The overlap between the dreamers of space exploration who see Sci-Fi as a potential source of inspiration and the American mythos built around the idea of the Frontier is not near as strong as many pretend. After you stretch the analogy of space and the American Frontier, or America in general as the New World, about as far as you can on emotionally, you find a much longer list of tendencies in the American psyche that currently work against the advance of large scale space exploration, and even against Sci-Fi as connected to space exploration.

The particular flavor of American anti-intellectualism is strong, getting worse as aerospace, unlike the American Frontier, is stuffed full of the educated. When there is that emotionalism of the Frontier, about pushing boundaries as a proper and useful drive, American attitudes gather up and favor space exploration. Sci-Fi loves this emotional angle. Inversely, to the degree that the inspiration from space exploration becomes about humanities growth, the crowd as quickly draws away, repelled by the Kubrick-like advance of the mind. Our visions of the space frontier are complex undertakings, filled with scientists and engineers. Sci-fi that makes a “wild” west of the setting, in this respect avoids how we get from here to there, from today’s astronaut and engineer and physicist, to the dirt farmer on the uncharted planet. Instead, we get stories about a Federation, with no interest in what is happening back on Earth. Alternately, sci-fi takes us warts and all, adds in spaceships as if they’ll be around one day and taken for granted just like airplanes, only to be fleeing from our own creations run amuck.

The best Sci-Fi stories have come to tell us we can change, we can step out of the pattern that cut us, and what happened before doesn’t have to happen again. Yet this applies as much on Earth, as out there. The Sci-Fi that’s rarely written would get people to see the possibilities in this moment in time, connecting how we get from here to there – stories where life on Earth is so much better because of how we must change along the way, and space exploration is the result, not the cause.

Your confusing the numbers on budget data again.

Look at the budgets BEFORE Columbia. After that the transition began to wind down the Shuttle's continuous upgrades budget's where were an additional $2.5B a year atop the operational $4B a year. Totals usually get rounded by people to around ~~ $1B a flight for Shuttle because they are taking this type number and dividing by an average launch rate of ~~ 5 a year.

These were not development dollars per se like the 70's, but rather dollars necessary to continued ownership and operation. Had we kept the Shuttle, we would have had to keep that large "upgrades" bucket of ongoing continuous developments for upgrades. i.e., Shuttles annual "expenditures" would be a good way to put it (consisting of operations AND upgrades) was about $5B a year back in the day, and would have been more now had we kept the system, adjusting for inflation, etc.

Your confusing Shuttle's recurring vs. non-recurring dollars here again.

One FIRST type of Shuttle dollar was up-front, back in the 70's. Mostly an ops comparison would not include these numbers against another systems prices (albeit, a company like Orbital or SpaceX may be recovering some up-front private investment dollars in downstream pricing).

The SECOND type dollars is what you'll find in budgets as "Ops, Manufacturing" etc. This was yearly, consisting of assorted fixed and variable costs from Utah to MAF to Rocketdyne to USA at JSC and KSC, etc. This was about $4B a year at the end of Shuttle (going to years like 2003-2006, before the layoffs started, knowing the program would end and flight rates were winding down).

The THIRD type of Shuttle dollars was also EVERY YEAR, the "upgrades" budget that was part and parcel of owning. This was about ANOTHER $2.5B a year.

Usually, to draw a strict $'kg comparison for Shuttle against anything, a recurring cost comparison should use the 2nd and 3rd types of dollars, and a full life cycle everything comparison would include the first kind of dollars.

Lets try and keep numbers and their content straight here...and then add in context.

Many of the responses here are correctly noting the difference, at times lost like a nuance, between the Falcon 9 price to the private sector, delivering a payload/sat to LEO (say 200nm, 28.5 circ.) and the Falcon 9 PLUS Dragon price to NASA, for delivering NASA's goods to the ISS (400nm. 51.6 circ.). The Falcon 9 "price" of $61.2M for a fully private customer does come in at $61,200,000/13,150kg = $4,654/kg. To ISS the proposition could subtract the prior for simplicities sake, leaving an UPPER bound on a Dragon (cargo version) for NASA of 133M-61.2M = $71.8M. That is, the remainder, Dragon added in and used by NASA, can be surmised.

The goods inside Dragon (NOT Dragon itself) would be about the 3,300kg of NASA goods. So NASA pays 133,000,000/3,300kg to the ISS or about 40,000/kg to the ISS - "in the door" (or hatch so to speak).

By way of a side-track-For the Atlas, one might work out the cost per kg to ISS by starting with prices to NASA of $187M for MAVEN in 2013 for a 10,000kg variant of Atlas. The Atlas operation would have to add a spacecraft to their price to figure the final number. By comparison to a Science payload, NASA did pay $87m for a Falcon 9 (Jason-3 mission). So that lends some insight into what the LOWER bound might be on Dragon (133M - 87M) leaving $46M for Dragon. It's reasonable to assume that Atlas as currently priced would be many times the $/kg to ISS vs. SpaceX. Perhaps Vulcan will change this (we hope!).

The Shuttle is another beast entirely, as you can't really "buy-by-the-yard". Two Shuttle launches could take up the entire slated cargo load contracted with SpaceX or Orbital. The Shuttle would take up between 7,500 to 10,000 kg per mission to ISS - remembering we have subtracted the MPLM here and adjusted for that Shuttle to ISS would be much less than the max Shuttle payload to LEO.

The Fixed costs alone per year for Shuttle would have been a few billion, plus the variable for the launches. Say we had kept the Shuttle, operating at $4B a year at the end (and lets NOT count the 2.5B a year in Shuttle upgrades - a continuous recurring yearly expense of ownership). Then we have about a billion per every 10,000kg to ISS, or $100,000/kg for Shuttle to ISS (goods "in the hatch").

Now to be fair, the Shuttle would have taken crew up too-along with any cargo. But also to be fair, there's that upgrades yearly recurring cost of ownership at $2.5B a year at the end of Shuttle (add that to the $4B a year). The advantages of Shuttle (flexibility, non-payload related capability, remote arm, crew, etc.) would have had to have been weighed against the flexibility "cost" wise of a commercial approach where a "buy-by-the-yard" approach to cargo was possible (not so with Shuttle).

Essentially, lacking this cost flexibility of moving payload to systems that were commercial, and that we could "buy-by-the-yard" we would have had NO funding for other things like SLS, Orion, or any major new developments, or STMD, or ISS R&D and so on.

Context is important.

While it will be curious to see the JPL analysis, some wording here gives me pause.

First -- The budget = inflation assumption is a poor one. A Mars exploration plan is needed that has a realistic budget increase and cost inflation outlook for the next decade and more. Should the actual be more generous, you increase the tempo, reduce the schedule, etc. But starting with the budget = inflation assumption is naive.

Second -- The timeline stretching first milestones almost as far out as the actual Mars exploration is a formula for showing how irrelevant all the interim activity is, year after year. It will not hold public or stakeholder or political interest. There have to be tangible achievements in the near term and throughout. Meaning actual accomplishments, NOT of the "welded this" or "completed study of that" or "went to CDR" type. Those are not tangible to stakeholders. The monolithic gathering up of all the small pieces till it all comes together decades from now is a formula for ever increasing irrelevance. That will lead to canceled programs and wasted time-again.

Third -- The NASA presence in LEO will likely not end with ISS, if anyone thinks of freeing up 100% of those funds. Many of these funds will not be available for new work anyway, as they are really Mission Control, JSC Mission Ops, etc. Other portions may go toward time on private space stations-as has been stated by Gerstenmaier. The cost of getting our crew and their needed cargo to those stations as we buy time by the day may not be part of the advance in the affordability equation of private stations. So again, not as much funding frees up from ending ISS in 2024 or 2028-and that's according to official NASA statements about a continued low Earth orbit presence post-ISS.

Fourth-when pushing dates out as far as these plans, the question must be answered, what is SLS doing in the years before Mars missions? Yes, again, take some of the post-ISS funding for some SLS flights here or there that may just keep that burner warm. Orion too. Unless the assumption is that somehow SLS and Orion have ready Mars related hardware in 2025 right after the end of ISS, using funding freed up there. Again, unlikely.

So all in all any Mars plan has lots less latitude to work with than the apparent rosy assumptions of many a Mars study. Budgets matching inflation (unlike anyone else in the federal government's outlook). Infinite stakeholder patience. Some extra money immediately so SLS and Orion have something to do between 2021 and possibly 2028. An immediate declaration that NASA will not use any private space stations or help these or keep any LEO presence or buy any time in LEO after 2024.

What else do you want- and a pony too!

It's doable, don't get me wrong. just not as a fantasy novel. Work the numbers. Get real. Keep it in the non-fiction shelf. And we'll all get to see Mars exploration sooner rather than never from having been lost in denial about how and what to plan around.

Yes...about the Orion service module...see my lengthier observation and more numbers below.

Putting aside the adjectives in the article that imply some bias, see my other post on costs numbers for Orion vs. Dragon, including a mod cost to make Dragon's cis-lunar capable. Can NASA sustain these types of Orion costs, and be oblivious to these potential opportunities? Want to go to Mar's or want to keep things, and way's of doing things, like Orion? This may end up as THE question over time.

Orion will not go to Mars. Any crew spacecraft in recent studies will probably remain in cis-lunar space, awaiting the return of the Mars craft, after which crew would get in the crew spacecraft for a return to Earth. In either case, the design of Dragon for crew is heading to be more capable of a direct Mars entry by choice of PICA, whereas the Orion Avcoat is inherently incapable of this. (Orion's choice of AvCoat was mostly to sustain the AvCoat industrial capability, nothing more).

Even spending the same again to upgrade Dragon crew to cis-lunar capable (about $1.5B, see my other post) would still be just about one and a half years of Orion's funding profile. Worse yet, once in missions, you can get Commercial Crew Dragon's into a simple set of variants, cargo, crew, etc. for much less per mission than Orion. Remember, when Orion is not flying it has NASA's full ownership, and that whole years FIXED costs. All that will go into a reflection of per mission costs as time goes by. For example, if Orion completes development in 2021 and then we fly Orion once again in 2023, the Orion per mission cost to that date is it's whole 2022 and 2023 budget, divided by ONE. (That example would be about $2B per Orion unit, as the burn rate will be about $1B a year).

A modified commercial crew Dragon, modded for cis-lunar space, from the recent contracts, shows every indication it's per unit costs would be much less than Orion's. In the range of a few hundred million a unit. BUT here's the kicker-you pay for the onsie-twosie NASA buys, and don't own the whole line as with Orion, having to keep that fixed Lockheed Martin cost in place even when not flying. The commercial crew experience already shows that even at just one a year these numbers will hold. Orion's one a year, or zero a year, burn rate, will be in the high hundred's of millions a year. Unsustainable.

Yep, I did the math. The Apollo Command and Service Module (CSM) came in at a LOW (there is uncertainty with very old cost data) of $20B in development in 2015 dollars. Orion is heading to $16.1, also in 2015 dollars, if development is complete by 2021, with every year of delay about an added billion. So Orion may not be the most expensive space capsule ever built, but it will be awful close!

About the cost comparison, the $2.6B you quote for crew Dragon is an amount that includes both development and use. The use would be about 5 flights, then you get the $2.6B tally (through 2022). To boot, that $2.6B contract amount includes the apx. 5 Falcon 9's for those 5 flights. The apples to apples with Orion's number (the $16.5B) would exclude the apx. 5 Dragon units, and the 5 launchers -unless you wish to add these as the first 5 flights of Orion (into the 2020's) atop SLS costs? The development of Commercial Crew Dragon, no service flights, no launcher, procurement, from contractual data, is actually just $1.5B.

For Orion, the $16.5 for development only, assuming complete by 2021, procurement only, is a good number. (I get $16.1B, with everything taken to 2015 dollars.)

That's 16.1 / 1.5 = 10.6 times as much up-front development cost, procurement dollars.

But here's the kicker, when the European trade was made for them to provide Service Modules to Orion, you'll see you can't find a mod to the prior Lockheed Martin contract on cost per year. In effect, the Orion scope was reduced to half the vehicle at the same prior development cost profile! So really the Orion number of $16.5B development is $16.5 PLUS some unknown amount of $ as European contributions; contributions that otherwise would have gone to ISS. So NASA basically lost money, or gave it away, as the service module was in the original Lockheed contract from what I can find, and we had that ISS contribution coming anyway. After the agreement, we still had service modules to come, but NO ISS contribution. Arguably, it's possible that holding Lockheed to the difference would have had to be made up somewhere, due to a time delay. Sad to say.

A mod to commercial crew Dragon for cis-lunar application? Suppose the upgrade cost the same again as the ENTIRE original development. That would only require about the $1.5B, which is a highly unlikely worse case. That's roughly the funding Orion will receive in any year and a half of it's usual cost runout.

And on recurring per flight or per unit or per mission costs...oh...don't get me started.