Making Rocket Propellants from Martian Air (1978)

Water frost on Utopia Planitia as imaged by the Viking 2 lander. The horizon appears tilted because Viking 2 alighted with one of its three foot pads on a large rock. Image credit: NASA.
In the late 1970s, through the initiative of its director, Bruce Murray, the Jet Propulsion Laboratory (JPL) studied a range of possible Mars missions, including Mars Sample Return (MSR). Murray and others at the Pasadena, California-based lab were aware that funds for new Mars missions would be hard to come by; the U.S. economy was under strain and NASA, JPL's main customer, was devoting most of its resources to developing the Space Shuttle.

In addition, equivocal data from the astrobiology experiments on the twin Vikings, the first successful Mars landers, had been interpreted as negative, helping to damp public enthusiasm for the Red Planet. Would-be Mars explorers reasoned that, if an MSR mission would stand a chance of being accepted, then they would need to find technologies and techniques that could dramatically cut its anticipated cost.

In July-August 1978, two years after the Vikings landed and looked for life on Mars, three engineers at JPL — Robert Ash, a visiting faculty fellow from Old Dominion University in Virginia, and JPL staffers William Dowler and Giulio Varsi — reported on a small study they had conducted of one such cost-saving technology: specifically, making MSR Earth-return rocket propellants from martian resources. Using Earth-return propellants made on Mars would greatly reduce the mass of the MSR spacecraft at launch from Earth, permitting it to be launched on a small, relatively cheap launch vehicle.

Earlier researchers had proposed using Mars resources to make rocket propellants, but Ash, Dowler, and Varsi were the first to base their study on data collected by spacecraft on and in orbit of Mars. The Viking landers had confirmed that martian air is made up almost entirely of carbon dioxide, and had found that the planet's rusty red dirt contains an appreciable amount of water. The Viking 2 lander, at rest on the northern plain of Utopia Planitia, had imaged water frost on the surface in winter (image at top of post). In addition, the twin Viking orbiters had imaged water ice clouds high in the atmosphere and polygonal terrain resembling that found in near-polar permafrost regions on Earth.

Ash, Dowler, and Varsi examined three propellant combinations that would exploit resources the Vikings had found on Mars. The first, carbon monoxide fuel and oxygen oxidizer, could be produced by splitting ubiquitous martian atmospheric carbon dioxide. They rejected this combination, however; while easy to produce, it could yield only mediocre performance.

Hydrogen/oxygen, on the other hand, was a high-performance propellant combination with more than three times the propulsive energy of carbon monoxide/oxygen. It could be produced by collecting and electrolyzing (splitting) martian water, but Ash, Dowler, and Varsi rejected the combination because a heavy, electricity-hungry cooling system would be needed to keep the hydrogen in usable liquid form. This requirement would, they estimated, negate the mass-savings of making Earth-return propellants on Mars.

The third combination they examined was methane/oxygen, which could be produced on Mars using a process discovered in 1897 by Nobel Prize-winning chemist Paul Sabatier. Combining a small amount of hydrogen brought from Earth with martian atmospheric carbon dioxide in the presence of a nickel or ruthenium catalyst would yield methane and water.

The methane would be pumped to the MSR Earth-return rocket stage fuel tank and the water would be split using electricity to produce oxygen and hydrogen. The oxygen would be pumped to the MSR Earth-return oxidizer tank and the hydrogen would be reacted with more martian atmospheric carbon dioxide to produce more methane and water.

Ash, Dowler, and Varsi favored methane/oxygen because it would provide 80% of hydrogen/oxygen's propulsive energy, and because methane remains in liquid form at typical martian surface temperatures. They estimated that launching a one-kilogram Mars sample directly to Earth (that is, with no stop in Mars orbit to rendezvous with and transfer the sample to a Earth-fueled Earth Return Vehicle) would require manufacture of 3780 kilograms of methane/oxygen, and calculated that a Mars surface stay-time of at least 400 days would be necessary to allow sufficient time to manufacture adequate quantities of propellants.

The 1978 JPL study would inspire many other mission designers to tap resources the twin Vikings had confirmed exist on Mars. At the 1982 AIAA/AAS Astrodynamics conference, for example, Science Applications Incorporated engineers presented a paper on use of Mars resources to make propellants for automated rocket-propelled ballistic hoppers and propeller-driven airplanes. The Mars base scenario developed at the second The Case for Mars Conference (1984) relied heavily on extraction of resources from the martian atmosphere for both life-support consumables and rocket propellants.

Conceptual design of a large system for extracting propellants and life-support consumables from martian air. Image credit: C. Emmart/Boulder Center for Science and Policy.

"Feasibility of Rocket Propellant Production on Mars," R. L. Ash, W. L. Dowler, and G. Varsi, Acta Astronautica, Vol. 5, July-August 1978, pp. 705-724.

"In Situ Propellant Production: The Key to Global Surface Exploration of Mars?" AIAA-82-1477, S. Hoffman, J. Niehoff, M. Stancati; paper presented at the AIAA/AAS Astrodynamics Conference in San Diego, California, 9-11 August 1982.

The Case for Mars: Concept Development of a Mars Research Station, JPL Publication 86-28, NASA Jet Propulsion Laboratory, 15 April 1986.

More Information

A Forgotten Pioneer of Mars Resource Utilization (1962-1963)

Gumdrops on Mars (1966)

Astronaut Sally Ride's Mission to Mars (1987)


  1. This is interesting because that it shows that methane/oxygen ISRU on Mars had been discovered long before Bob Zubrin imagined Mars Direct in the late 80's.
    The link between those early studies and Bob Zubrin is Martin Marietta. The company build the Viking landers in the 70's, and ten years later (in 1988) a young Zubrin landed a job there. Zubrin worked with seasonned engineers that had created the Viking spaceship a decade before, and that explain why and how he had the idea for Mars Direct.

    1. The connection with Viking is not that Martin Marietta engineers built the Viking landers and Zubrin eventually worked at Martin Marietta, it is that scientists using Viking data confirmed the existence of and wrote about potential resources on Mars. Ash, et al, were first, but there were many, many others. I cite two others in the sources for my post.

      Please note that the concept of using Mars resources was first put forward in the early 1960s. The difference in 1978 was that scientists and engineers had real data on Mars conditions from Mariner 9 and (especially) the Viking landers and orbiters.

      David Baker & Robert Zubrin's Mars Direct concept was a synthesis of concepts going back decades, just like most plans for Mars expeditions. Baker & Zubrin first presented Mars Direct in 1990, about a year after the Space Exploration Initiative got its start, when the Mars planning environment was in ferment. The concept of using propellants made on Mars was "in the air," so to speak.

      One must approach Mars Direct with some circumspection. On the one hand, it was actually unworkable as first proposed. On the other hand, it helped to spawn workable concepts, some of which Zubrin helped to develop. I've heard Mars Direct called a very useful "parable." It definitely shook up the Mars planning trade space.

      Incidentally, the concept of reaching Mars without propellants for the return trip first appeared in SAIC's split/sprint mission plan for Sally Ride's LEADERSHIP AND AMERICA'S FUTURE IN SPACE report in 1987. That is sometimes mistakenly attributed to Mars Direct. In both cases, it aroused anxiety, though both SAIC and Baker & Zubrin took pains to provide abort and back-up modes.

      Mars Direct remains known - unlike other equally significant concepts of the 1980s and 1990s - mainly because Zubrin has remained its aggressive champion. He has turned promoting it into a business.


  2. Great article! I've been a huge fan of yours for many years. I wrote a similar article to this one as a high school math lesson over at RocketSTEM magazine that uses the Sabatier method for making methane fuel. It's based on the movie "The Martian" ( Thanks for the inspiration, and keep up the excellent work. I always look forward to your articles.

  3. Joe:

    Your article is nothing like mine - yours has all the hard stuff in it! :-) Thanks for the kind words.


  4. The early economic model for making things on Mars will be overwhelmingly dependent on non-air-breathing robotics. Expensive, heavy, slow all-in-ones that will need to mine, refine, inventory, craft, and manufacture. I personally don't know with today's technology if this is do-able in 1mt packages, such as land now (Curiosity). In the Emmart painting above, I'm guessing that's about 10mt minimum of "stuff".

    In addition to fuel, water, and gas production, you'll need...
    Hard manufacturing:
    1. Mine - what metals can be pulled from the regolith and inventoried?
    2. Refine - separating the useful metals/alloys from silicon and other regolith.
    3. Inventory - a transport robot that will stack up ingots for use.
    4. Craft - 3D printing, forming, lathing, etc?
    5. Manufacture - assemble and make useful things. Like dome components.

    Farming of a sort, harvesting of biologic materials (nitrogen, carbon, etc) and mitigation of the perchlorates.

    This really is the ultimate engineering problem. If the economics can ultimately make it happen...

  5. Propellants are typically the heaviest part of any Mars expedition. Life support consumables are also non-trivial. If you can get at those on Mars, that's significant.

    The gadgetry in the Emmart art is an artists' concept and assumes a really big base; 35 people, permanently staffed, cyclers, all Mars-to-cycler propellants (CO/LOX) made on Mars. I should write about it.


    1. Yes please! The Case For Mars study is nearly unfindable aside from the Emmart art. So a comprehensive article about it would be appreciated.

  6. This is why I somewhat obsessively read your history posts. Each of these ideas was a concept given the knowledge of the day. Many early ideas were based on the assumption of Mars having a much thicker atmosphere and even higher oxygen content. So as our knowledge increases, the model adapts for future planetary expansion.

    Emmart's artwork would have to assume really HUGE lift capacities with our Earth-originating rockets. Which means, there's an equally HUGE gap in what we could lift in his day, and what he has pictured. Werner Von Braun depicted almost invasion-force armadas going to Mars and nearly hundreds of launches. (You will know the count better than I).

    Several administrations and depressing fiscal years later, all this seems laughable.

    But now we know a few more things than in Emmart's day.
    1. Robots will help us, and improve every year (ask DARPA)
    2. Computer processing power is off the charts.
    3. New manufacturing methods are in play (3D printing, the most popular these days)
    4. We can build dome stadiums, which are analogs (structurally) for dome habitats on Mars.

    It's exciting because we now know Mars is much, much harder to reach, but following tech trends, the gap gets smaller every day. Once we can manage Mars, what stops us from the Moon, Callisto, Europa, Titan, etc? It's crazy to imagine. Versatile robotics should drive down the costs with replication much like the personal computer boom did of the 80s and 90s. Next big hurdle is throw weigh vs. cost. Propellants, to your point.

  7. Ben:

    This makes me think back to the old argument that we did Apollo too soon. I think that's slightly silly - stuff happens when it does - but also profound. We weren't ready as a society to make a long-term commitment. Our internecine conflicts pushed us to the moon as surely as did the Saturn V rocket. And the tech was just barely capable of reaching to moon and giving us three days to look around.

    Now we're creeping up on an era when - if we do it right - space won't be that hard. A few problems I see now that keep that farther in the future (and harder) than it would be otherwise: fixation on Mars (picking up where we left off in 1972 would be a walk in the park by comparison); fixation on commercial space (the modern-day equivalent of Space Shuttle's false promise); fixation on cost (space is cheap as measured by any reasonable yardstick); fixation on astronauts (the human side of the house has fought progress every step of the way to the point of endangering crews and programs); fixation on pure robotics (teleoperated robots will be more effective); stupidity and ignorance; issues on Earth that will tie us down (climate change, etc.).

    Don't get me wrong, I like Mars a lot. Hints that life might exist there, which seem to mount up everyday, make me think we should be circumspect in how we approach it, however. Our history of interacting with new (to us) living things is not encouraging. I say explore Mars by robots guided from Earth, by teleoperations from inhabited spacecraft in Mars orbit, but keep people off until we have a clearer understanding of what we're dealing with on the Red Planet.

    BTW, von Braun expected we'd need more than 900 launches of his big ferry rocket to get his big Mars expedition into space. Most of that would be propellants.


    1. Wow. I thought when I said "hundreds" I might be a little long. 900. Just wow.

      Agree your points - particularly Moon.
      Analogous to my background, pharmaceuticals, you have 3 phases to bring a drug to market.

      FDA Model
      Phase-I Don't kill anyone.
      Phase-II Make sure it does what it claims.
      Phase-III Make sure you can repeat the benefits across genotypes, ages, genders.

      So for Moon I think like this:
      Phase-I outpost in Antarctica and so on. (These already done/in progress)
      Phase-II outpost in Near Earth (ISS)
      Phase-III outpost on Moon

      For Mars, similarly:
      Phase-I Antarctica or Near Earth
      Phase-II Moon
      Phase-III Mars

      If we aim to colonize Mars with (US) governmental backing, we'll follow a low-risk process like this, which NASA pretty much articulates in their recent plan. High-risk plans like Mars-One could kill some people (volunteers, have at it) as evidenced by the MIT study which refuted many of it's assumptions.

      The exciting thing is, it's technologically plausible and planned (ok, yeah, about the 20th plan, but still).

      I am still curious why NASA is not planning a centrifuge partial-gravity demo. A bucket-load of engineering work-arounds with human health would be resolved. (food, waste, muscle degradation, etc) Do one hard thing, solve five.

  8. Great article as always. History always has lessons for us.


I like hearing from my readers. No rules except the obvious ones - please keep it civil and on topic.

Advertiser comments have led me to enable comment moderation.