Last month I was out in Washington DC for a satellite industry conference (the 2017 Satellite Show). My startup has been working on a cool satellite servicing related project that I need to blog about soon, and we were out there to meet with potential customers, visit friends, and try to find new opportunities to pursue. While having dinner with one friend, who is a lot better connected with the satellite industry than I am, we got discussing an interesting thought about future market dynamics, that got me thinking. I’d like to share an abbreviated version of this thought, to see what people think.
Anyhow, by way of introducing the topic, right now there are several companies proposing massive fleets of LEO telecomms spacecraft. If you believe all the me-too filings with the FCC, there are nearly 20,000 new proposed spacecraft, all that would theoretically be launching within the next 10 years. That’s over an order of magnitude more spacecraft than are currently flying. Now, many of those are going to fail, but I think there’s a non-zero chance that we could see at least a few make it into successful operations. We had a similar wave in the late 90s, right as I was starting to get seriously interested in the industry, and in the end only a few of the proposed constellations of that time panned out (Iridium which had to go through bankruptcy but is now launching its second constellation, and Globalstar, and Orbcomm), but I do think there are some fundamental differences that give this round a better shot. There are a lot more affordable launch options coming up–not the least of which is SpaceX, technology especially on the electronics side has dramatically improved, there are shifts in how people consume telecomms products/services that make it easier for comms providers to reach customers, and some of the providers have raised very serious amounts of capital. That still doesn’t prove that SpaceX or OneWeb or Boeing will succeed, but it gives me some hope.
For the first round of megaconstellations, I’d give OneWeb the best shot of being first to market with a big LEO constellation (>100 satellites). By way of disclosure, I’m a bit biased since we’re actively talking with them and some of the others about some services we’re trying to provide, but I still think it’s not much of a stretch to see them as likely being first to market. They’re going after an ambitious but IMO realistic technical approach, that doesn’t have all the bells and whistles SpaceX is trying to cram into things. They’ve raised $1.7B of the ~$3.5B they need to make to market. If I’m not mistaken, they’ve raised enough to cover the development and production of their spacecraft, and most of what they still need to raise is just for the launch of the main fleet. They also have a lot of people who have been working on this idea for years, and who know their way around this industry. No guarantees, and no guarantees they’ll be successful in the long run, but if I were a betting man, I wouldn’t be surprised if they actually made it into operations.
But their biggest competitor is SpaceX. And while OneWeb may have the first mover advantage, SpaceX has an important advantage for second generation systems–they also own their own in-house low-cost launch system. You don’t have to fawn over every tweet Elon types to believe that they’re likely one of the cheapest launch providers in the world right now, and that their launch costs are probably significantly lower than anyone else’s launch price1. This gives them an inherent advantage, especially for a second generation. Even if other megaconstellation developers fly on them–they’re going to be paying the full price, while SpaceX only has to charge themselves their internal launch costs.
Is this an insurmountable challenge for OneWeb, Boeing, Kaskilo, and the others? I don’t think so. But it does suggest that in the long-run they may need to find a way to get a similar advantage. What that suggests to me is that for the second generation of LEO megaconstellations, the ones most likely to stay competitive with SpaceX will be the ones that find a way to duplicate their in-house launch cost advantage–by pairing-up and vertically integrating with their own low-cost launch provider (preferably an RLV developer). By doing so they can negate one of SpaceX’s biggest advantages in a way that might give Elon a serious run for his money.
So if this hunch is right, what are the options? Basically you need to do something that gets your launch costs below SpaceX’s published launch prices, even with reuse starting to have at least a modest price impact. I can think of a few approaches:
- Approach Blue Origin about a merger. While Blue Origin is seen as the most likely RLV competitor to SpaceX, this may be a challenging deal. You’re basically trying to merge with a company run by one of the richest people in the world, so it would likely be more of a “please Jeff, buy us out and let us keep at least a part of the resulting company”. It’s possible, but seems like a bit of a meh deal for a megaconstellation developer. Especially if they were able to get a successful first generation system up and are trying to stay competitive for the future.
- Buy out one of the small launch vehicle developers (VirginOrbit, RocketLabs, Ventions, Generation Orbit, Vector Space), and try to get them to scale up a bit and get into reuse. This one seems a little more plausible. You’d be taking a team that hopefully by the point where you’re acquiring them, they’re starting to fly regular space missions, and then doing what Elon did, and trying to convert an expendable vehicle into a more and more reusable vehicle. You don’t necessarily have to go to the full scale Elon did–if you’re just trying to have a way of launching your own satellites, you don’t necessarily have to go after an EELV class vehicle. In fact a ~0.5-2mT to LEO semi-reusable vehicle might be competitive enough with SpaceX to at least be cheaper costwise than paying their launch price, and would still allow you to launch several LEO megaconstellation birds per flight. One challenge is that at least the bigger of this group (Virgin Orbit and RocketLabs) have already raised a significant amount of money, and have pretty high valuations–you’d be paying pretty dearly for one of them.
- Roll your own RLV company. If you’re a successful 1st generation megaconstellation developer, you’ll hopefully have some non-trivial profits coming out of your operations going forward. These constellations are hoping to be chasing multi-billion dollar market opportunities. Which means that you could eventually spool up your own in-house launcher capability. But you’d be starting from scratch, trying to hire and build a team. It’s definitely possible given enough money, but seems like it would likely be a slow and painful process.
- Buy someone like Masten. I listed this one last, because as a shareholder of Masten, I’m obviously not an unbiased source2. But there’s a lot to be said for buying a company that has significant rocket development and operations experience. Sure, Masten has mostly been focused on low-altitude EDL work till now, but a lot of that has been due to being too undercapitalized to make a serious go at full suborbital flight. They have significant rocket development and flight operations experience. They haven’t done multiple VC rounds at unrealistically high valuations. And their focus has been on developing fully-reusable, gas-and-go launch vehicles. Not expendable vehicles with reusability powder sprinkled on them. With something the size of their Xephr XS-1 concept vehicle (or potentially a little smaller), if they could pull off a fully-reusable vehicle like they think they could, that could potentially be very competitive with SpaceX. Definitely cheaper than their launch price, maybe even creeping up on their launch cost.
Notice I didn’t say “go start another expendable small launch vehicle company” like far too many people are doing (and somehow raising money to do). I think you’re going to have to go in for at least partial reuse, if not full reuse to have a shot at getting close enough to stay competitive. I think there’s a lot to be said for looking at one of the smaller launcher companies that doesn’t yet have VC-inflated valuations (Masten, Generation Orbit, and maybe Ventions or Vector Space). A lot more of the money can go into scaling up and developing a competitive system, and while most of those companies are small, they all have at least some critical mass to them.
And if you had a small fully reusable launch vehicle, there are also a bunch of other markets you could serve above and beyond launching and replenishing your own constellation. Stuff like propellant launch for companies like ULA who want to do Distributed Launch, launching cargo or people to commercial LEO facilities (that might not be able to afford COTS-scale vehicles, or who might be interested in more frequent up/downmass opptys), space tourism, or launching materials for building large LEO, MEO, or GEO spacecraft and space facilities.
Anyhow, it’s food for thought, but while LEO megaconstellation developers start getting traction, I think it would be well worth it for them to start thinking about how to pair up with a promising launch vehicle developer that can help them stay competitive with SpaceX. Unless they really want to have SpaceX eat their lunch on the second generation megaconstellations.
Jonathan Goff
Latest posts by Jonathan Goff (see all)
- NASA’s Selection of the Blue Moon Lander for Artemis V - May 25, 2023
- Fill ‘er Up: New AIAA Aerospace America Article on Propellant Depots - September 2, 2022
- Independent Perspectives on Cislunar Depotization - August 26, 2022
- Since the launch price is more based on what the market will bear, and will include at least some profit in addition to the actual costs of the given launch.
- Though I can definitely say they did not put me up to writing this blog post. Hopefully I don’t give any of them a heart attack by suggesting this.
Does SpaceX face antitrust problems if they exploit their launch capabilities internally to compete with other LEO players?
Paul,
Not being a lawyer, I have no idea if that argument would stand up in court or not.
Jon
Do we know if Boeing as a partner in ULA receives discounting on their launches which would give them a similar advantage when launching such a constellation? They selected ULA as their launch provider for CST-100 so I figure there has to be some level of discounting occurring.
Intriguing points. I also wish someone would pour some of that exciting LEO-constellation money into scrappy companies like Masten, as long as they could stay true to their roots. I don’t particularly like the idea of satellite providers buying out any of the folks building humans-to-space launchers, though, because it’s too easy to mothball your people program in favor of sweet sweet satellite cash. (cough xcor cough)
Is the cost-vs-price difference such a slam-dunk, though? I agree that SpaceX can charge themselves cost, but doesn’t the opportunity cost of pushing off paying customers’ flights count as well? On the flip side, how does a deal like OneWeb’s with Virgin Orbit compare, when the operator has a vested interest in keeping the launch provider afloat and the provider can afford a steep discount to book as many launches as possible in a big block? Obviously you still can’t do better than the launch provider’s cost, but do they have to be purchased outright?
CA,
Boeing might in theory be able to work such a deal with ULA (basically foregoing their part of the profits from the launch to keep costs down), but even with Vulcan, I’m not sure if that would all the way compete with SpaceX for LEO launch. I could be wrong, and it would help, but it’s not clear whether or not that will be sufficient.
~Jon
Chris,
I’m pretty sure XCOR’s backburnering of Lynx was out of desperation and survival instinct, not distraction. Their RLV ended up costing more to develop, and fundraising went too slow, so they had to raise more money than they realistically could for the space tourism market they were pursuing. IMO, YMMV, etc, etc.
I do think the cost vs price difference can matter a lot in determining commercial viability. Could a megaconstellation developer still stay in the game while ignoring my suggestion? Sure. But I think they’d have a better chance if they at least made a strategic investment in a Masten or someone else pursuing reusability.
~Jon
Similar to Chris’ point, it seems that a lot of the focus of “commercial space” is on constellations for Internet and Earth imaging. With the seduction to head down that path, I’m concerned that the desire for lunar development on the part of Masten and perhaps Blue could be diverted while those dollars are pursued. e.g. the Xeus terrestrial demonstrator sits in Masten’s warehouse collecting dust. It’s like how GEO comm satellites are great and all but it didn’t necessarily lead to HSF. I’m not worried that SpaceX will be so seduced. Elon stated that the purpose of their satellite constellation company in Seattle is to fund the development of a Martian colony.
Addendum: It’s not just the secuction of the companies that I’m concerned about. In particular, in NewSpace policy circles the trend, as I see it, is that commercial space in all its forms is promoted whether it has a direct connection to BLEO HSF or not. So, for example, I could see that being the motivation for NASA funding for the transition to a commercial LEO station for commercial purposes that don’t particularly advance the ball when it comes to off-Earth settlement. So, $1-2 B/year for a “gapless transition” to a commercial station means no money for lunar lander development, telerobotic ice-harvesting, FH cross-feed, ACES development, etc.
Though it annoys me to admit it, I haven’t seen a convincing proof that SpaceX isn’t charging customers at or below cost right now, as argued by Bill C. So there may be no significant market advantage to SpaceX on the price of launching their own satellites, just the advantage of the “customer” having more insight than normal into the “supplier” and vice versa, as is standard in vertically integrated companies.
I guess this is me being contrary.
I think whether or not a constellation owner needs to get involved in launch companies depends on where the industry is at the decision time. If it is SpaceX or the expensive ones, it makes good sense. OTOH, if the industry has advanced to the point of true competition, maybe not. If competition has driven prices down to less than say, 10% over cost, then getting involved in launchers may be just a distraction. You don’t buy a trucking company unless you want to be in the trucking business.
I also disagree with the point Doug makes. Comsats and other legitimate space business promotes human spaceflight by keeping the technology active. Zero to manned spaceflight is just too big a jump without solid justification and long term subsidy. Revenue along the way is part of what makes the current progress possible. Distraction by one company is not a halt by all other interested parties. Engineers move from one company to another all the time. That is only possible with multiple players.
Small launchers are competing on flexibility and short payload integration periods. If you’re launching hundreds to thousands of satellites, that doesn’t seem very important. What you want is scaling, which implies putting as many satellites as possible on a single launcher.
That would seem to give a huge advantage to SpaceX, if they can get Falcon Heavy operational. It’ll need a bigger fairing and perhaps more stage 2 restartability on-orbit, but both of those seem to be pretty modest evolutions.
If you arm-wave 30t to LEO to recover all three stages and still have enough delta-v to do a couple of plane-changes to place a lot of satellites in interesting orbits, and you assume that an FH can reuse first stages 15 times, you get a launch cost of about $19M/launch (this based on a slightly more realistic version of the Jefferies analysis from last year), which gives you $633/kg.
The orbital lifetime section of the SpaceX FCC application lists the satellite mass at 386 kg, which would mass-limit the number of sats to 77. However, their dimensions are 4.0 x 1.8 x 1.2 m, which (if you assume a rectangular prism) would give a volume of ~8.7m^3. The usable volume inside the current fairing probably isn’t more than 120 m^2 (it’s 145 m^2 as a theoretical maximum), so they’re volume-limited to 13 sats, which takes the launch cost up to $3786/kg–unless they do a bigger fairing.
I’d think that they could probably get away with a 6m diameter fairing that was another 2m high pretty easily for the FH. The 3-stick first stage is going to move the center of pressure back quite a bit, so I’d think they could tolerate the extra 7 m² of aspect. That would take them up to about 230 m^3 of usable volume, which would accommodate 26 sats, taking launch cost down to $1893/kg.
At either of those specific costs, it’s hard to see how a small launcher gets within a factor of 10.
New Glenn obviously gets a lot closer, especially with a 7m diameter fairing. And if Elon’s got a satellite constellation, I can’t see Jeff being able to tolerate that without responding. So a merger makes a lot of sense. (Of course, a merger with ULA also makes a lot of sense…)
Interesting pointing out single small sat launcher companies as acquisition targets. A megaconstellation raw launch pace demands are harsh, which would behoove a full RLV, but the refurb time means you have a fairly large fleet in various stages of refurb to keep up with the pace.
Air launch boost forward means you could have first stages return to your launch ops location, which helps the refurb pace. So VG and GG (and Stratolaunch…) are probably not ruled out (downrange first stage recovery would kill your refurb pace due to return logistics). Ventions/Vector/Rocketlabs currently don’t have first stage recovery roadmapped but could if they brought in landing skills. Masten covers the landing well but have yet to go higher.
You would almost desire something like a mini Kistler-K1 form factor, with RocketLab’s Rutherford electric boosted engines, and Masten landing skills. Though you would need to do a swoop-of-death move to land the second stage propulsively after reentry.