SpaceX progress towards a 100 launch year

[Robotbeat]

  Has anybody figured out the Starlink payloads if they did RTLS?
 Any chance the lost sats could be more than made up for by more frequent launches?
 I'd guess they'll have to go that route in any case if a barge ever goes down for repair or scheduled maintenence/inspections.

It might actually make more sense to go the other way and start launching Starlink satellites on Falcon Heavy with 3 RTLS boosters.

Are there pads to land three RTLS boosters?

The pads are so huge they could fit all 3 boosters on one of them.

[Robotbeat]

I just did a quick calculation with Silverbird Astronautics’ LV calculator, and it gave me about 28t payload with all 3 cores recovered RTLS and the long fairing to the 525km, 43deg inclination launch for V2 minis, allowing 35 v2 minis per launch and no droneship needed.

36.5t with ocean recovery for the core. 45.3t with expendable core.

Now I don’t think spaceX is really working on the octograbber mods needed for droneship recover of the center core anymore, but they could do RTLS recovery still I think. I’m not sure that you can fit 45t of v2Minis even in the long fairing, but probably can fit 28t of them in the long fairing.

(41.4t for ocean recovery of all 3 cores. Probably the cheapest way to launch payloads per kg until Starship comes online as the only thing you’re expending is the upper stage but you get 2.7 times the performance—droneship recovery for block 5 F9 to this orbit is 15.25t according to this calculator—but they’d need to build more droneships. Might be able to halve the cost per kg to orbit this way, from the current $1500/kg down to about $750/kg.)


https://silverbirdastronautics.com/cgi-bin/LVPcalc.pl

[redneck]

  Has anybody figured out the Starlink payloads if they did RTLS?
 Any chance the lost sats could be more than made up for by more frequent launches?
 I'd guess they'll have to go that route in any case if a barge ever goes down for repair or scheduled maintenence/inspections.

It might actually make more sense to go the other way and start launching Starlink satellites on Falcon Heavy with 3 RTLS boosters.

Returning the center core to land doesn't make much sense. I order to make the way back, you are hard limited on speed and distance at MECO. Even a Mega-Heavy with 10 side boosters would still have to let the upper stage go at 1900m/s or so on a heavily lofted trajectory, and that means 15t of LEO payload.

That isn’t the upper limit. The upper stage of Falcon has a wet mass of about 116t tonnes, the payload being on the order of 20-25t. The dry mass of a booster is about 27t. I think with a Falcon 9 RTLS payload is about 12t now, so a single booster is burnout at 12+116=128. 3 boosters, even without sequential staging, would be able to push a booster to the same delta-v that had a mass of 384t, so that’s 243t of extra propellant to work with if the payload is 25t instead. Falcon upper stage is about 4.5t burnout.

Anyway, I think 3 core Falcon Heavy could probably do about 2-3 times the RTLS payload of a Falcon 9, particularly with sequential staging and maybe an upper stage stretch.

There was a thread a few years backs on using a single Raptor expendable center in the FH configuration.  Has anyone run numbers on such a thing recently?   I seem to recall that it made for a very large high energy upper stage.    Seem to recall being the problem. 

[whitelancer64]

There was a thread a few years backs on using a single Raptor expendable center in the FH configuration.  Has anyone run numbers on such a thing recently?   I seem to recall that it made for a very large high energy upper stage.    Seem to recall being the problem.

The main problem is that would turn the center core into a brand new design, which would also require new GSE equipment on the pad to fuel and launch it.

[niwax]

I just did a quick calculation with Silverbird Astronautics’ LV calculator, and it gave me about 28t payload with all 3 cores recovered RTLS and the long fairing to the 525km, 43deg inclination launch for V2 minis, allowing 35 v2 minis per launch and no droneship needed.

I'm not sure how that page gets its numbers, but that would increase staging velocity by 1300m/s, from ~6100m/s to go for S2 to ~4800m/s. You're not turning around after that. That's a 70% increase, not to mention being far further downrange than even the furthest droneship landings. Arabsat 6A staged just short of what you'd need for 28t even running S2 completely dry.

[Robotbeat]

I just did a quick calculation with Silverbird Astronautics’ LV calculator, and it gave me about 28t payload with all 3 cores recovered RTLS and the long fairing to the 525km, 43deg inclination launch for V2 minis, allowing 35 v2 minis per launch and no droneship needed.

I'm not sure how that page gets its numbers, but that would increase staging velocity by 1300m/s, from ~6100m/s to go for S2 to ~4800m/s. You're not turning around after that. That's a 70% increase, not to mention being far further downrange than even the furthest droneship landings. Arabsat 6A staged just short of what you'd need for 28t even running S2 completely dry.

So… Falcon 9 gets over 12.5tonnes to orbit even with RTLS. They do this via a lofted trajectory, so you cannot use long range droneship trajectories as a baseline.

I’d trust those numbers on that calculator more than I would “you’re not turning around after that, trust me bro.”

A Falcon 9 booster is essentially a SSTO capable stage, with several tons of payload. You can get back after a pretty aggressive launch.

[redneck]

There was a thread a few years backs on using a single Raptor expendable center in the FH configuration.  Has anyone run numbers on such a thing recently?   I seem to recall that it made for a very large high energy upper stage.    Seem to recall being the problem.

The main problem is that would turn the center core into a brand new design, which would also require new GSE equipment on the pad to fuel and launch it.

That’s the main problems, plus the unk -unks.  Assuming Starship takes long in development, I wonder if it’s a theoretical win.  I may try to go hunt up that discussion this weekend time permitting. 

[xyv]

Time for half year update.  Normally I would have posted yesterday with 43 launches but since there was a launch today and it is still in the first half of the year I included it.  The slope of the linear fit intercepts the end of the year at just over 87 launches.  This has been slowly creeping up from about 84-85 over most of the first part of the year.

Last year at this time we had 27 launces so Space X managed to increase that to 34 in the second half of 2022.  The same feat this year would  mean 55 launches in the second half for a total of 99 launches in 2023.

As usual, I have posted the updated figure and at the top of the thread along with the updated spread sheet.  One other interesting figure is 2022-2023 plot that "looks" like an uptick in rate at just before the end of last year - I'm sure Robotbeat will argue it looks like an exponential 

[meekGee]

Hey, for all the "flat liner" predictors:

I'm asking this without looking at the data first.

When was the last time that a half year had less than a 10% increase over the previous half, in terms of number of launches?

I do remember that when they hit 60 in 2022, the second half had more launches than the first.

Definitely first half of 2023 had more launches than second half of 2022.

How far back does this go?

While I agree, at a point they will be limited by the ASDS cycle times, range availability and weather.  The further they go the closer they get to those barriers pushing back.

I think they can hit 100 but some luck will be required.

Neither of us knows where the limit is, but assuming a fixed rate has so far proven to be the poorest of predictors, so why jump the gun?

For all the reasons RB has posted here time and time again, and for all the reasons he hasn't, using "fixed rate" as a predictor is not a conservative approach, it's just a silly one.

[Perchlorate]

Could have posted this in a number of places, but...

"SpaceX, it's time to update https://www.spacex.com/vehicles/falcon-9/.  It shows '234 total launches, 192 total landings and 168 total reflights'."

Obviously not a high priority, and the fact that they don't update it after every launch shows how busy they are, pushing forward in many areas.  But I thought they might have gotten it after "landings" passed the big 200 milestone.

I'm surprised nobody called me out on this total brain fart.

SpaceX is doing a perfectly fine job of keeping their website up to date.

I was just failing to add together the F9 and FH landing totals!

[meekGee]

Could have posted this in a number of places, but...

"SpaceX, it's time to update https://www.spacex.com/vehicles/falcon-9/.  It shows '234 total launches, 192 total landings and 168 total reflights'."

Obviously not a high priority, and the fact that they don't update it after every launch shows how busy they are, pushing forward in many areas.  But I thought they might have gotten it after "landings" passed the big 200 milestone.

I'm surprised nobody called me out on this total brain fart.

SpaceX is doing a perfectly fine job of keeping their website up to date.

I was just failing to add together the F9 and FH landing totals!

This is a charitable forum. We let people digest and introspect before piling on them.

[rfdesigner]

Time for half year update.  Normally I would have posted yesterday with 43 launches but since there was a launch today and it is still in the first half of the year I included it.  The slope of the linear fit intercepts the end of the year at just over 87 launches.  This has been slowly creeping up from about 84-85 over most of the first part of the year.

Last year at this time we had 27 launces so Space X managed to increase that to 34 in the second half of 2022.  The same feat this year would  mean 55 launches in the second half for a total of 99 launches in 2023.

As usual, I have posted the updated figure and at the top of the thread along with the updated spread sheet.  One other interesting figure is 2022-2023 plot that "looks" like an uptick in rate at just before the end of last year - I'm sure Robotbeat will argue it looks like an exponential 

good work.

I downloaded your excel data, worked out an exponentially weighted rolling average (so the daily average isn't impacted by launches dropping out of the other end of the window).

I then used the predictive trendline to guesstimate the launch rates for the rest of the year.

Month, trend Launches/month
july   8.1
august   8.5
september   8.8
october   9.2
november   9.6
december   9.9
total   54.1

That would make 98 for the year...   a LOT of assumptions included there, but that seems to be where the data is headed.  Very much in alignment with what you were saying about 55 launches for H2.

[xyv]

Here is another way of looking at a linear projection.  A rolling linear fit projects an end of year total and the plot of that versus launch date is a heavily filtered end point.  While Excel provides a function to do curve fitting, it does not seem to work with sparse data - unlike the same function in the plotting routines.  This means I have to update it as x-y pairs but it still gives a very intersting comparison from last year.

Note that as last year progressed the offset part of the fit became increasingly negative due to the backloaded increase that was observed.  The net result is the linear fit for the entire year intersected at 59.6 launches versus 61 actual.  the filtering of the curve fit shows the 'inertia' of the goal and the asymptotic approach to it after mid year.

[Comga]

And here, on just another random day, are other ways to look at SpaceX's prospects for achieving 100 launches in 2023:
Four data sets.
The first is launches to date this year.  It's up to 48.  Linear extrapolation to the year end is 89.
The second is launches in the past 365 days.  That's up to 79, where it looks like it's stalling. However, linear extraploation of that gets to 97.
The third takes that 48 launches so far, and assumes the average rate of the last ten launches, which is now 87, down from a peak over 110.  That also projects to 89, down from around 100 in May and June.
The fourth is a linear extrapolation of the launch rate in 2023.  Having started the year with a bang, this shot off the scale, but is now projecting around 110 launches in 2023.
Take your pick.  I still think launching 100 times is a close call.

[Bob Shaw]

A lot depends on the weather over the next few months. While SpaceX certainly looks like it has the ability to launch around a hundred times (let’s not quibble if they just hit 97!) they still mostly depend on the weather to reach this target and indeed are more sensitive to weather conditions than other operators because most F9 flights demand decent conditions both in Florida *AND* hundreds of miles out in the Atlantic. Nobody else works under these unforgiving circumstances (yet).

[meekGee]

My quick tracking estimate was:
100/yr requires 8/month on average, plus 4 some of which will be Starship.

Since we started with 7, then for every 7 we need a 9.

I was hoping for 7,7,7,7,8,8,8,8,9,9,9,9.

By mid year it had to be:
7,7,7,7,7,7,9,9,9,9,9,9.

Right now it'll have to be:
7,7,7,7,7,7,8,9,9,9,9,10

They need to be at 9 asap, It's getting harder by the week.

(Soon the usual suspects will throw a fit how they're sick and tired of missed goals...)

[Robotbeat]

My quick tracking estimate was:
100/yr requires 8/month on average, plus 4 some of which will be Starship.

Since we started with 7, then for every 7 we need a 9.

I was hoping for 7,7,7,7,8,8,8,8,9,9,9,9.

By mid year it had to be:
7,7,7,7,7,7,9,9,9,9,9,9.

Right now it'll have to be:
7,7,7,7,7,7,8,9,9,9,9,10

They need to be at 9 asap, It's getting harder by the week.
...

Starting at 7 and ending at 10 is just a smooth exponential curve similar to the curve they got last year. It's expected, unless they stop trying to improve launch rate (in which case the launch rate will plateau) or have a Falcon 9 launch failure.

I get an estimated number of launches of 102-103 as of today, assuming gradual, compounding improvements (no sharp increases anywhere).

This estimate has not changed more than about 1% in the last 2 months. Meanwhile, linear extrapolations keep changing over time.

I expect around 100 launches per year as the mode or median, but the "average" value would be lower than that as there's still a non-insignificant chance of a Falcon failure (say, 10-25% cumulative through the end of the year), which would reduce the "average" to about 90 or so, mitigated by the fact that the failure is slightly more likely to occur later in the year and that if it occurs earlier, they may get back to launching in a couple months.

[meekGee]

My quick tracking estimate was:
100/yr requires 8/month on average, plus 4 some of which will be Starship.

Since we started with 7, then for every 7 we need a 9.

I was hoping for 7,7,7,7,8,8,8,8,9,9,9,9.

By mid year it had to be:
7,7,7,7,7,7,9,9,9,9,9,9.

Right now it'll have to be:
7,7,7,7,7,7,8,9,9,9,9,10

They need to be at 9 asap, It's getting harder by the week.
...

Starting at 7 and ending at 10 is just a smooth exponential curve similar to the curve they got last year. It's expected, unless they stop trying to improve launch rate (in which case the launch rate will plateau) or have a Falcon 9 launch failure.

I get an estimated number of launches of 102-103 as of today, assuming gradual, compounding improvements (no sharp increases anywhere).

This estimate has not changed more than about 1% in the last 2 months. Meanwhile, linear extrapolations keep changing over time.

I expect around 100 launches per year as the mode or median, but the "average" value would be lower than that as there's still a non-insignificant chance of a Falcon failure (say, 10-25% cumulative through the end of the year), which would reduce the "average" to about 90 or so, mitigated by the fact that the failure is slightly more likely to occur later in the year and that if it occurs earlier, they may get back to launching in a couple months.

I know.

At this resolution though, it's really hard to tell the difference between different curve types, since the individual variances are too high.

What's expected is a generally monotonous increase, and right now it's a bit overdue.

I would be surprised if they stay at 7 per month.  I'd expect it to go up to 10 with time, but whether it'll happen fast enough to hit 100 this year is TBD.

It'll be hella cool if it happens, but honestly suppose they hit 100/year only in the last 6 months - I'd be ok with that too

[DanClemmensen]

What's expected is a generally monotonous increase, and right now it's a bit overdue.

Yes, math is boring, but I think you meant "monotonic".  And also yes, the difference between linear and exponential is currently below the level of the variance.

[Robotbeat]

What's expected is a generally monotonous increase, and right now it's a bit overdue.

Yes, math is boring, but I think you meant "monotonic".  And also yes, the difference between linear and exponential is currently below the level of the variance.

Only if you arbitrarily ignore the previous year.