The Space Reviewin association with SpaceNews

Mars exploration illustration
A number of significant technological hurdles have to be overcome before humans can successfully explore Mars. (credit: NASA)

The challenges of manned Mars exploration

Why 2040 is the earliest NASA can hope to send humans to the Red Planet

If one examines the many specific occurrences where NASA’s Exploration Systems Architecture Study (ESAS) Report mentions the word “Mars,” it is found that essentially all of them are very thin and lacking in content. It seems quite apparent that NASA has not yet made the effort to revisit the 1990s legacy of Mars human mission analysis to any depth at all, and seems to be content (at least for now) with handwaving, platitudes, and goal statements. The likely reason for this is that NASA has its hands full right now attempting to deal with lunar missions, and Mars missions have been pushed back beyond the back burner, to the point where they have fallen off the stove.

NASA has a very long road ahead to develop feasible, affordable approaches for human missions to Mars (if indeed such is possible—which seems unlikely at this juncture) and I have concluded that there is no way that NASA can send humans to Mars before 2040, and probably 2080. Nevertheless, NASA continues to express optimism about human missions to Mars, inferring that such missions will be implemented directly after human missions to the Moon, and, in fact, the ESAS Report refers to “Mars and beyond” several times. This disingenuous approach obscures the fact that NASA does not seem to have a plan for how to implement credible, feasible, affordable human missions to Mars.

There are many challenges in devising human missions to Mars. One of the overriding factors that makes Mars missions fundamentally different from lunar missions is the fact that there is very little opportunity to abort the mission. This drives up the requirements and the cost of systems, and since Mars missions will tend to be over 2.5 years roundtrip, extensive (and expensive) life-testing will be required.

NASA has a very long road ahead to develop feasible, affordable approaches for human missions to Mars (if indeed such is possible—which seems unlikely at this juncture).

As in most space missions, the greatest challenge for Mars missions is getting there and back. The huge masses of propellants needed for the legs of a space mission are significant limitations to feasibility of the mission. It takes about 20 metric tons (mT) on the Earth launch pad to put 1 mT of payload to Low Earth Orbit (LEO). For most space missions, even the 1 mT of “payload to LEO” is made up of mostly propellants to send a smaller payload to a distant destination. For example, in order to send a 1 mT payload (that includes an Earth entry system) to the Mars surface and return to Earth, it may require about 180 mT in LEO, and consequently 3600 mT on the Earth launch pad.

It is very difficult to estimate the overall cost of a complex space mission, and so it is common to use the initial mass in LEO (IMLEO) as a surrogate. Therefore it is common for analysts to compare IMLEO for various alternative space mission architectures. For a given launch vehicle capability, the IMLEO determines the number of launches required, and this clearly has a significant impact on mission cost. However, many problems arise as the required number of launches increases. These include difficulty in scheduling repetitious launches at sufficient frequency, problems in assembly in orbit, and even more difficult problems in landing or orbiting behemoths at Mars.


Chemical propulsion is likely to be the mainstay of foreseeable Mars missions. The lunar program will have developed launch vehicles capable of sending 125 mT into LEO, and Earth departure systems using hydrogen/oxygen propellants with a specific impulse of roughly 450 sec. There still remains considerable uncertainty as to which propulsion systems will be feasible for use beyond Earth departure. Whether it will be feasible to transport hydrogen to Mars and store it there (in orbit or on the surface) remains doubtful. Whether the complexities of cryogenic propulsion would negate the benefit of higher specific impulse of methane/oxygen propulsion (as compared to space storables) is also unclear. And, NASA has apparently decided not to develop the methane/oxygen rocket for the lunar program because of cost, schedule, and technology risks. Depending upon whether in situ resource utilization (ISRU) is used, and in what form, there will likely be advantages in utilizing chemical propulsion that can make use of propellants produced in situ. The development of space propulsion systems requires considerable funding, continuity over at least a decade (and probably more), and a determination to see it through to completion—characteristics that have been in short supply for NASA technology development for some considerable time.

One of the largest propulsion requirements in a Mars mission is Earth departure (aka trans-Mars injection (TMI)). Using hydrogen/oxygen propulsion for TMI, only one-third of the mass in LEO is the payload sent on its way toward Mars. Evidently, the mass of the TMI propulsion system is an important factor in driving up the required IMLEO for Mars missions.

A nuclear thermal rocket (NTR) with a specific impulse (Isp) of about 900 sec has the potential to significantly reduce the roughly two-thirds portion of LEO mass allocated to TMI propulsion. However, there are two major issues involved in using the NTR: estimating realistic performance benefits, and estimating the cost, duration, practicality, and political consequences of developing the NTR.

The financial, infrastructure, safety and political requirements to develop, test, validate, and implement the NTR are difficult to imagine at this juncture, except for the simplistic notion that it would take a gigantic effort spread over at least a decade (and, more likely, two decades) at a cost of many billions of dollars.

In regard to performance estimates, there are two important factors that need to be taken into account. One is the required “dry mass” of the power and propulsion system. This dry mass is a dead weight that is accelerated along with the payload, and reduces the performance of the NTR. The second critical factor is the altitude at which the NTR is turned on. If the NTR is fired up from LEO (as assumed by previous design reference missions (DRMs)) then its performance will be maximized. If, on the other hand, safety and political factors require that the NTR be lifted by chemical propulsion to a higher Earth orbit prior to initiation of the NTR, its effectiveness will be reduced due to the extra mass of chemical propellants used to raise the NTR orbit.

The dry mass of an NTR was estimated by several groups, but these were enthusiasts or advocates and therefore the estimates are likely to be on the optimistic side. In one analysis, the dry mass was parameterized as proportional to the propellant mass, with the proportionality factor treated as variable. If the NTR must be lifted to about 1,000 km altitude (or more), much of its benefit disappears compared to a chemical propulsion system that can be fired up in LEO at 200 km. It is noteworthy that the ESAS Report implies starting the NTP at 800–1,200 km altitude.

The financial, infrastructure, safety and political requirements to develop, test, validate, and implement the NTR are difficult to imagine at this juncture, except for the simplistic notion that it would take a gigantic effort spread over at least a decade (and, more likely, two decades) at a cost of many billions of dollars. Past history indicates that NASA is unlikely to have the fortitude to carry through such a program to completion.

The JSC “Dual-Landers” DRM proposed use of solar-electric propulsion (SEP) to raise the orbits of vehicles prior to Earth departure to reduce the propellant loads for trans-Mars injection. Documentation is sparse and the JSC estimates of mass requirements are unclear.

Woodcock describes a hypothetical SEP system for orbit-raising of heavy loads. This reference utilizes a payload of 50 mT driven by a 500 kW solar electric propulsion system with a specific impulse of 2000 sec. The trip time (up) is 240 days and (down) is 60 days. The required amount of xenon (Xe) propellant per transfer is 41.2 mT. According to estimates on the Internet, world production of Xe is presently 10 x 106 liters/yr = 53 mT/yr. Thus, one transfer would require approximately the present annual world production of Xe. Furthermore, Xe presently costs about $10/liter so the cost of Xe for one orbit transfer could be $100M. While it may be possible to increase world production significantly, recent articles on anesthesiology suggest difficulties. The viability of the SEP tug concept depends critically on use of a hypothetical high-efficiency lightweight solar array that is likely to be difficult to develop, and lightweight propulsion components. Furthermore, it seems unlikely that the required amount of xenon propellant could be obtained, and if obtainable, whether the cost would be affordable. Radiation would gradually diminish the efficiency of the solar arrays with each passage through the radiation belts. The total cost of the system includes the SEP tug, the mission operations involved, and the fast transit vehicle to take the crew up to HEO for rendezvous with the Trans-Habitat Vehicle. Because of the long time required for transfer, several of these “tugs” might be needed. At this point, it seems unlikely that such a scheme would be viable, but prior to the recent ESAS Report, JSC has argued for SEP orbit-raising as a likely mode for transfers from LEO in Mars missions. At this point, it appears that SEP for orbit-raising has been discarded by ESAS—and that is to their credit.

All of the previous Mars DRMs assumed that some form of aero-entry systems were used for Mars orbit insertion or Mars entry descent and landing (EDL). In order to achieve Mars orbit insertion and/or descent to the surface, rather large decelerations are required, which if implemented via chemical propulsion, result in very large propellant masses being required. This, in turn, leads to high IMLEO. Use of aero-entry systems typically depend on an ablative shield to slow down the incoming spacecraft and the shield is eventually discarded. Using optimistic estimates for the mass of the entry system, past DRMs have concluded that large mass savings result from use of aero-entry rather than propulsion. DRM-3 assumed that the mass of the entry system is 15% of the mass placed into Mars orbit. However it may more likely be 50%. Note that the MSP 2001 design used 60%.

Life support

The requirements for life support consumables and technology for providing them have been discussed extensively in great detail. Fundamental requirements for air are fairly well established, whereas requirements for water for lengthy Mars missions remain a matter of conjecture. For a Mars mission with 200-day transits to and from Mars and 560 days on the surface, the total amount of consumables needed for a crew of six is almost 200 mT. If no recycling or use of indigenous Mars resources are used, the required IMLEO would be at least 1400 mT, depending on assumptions made regarding propulsion and aero-assist. At 125 mT per launch, this would require more than a dozen launches merely for consumables.

Because the biological effects of exposure to space radiation are complex, variable from individual to individual, and may take years to show their full impact, definition of allowable exposure will always include considerable subjectivity.

Clearly, some degree of recycling will be possible based on current and evolving technology. Inevitably, this will require a physical plant and a back-up cache to account for losses in the process. Current estimates of the total mass of the environmental control and life support system (ECLSS) are estimated to be in the range 15–20% of the total mass of consumables provided. However, there is no certainty that such systems will be developed with the longevity and reliability needed for Mars missions, which differ from ISS and lunar applications where replacement units can be readily provided when the system breaks down, or where abort strategies allow rapid return to Earth in emergencies. Even under the most optimistic projections it seems likely that several launches will be required merely to supply life-support consumables.

NASA programs to develop recycling systems for life support seem to be focused on achieving over 99-percent recycling, at least for water, if not air. However, for Mars missions, where replacement units are not a consideration and the survival of the crew is at stake for over 2.5 years without hope of servicing, the emphasis should be on longevity, reliability, and proven lifetime. A system that recycles at, say, 90% that is reliable may be far more valuable than one that recycles at 99.8% but is likely to break down.


The allowable limits to radiation exposure, and technology for providing radiation protection are discussed in great detail in this report, in which estimates were made of the allowable radiation levels in lunar and Mars missions, and these were compared with estimated doses for various levels of shielding. There are two major sources of radiation in space: galactic cosmic radiation (GCR) and solar particle events (SPE).

Because the biological effects of exposure to space radiation are complex, variable from individual to individual, and may take years to show their full impact, definition of allowable exposure will always include considerable subjectivity. Aside from the difficulty in quantifying the biological impacts of exposure to radiation in space, there is also subjectivity in defining how much risk is appropriate. The standards presently adopted by the NCRP for LEO are based on the assumption that the allowable dose for excess risk for fatal cancer due to radiation exposure is 3%. A number of investigators have prepared point estimates of dose under various circumstances and have compared these with the allowable limits. However the current trend is to replace the point estimate by the 95% confidence interval (CI), which is likely to increase the dose by up to a factor of three to four. There are no established guidelines for allowable radiation exposure in space beyond LEO, but it is common to extrapolate the LEO limits to deep space.

A comparison of GCR and SPE is as follows:

  • The constant bombardment of high-energy GCR particles delivers a lower steady dose rate compared with large solar proton flares which can deliver a very high dose in a short period of time (on the order of hours to days).
  • The GCR contribution to dose becomes more significant as the mission duration increases.
  • For long-duration missions, the GCR dose can become career-limiting or year-limiting.
  • In addition, the biological effects of the GCR high-energy and high-charge particles are not well understood and lead to uncertainties in the biological risk estimates.
  • The main threat of SPEs is against the 30-day exposure limit.
  • SPE energies are far lower than GCR and are more amenable to mitigation by shielding.
  • The amount of shielding required to protect the astronauts will depend on the time and duration of the mission, and the effects of shielding are complex due to generation of secondaries.

The ESAS Report concluded that no shielding is needed on the CEV but this was based on a nine-day mission and use of point estimates. With a longer mission and 95% CI doses, this conclusion would no longer be supportable. In addition, radiation exposure in the Lunar Surface Access Module (LSAM) was not considered. If the mission duration is increased to 22 days (to include both the CEV and the LSAM, allowing for “loitering” and other maneuvers) and 95% confidence interval (CI) doses are adopted instead of point estimates, this picture changes significantly. In this case, without supplemental shielding, the 30-day limit would be exceeded even for a “1X 1972” SPE. The probability of such an event in a 22-day mission during solar maximum is about 0.4%.

For lunar outpost missions, the transits to and from the Moon are very short and as in the case of sortie missions, shielding would reduce the risk of shortened life due to an SPE during solar maximum. However, even with over 30 g/cm2 of regolith shielding, the 95% CI dose from a major SPE would exceed the 30-day limit on the lunar surface. The probability of encountering such a SPE during solar maximum in a six-month rotation is six times higher than those for a one-month sortie mission. The GCR during solar minimum is marginal against the annual limit, but this can be mitigated somewhat by use of regolith for shielding the habitat on the surface.

Radiation protection for Mars missions did not get much attention in the ESAS Report. Because crew transits to Mars are likely to require about 200 days in space, the probabilities of encountering major SPEs during solar maximum for each leg of the round trip are significant: occurrence of a “4X 1972” SPE is about 1.2% probable and occurrence of a “1X 1972” SPE is about 10% probable for each leg. For the round trip the figures would be 2.4% and 20%. Even with 10 g/cm2 of aluminum shielding the biological impact of a large SPE would be excessive.

On the surface of Mars, over the course of a year, the accumulated GCR 95% CI dose is about 77 cSv, which exceeds the annual allowable of 50 cSv. For a 560-day stay on Mars, the cumulative 95% CI dose is about 120 cSv. This would exceed the career allowable dose for most females and younger males. The 95% CI dose from a major SPE would exceed the 30-day allowable dose. The probabilities of encountering a large SPE are about 3.4% for a 4X 1972 SPE and about 28% for a 1X 1972 SPE for 560 days on the surface during solar maximum.

page 2: gravity >>