SpaceX Fundraising Exactly Covers 800 Satellites for an Operational Starlink Service

SpaceX is raising $500 million at a $30.5 billion valuation to get the Starlink Satellite network deployed.

SpaceX will get money from existing SpaceX investors and new investor Baillie Gifford and Co, who is also the third-largest shareholder in Musk-led electric carmaker Tesla.

How Much Starlink Could You Get With $500 million?

The SpaceX Super Heavy Starship would make the Starlink Satellite network very cheap to deploy.

In December, SpaceX proved that they should be able to reuse the first stage four times or more. Previously they were able to reuse two times. They flew the first stage once and then reused it once. They can now fly it once and re-use three or more times. SpaceX believes they can reuse it nine or ten times and then give them an overhaul and continue using them.

In 2019, let us suppose the average first stage re-use is four times. I think the first stage is about $30 million in cost. All of the Starlink launches would be to low-earth orbit and would be reusable.

$62 million price for SpaceX Falcon 9 block 5 launch. 30% operating margin.
$43.4 million in cost for one-time launch.
$30 million for first stage. A first stage reused four times is $8 million per launch. A first stage that is reused ten times is $4 million per launch.
$6 million for faring or nose cone payload covers. Those are now waterproof and can be reused. $1 million per launch including recovery and clean-up.
$7.5 million for second stage.

Each SpaceX Starlink Falcon 9 launch is $12.5 million to 16.5 million in cost.

A highly reusable Falcon 9 block 5 would cost about $12.5 million for launching 20 Starlink Satellites with each launch. This would be about $1 billion in launch costs for 1600. This is for 80 launches.

Getting to 1600 Starlink satellites would service the northern hemisphere. This is where New York, London, Tokyo, Paris, Zurich, Shanghai and other world financial capitals are located.

SpaceX anticipates starting service with around 800 satellites in 2020 or 2021. This is 800 Starlink satellites using 40 launches. This is the $500 million SpaceX is raising.

The numbers from Elon Musk tweets about the first stage and faring costs and the fundraising all tick and tie. Tick and tie is corporate finance jargon.

This is enough to get SpaceX to a lot of revenue. This would not be the full revenue from the full first phase of 4425 satellites or after both phases with 12000 satellites. It would still billions per year. It could be enough to pay for the remaining Starlink launches or to build the Super Heavy. As they get close, SpaceX can go back and raise a few billion at a $100 billion valuation.

Minimum Starlink Before the SpaceX Super Heavy

It would only take seven launches of the Super Heavy Starship to deploy the first 1600 Starlink satellites.
This would be about $70 million in launch cost. However, the Super Heavy will cost at least another $2 billion to develop. The SpaceX Super Heavy will not be flying payload to orbit until about 2022.

Falcon Heavy and Orbital Planes

I have seen estimates of the 20 satellites per Falcon 9 launch and the 200 for the Super Heavy. I have comment critique that this is too optimistic. It could be optimistic especially based upon fuel, reuse and which orbit different satellites need to be placed.

The Starlink satellites are not finalized and neither is the Super Heavy. My assumption is they will try to make the first 800 or even 1600 Starlinks operational before the Super Heavy. I think SpaceX moves forward on Starlink as much as possible without Super Heavy. The estimates could indeed be high based upon how many can be launched in one and still return to land for reuse. There is different fuel needed to get to different orbital planes.

Some tough orbits may need more costly Falcon Heavy launches. Falcon Heavy’s would be more reusable than a F9. Three first stages and faring are reused in the Falcon Heavy. If SpaceX masters Heavy reusability, then actual SpaceX costs are not badly impacted using the Heavy. Costs can even improve as Heavy launches can put up about double the satellites of an F9 launch. I did not want to go into the variance on orbital planes or using Heavies. I do not think those things throw off the analysis that much.

The Falcon Heavy reuses three first stages. This would be $12 million for the ten reuse assumption. The $7.5 million second stage is thrown away and we have the $1 million reusable faring. This is $20.5 million with perhaps double the satellites. The fuel costs are only about $200,000 to $400,000 per launch. The Falcon Heavy can reach tough orbits. The satellites will not need to use up a lot of fuel to reposition.

SpaceX might even only get 600 satellites of the operational 800 before they have to dip into assumed profits in 2019 and 2020. Alternatively, Spacex could need a new cash raise or loan at a higher valuation at some point.

22 thoughts on “SpaceX Fundraising Exactly Covers 800 Satellites for an Operational Starlink Service”

  1. As it turns out, your 20 number for a Falcon 9 heavy was WAY ON THE LOW SIDE.
    60 sats launched in the first semi-production load, and the plan appears to be for 60 more per launch in the future.

  2. Try 60 Starlink satellites per launch… I didn’t see you adding the cost for the satellites themselves.

  3. That is a neat trick. I was thinking something like that could be done, but I hadn’t figured it out. No expert here. Learned something today. Thanks.

  4. You can’t get to any of the orbital inclinations directly from Boca Chica. They would require overflight of heavily populated areas well within 1500 km downrange. You have to launch to about 33.2 degrees (down the Yucatan Channel), refuel, and do a plane change. When you amortize the refueling launches, I get between 30 and 86 birds per launch.

    Of course, if SpaceX changes its mind and builds a BFR pad at Canaveral, that’s a different story. But so far they have no public plans to do that.

    For FH launched from the Cape, I get between 53 and 62 birds per launch on mass to orbit, but that would require stretching the fairing to Category C size (about 16 x 5 m). I kind of expect them to do this–USAF would be very happy with a Cat C FH. I can make 30 fit in the current fairing if they’re only 1.3 m in length. (Deployed length is 4 m, but if 2/3 of that is folded up as antennas…)

    Note that there’s a neat trick that removes the need to fill orbits with an integral number of launches: You just deploy into a somewhat lower orbit, and the orbits will precess faster at the lower altitude. When you get to the proper RAAN, you boost up to the operational altitude. As an example, launching the birds destined for 550 x 53 degrees into 400 km x 53 will have them precess by the 15 degree orbital plane spacing in about 6 weeks.

  5. I pretty much agree that $500M will cover the first 800-1000 birds, which should get them operational. But of course I have some quibbles:

    1) I have a real hard time making the fairing cost $6M in my model ( https://docs.google.com/spreadsheets/d/1vYJyhajR3EcGVm0jRoatsU5YJqQsMrStMsAUpFAOFjk/edit?usp=sharing ). However, if you use 50% gross margin, then the fairing actually costs about $3M, which is pretty close to what I get.

    2) I get about $370M to build and launch the first 800 birds, assuming that the manufacturing cost is $100K/bird. That excludes R&D and ops money, so $500M is about right. But I’m using 50% gross margin, not 30%.

    3) I know what you were trying to say with this paragraph, but it’s wildly misleading:

    “Getting to 1600 Starlink satellites would service the northern hemisphere. This is where New York, London, Tokyo, Paris, Zurich, Shanghai and other world financial capitals are located.”

    Everything is accessible in the Southern Hemisphere as well. The polar Starlinks are well within the launch capacity of F9s operating out of Vandenberg. With those, you have the whole world covered–the 1275×81 degree orbit is well above the horizon even at the poles. After you have those polar birds up, the only real constraint is throughput.

  6. I have seen estimates of the 20 satellites per Falcon 9 launch and the 200 for the Super Heavy. Yes the Starlink satellites are not finalized and neither is the Super Heavy. My assumption is they will try to make the first 800 or even 1600 Starlinks operational before the Super Heavy. I think SpaceX moves forward on Starlink as much as possible without Super Heavy. The estimates could indeed be high based upon how many can be launched in one and still return to land. There is different fuel needed to get to different orbital planes. Some tough orbits may need more costly Falcon Heavy launches. Falcon Heavy’s would be more reusable. Three first stages and faring reused. Actual SpaceX costs are not that badly impacted using the Heavy if they master Heavy reusability. Heavy launches also can put up about double the satellites. I did not want to go into the variance on orbital planes or using Heavies. I do not think those things throw off the analysis that much.

    They could even only get 600 satellites of the operational 800 before they have to dip into presumable profits in 2019 and 2020 or they do a new cash raise or loan.

  7. Let’s see… SpaceX was never going to be able to get anything into orbit. And they could never recover a booster. If NASA couldn’t manage it, how could SpaceX?

    Heh.

  8. Starlink is planned to have 50 or 75 satellites per orbital plane. Most orbital planes will have 50. With its great lift capacity and payload volume Super Heavy Starship might be able to launch all 50 or 75 satellites plus maybe some inactive orbital spares into each orbital plane in a single launch. Assuming 50 satellites per launch that is 32 launches for the first 1600 satellites not 7 as you claim. There is a similar flaw in your math if launching on falcon 9. You are underestimating the number of launches that will be needed. Plane changes in low earth orbit require a lot of delta v. You cant simply divide the total number of satellites to be deployed by the number that can fit in the payload shroud. We don’t even know the payload bay dimensions of the Super Heavy Starship yet.

  9. What type of antenna will people need to receive starlink? Since the satellites are constantly moving does that mean they need a dish with a tracking motor built in?

  10. Although I am sure there are some, how many major Elon Musk projects have failed? And how many times did major players say they would fail? 🙂

  11. I’m getting the vibe of risk taking here like old school SpaceX, but with a little bit less to chop off, if you know what I mean. Hope they pull it off.

  12. Seems like a progressive financing scheme.

    They get some money, launch a batch of satellites, prove the capability and then they get more to complete the set.

    Risky, but I assume they are confident they can deliver it in parts.

  13. Nope $500M won’t get 800 state-of-the-art broadband comm sats with interlinks into the sky. Iridium Next with 64 LEO sats cost $2.9B for a service that is about 700kbits/sec (far less than broadband). While I expect some real cost savings from 10 time F9 reuse, lighter sats (1/3-1/2 mass) and less expensive sats with such a large run I don’t see how they can build and place 800 for less than $2B. There seems to be a $20M per F9 launch non-flight hardware and labor cost that gets lost. With a lot of identical payloads in a row you might be able to cut that down … but not less than 1/2. Basically you can’t have your Starship/Super Heavy and your Starlink at the same time without $5B in outside money soon.

    Recall that these LEO high bandwidth Ku band services like Starlink are subject to occasional rain fade/blackouts (like on Dish network) and probably have the rare sun-behind blackouts when inside +/- 23 deg lat. This is fine for many applications but would require high frequency traders to maintain their existing capacity as backup.

    Also, you state that “Getting to 1600 Starlink satellites would service the northern hemisphere. ” The constellation has to be symmetric to the North and South hemispheres with each sat spending exactly 1/2 of it’s time in either. With many more customers in the North Hemisphere one might argue that the Southern Hemisphere will have more than enough coverage as well.

Comments are closed.