Why Aren't All New Very High Throughput GEO Satellites Using Falcon Heavy?

[TheRadicalModerate]

The recent Eutelsat launch got me wondering:  If you're launching a Very High Throughput Satellite (VHTS) to GEO, why don't the economics of the VHTS make using an FH2R (two-stick reusable) a no-brainer?

Eutelsat apparently shaved about 10 days off of a 5-6 month circularization campaign by expending the F9.  If they'd used an FH2R direct to GEO, they would have shaved 5-6 months off of that circularization campaign.  Why didn't they?

I can't lay my hands on the exact price for an FH2R after the price hike.  I'm assuming its about... $115M?  And the F9R is $67M. 

So we can get our VHTS commissioned 5 months earlier on an FH2R than an F9R for about $48M more.

VHTS birds these days have more than 500Gbps of throughput.  That means that it should be more advantageous to use the FH2R over the F9 if the revenue generated by the VHTS is worth more than about $20/Mbps/month.  That's well under the going rate, isn't it?

Some possible explanations:

1) The cost/Mbps/month is a lot lower than I think it is, and the business case doesn't close.

2) SpaceX might have been squirrelly about scheduling FH flights soon enough to be attractive.  (The biz case is based on how long it takes from when the bird is ready to be integrated to when it's commissioned, not from when it's launched, so schedule risk is a big deal.)

3) The mass of a VHTS is too high for an FH2R to put it directly into GEO.  But even if it can't get there directly, it ought to be able to get it down to a month or so to circularize.

4) Lead times on VHTS birds are so long that FH wasn't a safe option when they made design decisions that locked in the launcher.

5) Everybody's waiting for Starship.  I tend to discount this, since Starship can only get to GEO if it's either expendable or refueled.

Anybody know what's going on?

[vaporcobra]

The easiest explanation is that SpaceX demonstrated its direct GEO launch capability for the first time in November 2022. Now knowing that it works as advertised, wouldn't surprise me if we see an increase in direct-to-GEO FH contract announcements.

And despite the above, ViaSat and EchoStar were already planning to launch two VHT sats direct to GEO months prior to USSF-44. So I suspect that what you've described is actually already happening, at least for the largest and most expensive satellites.

Edit: and FWIW, Eutelsat 10B is only 35 Gbps. I think the value proposition is much rockier a magnitude below Jupiter 3 and ViaSat-3.

[gongora]

Eutelsat 10B is not a VHTS.

[edkyle99]

The recent Eutelsat launch got me wondering:  If you're launching a Very High Throughput Satellite (VHTS) to GEO, why don't the economics of the VHTS make using an FH2R (two-stick reusable) a no-brainer?

Because the Fukushima Hydrogen Energy Research Field (FH2R) is not a launch vehicle?

 - Ed Kyle

[TheRadicalModerate]

The easiest explanation is that SpaceX demonstrated its direct GEO launch capability for the first time in November 2022. Now knowing that it works as advertised, wouldn't surprise me if we see an increase in direct-to-GEO FH contract announcements.

The thing is, though:  The 2018 test launch did an engine restart at T+6 hours.  That's long enough to do a GEO insertion from GTO apogee.  So there really shouldn't have been much doubt about the ability to do the maneuver.¹

Edit: and FWIW, Eutelsat 10B is only 35 Gbps. I think the value proposition is much rockier a magnitude below Jupiter 3 and ViaSat-3.

Fair point, and I agree.  I think the fact that expending a core just got a 10 day reduction got me thinking that using an FH2R would get you a lot more than that.

We know that the center core is a sturdier (heavier) structure than the side boosters, which are almost generic.  Do we know whether the center core is a completely different build, or can the reinforcement be added with some kind of a "kit"?  ISTM that SpaceX probably wants to plan FH cores pretty far in advance if they have to be built from scratch.

_____________
¹Weasel words:  Presumably the burn to C3>0 occurred at perigee, so the "VA Belt Experiment" orbit wasn't GTO (a 10.5-hour orbit), but a 6-hour orbit, which would make it something like 200 x 20,500km.  As such, the S2's kerosene probably wasn't exposed to temperatures as cold as they would be at GTO apogee.  So there was still a little uncertainty. 

Still, it's pretty good odds, and we're potentially talking about >$100M in extra revenue--and that's just on the front end.  On the back end, you've got a lot more prop available, which will usually extend the bird's life, which is still more revenue to feed into the IRR calculation.

[DanClemmensen]

We know that the center core is a sturdier (heavier) structure than the side boosters, which are almost generic.  Do we know whether the center core is a completely different build, or can the reinforcement be added with some kind of a "kit"?  ISTM that SpaceX probably wants to plan FH cores pretty far in advance if they have to be built from scratch.

Just looking at the booster build history and with no inside info whatsoever, it appears that FH cores are built to order. They are available for use on the originally-scheduled launch date, and then sit in inventory until the payload is actually ready to launch after a lengthy slip (in 100% of the cases so far). There is no evidence that an FH core can be converted to a standard F9/FHside  of vice versa.
    https://en.wikipedia.org/wiki/List_of_Falcon_9_first-stage_boosters

[Barley]

Doesn't direct to GEO leave the rocket as space junk somewhere near GEO, which really ought to be discouraged?

(Or as space junk in a graveyard orbit, which is still pretty close to GEO).


Also
Another way to look at the shorter transfer time is not that you get your satellite in place 6 months earlier, but that you can start building the satellite 6 months later.  The savings are the six months of carrying cost, which were quite low during the zero interest rate period.  A lot on how far in advance you can predict demand and exactly what the short pole in the tent is.

[hkultala]

Doesn't direct to GEO leave the rocket as space junk somewhere near GEO, which really ought to be discouraged?

(Or as space junk in a graveyard orbit, which is still pretty close to GEO).

Leaving space jump in graveyard orbit above GEO is not a problem.
The space is BIG and air resistance is practically non-existent at those altitudes, it will stay there in the graveyard orbit longer than there are humans alive on the planet surface.

[Barley]

Doesn't direct to GEO leave the rocket as space junk somewhere near GEO, which really ought to be discouraged?

(Or as space junk in a graveyard orbit, which is still pretty close to GEO).

Leaving space jump in graveyard orbit above GEO is not a problem.
The space is BIG and air resistance is practically non-existent at those altitudes, it will stay there in the graveyard orbit longer than there are humans alive on the planet surface.

A rather pessimistic view of the future of humanity. 

Graveyard orbits are not fully closed and wander secularly due to effects such as the Moon's gravity.  Sooner or later they will start bumping into each other and generate an expanding band of debris.  How soon that happens is subject to debate, and depends on how much material you put there, but it is not geological time scales.

[TheRadicalModerate]

Graveyard orbits are not fully closed and wander secularly due to effects such as the Moon's gravity.  Sooner or later they will start bumping into each other and generate an expanding band of debris.  How soon that happens is subject to debate, and depends on how much material you put there, but it is not geological time scales.

I'd be pretty comfortable assuming that the graveyard will be stable long enough that some future technology will be able to clean it up.  If not, it probably doesn't say good things about our future need for GEO satellites.

Another way to look at the shorter transfer time is not that you get your satellite in place 6 months earlier, but that you can start building the satellite 6 months later.  The savings are the six months of carrying cost, which were quite low during the zero interest rate period.  A lot on how far in advance you can predict demand and exactly what the short pole in the tent is.

Irrespective of interest rates, you're always going to get better IRR if you minimize the time from when the bird is ready for integration and when it goes into service.  Anything that shortens that time (integration and encapsulation efficiency, launch inventory, launch slots, and time to get on station) improves your return.  But as carrying costs rise, it becomes even more important.

I guess you could make a case that carrying costs have been so low for the last ten years that nobody cared very much.  If so, that's over now.

It is also true that the rate at which GEO birds have been launched in the past few years is quite a bit lower than it was in the mid-teens.  That's certainly another reason we haven't seen a flood of FH launches.

[Barley]

Another way to look at the shorter transfer time is not that you get your satellite in place 6 months earlier, but that you can start building the satellite 6 months later.  The savings are the six months of carrying cost, which were quite low during the zero interest rate period.  A lot on how far in advance you can predict demand and exactly what the short pole in the tent is.

Irrespective of interest rates, you're always going to get better IRR if you minimize the time from when the bird is ready for integration and when it goes into service.  Anything that shortens that time (integration and encapsulation efficiency, launch inventory, launch slots, and time to get on station) improves your return.  But as carrying costs rise, it becomes even more important.

Shortening the time only guarantees a higher IRR if it comes for free.  If it costs $60million extra for six months faster you have to do the calculation.