Site-to-site travel on Mars with the ITS

Ever since Elon Musk unveiled his plans to send humans to Mars, one important line of discussion revolves around the logistics of the eventual colony. How will it extract water from the regolith? How will it generate power? What form will the habitats take? And so on, and so on. Today, we will look at one small aspect of Martian habitation that can be simulated using what we know today: getting around Mars.

“However well the base location is chosen, it is certain that some essential resources needed for its development will be available only at sites tens, hundreds, or thousands of kilometers distant. Global exploration for and transport of these resources will be an essential capability necessary for the growth of the base.” – Robert Zubrin, The Case For Mars [1].

As Zubrin said, in the medium term the capability to travel long distances on the Martian surface will be essential for any future Mars base to have. In addition to resource retrieval, scientists will certainty want to study parts of Mars other than the one where the first base is built, problems with aerocapture can leave arriving crews stranded on the wrong side of the planet, and any Mars Base Alpha will certainly want to be followed up with Mars Bases Beta and Gamma, if our species is to truly colonise the Red Planet.

There are certainly many possibilities for such a transportation system, but the one we will look at today is in many ways the simplest: just take the ITS (Interplanetary Transport System – the upper stage of the Mars rocket) which the crew arrived in, fill it partway up with fuel, and fly on a suborbital trajectory to the destination, using retropropulsion and aerobraking to land.

We have built an approximate model of what the ITS would need for various such journeys. The calculations going into the model are here [2], while the model itself is here [3]. The model is based primarily on ballistic trajectories between different points on a sphere which minimise the total ΔV (change in velocity) of the mission. With these trajectories as baselines, we then consider the effect of corrections from atmospheric drag (including aerobraking), and various other sources, to confirm that these don’t drastically change the final result. We found that adding aerobraking reduces the total ΔV required by a little, but not a lot – because the ship is going at relatively low speed to start with (at least, low compared to aerocapture). The model is not built for extremely precise answers, but is designed for establishing the basic feasibility of the scheme.

The model tells us how much ΔV a ship (any ship, not just the ITS) needs to get various distances across the planet. dv-vs-distance


Unsurprisingly, the further you wish to go on your journey, the more velocity you need. Also unsurprisingly, if you want to go most of the way around the planet, you need to get most of the way to orbital velocity (~3.4 km/s on Mars [4]) and back for a total of 2*3.4 = 6.8 km/s. The important points are the distances in between. To make this a little more real (and specific to the ITS), we can plot the craft’s fuel requirements for various mission profiles.


We can see immediately that the ITS doesn’t possess the range, even with little or no payload, to go to the other side of the Red Planet from the surface and back again without refueling. If there was a base already there, possibly, but even then, the round trip burns as much fuel as returning to Earth, even with the payload limited to a few tens of metric tons.

As for shorter distances, the story is a little better. It’s certainly possible to go many hundreds of km and back even without refueling at the remote location, but the fuel requirements are significant. The ITS is fueled by methane and liquid oxygen, produced using the Sabatier reaction (and electrolysis) from carbon dioxide in the Martian atmosphere and water in the soil. Being all around the habitat makes the COsimple to obtain, but the water is likely to be much harder to produce on a large scale. That said, early missions must produce enough to return the spacecraft, which needs close to a full tank, so sparing 30% of that capacity for an important sortie may not be out of the question. Note that you would certainly want to wait until there were multiple ITSs at your base before sending one out like this, in case the one which is sent out crashes, and to ensure that you still had a fuel tank for your propellant plant to keep filling for the duration of the trip.

Keep in mind, though, that this scheme competes with ground-based electric vehicles (on which SpaceX’s CEO is an expert!) on the scale of hundreds on km – after all, a Telsa Model S has about enough battery power for a round trip to a location 250km away, at least on Earth. On Mars, this would be lengthened by the much thinner atmosphere providing less drag, but diminished by the fact that there are no roads – however, it’s also likely that such a vehicle would bring solar panels with it to recharge.

This brings us to the biggest advantage of such a scheme, however – speed. This is a graph of transit time for various distances of suborbital hop.distance vs travel time.pngNote that the scale of the y-axis is in minutes. In an emergency, the ship could get anywhere in a 1000 km radius in a quarter of an hour (plus however long it takes to prepare the ship for takeoff, of course). This would likely be most useful in emergency situations, allowing Mars Base Alpha to very quickly render aid to Mars Base Beta in the case of some catastrophe there, or to retrieve injured crew from a remote location which they originally reached by other means.

In summary, the ITS is not likely to be used as a workhorse for point-to-point travel on Mars any time soon. It remains an option, however, for certain niche scenarios, and has the great advantage that it will already exist on the Martian surface without the need for any design or manufacturing work. Only time will tell if this ever comes to pass.




[1] Chapter 7, page 229

[2] This document also contains a much more detailed look at all of the assumptions that go into the model, and the extent of their validity.

[3] This is the spreadsheet that contains the simulations used to produce the graphs shown here.


5 thoughts on “Site-to-site travel on Mars with the ITS

  1. This is an excellent piece of quick analysis!

    Two comments, First: You state in the supplementary calculations that “In our model, there are no forces retarding the azimuthal motion – in practice, only air resistance does this.”

    I completely understand and agree with the assumption to neglect atmospheric losses on Mars in a model like this; but if we are speaking in practice, it isn’t true that only air resistance contributes to azimuthal motion. In practice, Coriolis effects from the planet’s rotation are a significant contributor to the propulsive requirements of suborbital ballistic hops over the distances analyzed in this study. In a situation like this, Coriolis effects can actually be a significant fraction of atmospheric effects. For example, you estimated ~1800 m/s of atmospheric losses on landing for a one way antipodal hop. Assuming that means ~900 m/s for a one way equator-to-pole hop, Corriolis effects would end up being ~25% of that (~235 m/s).

    Second: This has no relevance to a model like this, but for any future work you do, you should constrain your landing accelerations. In this model, ITS is pulling +5 g’s at touchdown even before it starts its terminal burn. Your assumption of ignition at 12.5 km altitude would leave you only 15 seconds to kill ~1500 m/s, which is another 9 g’s of deceleration on top of that. 14-to-9 g’s for >10 seconds is not a tolerable landing for a crewed vehicle. If you do end up doing any terminal burn modeling in the future (which I realize isn’t a critical part of this model) you should evaluate your terminal burns with a throttle setting subject to a realistic g-limit contour.

    Liked by 1 person

    1. Thank you!

      I should point out that if we’re being picky, the Coriolis effect acts perpendicular to the azimuthal travel direction, so it doesn’t technically retard the motion… but, of course, you’re absolutely right – I should have at least modeled the effect to the extent of knowing how much it would matter. It’s probably not that hard to take a first pass at it – I might try and update the post with that at some point, and now I can check my answer against yours!

      Regarding acceleration, yes, you make great points. Designing the actual burn profiles for such a hop was well out of scope for a weekend’s work, but it’s definitely something to be mindful of in general. Slide 39 of the IAC talk has the ITS going through 4-6 g’s on landing, so we know that the airframe can survive that (and so if I was designing burn profiles, that would probably be my starting point), but the crew provide a strong upper limit not much greater than that, as you say. A light ITS burning its last drops of fuel with all engines at full throttle would pull about 20 g’s, so it’s not surprising that the raptors have such deep throttle capability (of course, the multi-engine design helps with that as well).


  2. As you say, the use of ISRU for a return trip from Mars to Earth requires hydrogen, oxygen and carbon. But I wold like to know more about your planned energy input. Unless you are using some other power source (such as…), this depends on available solar power at a defined lattitude in a given season. This, working with a given surface of solar panels at a given efficiency, should yield a “charge time”. We would then need to add installation time and consider the fact that the solar panels would have to be transported at least once.

    This type of energy budget is going to be important for calculating time factors in any project on Mars. Good luck with your courageos modelling effort ! Paul_Wi11iams 2016-12-01


  3. sorry, but am addintg to my above reply: Part of a phrase was lost in an edit: I’m replying about ISRU as it would also apply to moving around *on* Mars. The criteria are similar to those for a return *from* Mars.
    I’m wondering if the Red Dragon demonstrator could be used as a testbed for short hops. This would be later scaled up to ITS or intermediate-scale vehicules.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s