Q&A: Pegasus Designer Dr. Antonio Elias

• #140 by Chris Bergin on 30 Aug, 2007 03:14
• Quote

  tnphysics - 30/8/2007  4:13 AM
  
  Sorry Chris,
  
  My reasoning was that a Falcon 9 launch costs $35 million. A Pegasus launch costs about the same amount. But the Falcon 9 is much more powerful. I think that the Delta II class launcher will cost more than Pegasus. If this is correct, the conclusion follows.

  
  Thanks for expanding on that.
• #141 by Jim on 30 Aug, 2007 11:43
• Quote

  tnphysics - 29/8/2007  11:13 PM
  
  My reasoning was that a Falcon 9 launch costs $35 million.

  
  That is not proven yet
• #142 by antonioe on 30 Aug, 2007 19:53

• Look, guys:  It boils down to a matter of confidence in the future.  If we were confident that SpaceX would be able to provide a Delta II class service (with whatever size vehicle) for around $50M, with 90%-95% reliability, in the 2010-2012 time frame, we would be very happy to save our money and concentrate on developing and selling to NASA and the DoD Mid-class missions with its accompanying spacecraft, ground systems, some special instruments we build, etc.  We certainly were happy to have MDAC/Boeing/ULA provide $50M Delta II's (we flew on one last June 21; DAWN, our first deep-space spacecraft, will also launch on a Delta II, hopefully on September 26).

  Problem is: given a) SpaceX's Falcon 1 performance to date b) the leap that Falcon 9 represents over Falcon 1 c) our own experience developing, building and launching low-cost (or at least as low as we know how to) launch vehicles and d) our complete ignorance of the methodology that SpaceX uses to achieve their advertised pricing, we simply cannot rely on SpaceX being able to deliver.

  In a previous posting I related how Pegasus was started, in part, by the cool reception that George Koopman (founder of AMROC) gave to our bona fide inquiries as a potential customer.  If he had tried, even so slightly, to sell us his LV, we would not only not developed Pegasus, but would have signed up as the anchor customer for HIS rocket!!! (we are a little more savvy now - had we done that, and given what happened to AMROC, Orbital would probably have been bankrupt by the early 90's).  Point is, we had no confidence that we would find launch services for our ORBCOMM system.

  Similarly, if the Delta II service would have continued to be offered at $50M - $70M, we would NEVER dream of competing with it.

  I wish, I really wish, that Elon could build a 10MT launch vehicle and sell it, at a reasonable profit, for $35M.  (actually, I'd prefer that he sell a 5MT service for $17M - no need to help the big guys' markets with cheap EELV's!)  I just don't see how.  I look at his industrial infrastructure, which is starting to rival Orbital's - but we build almost $1B worth of spacecraft and rockets a year with it!  I look at his vertical integration, which means he has to fund and capitalize (and then amortize) a very large number of resources that suppliers may already possess.  I read somewhere that he now has an 11-acre plant (bigger that our Chandler facility, although ours is filled to the brim and his may be a little empty at the moment) and 350 employees - more than Orbital employs on its Pegasus, Taurus and Minotaur programs PUT TOGETHER.  We developed Pegasus with a team that PEAKED at around 45-50 people (including yours truly) - the largest Orbital launcher development team, Taurus, peaked around 75 (aero313 - is my memory correct?  You were the PM...)

  If they need over four years, hundreds of people and over $100M before the first sucessful launch of a Pegasus-class LV (we spent $42M in 1990 $'s to get to a sucessful two-satellite first flight in less than three years), how many people and how much money will they need to keep a Falcon 9 line going?  Who is going to pay for that marching army?  12 launches per year?  Now, don't tell me that lower costs will cause dozens of customers to come out of the woodwork magically - I'd rather believe in a "late 2008" Falcon 9 first flight.

  Launch demand is very inelastic in an "absolute demand" sense (i.e., a 10% reduction in average supply cost increases total market demand by a lot less than 10%) even if is it "competitively elastic" (a launch service which is $1 cheaper than an identical service gets all the market) - the sale of 52 flights to half a dozen customers, and the sale of two dozen commercial GeoComs to another half a dozen customers (and participating in their buy of Ariane, Soyuz and Zenit launch services) backs this statement.

  Then, for a privately-funded development such as Pegasus, Taurus or the new Delta-II class one we are developing, there is the small issue of amortizing the development costs.  Admittedly, SpaceX may not have to pay back Mr. Musk's $100M rumored investment and the $270M no-strings-attached NASA COTS money (or at least the $198M of it that is only attached by VUgraphs).  But what of the additional, outside investors?  the NASA/SpaceX COTS Space Act Agreement mentions :"financing milestones" - I assume they are not exclusively's Mr. Musk's money.

  I sincerely mean no disrespect for SpaceX or Mr. Musk.  I admire and salute what they are trying to do.  Our business would benefit from their success.  But, given my 20 years and 52 launches experience trying to develop, build and fly LVs on the cheap, I don't see, I simply cannot see how they can do it. 

• #143 by jimvela on 30 Aug, 2007 21:40
• Quote

  antonioe - 30/8/2007  1:53 PM
   Admittedly, SpaceX may not have to pay back Mr. Musk's $100M rumored investment and the $270M no-strings-attached NASA COTS money (or at least the $198M of it that is only attached by VUgraphs).  

  
  There's the root of my issue with COTS I.  How much launch vehicle development could you have done with a couple of hundred million bones?
  
  At least SpaceX has some credibility in that they've built and flown two F1, whereas the other COTS I finalist has done nothing well except for end up bankrupt-  and they've historically been pretty good at doing that.   I don't expect that vehicle to ever fly.
  
  I'd far rather the money go to reasonably viable providers.  [Disclaimer: My employer bid as a teamed partner on  COTS I].
  
  With respect to a D-II replacement, I agree that $35M isn't a realistic price point for an F9.  I'm thinking it'll be more like $85M when it all shakes out.  Power to SpaceX if they succeed, my interest is in getting payloads like the ones my employer builds into space.  Viable but inexpensive  launchers are really good for me....
  
• #144 by antonioe on 30 Aug, 2007 22:13

• Quote

  jimvela - 30/8/2007 4:40 PM At least SpaceX has some credibility in that they've built and flown two F1

  Yes, but SpaceX has to

  1. make it to orbit
  2. grow the launch vehicle by a factor of 20
  3. develop a 5 to 7 MT spacecraft with a human-rated interior volume (including micrometeorid shielding, etc)
  4. make that spacecraft dock with ISS passing all of JSC's safety hurdles and
  5. make said spacecraft reenter intact (COTS Capabilities A through C)
  6. do 1 through 4 (only, if I read the agreement correctly) before the end of 2010.

  At least the NASA/SpaceX COTS agreement does not require SpaceX to demonstrate capability D, thank God.  My estimate to do all of the above is at least $500M - quite a challenge!  On the other hand, the agreement specifies that SpaceX will receive $198M just by holding design reviews and passing "financial milestones" - that is, before the first hardware milestone or test, so they have a good chance of getting that money without having to spend a lot on the more advanced capabilities.  That allows them to concentrate on the LV if they so choose.

  Quote

  $35M isn't a realistic price point for an F9. I'm thinking it'll be more like $85M when it all shakes out

  A 10MT LV for any number under $100M would be a very fine and worthwhile acheivement, indeed!!!

• #145 by STS Tony on 31 Aug, 2007 17:32
• Is OSC interested in taking on Sea Launch market? This new vehicle sounds like it's in that sort of payload range?
  
  I say that as they unfortunatly seem to be in a fair amount of uncertainty until they are a few flights into return to flight and it may take years for customers to trust them.
  
  Sorry if that's a dumb question, I know the obvious payload range is taking over what is missing when Delta II goes, but just trying to understand the market.
• #146 by antonioe on 31 Aug, 2007 22:29

• Quote

  STS Tony - 31/8/2007 12:32 PM Is OSC interested in taking on Sea Launch market?

  Whoa!  Hold your horses!  Now, you're talking about a BIG step from our current LV offerings!  Plus, although it's true that Ariane has had some economic difficulties, and that the Russian situation is, to say the least, muddy (and the mud is splattering the poor Ukranians,) Zenit and Ariane are still two formidable competitors.

  ON THE OTHER HAND... One of Orbital's great unsung success stories, our taking the under-5KW GeoCom market by storm, opens up the possibility of extending our Delta II-class LV not to the Sea Launch class, but to something just a bit north of Soyuz and just a bit south of Land Launch, i.e., about 3MT into GTO.  This would require, of course, something like a full cryo upper stage, which is something we're not ready to bite into, although it would be a good PPI for our MLV.

• #147 by meiza on 31 Aug, 2007 22:39
• Will the design of the new booster be done in imperial or SI units?
• #148 by bigdog on 01 Sep, 2007 01:52
• Thanks Dr. Elias for that terrific explanation of how the SpaceX prices really don't make sense.
  
  I really wonder if Musks prices are based on the big picture.  I think it's based on only the cost to build and in small part launch an F1 or F9.  I really doubt he's accounting for all the costs of the buildings, salaries, launch site, etc. etc. costs.  His personal financial support hides much of the true cost I think.
  
  I mean think about the cost of salaries.  If he gets to the 600 employees he talks about the payroll is huge.  If the average salary is 50K (I think I'm low there) that's 30Mil a year.  That's a couple F1's alone.  His new building will rack up big energy bills being in California and maintenance on SLC-40 won't be cheap either.  I think the cost for a F9 will get to at least 100mil which still not being full EELV class is no bargan.
  
  There just aren't enough payloads out there and they sure can't be ready to launch fast enough to support the rate he'll need to keep prices anywhere near what he's talking now.
• #149 by Carl G on 01 Sep, 2007 04:10
• Changing tacts, what was it that got you interested in designing space vehicles, Antonio? I know it'd need a load of qualifications, but something must have driven you to want to take that route before that.
  
  Would find that interesting to know, as we're all a bunch of armchair engineers on here, apart from the big group of actual USA and NASA engineers etc here.
  
  But I'm sure you and maybe Dr Stanley are the only ones here who are at the sharp end of thinking vehicles up, with your story sounding rather like Richard Prior drawing his plans for a super computer in Superman III  :cool:  (Funnily, an evil Pegasus was featured in that movie! )

  ---

  
• #150 by pippin on 01 Sep, 2007 22:04
• Quote

  bigdog - 1/9/2007  3:52 AM
  
  Thanks Dr. Elias for that terrific explanation of how the SpaceX prices really don't make sense.
  
  I really wonder if Musks prices are based on the big picture.  I think it's based on only the cost to build and in small part launch an F1 or F9.  I really doubt he's accounting for all the costs of the buildings, salaries, launch site, etc. etc. costs.  His personal financial support hides much of the true cost I think.
  
  I mean think about the cost of salaries.  If he gets to the 600 employees he talks about the payroll is huge.  If the average salary is 50K (I think I'm low there) that's 30Mil a year.  That's a couple F1's alone.  His new building will rack up big energy bills being in California and maintenance on SLC-40 won't be cheap either.  I think the cost for a F9 will get to at least 100mil which still not being full EELV class is no bargan.
  
  There just aren't enough payloads out there and they sure can't be ready to launch fast enough to support the rate he'll need to keep prices anywhere near what he's talking now.

  
  Mr. Musk has run a business or two, he didn't get his money thrown at him, so I'm pretty sure he knows how to calculate costs, especially regarding such simple things.
  
  Whether his assumptions about market growth etc. hold true  is a different story, however...
• #151 by Andy L on 01 Sep, 2007 23:37
• Is SpaceX OSC's biggest rival (potentially)? Some of the OSC guys aren't very nice about SpaceX on their threads here.
• #152 by antonioe on 02 Sep, 2007 03:14

• Quote

  Andy L - 1/9/2007 6:37 PM Is SpaceX OSC's biggest rival (potentially)?

  I wish... but no - only 8% of Orbital's revenue last year came from space launches. 27% came from Commercial GeoComm spacecraft, 22% from MDA through Boeing, 17% from NASA and DoD scientific and research satellites and, of course, the Orion program.  Our main competitors are Alcatel and Lockheed-Martin Newtown in the GeoCom market (Loral is waaay above our market in size and Boeing-ex-Hughes has essentially abandoned the market), nobody in the missile defense interceptor market (Lockheed used to be, no longer,) Lockheed-Martin Denver in the DoD spacecraft market (also, coincidentally, one of our biggest customers through Orion) and, of course, Ball and Spectr... I mean,  GD.  As far as I know, SpaceX is not a player in 92% of our market.  To put things in perspective, I'm enclosing a picture of our Satellite Manufacturing Facility in Dulles, VA, showing five - count 'em, five! - GeoComs in various stages of assembly; over $300M worth of satellites.  They are BIG!  You could camp comfortably for a week inside one of them.  And they have to last for 15-16 years! 

  It's nice to work on rockets and satellites at the same time.  They are very, VERY different beasts indeed.  By building both, you gain a better appreciation of both sides of the "space thingy".

  Quote

   Some of the OSC guys aren't very nice about SpaceX on their threads here.

  Well, shame on us.  If anybody from Orbital has been "not nice" with SpaceX or anybody else, I apologyze in Orbital's name.  There is no need to be un-nice to anyone, friend or foe, and, specifically, I don't consider SpaceX our foe, even if I don't understand some of the things they do and say.  You may not agree with someone, but you can disagree politely.

  BTW, we were once know by the initials "OSC", but it's been years since we changed our nickname to "Orbital" (for better or worse - at least we shed that horrible NASA-worm-like logo that everybody thought spelled "CSC"

  ---

  
• #153 by antonioe on 02 Sep, 2007 03:24

• Quote

  bigdog - 31/8/2007 8:52 PMIf he gets to the 600 employees he talks about the payroll is huge. If the average salary is 50K (I think I'm low there) that's 30Mil a year.

  *SIGH* well, let be disabuse you; the average aerospace engineer salary in the L.A. area is probably closer to $100K/year - but that's only half of the story.  The cost to a company of a $100K employee must also include insurance, vacation time (hmm... vacation... the word sounds familiar, but I can't quite place the meaning...) and other costs that add anywhere between 80% and 120% to the basic salary, depending on the company, location, etc.  So I would not be surprised if the average cost per labor-year in the SpaceX case is closer to $150K than $50K.  By the way, this figure is historically one of the most guarded secrets of aerospace companies - I don't know why, since it's quite easy to reverse-engineer (I lie - yes, I know why: many "classical" Cost-plus government contracts are won or lost based on these "fully loaded" labor rates.)

  Quote

  I think the cost for a F9 will get to at least 100mil which still not being full EELV class is no bargan.

  I disagree - if a 10MT-payload F9 demonstrates a better than 90% reliability and can be bought at $100M/launch, I would consider it a real bargain.

• #154 by antonioe on 02 Sep, 2007 03:42

• Quote

  Carl G - 31/8/2007 11:10 PM what was it that got you interested in designing space vehicles, Antonio?

  As long as I can remember I've been in love with flying and airplanes (not space!)  Thank goodness for my poor vision, or else I would have become a professional pilot... but I could not, and ended up an undergrad at MIT's Aero&Astro Dept.  Then one day around 1970 a fellow by the name of John McCarthy ("Big John" - former North American Aviation executive, later Director of NASA's Lewis Research Center) gave a lecture at the Dept. on this fascinating new project called the "Space Shuttle".  Two things happened to me that day:  First, I was for the first time exposed to these things called "Viewgraphs" - and with lots of engineering data in them, the likes of which I had never seen!!!  Second,  I began to be interested in space vehicles... look, some of them have WINGS!... :laugh:

  Quote

  we're all a bunch of armchair engineers on here

  So was I until a certain David W. Thompson called me on the phone one fine spring day in 1986.  The fact that he did, and hired me, shows how desperate he was.  How I managed, with NO hardware experience at all - all I had done professionally at that point was design guidance equations at Draper - to pull together the Pegasus team still amazes me (well, I had lots of help).  That's why I'm so leery of pre-judging people that attempt to build a launch vehicle with no previous experience.  I'm probably much wiser and experienced now, but I certainly was neither then.

  Quote

  But I'm sure you and maybe Dr Stanley are the only ones here who are at the sharp end of thinking vehicles up

  Reserve your judgement until you see a picture of Taurus II - it will appear extremely dull to you.  But underneath that dull-looking skin, there's not 20 years, but more like 50 calendar years of experience by a lot of people much smarter than me!...

  ---

  
• #155 by antonioe on 02 Sep, 2007 04:25
• Ohmygod!... I just realized... I joined Orbital 30 years after Sputnik... and that was TWENTY years ago... look what was accomplished in the first 30 years... and what was done in the last 20... how depressing... pardon me while I go to a corner to sulk...
• #156 by antonioe on 02 Sep, 2007 13:40
• Quote

  Andy L - 1/9/2007  6:37 PMSome of the OSC guys aren't very nice about SpaceX on their threads here.

  
  *DOUBLE SIGH* - I found this in another thread (notice the signature line):
  
  

  Quote

  Quote

  pad rat - 8/3/2007 11:07 AM "Some"?!!! Falcon 9 is an EELV-class booster with over 30% more capacity than a 7920H to a reference LEO! That's why we were disappointed to hear about the Falcon 5 being shelved. It did a good job of hitting the Delta II sweet spot and competed with the K-1 for a lot of Delta II-class missions.

  My sentiments exactly. It would seem that F9 superceded F5 (first F5 evolved to a wider tankage that could be stretched for F9, then it was dropped entirely.) I assume that SpaceX could develop F5 if somebody wanted to pony up the dough, but the development costs wouldn't be worth it. Elon wants to make his money off COTS & Bigelow, and F9 is the way he's going to do it. Then again, Elon is very wise to say "The way to make a small fortune in the launch business is to start with a large fortune." Hopefully he will prove himself wrong by the time he's ready to cash out at SpaceX.
  
  -----
  Minotaur, f**k yeah!

  
  BTW, the individual with that signature line seems to be somebody in New Mexico NOT associated with SpaceX.
  
  Look, guys, we all have enough problems already...
• #157 by Seer on 02 Sep, 2007 14:30
• Hi Antonio, great to hear your plans for Taurus 2 and your thoughts on Spacex.
  
  I was also mysterified about spacex's pricing, as Spacex is leaving over $50 million on the table for some geo launches compared to Ariane 5.  $80 million a launch would probably put Ariane out of business. So either Elon is feeling really generous toward the satellite makers or he is going to raise prices later on. But why publish a set of prices now if your only going to raise them later?
  But you prompted me to realize that those low prices are detering potential competitiers from developing vehicles. I don't think Elon had you in mind, but it can't have helped Kistler win over investors for the k1.
  Low prices for falcon 9 must have been a factor for NASA choosing Spacex for COTS. Kistler was a shoein for one of the places because they were the furthest along in development, and promised the most bang for buck: the first fully reusable vehicle. That left Spacex to fight it out with the rest of the proposals, including all the proposals that included using existing boosters, such as the Atlas 5, to launch a capsule. Using an existing booster would mean less development  risk and quicker deployment. To beat that Spacex needed to make its booster much cheaper than anything else.
  However, that price of $35 million to leo only holds for government launches - spacex can charge as much as it likes for commercial contracts to leo and geo. Thats a market of well over a $1 billion, most of which is profit. All Spacex loses out on is the money it could have charged NASA and the DoD for Atlas 5 and Delta 2 class missions. How much is that? Half a dozen launches a year? Perhaps $300 - $400 million dollars profit. That assumes both the delta 4 and atlas 5 lines are shut down, which probably wouldn't happen.
  All in all, it shows what a shrewd businessman Elon is.
  
  
  
  
  
  
  
  
  
  
  
• #158 by Jim on 02 Sep, 2007 15:17
• "But you prompted me to realize that those low prices are detering potential competitiers from developing vehicles."
  
  Not true, the glut of available LV's is detering development of new ones
• #159 by Seer on 02 Sep, 2007 16:14
• err...Jim, why then is Antonio thinking of developing a new vehicle?