SpaceX Manifest Discussion Thread

[Jansen]

I looked up on US launches posted by Salo a few hours ago.  SpaceX has 39 scheduled for 2021.  Two of which are orbital Starships.  There were 2 or 3 Falcon Heavies, don't remember exactly right now.  This is more than all others combined.  Most of which are Starlink launches. 

How many used boosters are still in use?  How many new boosters will they make?  I know they have to make new FH cores if they can't manage to save them.

SpaceX is targeting 48 launches for 2021. The priority is NSSL and NASA launches, followed by commercial launches. Starlink will launch in any available gaps.

B1049 - Reprocessing (mid Jan)
B1051 - ASDS unloading (late-Jan)
B1058 - Reprocessing (late-Jan)
B1059 - LZ-1 (early Feb)
B1060 - Turksat 5A
B1063 - Reprocessing (mid-Jan)

B1061 - Reserved Crew-2
B1062 - Reserved GPS III SV05

B1064 FH
B1065 FH
B1066 FH

The plan was for 10 new Falcon 9 first stages in 2020. There is a need for at least 6 new boosters due to FH requirements for 2021, 3 of which are probably already built (64-66).

Where do you get the target to build 10 in 2020?
They have build 1060-1066, with 1066 unseen. That is 7, so either they already also build 1067-1069 (FH too?), or they fell short of their target.
Somehow the test flow in McGregor seems halted for S1’s as none have been seen vertical there since the arrival of presumably 1065.
And no other has been reported being transported. Interesting.

https://spaceflightnow.com/2019/12/20/spacex-poised-to-accelerate-launch-cadence-with-series-of-starlink-missions/

The company says it plans to build around 10 new Falcon 9 first stages in Hawthorne next year. That’s down from around 16 to 18 new first stages that SpaceX manufactured a couple of years ago.

B1064 was seen in McGregor in September. Everything past that is unconfirmed.

There has a been a lot of testing at McGregor, it’s just that there are more restrictions there and people are unable to get close enough for booster confirmation.

A booster was spotted horizontal on Nov 10, and a new booster left Hawthorne and should’ve arrived around Nov 20. That might’ve been B1067.

[Jakusb]

I looked up on US launches posted by Salo a few hours ago.  SpaceX has 39 scheduled for 2021.  Two of which are orbital Starships.  There were 2 or 3 Falcon Heavies, don't remember exactly right now.  This is more than all others combined.  Most of which are Starlink launches. 

How many used boosters are still in use?  How many new boosters will they make?  I know they have to make new FH cores if they can't manage to save them.

SpaceX is targeting 48 launches for 2021. The priority is NSSL and NASA launches, followed by commercial launches. Starlink will launch in any available gaps.

B1049 - Reprocessing (mid Jan)
B1051 - ASDS unloading (late-Jan)
B1058 - Reprocessing (late-Jan)
B1059 - LZ-1 (early Feb)
B1060 - Turksat 5A
B1063 - Reprocessing (mid-Jan)

B1061 - Reserved Crew-2
B1062 - Reserved GPS III SV05

B1064 FH
B1065 FH
B1066 FH

The plan was for 10 new Falcon 9 first stages in 2020. There is a need for at least 6 new boosters due to FH requirements for 2021, 3 of which are probably already built (64-66).

Where do you get the target to build 10 in 2020?
They have build 1060-1066, with 1066 unseen. That is 7, so either they already also build 1067-1069 (FH too?), or they fell short of their target.
Somehow the test flow in McGregor seems halted for S1’s as none have been seen vertical there since the arrival of presumably 1065.
And no other has been reported being transported. Interesting.

https://spaceflightnow.com/2019/12/20/spacex-poised-to-accelerate-launch-cadence-with-series-of-starlink-missions/

The company says it plans to build around 10 new Falcon 9 first stages in Hawthorne next year. That’s down from around 16 to 18 new first stages that SpaceX manufactured a couple of years ago.

B1064 was seen in McGregor in September. Everything past that is unconfirmed.

There has a been a lot of testing at McGregor, it’s just that there are more restrictions there and people are unable to get close enough for booster confirmation.

A booster was spotted horizontal on Nov 10, and a new booster left Hawthorne and should’ve arrived around Nov 20. That might’ve been B1067.

How do you make that nov20 booster 1067?
Boosters being transported have, sofar, almost always been spotted somewhere along the route.
In L2 we have information implying that since 1064 no booster has been vertical on the test stand there, where 1064 is also implied, as I believe there is no photo yet of the booster explicitly showing 64.
So either several boosters have been transported unseen, and tested at McGregor unseen, and in the very short windows in where there were no eyes on the test stand, I have to conclude that 1065 is at McGregor and 1066+ are all still at Hawthorne.

[ZachS09]

The 3 scheduled Falcon Heavy missions for 2021 have been eating up booster production capacity in the last quarter of 2020 and may eat up significant launch operations capacity in 2021. Starlink will have LC-40 mostly to itself for the first half of the year for lack of customer payloads for mid/low-inclination F9 launches, but LC-39A is going to be tied up for chunks of the year, and those FH missions will require both droneships, which will block Starlink missions from the other pad. FH is a bit of a cadence-killer for the Cape operations.

Fortunately, behind those three FH boosters on the production line, there should be badly-needed F9 boosters to replenish the fleet, hopefully by springtime. They need to replace the two life-leader boosters that are approaching well-deserved retirements, and they need to replace the two boosters which were unexpectedly lost in 2020.

To be honest, I'm tired of B1052 and B1053 being stuck in storage for a year and a half. They only have two flights under their belts and have had all those opportunities to get up to where B1049 and B1051 are right now.

[Jansen]

SpaceX is targeting 48 launches for 2021.

Yes, I heard this too. However I doubt they can get close to such target.

You need to look at the launch cadence in November and December after Elon’s site visit. There is a new drive to speed things up with overnight operations, faster payload processing, and more hiring for refurbishment and LV integration.

November already demonstrated that 4 launches a month is doable, and 5 should be sustainable if you consider current pad turnaround times at around 9 days and 12 days.

Don’t forget that 2020 has been full of anomalies  such as COVID-19, several engine issues, and significant launch delays from customers. The target for 2020 was 38, and that would’ve been reachable.

I concur that the current constraint is booster availability, and I believe that booster production has increased as a result of NSSL2, Starlink launches, Artemis, the FH backlog, and recovery losses.

[abaddon]

And as for pads... it was once said SX were aiming for 24hr turn around.

In fact, that goal was reiterated in the most recent launch broadcast.

[smoliarm]

...

From your graph Oct Nov Dec, ballpark average is about 8 days between launches. 365/8 =~ 45 Launches per year.

...

Well, they can launch 10 missions in 3 months, there is no doubt about it.
But classic proportion - which works so nice in arithmetic - just does not work here.
Unfortunately...
Yes, [10 missions / 3 months * 12 months] = 40 missions per year
- that's true in theoretical math,
but in real life -
you typically have - sometimes:
- bad weather (up to hurricane season)
- bad luck - all fashions and colors - on your side, or supplier side, or customer side, range side etc.
- and don't forget about NASA's missions and USSF missions which always have priority.

Bottom line:
Add September 2020 (with a single launch) to the last quarter 2020 (with 10 launches) - and you'll get more realistic estimate.

[Jansen]

...

From your graph Oct Nov Dec, ballpark average is about 8 days between launches. 365/8 =~ 45 Launches per year.

...

Well, they can launch 10 missions in 3 months, there is no doubt about it.
But classic proportion - which works so nice in arithmetic - just does not work here.
Unfortunately...
Yes, [10 missions / 3 months * 12 months] = 40 missions per year
- that's true in theoretical math,
but in real life -
you typically have - sometimes:
- bad weather (up to hurricane season)
- bad luck - all fashions and colors - on your side, or supplier side, or customer side, range side etc.
- and don't forget about NASA's missions and USSF missions which always have priority.

Bottom line:
Add September 2020 (with a single launch) to the last quarter 2020 (with 10 launches) - and you'll get more realistic estimate.

Consider that October-December already had many delays for weather, as well as delays from customers, range delays from ULA, and engine issues that forced more delays.

If anything, that period is as real as it gets.

[wannamoonbase]

Consider that October-December already had many delays for weather, as well as delays from customers, range delays from ULA, and engine issues that forced more delays.

If anything, that period is as real as it gets.

Yep, and those things come up all the time right, so they need to drive for every opportunity they can and fill things with Starlinks.  One never knows when weather, payload issue or pandemic impacts things.

48 is possible, with some luck.  They probably need to be trying for launch at least every 6-7 days and those impacts will slow you down.

It's going to be a hard push for sure, Vandenberg will have to do some lifting for sure.

[kdhilliard]

The 3 scheduled Falcon Heavy missions for 2021 ...

Three Falcon Heavies for 2021?
I only see two on our manifest: USSF-44 (2021 late spring) & USSF-52 (2021).
What is number three?

And are neither of the two (none of the three?) expected to RTLS the side boosters?

[Jansen]

The 3 scheduled Falcon Heavy missions for 2021 ...

Three Falcon Heavies for 2021?
I only see two on our manifest: USSF-44 (2021 late spring) & USSF-52 (2021).
What is number three?

And are neither of the two (none of the three?) expected to RTLS the side boosters?

USSF-44 will recover both side boosters via ASDS. Probably all three for USSF-52.

There is the possibility of a Viasat-3 launch in Q4 2021, but unconfirmed. It is likely to recover all boosters.

[gongora]

Recent NLS II contract mods:

NASA LAUNCH SERVICES II - MOD 146: Revises the launch dates for IXPE, IMAP, Psyche and Sentinel-6 missions.  Also, to adds funding.

(MOD 145)NASA LAUNCH SERVICES II - SPACE EXPLORATION TECHNOLOGIES.  This modification is to add a Non-Standard Service (NSS) 27.1 CLA for the Psyche/Janus Mission.

[wannamoonbase]

SpaceX is targeting 48 launches for 2021.

Yes, I heard this too. However I doubt they can get close to such target.
Below I attached graph showing SpaceX cadence during 2020 in the form "Date vs # days between launches".
Assuming SpaceX was REALLY busy with Dragon Crew DM2 in the first months of 2020, therefore they had slow launch cadence in this period.
So I calculated the average tempo for the rest of the year - from June to Dec. It gives 11.3 days between launches on average, which translates in (365/11.3) = 32.3 launches per year.
This is what one should expect IF SpaceX will keep the pace.
As we see from this year scheduling history, the launch tempo depends on (1) number of cores in rotation and on (2) refurbishing time.
As far as we can see, the refurbishing time remained the same during 2020
The number of cores did not go up significantly, as two new cores turned as GPS/Crew-assigned.
Actually one could expect number of cores will go down when B1049 and B1051 reach 10 flights. At that point SpaceX may want to take the core out of rotation for overhaul.

Bottom line - I don't see a way to boost the flight rate 50% up.
And to be honest - the flight rate of one launch every 11 days - that's already awesome !!
 

You have your graphs.  I have mine.
As for the line bolded above, my calculations include the annualized pace for the last ten launches.
As of the Dec 17 launch of NROL-108 it was 34.7, vs your 32.3 for the last half of the year.
48 launches is barely more than a third higher than that.   [/sarcasm]

Edit: Seems that to get to 48 launches per year SpaceX needs more out of VAFB or another east coast pad, more HIF capacity and a 3rd ASDS, a 4th ASDS if they move one to the west coast.

With all the things that can delay launches, then to get to 48 launches a year we'll eventually see 2 launches scheduled within 24-48 hours of each other on a routine basis. 

[Comga]

SpaceX is targeting 48 launches for 2021.

Yes, I heard this too. However I doubt they can get close to such target.
Below I attached graph showing SpaceX cadence during 2020 in the form "Date vs # days between launches".
Assuming SpaceX was REALLY busy with Dragon Crew DM2 in the first months of 2020, therefore they had slow launch cadence in this period.
So I calculated the average tempo for the rest of the year - from June to Dec. It gives 11.3 days between launches on average, which translates in (365/11.3) = 32.3 launches per year.
This is what one should expect IF SpaceX will keep the pace.
As we see from this year scheduling history, the launch tempo depends on (1) number of cores in rotation and on (2) refurbishing time.
As far as we can see, the refurbishing time remained the same during 2020
The number of cores did not go up significantly, as two new cores turned as GPS/Crew-assigned.
Actually one could expect number of cores will go down when B1049 and B1051 reach 10 flights. At that point SpaceX may want to take the core out of rotation for overhaul.

Bottom line - I don't see a way to boost the flight rate 50% up.
And to be honest - the flight rate of one launch evry 11 days - that's already awesome !!
 

You have your graphs.  I have mine.
As for the line bolded above, my calculations include the annualized pace for the last ten launches.
As of the Dec 17 launch of NROL-108 it was 34.7, vs your 32.3 for the last half of the year.
48 launches is barely more than a third higher than that.   [/sarcasm]

{edit: Reposted with better version of graph after wannamoonbase's reply }

[vaporcobra]

SpaceX is targeting 48 launches for 2021.

Bottom line - I don't see a way to boost the flight rate 50% up.
And to be honest - the flight rate of one launch evry 11 days - that's already awesome !!

You have your graphs.  I have mine.
As for the line bolded above, my calculations include the annualized pace for the last ten launches.
As of the Dec 17 launch of NROL-108 it was 34.7, vs your 32.3 for the last half of the year.
48 launches is barely more than a third higher than that.   [/sarcasm]

{edit: Reposted with better version of graph after wannamoonbase's reply }

Both some reasonable and data-driven arguments. At the same time, if we assume that a single launch in September was an outlier, as it certainly appears to be, SpaceX launched 10 times in Q4 2020 or once every 9.1 days, which averages out to 40.1 launches annually.

And that's despite the fact that there was only one Starlink launch total in two of those three months. IMHO, the high-30s or low-40s are well within reach in 2021, though that's obviously only the case if there are no major hiccups. I'd go so far as to say that had there had been no Starlink-4 landing failure, Starlink-5 engine-out/landing failure, or GPS III SV04 static fire anomaly, 2020 may have easily reached the low 30s.

[smoliarm]

At the same time, if we assume that a single launch in September was an outlier, as it certainly appears to be

Yes, statistically - the September point is an outlier - there is no doubt about it.
But the bitter truth is: we WILL have such "outliers" - from time to time - inevitably.
It represents the fundamental Law of Nature which says that "life is full of unpredictable events", or "C'est la vie" in French.
Or in plain English - "Shit Happens".
And it always looks like "outlier", but these outliers will happen more or less systematically.
There are hundreds of causes for delays, and most of them are quite unique and would not repeat in decades - but two or three WILL happen on the scale of a given year. Just because there are hundreds of them out there

Also - an important note:
Beside these hundreds of causes for delays with unique nature - there are a few with SYSTEMATIC  nature, e.g. - hurricane season. There are heavy hurricanes hitting Cape and sometimes they cause pretty long stand-downs.
Another example - winter storms in Atlantic, which may not affect Delta or Atlas launches at all, but they WILL affect ASDS operations for Falcons.

Strictly statistically:
If SpaceX had 10 launches in 3 months - and historically it's a rare (or lucky) event -
- one MUST assume probability less-than-1 for the repetition of this event. Let's call this probability "P(q)".
Now, if we want 40 launches in a year - we NEED this lucky event happen FOUR TIMES IN A ROW.
The probability of such year-long string - P(y) - will be equal to P(q)^4.
That's the bitter truth -
P(y) = P(q)^4
(yes, statistics is a bitch)

or in plain numbers
P(q) = 0.80
P(y) = 0.41
or
even if P(q) is quite high = 0.90
P(y) = 0.66

The bottom line:
If you take SHORT period with the best performance and propagate it to a much longer period - you are bound for some disappointment (statistically bound).
What it's gonna be in numbers? - hard to tell
This is why we use CI (confidence intervals) marked with "±".
So judging from the available data, next year SpaceX will have 33 ± 4 launches of Falcons (both 9s and Heavies) - if no major bad thing happen.

[DistantTemple]

At the same time, if we assume that a single launch in September was an outlier, as it certainly appears to be

Yes, statistically - the September point is an outlier - there is no doubt about it.
But the bitter truth is: we WILL have such "outliers" - from time to time - inevitably.
It represents the fundamental Law of Nature which says that "life is full of unpredictable events", or "C'est la vie" in French.
Or in plain English - "Shit Happens".
And it always looks like "outlier", but these outliers will happen more or less systematically.
There are hundreds of causes for delays, and most of them are quite unique and would not repeat in decades - but two or three WILL happen on the scale of a given year. Just because there are hundreds of them out there

Also - an important note:
Beside these hundreds of causes for delays with unique nature - there are a few with SYSTEMATIC  nature,
snip...

Strictly statistically:

Snip...
The bottom line:
If you take SHORT period with the best performance and propagate it to a much longer period - you are bound for some disappointment (statistically bound).
What it's gonna be in numbers? - hard to tell
This is why we use CI (confidence intervals) marked with "±".
So judging from the available data, next year SpaceX will have 33 ± 4 launches of Falcons (both 9s and Heavies) - if no major bad thing happen.

(my snips of maths etc...)
All of what is said is of itself valid. However you are talking about it all as if its a static situation, and the risks and response of SpaceX to that risk are both constant.
Elon is strongly driving the whole of his space program towards the "airline like" reliability he frequently touts. We see his success in his other businesses, and his success with SpaceX has seen it move from a joke to the premier world launch business. (IMO). There is not much reason to think he will now rest on his laurels, and coast on F9's current cadence. Conversely it seems his attention has returned to the bottle necks in the F9 system.

When we pick the best 3 months from the data, we didn't cherry pick from the whole year, but took the last three, which arguably show SX's latest and most improved performance. One reason that is the best, is that more of SX's systems are optimised.

If we assume (as you implied) that this is the best possible then the disruptions like hurricane, landing weather, accidents, competitors, etc will at times make this cadence unobtainable. However that is an incorrect assumption.

For example, if SX had enough boosters ready to fly, pad turn around and readiness down to 48 hrs, and payload processing and encapsulation lined up smoothly, they could take advantage of good weather and no range contention to get 6 launches off in two weeks!  Then outliers, ULA's hogging the range, or some bad weather might destry the month's figures, but not the year. This is not outrageous, it is the sort of turn around SX has said they are aiming for. SX also have the need for it as they have al least a year before SS will be launching Starlink, and they want to get the constellation up as fast as they can manufacture the satellites.

The "best period" of the last 10 weeks, is the best because they DID have the payloads, and boosters ready. Earlier in the year like the 1st 1/4 they didn't. Therefore these last 10 weeks are MORE representative of their performance than earlier in the year.

(Also some SL from VAFB will increase overall cadence)

Conclusion. SpaceX will significantly improve on its best period by streamlining its processing, and removing bottlenecks. This will allow for some disruption. Overall Sx will increase its cadence greatly.

[Jansen]

I’m going to suggest that discussion about the number of flights in 2021 should continue with the poll here:

POLL: Number of SpaceX orbital flights in 2021
https://forum.nasaspaceflight.com/index.php?topic=52633.0

[FutureSpaceTourist]

Gwynne Shotwell talks about selling flight-proven rockets, Starship
"It was easier to sell 'flight proven' to customers than it was to sell Falcons."

ERIC BERGER - 1/4/2021, 8:45 PM

SpaceX enjoyed its most successful year in 2020. Amidst the pandemic, the company set a record for total number of launches: 26. All met their objectives. The Crew Dragon spacecraft flew humans—Doug Hurley and Bob Behnken—into orbit for the first time. And then it did so again, with the Crew-1 mission in November. SpaceX also made demonstrable progress on its next-generation Starship launch system.

https://arstechnica.com/science/2021/01/gwynne-shotwell-talks-about-selling-flight-proven-rockets-starship/

Some discussion in the article of SpaceX selling Starship to customers includes:

"We have signed deals where we can pick whether it's a Falcon or a Starship,"

[Jansen]

B1051.7 is the only other booster that would be remotely available. A launch on Nov 28 would mean a new booster turnaround record of 41 days. Possible payloads are SXM-7 or Starlink v1.0 L16.

Not SXM-7. B1049 and B1051 are destined to fly Starlink the rest of their life. I dunno why you keep insisting they will take on commercial payloads.

Most of the posts saying how wrong I was about commercial launches have mysteriously disappeared. But here is a quote from Shotwell saying how it works:

Unless a customer has a strong argument one way or the other, the decision on what booster to use is left up to SpaceX. "You're buying a launch service, and we will provide you the best vehicle that we can in the timeframe that you need to fly," she said. "And we basically put the control for the most part in our hands."

https://arstechnica.com/science/2021/01/gwynne-shotwell-talks-about-selling-flight-proven-rockets-starship/

Which jives with what I’ve been saying for months:

Commercial contracts are very different from government launch contracts. NASA, USSF, and NRO pay extra (a lot extra) to have control over all aspects of a launch. That results in a lot of specifics around workflow, testing requirements, and booster selection.  They can basically veto any decision because they’ve bought that right.

My understanding is that commercial contracts are based more on deliverables. Basically that SpaceX will utilize F9 to deliver this payload to this orbit on such date, with allowable delays for weather, governmental priority launches, etc.

SX has typically had control over which booster to use, because to have that specified so far out when the contract is signed cuts down on flexibility. If the customer wants control, they pay for it, which commercial customers tend not to do.

[Jansen]

https://twitter.com/julia_bergeron/status/1346503659996274690

Maintenance on OCISLY could affect launch cadence, might delay Transporter-1