Author Topic: Are re-used boosters less finicky?  (Read 3951 times)

Offline LouScheffer

  • Full Member
  • ****
  • Posts: 1959
  • Liked: 2641
  • Likes Given: 291
Are re-used boosters less finicky?
« on: 12/20/2018 02:45 pm »
Are the second and later launches of the same booster more or less likely to proceed through the countdown without encountering technical bugs causing a scrub?  There are reasons each direction might be possible - infant mortality vs wear-and-tear, for example:
these new boosters are always so finicky on their first launch...
[...]
I think we might begin to witness the positive effect of "flight proven" being turned around against all these "untested" boosters ;)

there's ever only so much you can test in component tests, dress rehearsal or even a hold down hot fire.
Not true.  [...]
c.  Can't say that the grid fin failure was due to infant mortality, poor workmanship (which still applies to a refurb booster), effects from second stage engine.
But now we are emerging from the wilderness of opinions to the department of data, though we are not there yet.  There have been (or will be after GPS) 10 launches of Block 5.  The number of technical scrubs after the countdown has started are:

Booster scrubs mission
1046.1       1 Bangabandhu
1047.1       0 Telstar 19
1048.1       0 Iridium/Grace
1046.2       0 Merah Putih
1049.1       0 Telstar 18
1048.2       0 SAOCOM
1047.2       0 Es'hail
1046.3       0 SSO-A
1050.1       0 CRS-16
1054.1       1 GPS-III


So there have been 6 new and 4 used launches, and overall 2 of 10 (20%) had a technical scrub during countdown.  If two scrubs were evenly distributed among the 10 launches, we would naturally expect both on new boosters 15/45 (33%) of the time, both on used boosters 6/45 (13%) of the time, and one each 24/45 (53%) of the time.

Clearly the numbers are small, and not statistically significant, but the evidence points slightly in the direction of used boosters having fewer scrubs.  By the end of next year we may have better data on this.  (On the other hand, SpaceX is presumably trying to reduce the number of scrubs in general.  If they succeed, then the number of scrubs will remain low and the statistical power will be limited).

Online Cheapchips

  • Full Member
  • ***
  • Posts: 347
  • UK
  • Liked: 199
  • Likes Given: 357
Re: Are re-used boosters less finicky?
« Reply #1 on: 12/20/2018 04:04 pm »

As you say, theres not statistical significance at the moment.

It would be nice to have more details on both scrubs.  It might give insight into if they're unproven hardware issues or just early life failures.

Offline vaporcobra

  • Full Member
  • ****
  • Posts: 1271
  • Tacoma, WA
  • Liked: 2640
  • Likes Given: 2832
Re: Are re-used boosters less finicky?
« Reply #2 on: 12/20/2018 05:18 pm »
Add pre-Block 5 launches and this could be a bit more statistically interesting :)

Online meekGee

  • Senior Member
  • *****
  • Posts: 9157
  • N. California
  • Liked: 5520
  • Likes Given: 942
Re: Are re-used boosters less finicky?
« Reply #3 on: 12/20/2018 05:25 pm »
Are the second and later launches of the same booster more or less likely to proceed through the countdown without encountering technical bugs causing a scrub?  There are reasons each direction might be possible - infant mortality vs wear-and-tear, for example:
these new boosters are always so finicky on their first launch...
[...]
I think we might begin to witness the positive effect of "flight proven" being turned around against all these "untested" boosters ;)

there's ever only so much you can test in component tests, dress rehearsal or even a hold down hot fire.
Not true.  [...]
c.  Can't say that the grid fin failure was due to infant mortality, poor workmanship (which still applies to a refurb booster), effects from second stage engine.
But now we are emerging from the wilderness of opinions to the department of data, though we are not there yet.  There have been (or will be after GPS) 10 launches of Block 5.  The number of technical scrubs after the countdown has started are:

Booster scrubs mission
1046.1       1 Bangabandhu
1047.1       0 Telstar 19
1048.1       0 Iridium/Grace
1046.2       0 Merah Putih
1049.1       0 Telstar 18
1048.2       0 SAOCOM
1047.2       0 Es'hail
1046.3       0 SSO-A
1050.1       0 CRS-16
1054.1       1 GPS-III


So there have been 6 new and 4 used launches, and overall 2 of 10 (20%) had a technical scrub during countdown.  If two scrubs were evenly distributed among the 10 launches, we would naturally expect both on new boosters 15/45 (33%) of the time, both on used boosters 6/45 (13%) of the time, and one each 24/45 (53%) of the time.

Clearly the numbers are small, and not statistically significant, but the evidence points slightly in the direction of used boosters having fewer scrubs.  By the end of next year we may have better data on this.  (On the other hand, SpaceX is presumably trying to reduce the number of scrubs in general.  If they succeed, then the number of scrubs will remain low and the statistical power will be limited).
We also had one in-flight anomaly, also on a new booster.

-----
ABCD: Always Be Counting Down

ABCD - Always Be Counting Down

Offline Lar

  • Fan boy at large
  • Global Moderator
  • Senior Member
  • *****
  • Posts: 11356
  • Saw Gemini live on TV
  • A large LEGO storage facility ... in Michigan
  • Liked: 8402
  • Likes Given: 6735
Re: Are re-used boosters less finicky?
« Reply #4 on: 12/20/2018 07:59 pm »
Bathtub curve tells us that IF component lifetimes are well over 2-3 launches, and IF infant mortality testing weeds out most infant mortality, the second launch OUGHT to be more reliable.

But we aren't really sure of either of the IFs yet (SpaceX almost certainly has more data than we do, and therefore may be more sure... ) I don't think. 

Fascinating topic. We had speculated that this might be the case, reflown boosters are "better"... I look forward to more data.
"I think it would be great to be born on Earth and to die on Mars. Just hopefully not at the point of impact." -Elon Musk
"We're a little bit like the dog who caught the bus" - Musk after CRS-8 S1 successfully landed on ASDS OCISLY

Offline groundbound

  • Full Member
  • **
  • Posts: 239
  • Liked: 202
  • Likes Given: 8
Re: Are re-used boosters less finicky?
« Reply #5 on: 12/23/2018 10:26 pm »
There is another effect that may cause what amounts to reduced statistical significance for the next year or two.

The first time any booster sees it's n'th re-use, launch ops are likely to be more cautious just because they are expanding the re-use envelope, even if only by a little.

Offline deruch

  • Senior Member
  • *****
  • Posts: 2395
  • California
  • Liked: 1933
  • Likes Given: 4881
Re: Are re-used boosters less finicky?
« Reply #6 on: 12/25/2018 02:32 pm »
Are the second and later launches of the same booster more or less likely to proceed through the countdown without encountering technical bugs causing a scrub?  There are reasons each direction might be possible - infant mortality vs wear-and-tear, for example:
these new boosters are always so finicky on their first launch...
[...]
I think we might begin to witness the positive effect of "flight proven" being turned around against all these "untested" boosters ;)

there's ever only so much you can test in component tests, dress rehearsal or even a hold down hot fire.
Not true.  [...]
c.  Can't say that the grid fin failure was due to infant mortality, poor workmanship (which still applies to a refurb booster), effects from second stage engine.
But now we are emerging from the wilderness of opinions to the department of data, though we are not there yet.  There have been (or will be after GPS) 10 launches of Block 5.  The number of technical scrubs after the countdown has started are:

Booster scrubs mission
1046.1       1 Bangabandhu
1047.1       0 Telstar 19
1048.1       0 Iridium/Grace
1046.2       0 Merah Putih
1049.1       0 Telstar 18
1048.2       0 SAOCOM
1047.2       0 Es'hail
1046.3       0 SSO-A
1050.1       0 CRS-16
1054.1       1 GPS-III


So there have been 6 new and 4 used launches, and overall 2 of 10 (20%) had a technical scrub during countdown.  If two scrubs were evenly distributed among the 10 launches, we would naturally expect both on new boosters 15/45 (33%) of the time, both on used boosters 6/45 (13%) of the time, and one each 24/45 (53%) of the time.

Clearly the numbers are small, and not statistically significant, but the evidence points slightly in the direction of used boosters having fewer scrubs.  By the end of next year we may have better data on this.  (On the other hand, SpaceX is presumably trying to reduce the number of scrubs in general.  If they succeed, then the number of scrubs will remain low and the statistical power will be limited).

The Bangabandhu scrub is listed in the log as being a ground system fault.  Seems like dubious reasoning to chalk that up to SpaceX using a new booster for that mission.  Should we also evaluate whether the likelihood of weather scrubs is higher/lower when using pre-flown boosters?
Shouldn't reality posts be in "Advanced concepts"?  --Nomadd

Offline rpapo

Re: Are re-used boosters less finicky?
« Reply #7 on: 12/25/2018 06:23 pm »
The Bangabandhu scrub is listed in the log as being a ground system fault.  Seems like dubious reasoning to chalk that up to SpaceX using a new booster for that mission.  Should we also evaluate whether the likelihood of weather scrubs is higher/lower when using pre-flown boosters?
FWIW, being that the Bangabandhu launch was the very first Block 5 launch, there were changes to the GSE.  Any time you have a change, there are bugs to root out.  As said before, this has nothing to do with the difference between the first launch and a subsequent launch.
An Apollo fanboy . . . fifty years ago.

Online meekGee

  • Senior Member
  • *****
  • Posts: 9157
  • N. California
  • Liked: 5520
  • Likes Given: 942
Re: Are re-used boosters less finicky?
« Reply #8 on: 12/25/2018 06:40 pm »
The idea from the first post was to "emerge from the wilderness of opinions to the department of data".

If it's GSE, then it will statistically effect new and used booster in the same manner, and factor out.  However if some unforeseen effect causes there to be a difference, it'll show up in the data as well.

FWIW, same goes for weather...

ABCD - Always Be Counting Down

Offline rpapo

Re: Are re-used boosters less finicky?
« Reply #9 on: 12/25/2018 07:27 pm »
If it's GSE, then it will statistically effect new and used booster in the same manner, and factor out.  However if some unforeseen effect causes there to be a difference, it'll show up in the data as well.
Not entirely true.  Every time they have introduced a new "block", they have made changes to the GSE, and have had to work through new problems as a result.  These problems faded with more launches, and so the first launches were disproportionately affected.
An Apollo fanboy . . . fifty years ago.

Online meekGee

  • Senior Member
  • *****
  • Posts: 9157
  • N. California
  • Liked: 5520
  • Likes Given: 942
Re: Are re-used boosters less finicky?
« Reply #10 on: 12/25/2018 07:39 pm »
If it's GSE, then it will statistically effect new and used booster in the same manner, and factor out.  However if some unforeseen effect causes there to be a difference, it'll show up in the data as well.
Not entirely true.  Every time they have introduced a new "block", they have made changes to the GSE, and have had to work through new problems as a result.  These problems faded with more launches, and so the first launches were disproportionately affected.
.. and so unless there's a block 6, since we're already 20 launches onto it, we will soon have some statistical idea.

-----
ABCD: Always Be Counting Down

ABCD - Always Be Counting Down

Offline mgeagon

  • Full Member
  • *
  • Posts: 102
  • Hong Kong
  • Liked: 105
  • Likes Given: 1
Re: Are re-used boosters less finicky?
« Reply #11 on: 12/31/2018 05:08 am »
This is a great thread. Please update the opening post after every launch and we'll see significant statistics emerge before our eyes. It may be instructive to include pre-block 5 numbers, as a side comparison.

Online FutureSpaceTourist

  • Senior Member
  • *****
  • Posts: 7214
  • UK
    • Plan 28
  • Liked: 7494
  • Likes Given: 2505
Re: Are re-used boosters less finicky?
« Reply #12 on: 12/31/2018 10:34 am »
Hereís an attempt to extend data back as far as first flight of first booster to be re-used. Iíll let others look at statistics - choices to be made about static fire scrubs vs launch scrubs, GSE vs vehicle issues, S1 vs S2 issues etc. Iíve ignored all fairing issues and have made a judgement about whether to include delays, not just scrubs, post static fire. Reason for each included delay noted below so you can choose to ignore if you wish. I have omitted all weather scrubs.

Irrespective of choices made, there are no known launch scrubs yet due to a flight proven booster. However, SES-11 had a launch delay due to a booster engine issue on flight proven 1031.2.

Booster scrubs mission
1021.1       0 CRS-8
1022         0 JCSAT-14
1023         1 Thaicom-8 (launch scrub S2 actuator issue)
1024?        0 Eutelsat 117W B & ABS 2A
1025?        0 CRS-9
1026?        0 JCSAT-16
1029.1       1 Iridium NEXT-1 (static fire scrub due to GSE)
1031.1       1 CRS-10 (launch scrub S2 actuator issue)
1030         1 Echostar 23 (static fire scrub unknown cause)
1021.2       0 SES-10
1032.1       1 NROL-76 (launch scrub S1 issue)
1034         0 Inmarsat 5 F4
1035.1       0 CRS-11
1029.2       0 Bulgariasat-1
1036.1       0 Iridium NEXT-2
1037         2 Intelsat 35e (launch scrubs: GNC & ground computer issues)
1039.1       0 CRS-12
1038.1       0 Formosat-5
1040         0 OTV-5
1041.1       0 Iridium NEXT-3
1031.2       1 SES-11 / Echostar 105 (launch delay, post static fire, due to S1 engine)
1042         0 Koreasat-5A
1035.2       1 CRS-13 (launch delay, post static fire, S2 fuel system particles)
1036.2       0 Iridium NEXT-4
1043.1       0 Zuma
1032.2       1 SES-16 (launch delay to replace S2 sensor)
1023.2   &
  1033   &   0 FH demo
  1025.2
1038.2       0 PAZ & Microsat 2a/2b
1044         0 Hispasat1F (30W-6)
1041.2       0 Iridium NEXT-5
1039.2       0 CRS-14
1045.1       1 TESS (launch scrub for more GNC analysis)
1046.1       1 Bangabandhu
1043.2       0 Iridium NEXT-6/GRACE-FO
1040.2       0 SES-12
1045.2       0 CRS-15
1047.1       0 Telstar 19
1048.1       0 Iridium/Grace
1046.2       0 Merah Putih
1049.1       0 Telstar 18
1048.2       0 SAOCOM
1047.2       0 Es'hail
1046.3       0 SSO-A
1050.1       0 CRS-16
1054.1       1 GPS-III


Edit to add: all data from the launch manifest and launch log threads, with many thanks to gongora & cartman respectively.
« Last Edit: 01/26/2019 09:22 am by FutureSpaceTourist »

Offline RDMM2081

  • Full Member
  • *
  • Posts: 164
  • Liked: 98
  • Likes Given: 245
Re: Are re-used boosters less finicky?
« Reply #13 on: 12/31/2018 07:10 pm »
Nice chart thanks!

Looking at the ~2017 era (model year?) boosters, I don't see any evidence that reused boosters were less finnicky.  1031.2, 1035.2 1032.2 all had various scrubs/delays, while 1031.1 and 1032.1 both had scrubs of their own, but 1035.1 actually had no scrubs on its "unproven" flight, but the same booster did scrub on its re-use attempt.  however 1029.1 did have a scrub, and 1029.2 did not. other boosters that were re-used successfully, 1036.1 and 1036.2 both launched scrub-free.  Hardly anything that demonstrates a pattern, yet.

What is more clear, is that 2018 overall had less scrubs.  In fact both 1045.1 and 1046.2 I personally attribute more to "overabundance of caution" type scrubs than actual problems.  1054.1 is really the first "issue" with block 5 boosters that has caused any launch delay (I am sure this is wrong, but speaking only to the scrub data in the chart).

Online FutureSpaceTourist

  • Senior Member
  • *****
  • Posts: 7214
  • UK
    • Plan 28
  • Liked: 7494
  • Likes Given: 2505
Re: Are re-used boosters less finicky?
« Reply #14 on: 12/31/2018 07:38 pm »
Itís important to look at the reasons for the scrubs:

[...] 1035.1 actually had no scrubs on its "unproven" flight, but the same booster did scrub on its re-use attempt. 

The 1035.2 scrub was a stage 2 issue so nothing to do with the flight proven booster. SES-11 is the only launch with a (post static fire) delay due to a booster issue on a flight proven stage.

Offline RDMM2081

  • Full Member
  • *
  • Posts: 164
  • Liked: 98
  • Likes Given: 245
Re: Are re-used boosters less finicky?
« Reply #15 on: 12/31/2018 08:02 pm »
Good point of course, I missed that in my first glance through.

Offline CorvusCorax

  • Full Member
  • ****
  • Posts: 514
  • Germany
  • Liked: 662
  • Likes Given: 567
Re: Are re-used boosters less finicky?
« Reply #16 on: 01/25/2019 08:28 pm »
Would it be fair to note that 1033 had an inflight issue ( run outof ignition fluid ) on its first flight, while 1023.2 & 1025.2 both landed successfully.

Do we have data about delays/issues in static fire campaigns? Should that be included?

Online FutureSpaceTourist

  • Senior Member
  • *****
  • Posts: 7214
  • UK
    • Plan 28
  • Liked: 7494
  • Likes Given: 2505
Re: Are re-used boosters less finicky?
« Reply #17 on: 01/26/2019 09:33 am »
I donít think counting 1033 landing failure is fair. It was trying to do something never previously attempted and so was a test rather than a proven operational capability. For the same reason I personally wouldnít have counted any failures associated with separation of the side boosters on that FH demo flight.

We do sometimes know when static fire dates have moved but we often donít know why (and certainly not in publicly available info). It would be great to know how easy/difficult preparing each flight proven booster is (eg extent of any refurbishment work, components/engines replaced if any etc) but obviously thatís not public. So I think the best we can do is like at how reliable things are post static fire, when SpaceX think they are ready for flight.
« Last Edit: 01/26/2019 09:34 am by FutureSpaceTourist »

Tags: