Because of layers of disagreement, I will present two cost breakdowns here, one skeptical and one rosy. Government cost accounting can be a pretty slippery subject, and much depends on the favorable or unfavorable realization of the project. Note that the numbers presented here derive from sci.space.policy newsgroup discussions between prb@clark.net (Pat), tabbott@intellex.com (Tom Abbott), and gherbert@crl.com (George Herbert)
Much of the disagreement rests with the cost accounting for the existing Space Shuttle program, as well as the historical record of cost overruns for large development projects (witness the current space station project). There are underlying currents of an argument between gov't funded and privately funded space development, as well as an debate on the leverage gained by utlizing infrastructure and manpower currently already in place for existing Shuttle operations.
HLV _ SHuttle Derived- Expendable.
DDTE $5G. The Augustine commision stated this could be done for $2G, but given their history of over-runs, I'd not expect anything less then 4G, and that would be with severe penalty clauses for non-performance. And with the lobbysists Lockheed owns, it is highly unlikely to expect real performance from them.
Mission Fly-away costs:So we can expect to drop $300 Million on hardware per vehicle, and $250 Million in servicing costs.
Flight test program:So DDTE looks to be anywhere between $5-$7 billion for a Shuttle-C.
Now assuming a space station, moon base and Mars base, we can expect no more then 15 missions, so the derived costs per mission run $250M + $7G/15 = $650 Million per mission.(100 Tons) $6,000/ton base costs. given an interest rate of 6% and 5 year developement program we can add about 15% in costs onto that for interest. the numbers of course get worse or better to some degree depending upon flight rate, etc... Now the abbott mind of course writes off developement costs and interest and insists that lockheed can do this on schedule and budget. HaHa. Abbott also insists that a HLV can be made re-usable, but in which case, the DDTE rises even higher, and the per mission servicing costs rise as there are more systems to service. I ballpark estimate a re-usable pod based system at $9G DDTE, with an extra $50 Million in mission service costs, but a savings of $100 million in consumed engines and a savings of $50 million in consumed electronics. I leave the mission costing for that model to any reader except Mr abbott. George Herbert ran these numbers for anyone interested. Now what would be interesting is a sensitivity analysis of an HLV costs to Mission rate, DDTE and servicing costs. My gut instinct is it is insensitive to all but mission rate and on mildly sensitive to that. I try to throw out of consideration ideas like writing off the DDTE of an HLV, psychotic ideas like KSC increasing launch rates or re-usable hardware being easy to develope from shuttle heritage equipment.Concerning your dismissal of their $2 billion (1990) estimate to build Shuttle-C, I would just say that $2 billion is what it cost to build the last space shuttle, and a shuttle-derived heavy-lift Shuttle-C cargo vehicle is a very abreviated version of the shuttle, so why should a Shuttle-C cost more than building another shuttle? By all rights, is should cost less, and the $2 billion cost to build the last space shuttle includes all that government waste you worry so much about.
Here are the figures from NASA's budget webpage:
SHUTTLE OPERATIONS
BASIS OF FY 1998 FUNDING
REQUIREMENT (Thousands of Dollars)
(FY 1996, FY 1997, and FY 1998 shown)
External Tank 327,500 (96) 339,000 (97) 359,700 (98) Space shuttle main engine 185,000 (96) 182,300 (97) 184,900 (98) Reusable solid rocket motor 395,700 (96) 427,000 (97) 434,600 (98) Solid rocket booster 153,300 (96) 150,400 (97) 157,700 (98)
Now according to these figures the ET costs $42 million per shuttle launch (8 shuttle launches per year divided into $339 million); and two solid rocket boosters come to about $72 million.
>Proportional overhead of KSC $250 Million. ($400 - $150M) due to simplerI'm not sure what you mean by "proportional overhead." Are you saying for example, that it cost more than some $60 million to relaunch the space shuttle Columbia?
A Shuttle-C heavy-lift vehicle would require about the same level of processing as the Columbia relaunch. Boeing was estimating they could turn around their reusable two-SSME Shuttle-C derivative in 24 hours, and I don't see why a three or four-engined Shuttle-C would be much different (other than the solids). Normal processing for a space shuttle launch is 90 days, so a Shuttle-C will require only a fraction of the processing a space shuttle requires, if Boeing's claims are in the ballpark.
>$50M consumed electronics (Sensors, communications, telemetry, GPC's)A reusable Shuttle-C would reuse such things.
>So we can expect to drop $300 Million on hardware per vehicle, and $250 Million in servicing costs.
Costs for Expendable Shuttle-C:So your worst case scenario costs $550 million per vehicle, and my best case scenario (ignoring estimates as low as $50 million) is $114 million per vehicle. Quite a discrepancy.
Flight test program:I can go along with four test flights, although it should be noted that the space shuttle was carrying payloads not related to shuttle testing, on its second flight into space.
Four Shuttle-C test launches would cost about $100 million each, for the hardware, for a total of $400 million.
Not even close. Consider:
Reference Original Page
08/01/87
Rockwell to build 4th shuttle for $1.3 billion
By MARK CARREAU
Houston Chronicle
In a critical sense, construction of the new orbiter began even before the Challenger accident. In early 1983, Congress authorized $400 million to build structural spare parts for the shuttle fleet, components that will become the backbone of the new spacecraft.
About 90 percent of the work on those spares is complete. Components include:
- A crew cabin, aft fuselage, engine compartment and payload bay doors, all built by Rockwell.
- A midfuselage built by General Dynamics Corp.
- Two wings, built by Grumman Aerospace Inc.
- A vertical tail fin with speed brake assembly, built by Fairchild.
- Re-entry rocket pods built by McDonnell Douglass Co.
Note that all these space shuttle components cost $400 million in 1983. They were built under those high-cost, inefficient government contracts Pat loathes so much. Nestled among those space shuttle components, is the engine compartment. Now, the engine compartment is just about the only piece of the space shuttle that Shuttle-C needs to turn the shuttle stack into a heavy-lift vehicle. We can surmise that the engine compartment must cost less than $400 million to build. If we take this engine compartment and add 3 space shuttle main engines to it at a cost of about $60 million, we have built Shuttle-C's propulsion module. All that's left is to attach a cargo container to the top of the propulsion module and attach this combination to the ET and solid rocket boosters and we have a Shuttle-C ready to launch.
Shuttle-C costs:Adding in 4 test flights at $400 million increases the cost to $1.4 billion. This leaves $1.6 billion of Shuttle-C's $3 billion development budget (1997 dollars) left for such things as program costs and making the boattail reusable. I think $3 billion to build a Shuttle-C is a reasonable figure.
Pat B.: Now assuming a space station, moon base and Mars base, we can expect no more then 15 missions.
Tom Abbott: If we change crews on a Moonbase every six months, we'll need a minimum of 4 heavy-lift launches per year. A ten year Moon program would therefore require 40+ heavy-lift launches. A Mars program would use similar numbers although producing fuel on Mars would probably cut that in half. I would say there is a market for at least 60 heavy-lift launches if we plan on exploring the Moon and Mars on any kind of sustained basis. The ability to produce fuel in space would reduce these numbers, since most of what Shuttle-C would launch would be fuel.
Pat B.: Abbott also insists that a HLV can be made re-usable.
Tom Abbott: Me and Boeing.
Pat B.: ...but in which case, the DDTE rises even higher...
Tom Abbott: Well, I have $1.6 billion to play with. That ought to cover it.
Pat B.: ... and the per mission servicing costs rise as there are more systems to service.
Tom Abbott: A short turnaround of about a week couldn't be too costly. Fish the Shuttle-C propulsion module out of the ocean, attach another payload to it, attach this combination to an ET and solid rocket boosters and we're ready to go again.
Pat B.: ... psychotic ideas like KSC increasing launch rates
Tom Abbott: They already did, it was called the Columbia relaunch.
Pat B.: ... or re-usable hardware being easy to develope from shuttle heritage equipment.
Tom Abbott: Boeing doesn't agree with you.
BTW, I looked up Kistler's numbers: They claim they should be able to put 4.5 tons in low-Earth orbit for $17 million. That would come to about $377 million for Kistler to put 100 tons in LEO (22 launches). Shuttle-C can do it for around $100 million (one launch).Pat B.: Kistler will recover their DDT&E.
Tom Abbott: If we put Kistler in charge of launching our Moon/Mars mission they will certainly recover their development costs because the American taxpayer will end up subsidizing them to the tune of billions of dollars. And that's just for the first mission. The bottom line is what it costs the taxpayers to develop space, and heavy-lift saves billions of dollars over any other method of doing so, including Kistler's.
George Herbert: I have said this before, and will say it again: $100 million for a Shuttle-C launch is at least a factor of 2 lowball from actual costing, based on the costs for the components. The SRBs are $60 million/pair; the ET is about $60 million each.
Tom Abbott: George, the last time I looked, the ET was about $46 million. I'll agree on your cost for the SRB's.
George Herbert: I got $59 million apiece dividing the total output of tanks per year by total payments to that factory from NASA (which exclusively makes ETs); the "price per" isn't the whole price paid for manufacturing them, which is a more fair price to use for comparing.
The SSMEs are around $20 million each. Even if you use used SSMEs near their lifetime limits...
Tom Abbott: A reusable Shuttle-C would purchase the space shuttle main engines (SSME) once, with the money coming from the construction budget, and then reuse them, so their costs are not figured into the marginal launch cost.
George Herbert: Again, the reusable Shuttle-C proposals are more R&D expensive than expendables of equal technology assumptions, and thus incur more R&D amortization load. It comes out a wash.
Phil Fraering: The shuttle has a fair amount of marginal costs on its SSME's and also has to replace them fairly often.
Tom Abbott: Phil, engine servicing is included in the fixed costs of the shuttle budget, not in the marginal launch cost. It is assumed that the workforce which currently services space shuttle engines can squeeze one or more engine sets into their work schedule without increasing the fixed costs, up to a certain number of extra flights. After that number is reached, then the fixed costs must rise.
George Herbert: The rest of the vehicle, integration, and operations costs appear to require minimal margainal costs of $200 million per flight. More realistic estimates cluster around $220-250m/flight.
Tom Abbott: These costs are considered fixed costs up to a certain number of additional launches, and are also not included in the marginal launch cost of launching an additional shuttle or shuttle-derived heavy-lift vehicle.
George Herbert: That's not including any amortization of the R&D cost for the Shuttle-C, which at $3 billion worth of R&D and 10 flights/year would be at *least* $30m/flight (zero ROR/10 yr payback) and more likely much more (10%/10yr gives you $45m/flight, 10%/5yr gives you $75m/flight, etc). If the flight rate is less than 10/year, scale appropriately. For example, if you only use two flights a year, 0%/10yr requires $150m/flight amortization, at four flights, it's $75m/flight.
Tom Abbott: Amortization in this instance just confuses the issue, IMO. We don't figure the space shuttle's development costs into the yearly cost of launching shuttles and the same should apply to a heavy-lift Shuttle-C. The $3 billion needed to build Shuttle-C is an investment which will pay the space program back many times over.
George Herbert: Shuttle-C only wins out economically if Kistler uses realistic commercial financial practices and NASA doesn't and flys 3 or more flights per year on the Shuttle-C, for which payloads appear nonexistent.
Tom Abbott: Shuttle-C would pay for itself in one Moon mission. Consider:
George Herbert: Not to mention that those Shuttle-C costs won't be coming down much,
Tom Abbott: I don't see how you can say that. USA is reducing cost of operations and many improvements for the shuttle are coming, such as liquid boosters which should lower costs and increase lift capacity, and we can always add a fourth SSME to the Shuttle-C propulsion boattail to increase the tonnage even more, and for the same marginal launch costs! In fact, we should start out with a four-engined Shuttle-C, like Zubrin describes.
George Herbert: Bunch of bad assumptions there, Tom. First of all, STOP using $100m/launch for Shuttle-C: you are killing your credibility. You do not win arguments by lowballing your own costs. Cost to launch those 500 tons to LEO is thus about $4.25 billion ($250m/flight minimum) not $3.5b.
Tom Abbott: The Department of Defense estimated the cost to add an extra space shuttle flight to NASA's schedule at $50 million per launch. NASA is putting out estimates of $50 million to $80 million to launch an additional shuttle, and claimed they launched the shuttle Columbia for $59 million. And now, my friend, Robert Oler, communicates to me over on Compuserve, that he is presently doing a study of space shuttle launch costs for an unnamed Senator, and Robert estimates the marginal launch costs at about $70 million per additional space shuttle launch. Dividing the number of ET's flown per year into the total ET budget does not give the true cost of adding one more ET to the mix. The same goes for solid rocket boosters. ET's and solid rocket boosters and the fuel that fills them are the marginal launch costs. At any rate, I think I'm being generous in quoting a $100 million figure for the marginal costs of launching an additional space shuttle or shuttle-derived heavy-lift vehicle, considering that several groups and individuals who should know, are quoting smaller figures.
I understand the opportunity costs to NASA and the taxpayers of choosing Shuttle-C: NASA's first Moon/Mars launch will cost them $3.5 billion using Shuttle-C. Each subsequent launch of 500 tons to low-Earth orbit (one Moon/Mars mission) would cost NASA $500 million using Shuttle-C.
I assume a basic human habitat which could probably be built for $500 million using an External Tank, but if NASA is going to handle it you know it will be more. The basic Laboratory Module and Habitat Module and their connecting Nodes cost about $5.6 billion. Of course, if you want to assemble at ISS, then you can rent space from NASA for a lot less, I would imagine. But you still have to have an orbital transfer vehicle, probably two, and NASA is talking about spending $1 billion on the space station control module, which is probably a good model for our orbital transfer vehicle.
And for reference, you need about 80 tons of fuel to put a useful payload on the Moon or Mars.