start ISRU. With ISRU, the market for Earth launches might shrink, eventually. But there’s a difference between “start ISRU” and “develop it enough to replace Earth launches”. The latter will take time, during which the Earth launch market may still grow.

    Even with ISRU, there’ll be a need to launch “vitamins” from Earth. Stuff that’s too difficult to make in space. The more ISRU, the more vitamins will be needed, though gradually of fewer and fewer types, until the volume starts to shrink too.

    Also, the more we develop space, the more traffic there will be to and from it. So gradually, the launch market will shift from cargo to people.

  107. So, are they going back to Kwajalein (site of the Falcon-1 launches)?

    I mean, as when we did open-air nuke testing, where but the mid-Pacific could you get away with the acoustic signature of that monster?

    If launched like Sea Dragon, that’s going to have a serious acoustic effect on marine life as well.

    And then there’s passengers…if they dare.

  108. A 1000T to LEO reusable ship is the kind of thing Gerard K. O’Neill would dream about for making space colonies. A humongous project requiring millions of tons of stuff sent to orbit in a reasonable time.

    And it would need an equally gigantic commitment and budget!

  109. I’d love to see it, but my inner skeptic tells me it won’t be that straightforward.

    Not that I think they can’t do it, mind you. I’m more concerned about the financing and success of the overall commercial endeavor, even if the launcher itself succeeds. This could pretty much be Elon Musk’s Spruce Goose.

    Making hardware and machinery that can and requires to be launched to space isn’t straightforward nor cheap. And there is no urgent need for making produced-in-series sats, ships or space habs.

    We would need to develop real production in series of such parts, with a strong commercial need behind it, in order to justify even a fully reusable Starship, let alone a hulking 1000T to LEO monster.

    I’m aware Starlink tries to bring production in series to space economics, but much more than that would be required for using the payload capacity of a reusable ship of that size.

    At least, I know Elon Musk is no Howard Hughes and he won’t make a boondoggle just to prove a point he made in public.

  110. And let’s see if there actually is a market for that.

    My hunch is they will eat up the existing launch market rather quickly, with a reusable 100+ tons to LEO launcher.

    Falcon Heavy already has some problems finding buyers.

    Everything after Starship assumes the launch market will grow, and Starlink is a partial answer to that problem.

  111. This is going to be I guess a rocket that can land on any point in the Solar system with less than half of earth gravity and return to earth including mercury and Jupiter’s moons without refueling and of course can also serve for building big structures in space.  

    I think that it will be smarter to focus on building huge shuttles of multitude of multiples of the most advanced Ion thruster design we have energized by nuclear power. They will be specially launched and deployed in space , will never land on earth or any other celestial body but will tow space crafts going between Earth LEO and even lower energy trajectory to that of other destinations in the solar system. They will be refueled by the crafts that they are carrying. That will also allow a multiplication of load that can be moved in deep space on a more economic scale.


Leave a Comment