Architecture, ISRU, Mars

In Situ Air Production (ISAP)

The primary constituents of the Martian atmosphere are carbon dioxide (CO2), nitrogen (N2) and argon (Ar), whereas Earth’s are mainly N2, oxygen (O2) and Ar. Therefore, all the elements required to make breathable, Earth-like air are available in Martian air.

A Mars habitat may include inflatable extensions to significantly increase habitat volume beyond the size of the initial spacecraft. If the Hab is landed along with the MAV during a pre-deployment phase, it will be remotely activated from Earth, initiating the ISAP system, and causing the inflatable extensions to inflate. If the Hab is sent one launch window earlier than the crew, we will have about 20 months to inflate the Hab and test its other systems before the crew leave Earth. In order to inflate the extensions it will be preferable to make the necessary air using ISRU technology rather than bring air from Earth, which would require tanks of compressed O2 and N2.

The mass of the proposed ISAP system is likely to be less than the mass of full O2 and N2 tanks.  More importantly, having the capability to manufacture breathable air from local resources will be a useful advantage for the mission, providing increased safety and reducing limitations on the mission. Locally manufactured air can compensate for air losses due to leaks, airlock cycling, atmosphere scrubbing, refilling of O2 and N2 tanks in pressurised vehicles and marssuits, punctures if they occur, and other potential causes.

The DRA proposes separating N2 and potentially also Ar from the Martian atmosphere for use as a buffer gas. However, it is actually much cheaper and easier to use the gas that remains after CO2 and dust is removed from Martian air, as this is mostly N2 and Ar, both of which are perfectly acceptable buffer gases. The Martian atmosphere may also contain undesirable toxic gases such as ozone (O3), carbon monoxide (CO), and nitric oxide (NO), that exceed safe limits (as specified in JSC 20584: Spacecraft Maximum Allowable Concentrations for Airborne Contaminants). However, these can be relatively easily scrubbed out using ordinary COTS catalytic converters, and this process is far simpler and energetically cheaper than trying to separate out pure N2 from the air. (Note that NO is not actually toxic, but when combined with O2 it rapidly oxidises to form NO2, which is.)

Flowchart for In Situ Air Production system for Mars habitat
Flowchart for In Situ Air Production system for Mars habitat (click to enlarge)

Step 1: Mars atmosphere is drawn into the ISAP system through a dust filter.

Step 2: Water is removed from the gas mix via zeolite adsorption. The captured water is stored in the Hab’s water tank and used to replace recycling losses, and for O2 production.

Step 3: Water is separated into H2 and O2 via electrolysis. The O2 is stored for habitat atmosphere.

Step 4: Microchannel adsorption or cryogenic separation (CO2 freezing) is used to separate CO2 (about 96%) from the gas mix.

Step 5: The CO2 is reacted with H2 via the reverse water gas shift (RWGS) reaction:

CO2 + H2 → CO + H2O

The H2O produced is returned to the water tank. The CO may be stored for use in fuel cells, or it may simply be vented.

The gas mix that remains after CO2 is removed is mostly N2 and Ar, with trace amounts of various gases, which may include neon (Ne), krypton (Kr), xenon (Xe), O3, CO, NO, methane (CH4), hydrogen peroxide (H2O2), and sulphur dioxide (SO2).

Step 6: This gas mixture is first passed through an ozone scrubber, which reduces the O3 to O2.

Step 7: The resultant gas mixture is then passed through an ordinary automobile 3-way catalytic converter, which converts any CO and NO into CO2 and N2.

Step 8: The result is a safe buffer gas comprised mostly of N2 and Ar, with small amounts of O2 and CO2, and traces of Ne, Xe and Kr. This mixture can be combined with additional O2 to provide breathing gas for  the Hab. The CO2 level in this gas mixture is slightly higher than the proposed upper limit for the habitat atmosphere, but this excess will be removed by the ECLSS.

The DRA describes several methods for making O2 from CO2, including:

  • SOCE (Solid Oxide CO2 Electrolysis)
  • Sabatier reaction
  • RWGS reaction

SOCE requires considerable energy, whereas the other two options require H2. Fortunately, H2 is available because the Hab will contain H2O for the crew, which makes SOCE’s less appealing than the other two alternatives.

RWGS, which produces CO and H2O, is preferred over the Sabatier reaction, which produces CH4 and H2O, because in RWGS all the H2 is converted to H2O, which is a valuable resource in its own right, and from which H2 can be easily recovered via electrolysis. Using the Sabatier reaction would either consume H2, or the H2 would have to be recovered from the CH4. In theory CH4 could be used in a fuel cell to provide additional energy to the Hab. Combustion of the CH4 would produce H2O, which could be captured. However, it remains to be seen if fuel cells will be relevant for the Hab, and this process would add a layer of complexity and inefficiency to H2 recovery. Choosing RWGS eliminates the need for any method to recover H2 from methane. An effective method for separating H2O from the gas stream existing the RWGS chamber is to adsorb it with zeolite 3A, as in step 2.

The DRA specifies that H2 be brought from Earth in order to make water for the crew; however, in the Blue Dragon architecture H2O is obtained from the atmosphere, and potentially also from the ground. Research into extraction of H2O from the Martian atmosphere (Grover et al, 1998; Williams et al, 1995) has shown how 3.3kg of H2O per day (enough to replace losses through life support regenerative processes) could potentially be obtained from the Martian atmosphere using adsorption into zeolite 3A. This idea has been incorporated into the ISAP process above, indicating that H2O necessary for the RWGS reaction need not necessarily come from the Hab’s supplies. The amount of H2O obtainable from the atmosphere may be small, but this doesn’t matter because predeploying the Hab allows plenty of time to collect the necessary H2O and make the air, and the H2 is recycled anyway.

More research needs to be done into this system. We need to investigate efficient gas separation techniques, the rate at which air can be manufactured, the unit’s mass, volume and energy requirements, and how it integrates with the ECLSS.

Architecture, Mars

Atmosphere design for the surface of Mars

The Hab

The advantages of a normal Earthian atmosphere apply equally well to the Hab as they do to the ISS and other spacecraft. However, there are several compelling reasons for a reduced atmospheric pressure for the Hab:

  • Lower Hab mass.
  • Less air to be manufactured, which therefore reduces the mass and energy requirements of the ISAP system.
  • Potential for a zero prebreathe protocol for EHA/EVA.

The higher the atmospheric pressure of the Hab, the stronger and therefore heavier it will need to be. This is contrary to our requirement of reducing the mass of the Hab as much as possible because of the significant challenge of landing such a heavy object on Mars.

Atmospheres in early spacecraft had low total pressure, low oxygen pressure, or both. However, not all of these atmospheres would be suitable for an 1.5 year stay on Mars. Research conducted in recent years has more clearly defined the limits for artificial atmospheres suitable for extended exposure.

The minimum partial pressure of oxygen required to support human physiology is considered to be 16kPa. However, for long-duration space missions, a minimum partial pressure of oxygen of 18kPa is recommended  (Duffield, 2003). This is based on a previous study about planetary surface habitats (Campbell, 1991), which reviewed 33 different considerations related to atmospheric pressure and composition.

From a physiological perspective, an O2 pressure of 18kPa is perfectly safe. This is equivalent to about 1370m altitude (approximately the altitude of Kathmandu, Nepal), which does not even qualify as “high altitude” in mountain medicine (1500 – 3500m). Acclimatisation to reduced O2 pressure at altitude is characterised by an increase in pulse and breathing rate. Most people can ascend to 2400m (where O2 pressure is about 16kPa) without difficulty, however, altitude sickness may occur above this level. Astronauts can be conditioned for an O2 pressure of 18kPa by training in a hypobaric chamber, or at a moderate altitude (e.g. Black Mesa, US). In a microgravity environment there would already be increased strain on the cardiovascular system , and it would be preferable not to cause any further strain; however, the habitat is in a gravity environment on the surface of Mars, and although this is still a reduced gravity environment compared with Earth, the increased load on the heart will be mitigated.

The next design question is how much buffer gas to include. A pure oxygen atmosphere introduces an unacceptably high risk of fire, such as the one that occurred in the Apollo 1 Command Module. The upper limit of oxygen concentration with regard to fire safety has not clearly defined, but 30% is considered a reasonable upper limit (Campbell, 1991). This gives us a total atmospheric pressure of 60kPa, about 60% of Earth.

Buffer gas refers to the component of the atmosphere comprised of metabolically inert gases, which usually means nitrogen (N2), plus the noble gases helium (He), neon (Ne), argon (Ar), krypton (Kr) and xenon (Xe). The buffer gas portion of the atmosphere of Earth is almost entirely N2 (99%), with about 1% Ar and trace amounts of He, Ne and Kr. As described in the section on In Situ Air Production, because we’re making buffer gas in an economical way by simply using the Martian atmosphere with dust, CO2 and contaminants removed, our buffer gas on Mars will be about half-half N2 and Ar, possibly with trace amounts of Ne, Kr and Xe.

Nominal atmospheric concentrations of CO2 and H2O must also be determined. According to JSC 20584 (Spacecraft Maximum Allowable Concentrations for Airborne Contaminants), the maximum CO2 concentration is 0.7%. A CO2 concentration of 1% can cause drowsiness, with more serious symptoms occurring at higher concentrations. A typical concentration in normal spacecraft operations is 0.5%, which is a reasonable design goal. This gives us a CO2 partial pressure of about 0.3kPa.

With regard to water vapour, NASA specifies a RH (Relative Humidity) of 30-70%, i.e. an average of about 50%. Our target temperature is 295K (about 72°F or 22°C, which is optimal for human comfort and productivity), and the saturated water vapour pressure at this temperature and pressure is about 2.6kPa. Our average water vapour partial pressure will be 50% of this, or about 1.3kPa.

Any other gases present in the atmosphere should be present in trace amounts only.

Proposed design for Mars habitat atmosphere.

Gas Partial pressure (kPa)
Oxygen (O2) 18.0
Carbon dioxide (CO2) 0.3
Water vapour (H2O) 1.3
Buffer gas (N2/Ar) 40.4
Total 60.0

This atmosphere will produce differences in sound quality that the crew will be required to adapt to. The higher density of Ar compared with N2 will have the effect of lowering audio frequencies, including astronaut voices. Sound will also not be as loud or travel as far due to the reduced atmospheric pressure.

Continue reading

ISRU, Mars

In Situ Food Production (ISFP)

Growing food on Mars is an exercise in efficiency. The facilities will not (at least initially) be available to grow every kind of fruit, vegetable, grain, nut and herb that we are used to. We may only be able to grow small amounts of a small number of crops. For a while there may be restrictions on the number of available ingredients and therefore the meal choices. The challenge is to find which crops deliver the most nutrition for the amount of volume, mass and energy required to produce them.

Top of this list would surely be spinach. Tomatoes, mushrooms, cabbages, garlic, kale, carrots would perhaps also make the list. What is the best way to grow each of these?
There are 3 primary mechanisms proposed for producing food on Mars:

  • In soil
  • Hydroponics
  • Aeroponics

Experiments are already underway to grow food in Martian soil simulant. Because all the chemical elements necessary for life are available in Martian soil, it should be possible to grow plants in it; however, it will be necessary to analyse the crops thus produced, in order to determine if they have healthful levels of vitamins and minerals. They will not necessarily have the exact same nutrient profile as their counterparts on Earth, because of the availability of those elements in the soil.

Some chemical processing of the soil may be necessary to prepare it for plant growth; for example, it may be too acidic or salty. Therefore the addition of a specially prepared fertiliser may be necessary. It may be beneficial to introduce worms to the soil and feed them with food scraps, so they can process the dirt grains and organic material together to make fertile soil. Of course, for this to happen we would initially need food scraps, which would have to come from somewhere, so this would not be an option for the first crops.

In any case, the first crops are more likely to be grown using hydroponics or aeroponics. These are similar setups in that the plants are grown in a dirt-free environment, fed with nutrient-rich water. In the case of hydroponics, the water flows through pipes in which sit the roots of the plants, so they can access the nutrients in the water. In the case of aeroponics, the plants are suspended, with their roots exposed to the air; nutrient-rich water is provided to the roots as a mist.

The primary advantage of aeroponics over hydroponics is that the water requirement is minimal, which will be important for a Mars base where water may be scarce in the early years. The disadvantage, however, is that the mist greatly increases the relative humidity of the greenhouse atmosphere. Relative humidity of controlled environments like spacecraft atmospheres should not exceed 70% (according to NASA guidelines), as this can interfere with electronics or cause build-up of mould. This could be addressed by separating the greenhouse environment from the main habitat environment by a gate.

A double-gate, whereby a person transitioning from the habitat to the greenhouse would open one gate, step through, close that gate behind them, open a second gate, step through and close that one behind them, may be an effective method of containing humidity in the greenhouse. However, this level of control is probably unnecessary, and a single gate will work fine if people don’t leave it open. The small amount of water vapour that would travel across from the greenhouse to the habitat would be easily soaked up by the habitat’s ECLSS.

One way to mitigate the migration of water vapour from the greenhouse into the habitat would be to place intake fans near the gate, which draw air form the region around the gate into the THC (Thermal and Humidity Control) subsystem of the habitat’s ECLSS, which will remove any surplus water vapour from the atmosphere.
Aeroponics may therefore be the preferred choice. The questions remain:

  • What crops would be good to commence experimentation with? (e.g. spinach and tomatoes)
  • What nutrients are added to the water provided to the system, if any?
  • What mass and volume of equipment is required to produce what mass, volume and nutrient value of food?

Another downside of aeroponics is that in some cases the crops will not grow as large as they would in a hydroponic system or if they were grown in soil. This is usually due to lack of available nutrients in the provided water. When using water enriched with sufficient nutrients, and with the proper equipment, aeroponic crops can grow to full size.

Therefore, if we are to experiment with an aeroponic system on Mars, it will be important to take with us a supply of nutrients optimised for aeroponic crop production.

Aquaponics combines hydroponics with fish-farming, and is another approach to food production that has been suggested for Mars. The fish can be fed food scraps, possibly supplemented with special food; the water in which the fish swim becomes nutrient-rich due to the metabolic outputs of the fish, and is then provided to the plants via the hydroponic system.

The main problem with aquaponics is the amount of water required, which is naturally much higher than for the other proposed methods. It is far more likely that Martians will be vegetarians, which is perfectly safe and healthy; many millions of people on Earth live healthfully on plants only. Producing food from plants is considerably more efficient in terms of energy and water, which will be crucial on Mars. In fact, it’s also crucial on Earth, but humanity is still learning this. Some scientists predict that by 2050 everyone (or perhaps the majority of people) on Earth will be vegetarian due to expansion of the population. In any case, if aquaponics is implemented on Mars it will probably not be for a few years, when water and energy production are much higher.


Mars, Video

Me presenting Blue Dragon at the Australian Space Science Conference, 1 October 2013

“Blue Dragon” is a humans-to-Mars mission architecture based on the NASA Design Reference Architecture, with various improvements to improve safety and reduce cost. It leverages COTS hardware likely to be available within the next decade, is designed as the seed of a settlement rather than a “flags and footprints” program, and is specifically intended as an international mission.

ISRU, Mars, Space Program

Green Dragon


Green Dragon is a mission concept to send integrated ISRU packages to function as resource factories to support human Mars missions. It builds on the accomplishments of NASA’s Ice Dragon mission, while functioning as an important precursor to the Humans-to-Mars mission architecture, Blue Dragon. The intention is to test and iteratively improve both ISRU and EDL systems.

The mission is named “Green Dragon” as, like Ice Dragon, it is based on the Red Dragon EDL system. It’s “green” because it demonstrates technologies and provides resources crucial to life support, and to match the pattern of red, green, blue commonly associated with Mars.

The initial Green Dragon program will land 1-3 Dragon capsules on Mars containing integrated ISRU hardware capable of manufacturing several products that will be of great value in exploration and settlement, namely LOX/LCH4 bipropellant, water and breathable air, from local Martian resources.

It will also serve as an opportunity to test and refine the Red Dragon EDL system in preparation for using it to land crew and cargo on Mars in the Blue Dragon mission. Our goal for landing humans on Mars is to maintain g-forces below an acceptable threshold for human health, and to land with a reasonably high degree of accuracy. Green Dragon will ideally serve to bringing the ISRU and EDL systems needed by Blue Dragon up to TRL-9.

Apollos 11-17 succeeded largely because the preceding missions in the program tested and proved each element of the target mission, i.e. the human landing. This approach of a program of precursor missions is not normally described in conjunction with H2M architectures, presumably because the large cost of sending stuff to Mars has typically resulted in very conservative programs. However, if the Red Dragon technology can be shown to reliably deliver a 1-2 tonne package to Mars surface for under $200M per launch-plus-capsule, a program of precursor missions that build the necessary confidence in a human landing will be affordable. Green Dragon would represent a key subset of those missions.

Entry, Descent and Landing

The Ice Dragon mission will inevitably stimulate ideas for improving the Red Dragon EDL system, and it will be important for the success of the Blue Dragon mission that we take the opportunity to implement these improvements and refine Red Dragon further. The more capsules we’ve landed on Mars before we attempt the landing of humans, the greater confidence we will have that the landing will succeed.

One of the EDL factors necessary to improve is a reduction in the deceleration forces experienced by the capsules. The first mission to land a Dragon capsule on Mars may involve deceleration forces that would be excessive for a human crew. While this is acceptable for a cargo capsule, the Green Dragon series of missions will aim to reduce deceleration forces to below about 10 g’s, and ideally lower, in preparation for safe landing of humans.

Another factor related to EDL that Green Dragon would seek to improve is the accuracy of landing. Green Dragon missions will make use of image recognition software, as well as signals from spacecraft, to land on preselected coordinates as accurately as possible. Having proven that Red Dragon capsules can be landed on a dime will make it possible to design a layout of the Blue Dragon base with much greater precision, thus improving safety and probability of success. For example, the DragonRider that delivers the crew to Mars surface can be landed exactly the desired distance from the Hab, which is to say, not so close that debris thrown up by the landing could damage the Hab’s inflatable modules or instruments (or any other part), yet not so far away that it would be unreasonably difficult or time-consuming for the crew to reach the Hab after landing.

The more Dragon capsules we land on Mars before sending humans, the more confidence we will have in the technology before using a DragonRider to land a crew on the surface of Mars. We need to know exactly how the capsule behaves during EDL to Mars, how it interacts with the Martian atmosphere and surface (especially dust), how to land one with precision, and where the strengths and weaknesses are in this approach. This knowledge can only be obtained definitively with practice.

By the time we land a crew on Mars in a DragonRider, we will ideally have landed a minimum of four uncrewed Red Dragons:

  • 1 for Ice Dragon
  • 1-3 for Green Dragon
  • 2+ pre-deployed cargo capsules for Blue Dragon

In Situ Resource Utilisation

The Green Dragon mission, or series of missions, is designed to test, demonstrate and develop the ISRU technologies necessary for successful execution of the Blue Dragon architecture. Previous architectures have been reasonably conservative in terms of inclusion of ISRU due to the low TRL of the concepts. Blue Dragon pushes the boundaries of what has previously been considered in terms of ISRU to support a humans to Mars mission, and, being mission critical elements, it’s essential to prove that they will work.

Green Dragon will allow us the opportunity to test a variety of ISRU experiments on Mars, including the manufacture of LOX/LCH4 bipropellant, water and breathable air from local Martian resources.

In particular, the demonstrated ability to obtain abundant water on Mars is considered crucial to exploration and settlement. This capability will reduce the cost and increase the likelihood of success of all future H2M missions by a significant margin, as neither water nor hydrogen will need to be brought from Earth.

A design goal of Blue Dragon is to not take any surplus hydrogen, water, oxygen or nitrogen to the surface of Mars. As all these things are available locally and can be obtained directly from Mars, and since utilising local resources significantly reduces mission cost, there’s no sense carrying them all the way from Earth; that is, if the necessary hardware can be developed to successfully and reliably obtain them, which it certainly can.

Proposals for ISRU tech for Mars have traditionally been very simple and conservative. Considering the success of MER’s, Phoenix and MSL – all very complex robots – this cautiousness is probably unfounded. Although the ISRU experiments encapsulated by Green Dragon are more complex than basic ISPP, they are also much less complex than any of these missions that have already been successful. Furthermore, the whole Green Dragon program should, in theory, cost less than MSL.

Green Dragon contains a set of three integrated ISRU experiments:

  1. ISPP – In Situ Propellant Production: Making LOX/LCH4 bipropellant.
  2. ISWP – In Situ Water Production: Extracting water from the ground.
  3. ISAP – In Situ Air Production: Producing breathable air.

This will include the following subsystems:

  • Storage tanks for water, LOX, LCH4 and buffer gas
  • Small, self-contained nuclear fission reactor, or solar panels
  • Electrolysis unit
  • Catalytic converter
  • Ozone scrubber
  • Heating unit
  • Refrigeration unit
  • Compression unit

The following colour codes are used for tanks:

Blue tank Red tank Yellow tank Green tank
Water LCH4 LOX Buffer gas
(mainly N2/Ar)

The following schematic illustrates the flow of materials through the system. The masses are calculated based on Mars Direct ERV propellant requirements, but can be scaled to suit the capacity of the Green Dragon (small or large version).

Integrated IISRU schematic
Integrated IISRU schematic (click to embiggen)



The concept of ISPP (In Situ Propellant Production), whereby LOX/LCH4 (Liquid Oxygen/Liquid Methane) bipropellant is produced by reacting carbon dioxide obtained from the Martian atmosphere with hydrogen brought from Earth, was most famously described by Robert Zubrin and David Baker as part of the Mars Direct mission architecture in 1990 [1]. They also constructed a small plant to prove that the technique was viable.

Our goal is to use hydrogen manufactured via electrolysis of locally collected water, rather than bringing it from Earth as in the original ISPP concept.

The process is as follows:

Step 1: Carbon dioxide is obtained from the Martian atmosphere by filtering out dust, removing water via zeolite adsorption, and freezing the CO2 out of the remaining gas mix, enabling it to be separated from the remaining gases. The small amount of water captured by zeolite adsorption may be cycled into step 3.

Step 2: The AWESOM rover traverses back and forth across a patch of ground, microwaving the regolith below and collecting released water. When the rover’s tank is full, it returns to the MAV to deliver its payload of water, then returns to work. This step requires some form of LPS (Local Positioning System) so the rover knows which ground it has covered.

Step 3: Water is separated into hydrogen and oxygen gas via electrolysis:

2 H2O(l)  →  2 H2(g) + O2(g)

The oxygen is stored cryogenically as LOX.

Step 4: Methane is produced by reacting carbon dioxide (CO2) obtained from the Martian atmosphere with hydrogen (H2) produced in step 3, via the Sabatier reaction:

CO2(g) + 4 H2(g)  →  CH4(g) + 2 H2O(v)

The Sabatier reaction occurs at high temperatures, optimally around 600K. The methane is liquified to LCH4 and stored in the red tank. The water produced by the Sabatier reaction is stored in the blue tank, where it may be used to make more hydrogen.

Combining the reactions in step 3 and 4, the overall result is:

CO2(g) + 2 H2O(g)  →  CH4(g) + 2 O2(g)

66 tonnes of carbon dioxide combined with 54 tonnes of water (a total of 120 tonnes) will produce 24 tonnes of methane and 96 tonnes of oxygen. This gives us a surplus of 12 tonnes of oxygen that may be used for air. As the Martian atmosphere is about 95% CO2, we need to process about 70 tonnes of it to obtain 66 tonnes of CO2, which therefore results in about 4 tonnes of buffer gas.


ISWP (In Situ Water Production) via extraction of water from the Martian ground is well understood to be a crucial enabling technology for human settlement of Mars, and would similarly be a boon for preliminary exploration. The robotic technology required to achieve this is likely to be simpler than that we have already deployed to Mars in the form of Spirit, Opportunity and Curiosity, and is therefore well within our capability.

The advantage of not having to bring hydrogen from Earth is of such value that developing a system for obtaining water on Mars is one of the most important outcomes of Green Dragon. Not only will we save the mass of the hydrogen needed for ISPP, but also the water required by the crew during the 1.5-year surface mission. This capability will greatly reduce the launched mass, landed mass, cost and complexity of any H2M mission.

Producing LOX/LCH4 propellant using hydrogen obtained from Martian atmospheric water vapour, instead of hydrogen brought from Earth, has only been discussed theoretically and not demonstrated. One goal of Green Dragon is to demonstrate that.

Water collected by the ISWP experiment will be collected in a tank. This will be connected to the electrolysis unit so that any amount can be converted to H2 and O2 as required. H2 will not be stored, but will be fed directly into the ISPP unit. In this way we avoid the need to store hydrogen gas, and thereby losing any through leakage.

Note also that water is also produced by the Sabatier reaction used in the ISPP experiment. This water will simply be captured and collected in the same tank, thus recycling the hydrogen.


ISAP (In Situ Air Production) will be an important aspect of Blue Dragon. The Martian atmosphere contains plenty of nitrogen and oxygen, which is readily accessible, and considering the benefits of minimising launched mass there is no justification for bringing air from Earth. Of course, the Hab, MAV and rover may be launched full of air, but we need more than that.

Because we intend to utilise inflatable modules to expand the Hab volume, our primary requirement for air is to inflate these modules. Also, if the rover or MAV becomes depressurised for any reason, they will need to be re-pressurised.

Air pressure within the Hab may be lost due to:

  • Airlock cycling (a daily occurrence during the surface mission).
  • Leakage.
  • Recycling/scrubbing.
  • Punctures or tears.

Oxygen is easily produced via electrolysis of water, as described above. Every kilogram of hydrogen manufactured from water for ISPP will produce eight kilograms of oxygen that may be used for air.

Earthian air is approximately 78% nitrogen, 21% oxygen, 1% argon, plus traces of other gases. Oxygen is needed by animals; nitrogen is needed by plants and certain microbes, and acts as a buffer gas. Argon is non-reactive and safe to breath as long as there’s sufficient O2 in the mix. Ideally in our habitats we will want a full 1 atmosphere of pressure, as on the ISS, because then we can use COTS equipment which is much cheaper. However, the buffer gas does not necessarily need to be comprised of the same ratio of nitrogen and argon as on Earth.

The ISPP process separates carbon dioxide from the remainder of the Martian air using fractional distillation, which means cooling the gas to below the boiling point of carbon dioxide, thus causing it to become liquid, in which state it can be easily separated from the remaining gases.

There are at least 12 gases present in the Martian atmosphere. As part of the ISPP process, dust and water will be removed from the atmosphere first, then carbon dioxide is separated from the remainder. This produces a residual gas mix with the following approximate composition, which is about 98% safe to breathe.

Gas Fraction Notes
N2 60% Preferred buffer gas.
Ar 35% Safe.
O2 3% Safe.
CO 1.8% Toxic.
NO 2200ppm When combined with oxygen this will oxidise to NO2, which is toxic.
Ne 55ppm Safe.
Kr 7ppm Safe.
Xe 2ppm Safe.
O3 0.7ppm Toxic.
CH4 0.2ppm Safe.

Creating an exact match for Earthian air by separating out the nitrogen and argon would be energetically costly due to the low boiling point of these gases (77K and 87K respectively). However, the above gas mix can be made breathable by the removal of toxic elements, highlighted above in green. Fortunately, this is a PSP (Previously Solved Problem).

Scrubbing CO and NO

A standard automobile catalytic converter (~$200 – $500) can be used to scrub CO and NO from the gas mixture.

Automobile catalytic converter
Automobile catalytic converter (click to embiggen)

The reduction catalyst (platinum and rhodium) in the catalytic converter converts NO into N2 and O2:

2NO → N2 + O2

The oxidation catalyst (platinum and palladium) in the catalytic converter converts CO into CO2:

2CO + O2 → 2CO2

It will also oxidise the tiny amount of methane:

CH4 + 2O2 → CO2 + 2H2O

One issue to address is that COTS catalytic converters operate at high temperatures, whereas the ambient temperature on Mars is comparatively very low. The nuclear power plant can provide ample electricity to heat the converter as required. We can simply position the catalytic converter somewhere near the Sabatier reactor, which operates at around 600K.

Scrubbing ozone

Ozone is relatively easily eliminated using a COTS ozone scrubber (or “ozone destruct unit”) (~$1200). Ozone scrubbers work best with dry gas, therefore, since the catalytic converter will produce a small amount of water in the process of oxidising the methane, it’s preferable to pass the gas through the ozone scrubber before the catalytic converter.

Ozone destruct unit
Ozone destruct unit (click to embiggen)

Breathing gas

A breathing gas comprised of 20% oxygen plus 80% of the resultant buffer gas mixture (which has a little O2) will contain about 48.3% nitrogen, 28.6% argon, 21.7% oxygen, 1.4% carbon dioxide, and traces of neon, krypton, xenon and water vapour.

This is a very good breathing gas mixture except that it’s a little high in CO2, and would cause drowsiness. This can be easily removed by the air recycler in the ECLSS (Environment Control and Life Support System). In other words, once the air has been manufactured in this way, it is first treated in the same way as stale air coming from the Hab and pumped through the air recycler to scrub out the excess CO2. In fact, it may even be better to incorporate the ozone scrubber and catalytic converter in the ECLSS rather than the ISAP unit, and simply treat all buffer gas as stale/dirty air.

At one atmosphere of pressure, this breathing gas is sufficiently similar to Earth-normal that no difference would be noticeable except perhaps a reduction in audio frequencies (including astronaut voices) due to the high percentage of argon.

However, it is not a firm requirement that the habitat air pressure be one atmosphere, although that would be ideal in some respects (the internal atmospheric pressure of ISS is one atmosphere, which simplifies design of other components). Green Dragon produces about three times as much surplus O2 as buffer gas, therefore we may opt for a lower concentration of buffer gas (say, 20-50% instead of the usual 80%), which would probably be acceptable.

With a source of oxygen plus a suitable buffer gas mix available, we can plug these into the Hab to maintain an optimised internal atmosphere. Whenever the partial pressure of O2 fraction drops below a certain threshold, additional O2 is pumped in; similarly, if the buffer gas partial pressure drops too low, more buffer gas will be automatically pumped in. This is how the ECLSS functions on the ISS.

Larger Green Dragons

The Mars One mission plans to land both cargo and crew on Mars using an upgraded Dragon capsule with a 5 metre diameter. At the time of writing this capsule has not yet been developed, although SpaceX has indicated that it will be developed within the next few years in order to be ready for the first Mars One cargo mission in January 2016. This larger Dragon will be capable of delivering a payload upwards of 2.5 tonnes to Mars surface.

For at least the first Green Dragon mission, the smaller capsule with the 3.66m diameter and the approximately 1-2 tonne payload will be adequate. However, when these larger capsules become available, larger Green Dragons based on these could be sent as resource factories to support Blue Dragon, Mars One, or other missions or bases. The primary advantage of the larger capsules will be the capacity to store greater volumes of their products.

The suggested approach therefore is to use the smaller capsules to develop and refine the design of a self-contained resource factory, then use the larger capsules to support missions.


The Mars Ascent Vehicle

The MAV will be an SSTO (Single Stage To Orbit) VTOL (Vertical Take-Off and Landing) vehicle with powerful LCH4/LOX engines, capable of pinpoint landing on Mars, refuelling from local Martian resources, and carrying up six astronauts to Mars orbit for transfer to the MTV. Ideally it will also be capable of then returning to Mars surface.

The MAV has two main sections:

  1. The lower portion is the ascent/descent stage, which includes:
    • the engines
    • fuel tanks (LCH4 and LOX)
    • a sophisticated GNC system
    • solar panels
    • ISPP plant (electrolysis unit and Sabatier reactor)
    • a water-mining robot called AWESOM (Autonomous Water Extraction from Surface Of Mars)
  2. The upper portion is a DragonRider (Pern-2) with trunk containing additional solar panels.
Mars Ascent Vehicle concept
Mars Ascent Vehicle concept


The MAV’s engines will need to be LCH4/LOX (liquid methane/liquid oxygen), in order that they can be refuelled from local Martian resources (see below). Several types of engines that burn methane fuel have been developed, however, SpaceX are currently developing a new LCH4/LOX engine called “Raptor” that is likely to be the preferred choice.

Based on SpaceX’s record, the Raptor design will presumably be more modern and advanced than existing methane fuelled rocket engines. It will also have a high degree of interoperability with other SpaceX hardware used in the mission, such as the Dragon. Furthermore, since SpaceX will already be involved, using more hardware from the same company should, in theory, reduce costs, if for no other reasons that a fewer number of engineers need to be paid, since engineers who understand the Dragon and Falcon are also probably the same people who understand the Raptor engine.

The engines must be powerful enough to carry the MAV from Mars orbit to surface, and back. In theory, if the MAV is capable of refuelling itself at Mars Surface, it should be able to do this repeatably (thus making it a SSTO VTOL RLS).


The MAV will have an integrated ISPP plant comprised of:

  • water mining robot (AWESOM)
  • electrolysis unit
  • Sabatier reactor
  • cryogenic propellant storage

The ISPP process currently envisaged for the MAV closely parallels that described in the Mars Direct architecture (1990). In Mars Direct, hydrogen (which represents the lightest fraction of the propellant yet also the hardest to obtain from local Martian resources) is carried from Earth, and reacted with CO2 obtained from the Martian atmosphere in order to produce methane (CH4) and oxygen (O2), which are liquified and stored cryogenically. Additional oxygen is obtained from carbon dioxide via the reverse water gas shift reaction, which produces water (H2O) that is then electrolysed to hydrogen (H2) and oxygen (O2).

The primary difference in Blue Dragon is that hydrogen is not carried from Earth, but is obtained by mining water from the surrounding regolith, and electrolysing it into hydrogen and oxygen. Every kilogram of Martian regolith is estimated to contain up to 40g of water (Slosberg), and 80% of this can be liberated from the top [x]cm of regolith using microwave radiation.

The optimal stoichiometric ratio for LOX/LCH4 bipropellant is 7:2. Mars Direct specifies a fuel requirement of 24 tonnes of CH4. Taking that as a nominal value (for now), the amount of LOX required is therefore 84 tonnes.

The process is as follows:

Step 1: Carbon dioxide is obtained from the Martian atmosphere by filtering out dust, removing water via zeolite adsorption, compressing the remaining gas mix to 700 kPa, and allowing it to equilibrate to ambient Martian temperatures (about 210K). The CO2 will condense, enabling it to be separated from the remaining gases.

Step 2: The AWESOM rover traverses back and forth across a patch of ground, microwaving the regolith below and collecting released water. When the rover’s tank is full, it returns to the MAV to deliver its payload of water, then returns to work. This step requires some form of LPS (Local Positioning System) so the rover knows which ground it has covered.

Step 3: Water is separated into hydrogen and oxygen gas via electrolysis:

2 H2O(l)  →  2 H2(g) + O2(g)

Step 4: Oxygen produced in step 3 is stored cryogenically as LOX.

Step 5: Methane is produced by reacting carbon dioxide (CO2) obtained from the Martian atmosphere with hydrogen (H2) produced in step 3, via the Sabatier reaction:

CO2(g) + 4 H2(g)  →  CH4(g) + 2 H2O(v)

The Sabatier reaction occurs at high temperatures, optimally around 600K. Water produced by the reaction is cycled back into the electrolysis unit to produce more hydrogen and oxygen.

Combining the reactions in step 3 and 5, the overall result is:

CO2(g) + 2 H2O(g)  →  CH4(g) + 2 O2(g)

66 tonnes of carbon dioxide combined with 54 tonnes of water (a total of 120 tonnes) will produce 24 tonnes of methane and 96 tonnes of oxygen. This gives us a surplus of 12 tonnes of oxygen that may be used to supplement air supplies for the Hab or the Rover.

Obtaining 66 tonnes of carbon dioxide requires processing about 70 tonnes of Martian atmosphere. The density of the atmosphere at the surface of Mars is about 0.02 kg/m3, which means 3.5 million cubic metres of atmosphere

If our goal is for the MAV to be fully fuelled within 20 months (the time difference between the arrival of the MAV at Mars and the departure of the crew from Earth), that means processing about 117kg (5830 m3) of atmosphere per day, or 80 grams (4 m3) minute. It remains to be seen how achievable this is. However, there is 44 months between the arrival and departure of the MAV, therefore, a slower processing rate may still be acceptable, although the goal of ensuring that the MAV is fully fuelled before the crew leaves Earth would not be met.


The ability to make use of Martian water is fundamental to the Blue Dragon architecture; therefore, it’s preferable to locate Marsopolis in a region where the water content is reasonably high, while still aiming for lower latitudes in order to simplify landing and to maximise solar power.

The diagram below shows the lower limit of water concentration across Mars.

Water on Mars
Water on Mars

The proposed location for Marsopolis (discussed in the Location section) is in Arcadia Planitia, a region to the north-west of Olypmus Mons. The water content at the landing site should be at least 10% by mass, and the presence of grooves and ridges in the region are indicative of ground ice in the near surface. Furthermore, Arcadia Planitia is extremely flat, which will support autonomous water mining by a mobile robot.

If we estimate that the average water concentration of the regolith is 10% by mass [ref], and that 80% can be extracted from the top decimetre of regolith by microwave radiation [ref], 675 tonnes of regolith will need to be processed.

675 tonnes of regolith = 67.5 tonnes water + 607.5 tonnes dry soil

67.5 tonnes of water * 80% captured = 54 tonnes

The density of dry Martian soil is about 1.4 g/cm3 = 1400 kg/m3 [ref]. If the regolith is 10% water, its density will be about:

10% * 1000 kg/m3 + 90% * 1400 kg/m3 = 1360 kg/m3

The volume of regolith to process is therefore:

675,000 kg / 1360 kg/m3 = 496 m3

Assuming we can only extract water from the top decimetre of regolith by microwave radiation, the area that must be covered by the rover is about:

496 m3 / 0.1 m = 4960 m2

If the AWESOM rover has a 1 metre wide catchment, it must therefore traverse a 100 metre long strip about 50 times (or cover a square about 70 metres on a side).

Specifying the same 20 month time constraint, AWESOM must collect 90 kg of water per day, which requires covering about 8.3 m2 of terrain per day on average. This seems like it should be easy. It will be preferable to collect the water as rapidly as possible, for peace of mind, and also because H2 is needed before CH4 can be produced. The design goal will be to build the robot to be capable of carrying up to 100 kg of water; it should therefore return to the MAV to unload at least once per day.

It may be that a single AWESOM is capable of collecting enough water for the MAV as well as for the Hab; although, considering the importance of water, having two at the site will still be preferable. Note that there’s no reason why the AWESOM must be delivered to Mars inside the MAV. It could be delivered separately in one of the cargo modules, and drive itself over to the MAV.

It’s true there there would be a benefit if the MAV and AWESOM were an integrated package (i.e. the AWESOM robot is carried inside the MAV), because then the MAV could refuel anywhere on Mars; it would simply land and deploy its AWESOM to collect more water. However, for our early MAV designs it will be preferable to reduce mass as much as possible.

It may even be preferable to keep all ISRU hardware separate, in, for example, a Green Dragon capsule, and transfer propellant to the MAV via hoses. There are several benefits to doing this, such as a more efficient integrated ISRU design, and reduction in the landed mass of both the MAV and Hab. However, connecting up hoses may be difficult to do autonomously, and it will be better if the MAV is fully fuelled before the crew leaves Earth.

Keeping water in a liquid form within the rover will require heat, perhaps more than can be reliably produced from solar cells. An RTG (Radioisotope Thermoelectric Generator), like the one used in Curiosity, will provide both heat and electrical energy.

Power system

One of the primary challenges associated with the MAV is the power system.


In Mars Direct, once the MAV (a.k.a. ERV) lands on Mars, a small robotic truck emerges with an nuclear reactor. The truck carries the reactor some way off, trailing electrical cable connecting it to the MAV in order to provide power to the ISPP plant.

The reason why it must be distanced from the MAV is because the reactor will be unshielded and the MAV is designed to carry people. Shielding is typically very heavy and would add a lot of mass to the MAV’s payload, making it much harder to land on Mars. If an unshielded reactor is operated inside or near the MAV, it would irradiate the vehicle, making it unsafe for carrying people. Therefore, it must be moved a safe distance away where radiation from the reactor will not reach the vehicle.

There are some issues with this design. If the reactor is unshielded, even if it’s moved some distance away from the MAV (and the Hab and other parts of the base), it will create a zone around it into which the astronauts must not venture. Perhaps the MAV will not be affected by radiation from the reactor, but the natural environment surrounding the reactor will. We should treat Mars a little more responsibly than this. What if the reactor is parked near something of scientific interest? What if the reactor needs attention – for example, the cable becomes loose?

If the reactor is shielded, it achieves two things. It means it does not need to be relocated from the MAV, thus eliminating the mass of the robotic truck and partially offsetting the mass of the shielding. It would also allow for the reactor to be kept contained within the MAV, enabling it relocate itself to almost anywhere on Mars and refuel (as long as there is sufficient water in the surrounding regolith). This would be an extremely useful capability, but not essential for the first H2M mission.

One of the advantages of using a nuclear reactor, in the Mars Direct design, is that it represents a reliable source of abundant energy that can be used to convert the hydrogen into methane as quickly as possible, in order to minimise boil-off and ensure that ample fuel will be manufactured.


However, if we commit to using local water as a hydrogen source, there will be no boil-off, and less reason for to manufacture all the propellant as quickly as possible. Considering this, as well as the difficulties associated with using a nuclear reactor to power the ISPP process, it will be better to use solar panels.

Solar panels represent a fluctuating energy source that will vary with the diurnal cycle, seasonal cycle, and atmospheric dust levels. As a rule of thumb, daytime solar energy on Mars is about half that of Earth. However, we have about 20 months between when the MAV arrives at Mars and when the crew leaves Earth at the subsequent launch opportunity, which should be plenty of time to make the necessary propellant if we wish to ensure that the MAV is fully fuelled before the crew leave Earth. Actually, there is about 44 months between when MAV’s arrival and departure, which should certainly be ample if that much time is needed.

The Dragon’s trunk contains solar panels, which can be used to provide some of the power necessary for ISPP, and the MAV’s ascent stage may include additional panels as required.

By choosing solar panels over a reactor for ISPP power, we eliminate the mass of the reactor, shielding, the robotic truck to relocate it, and the electrical cable that would connect the reactor to the MAV. We eliminate the risk of an unshielded reactor irradiating the MAV, or the surrounding terrain, or the crew.

If the AWESOM can be carried inside the MAV, using solar panels for power also provides for relocating and refuelling, because the solar panels can be folded up inside the vehicle before launch.

Are solar panels a reliable energy source on Mars? The Mars Exploration Rovers Spirit and Opportunity were powered by solar energy. Spirit was active for 6 years, and Opportunity has been going for 9.

Differences from Mars Direct ERV

  1. This is a much lighter vehicle. It does not include the mass of:
    1. 6 tonnes of hydrogen brought from Earth
    2. Nuclear reactor
    3. Light robotic truck for relocating reactor
    4. Electrical cable to connect MAV to reactor
  2. Instead, it includes:
    1. AWESOM water-mining robot (although this could be delivered separately)
    2. Solar panels
  3. No need for reverse water gas shift reactions or dissociation of CO2 to produce additional O2, as ample is obtained from water electrolysis.

Red Dragon

“Red Dragon” is a potential variant of the SpaceX Dragon capsule that will be able to land on Mars, currently being investigated by NASA and SpaceX.

The first series of Dragon capsules, including the one that historically became the first commercial spacecraft to dock with the ISS, were designed to splash down in water, like those used in NASA’s Mercury, Gemini and Apollo programs, and like NASA’s new Orion capsule.

The next generation of Dragon capsules, currently in development at SpaceX, are fitted with eight SuperDraco engines. These are a powerful new variation of the Draco engines used by the current Dragon RCS (Reaction Control System). Like the Dracos, they use non-cryogenic propellant (monomethyl hydrazine fuel and nitrogen tetroxide oxidiser), however, they’re much more powerful, each capable of delivering about 67 kilonewtons (15,000lbf) of axial thrust, for a total of about 534kN (120,000lbf). These engines will enable the Dragon to land on solid ground back at the launchpad, saving the time and expense of water recovery, and opening up the possibility for Dragon capsules to land on Mars. This is in alignment with SpaceX CEO Elon Musk’s stated purpose of establishing settlements on Mars.

Red Dragon will presumably be based on this or a similar engine configuration, with modifications to suit EDL on Mars. Variations may include:

  • Remove systems unique to LEO missions, such as berthing hardware.
  • Add deep space communications.
  • Modifications to SuperDraco engines to suit Martian atmosphere.
  • Reduction of heat shield mass.
  • Algorithms/avionics to enable pinpoint landing on Mars.

The gravity on Mars is lower, which reduces the acceleration of the capsule towards Mars; however, the capsule will be approaching from interplanetary space at a much higher velocity than if it were approaching from Earth orbit. Also, Mars’ atmosphere is much thinner (less than 1%) than Earth’s, so it will play less of a role in reducing spacecraft velocity during EDL. However, for the same reason, there is less heating due to atmospheric friction, and therefore less (or different) heat shield material may potentially be used. The different conditions will affect the forces experienced by the spacecraft, which may necessitate changes to thrusters, heat shield, avionics and other factors.

Red Dragon landing on Mars
Red Dragon landing on Mars

Another clear advantage is that a landed capsule can be repurposed as a storage unit, shelter or habitat.

Once the Red Dragon technology has been proven as a reliable mechanism for delivery of cargo, this approach may be used to deliver up to seven crew to Mars surface, by using a DragonRider modified in the same way.

Red Dragon represents a near term technology that can enable comparatively inexpensive and functional Mars missions. It’s an fundamental element of the Blue Dragon architecture, being utilised for both crew and cargo delivery to Mars surface.

Theoretically, Red Dragon will be able to land with a high degree of accuracy. From the SpaceX website:

SuperDraco engines will power a revolutionary launch escape system that will make Dragon the safest spacecraft in history and enable it to land propulsively on Earth or another planet with pinpoint accuracy.

This ability to land “with pinpoint accuracy” is supported by the Dragon Guidance, Navigation and Control (GNC) system. Due to the lack of GPS on Mars, alternate methods of achieving high-accuracy landings must be achieved using alternate methods; however, this problem is effectively solved. For example, Jeff Delaune at ESA has been developing a system known as “LION” (Landing with Inertial and Optical Navigation) that enables pinpoint landing on the Moon, Mars and asteroids using image recognition of major landmarks. Another impressive development is the Fuel Optimal Large Divert Guidance (G-FOLD) algorithm developed by JPL, able to autonomously calculate landing trajectories in real-time. This was recently tested with Masten Space System’s Xombie VTOL experimental rocket, very successfully, making a 750 metre course correction in real time. Considering these developments it’s safe to assume that the Red Dragon will be capable of pinpoint landings on Mars by the time we begin sending them, enabling a neat and optimised layout of the base to be designed beforehand.

It’s estimated that a payload of up to ~2 tonnes can be delivered to Mars surface via a Red Dragon. This will be a major breakthrough in the goal of human settlement of Mars.

Red Dragon potentially represents a mechanism for delivering cargo to the surface of Mars that is not only repeatable, but affordable. SpaceX currently charge $135M for a Falcon Heavy launch including the Dragon capsule. Making use of COTS and other pre-developed hardware, and simplifying missions, it may be possible to design and build a payload and send it to Mars for only $200 – $250M. Compare this with the $2.5B price tag of the Curiosity rover.

Once SpaceX have developed their RLS for the Falcon Heavy – a goal likely to be achieved within a few years, in light of the recent Grasshopper tests, and thus well before the first H2M mission – this price will come down even further.

Red SuperDragons

The architecture for the Mars One mission, which proposes to send up to 40 astronauts on a one-way mission to Mars, relies on a larger, 5-metre-diameter Dragon capsule for habitat modules. Although these are yet be to be built or demonstrated, their plan is to land the first two of these on Mars in 2020 – only 7 years from the time of writing.

Mars One settlement
Mars One settlement

Therefore, it can be inferred that a plan exists to have operational “SuperDragons” available within 7 years. This is well within the timeline of Blue Dragon.

Mars One is also planning to build a model settlement in a Mars analog environment in order to commence crew training in 2015, so it may even be possible that we will see what the SuperDragon capsules look like within 2 years or less.

At present the Blue Dragon architecture does not incorporate SuperDragon capsules. However, this may change as more information about them becomes available.


Ice Dragon

NASA have commenced studies of a mission to Mars based on the Red Dragon landing system, which may be flown as early as 2018. Known as “Ice Dragon”, it’s being developed in collaboration with SpaceX, and will deliver a science package to Mars including a drill that will penetrate up to two metres into the permafrost to investigate environmental conditions suitable for past or extant life.

There are 6 objectives currently envisaged for Ice Dragon[xx]:

  1. Determine if life ever arose on Mars.
  2. Assess subsurface habitability.
  3. Establish the origin, vertical distribution and composition of ground ice.
  4. Assess potential human hazards in dust, regolith and ground ice, and cosmic radiation.
  5. Demonstrate ISRU for propellant production on Mars.
  6. Conduct human relevant EDL demonstration.

Aside from the scientific outcomes of the mission, perhaps the most important contribution of Ice Dragon will be the demonstration of the EDL capabilities of the Red Dragon capsule.