Rob Davidoff has a degree in Mechanical Engineering with a concentration in Aerospace. Which makes me proud since he says my Atomic Rocket website inspired him to try for a career in rocketry. Currently he is printing turbines for GE Power. I had mentioned his alternate history Eyes Turned Skyward which I awarded the Atomic Rocket Seal of Approval.

Anyway, one fine day in March 2012, he mentioned a curious proposal for a tether from Mars-Phobos Lagrange to Phobos. It seemed of questionable utility, but it got him thinking about Phobos. Specifically about the possible resources of Phobos and Deimos.

We started brainstorming about it and suddenly we had a marvelous background for near-future history stories about piloted space exploration. We called it "Cape Dread" after the English translation of the word "Deimos."

The concept hinged on three factors:

  • Space exploration and industrialization is not going to grow until you can avoid lugging your propellant up Terra's gravity well. Easiest route is to use in-situ resource utilization, specifically for water ice.

  • Phobos and Deimos contain lots of ice

  • The delta V costs to go from Terra to Luna and back is only 1 km/s less than the delta V cost to go from Terra to Deimos and back (19 km/s as opposed to 20 km/s).


In the following, some of the suggestions are from Rob Davidoff and others are from Yours Truly Winchell Chung. I'm not going to bother distinguishing who said what since I figure you could care less.

Rob mentioned the proposal for a tether from Mars-Phobos Lagrange to Phobos, and noted it was interesting but had no advantage over conventional rockets. But it got him thinking about Phobos. Back of the envelope calculations suggested that if Phobos was a mix of water ice and the same material as Mars, Phobos' density of 1900 km/m3 would suggest it was a whopping 68.9% water ice! Or if Phobos has large subsurface bubbles, those voids make up 28% of the volume.

Either of which would make Phobos quite valuable, as either a source of water ice or as a ready-made colony/base site with free radiation shielding and low delta V transit cost. And it could be both. Water ice and a few cubic kilometers of volume? You could fit an entire civilization inside there. Who needs RGB Mars when you have Phobos?

And this would be already on top of the gravity well, "halfway to anywhere" so to speak. It could be the key to the outer solar system.

Yes, there is ice at the Lunar poles, but that's at the bottom of Luna's gravity well. And starting from LEO, it actually takes less delta V to go to Deimos than to go to the Lunar Poles.

delta-V for Transfers from LEO
Locale to LEOLEO to Locale
Localedelta-VTrip Timedelta-VTrip Time
Lunar Base6.2 km/s3 days3.2 km/s3 days
Deimos5.6 km/s270 days1.8 km/s270 days

In the table below the spacecraft is assumed to be using chemical propulsion and the Oberth effect:

FromToDelta-v (km/s)
LEOMars transfer orbit4.3
Terra escape velocity (C3=0)Mars transfer orbit0.6
Mars transfer orbitMars capture orbit0.9
Mars capture orbitDeimos transfer orbit0.2
Deimos transfer orbitDeimos surface0.7
Deimos transfer orbitPhobos transfer orbit0.3
Phobos transfer orbitPhobos surface0.5
Mars capture orbitLow Mars orbit1.4
Low Mars orbitPhobos1.4
Low Mars orbitDeimos1.9
Low Mars orbitMars surface4.1
EML-1Mars transfer orbit0.74
EML-2Mars transfer orbit0.74
Mars transfer orbitLow Mars Orbit2.7
Terra escape velocity (C3=0)Closest NEO0.8–2.0

There will probably be an orbital propellant depot at either Earth-Moon-Lagrange-1 (EML1) or EM2.

From Earth-Moon-Lagrange-1 (EML1), Mars transfer costs 0.74km/s. Mars capture costs 0.9km/s and the move to low Mars orbit costs 1.4km/s. Trip total is just over 3km/s.

Put an orbital propellant depot in Low Mars Orbit, and supply it with ice from Phobos and/or Deimos. Remember that Rick Robinson noticed that with access to an orbital propellant depot, most cis-Lunar and Mars missions are well within the delta-V capabilities of a sluggish chemical rocket engine, no nuclear rockets needed.

For that matter, shipping water from the Lunar poles to a propellant depot at either EML1 or EML2 takes 2.52 km/s of delta-V. Shipping water from Low Mars Orbit to EML1 or EML2 takes 3.04 km/s, which is only 20% more. So if it cost lots less to mine ice at Deimos compared to Luna, it might be cheaper to supply Terra's propellant depots from Deimos.

The heading of the table below are explained here.


The main thing I see in the table above is that a spacecraft with a maximum delta-V of 6 km/s starting from Terra cannot reach any of the asteroids. But if they travel to Deimos first and refuel, they can reach the asteroids easily.

Better still, although the origins of both Deimos and Phobos are yet unsettled, both appear to have the characteristics of dark carbonaceous asteroids, with anhydrous silicates, carbon, organic compounds, and ice (Bell et al. 1992). If this bears out, Deimos’ regolith would be able to provide water and other volatiles for life support and propellant. Besides silicates, its regolith will also likely contain metals and other valuable materials for construction and manufacturing (Norton 2002).

In this plan, I find it hard to see any good in wasting money and time on the Moon step. Deimos requires less delta-v to get to/from anyway. If it can be exploited for oxygen propellant, then it would be a better place to get it than the Moon (which has a rather steep gravity well).

Also, while I'm generally wary of tether schemes, I must admit that they might be useful for transferring stuff from Deimos to Earth. Deimos starts off in a somewhat inconvenient circular orbit. One or more rotating tethers made out of "waste" metal from oxygen extraction could be used to sling payloads into an elliptical orbit more suitable for navigation back to Earth.

That said, I really think it's better to get oxygen from Earth's atmosphere or maybe Venus or Mars. Much simpler processing, and uses mostly off the shelf satellite technology.

Isaac Kuo

From the standpoint of a science fiction writer or science fiction game designer this would be several orders of magnitude better than a puny little planetary colony or an L5 station like Babylon 5. Phobos docks, ice miners, longshoremen, local 235 ice-miner's union, a set of docking rings sort of sticking up from the ground where NTR spacecraft dock nose first in the microscopic gravity (have to hold a 1 kiloton spaceship, no weight but full inertia). Giant traffic control complex to keep the ships from colliding. Ship repair docks. Warehouses for merchant cargo in transit along with factors for various merchant corporations. Trans-ship point from the Terra-bound clippers to the reusable Mars shuttles. A place for independent asteroid miners to sell their hard found ore.

Tourist traps, luxury hotels. Not to mention the pawn shops, clip joints, bars and brothels that spring up around any spaceport or space station, in other words "Startown". Perhaps repo men ready to seize ships where the captain/owner has gotten too far behind on the ship's mortgage payment. The Phobos Port Authority would of course need a security squad. And there may be Lurkers.

PhobosPort would probably want to stay independent from Mars since free ports can make more money. There could evolve a similar situation to that found in the movie Casablanca. Various national governments (both on and off Terra) would want to seize control of the lucrative port. Meanwhile PhobosPort would be doing all sorts of shenanigans to maintain independence. It might end up corporation-dominated: Phobos Port Authority, Inc, a wholly owned subsidiary of the United Launch Alliance. Or it could start off corporate-dominated and undergo a revolution, like in the wargame Battlefleet Mars. And maybe the mayor could be Elon Musk, laughing all the way to the bank at all the short-sighted naysayers who dismissed the idea of actually making money from space. This is along the lines of the "If You Build It, They Will Come" scenario.

Now that's a setting. All based on the value of strategically located ice. There are eight million stories in the Naked Asteroid, this has been one of them...

PhobosPort was not a catchy enough name, so I suggested "Port Fear", based on the English meaning of "Phobos". Rob changed it to "Cape Fear". Cape Canaveral, Cape of Good Hope, Cape Fear ... it just fits.

Technically they'd only own the installation itself, due to the terms of the Outer Space Treaty. So they would charge for the service of mining the ice and pumping it into your propellant tanks. Any pilot who does not want to pay for the service is handed a shovel and told to do it themself.

There might also be wildcat Phobians setting up shop with their own ice processing gear in smaller internal bubble-caves. Old spacecraft too broken down to make the run back to Terra might be retired to serve as surface facilities. Break off the propellant tanks to melt'em down for metal, bury the habitat modules in regolith for protection, take the reactor and use it as a power generator. Use a Mylar bubble mirror with the dilute Martian sunshine to slowly crack water into hydrogen and oxygen, subsist on a diet of algae. And you'd have a habitat shack for an eccentric outer space mountain man. The Old Rocket Bar might actually be an old rocket.

Sadly, the eccentric mountain men will be a brutal experiment for everybody to find out exactly what the bare minimum is needed to survive in space. The crazy old coot hermits will empirically discover, the hard way, what is vital when living alone in your own cave.

Tony Zuppero called water-ice "Rocket-Fuel Ore". You melt it, you split it into oxygen and hydrogen by electrolysis. How hard is this?

Figure that the ice comes out of the ground at a chilly -100° C. It has to be heated to 22° to become liquid. Ice takes 2.1 J/gram to raise one degree C. So to get it to melting takes 210 J/gram. Then it needs an additional 334 J/gram to melt (the enthalpy of fusion). Finally another 20° at the specific heat of water at 4.182 J/gram per degree. So the total to turn our rocket-fuel ore into liquid water is about 627,000 joules per kilogram (627 kJ/kg).

Now we need energy to split the water into oxygen and hydrogen. Takes 237.13 kJ of electricity per mole of water. 1 mole of water is about 18.0152 grams, so 1 kilogram of water has 55.55 moles. So the electrolysis energy is a whopping 13,000 kJ/kg!

Grand total is 13,627 kJ or 13,627,000 joules. That's a lot. But this is about the same energy you get in the rocket engine by burning oxygen and hydrogen (less the ice melting energy). Everything has to balance. You are taking water and putting the energy into it that will later come blazing out when the oxygen and hydrogen is burned in the rocket engine.

So if a base had a 250 kW power plant devoted to melting/electrolysis of ice, and an unlimited amount of ice, in 120 days it could produce 190 metric tons of fuel (liquid oxygen and liquid hydrogen).

  • 250 kW power plant = 250,000 joules per second
  • 250,000 / 13,627,000 = 0.018 kilograms of ice converted to fuel per second
  • 120 days = 10,368,000 seconds
  • 0.018 * 10,368,000 = 190,000 kilograms = 190 metric tons of fuel

Since Phobos has microgravity, the direction of its core might not be considered "down." One might have an arrangement like the International Space Station, where the entire moon had Forward - Aft - Port - Starboard - Zenith and Nadir instead of up, down, north, south, east and west.

Stickney crater is Phobos' most impressive surface feature. Perhaps it conceals an entrance to the caves of Phobos. Or at least a shorter amount of rock to tunnel through. So operations would start there, eventually making Phobos City in the middle of Stickney. And with docked spacecraft spaced out along the rim of the crater. Initially there will be a few habitat modules inside the caves. Eventually some caves are sealed and they will start building inside the cave shirt-sleeve fashion. A crazy-quilt of structures and floatways to make an Terran civil engineer weep. More grown organically than laid out logically.

Big caverns sealed up except for a big door on the front. Surface-dock, dry-dock, and sleeve-dock. Ship that need repairs floats up. Pull the ship in and pressurize the entire cavern. With the ship owner crying as they sign the second mortgage gleefully offered by the local banker as the sleeve dock pressurizes. "Sorry about the terms to use the dock, but air ain't free, don't you know?" Dock owner swaps out the VASIMR drive, and makes a note to place an order to ULA to replenish his parts stock. Shipped from LEO to Phobos FOB. Meanwhile an engineer sees a business opportunity and opens a used propulsion system refurbishing shop. Or a engine factory. Sure it would cost more but they could corner the propulsion system market in cis-lunar space from one and a half AU away (about 227,000,000 kilometers).

The Phobos Hilton will have to have at least an observation dome, so that the guests can see the impressive view of Mars blotting out the sky. Mars will be about 42.5° or x85 the full moon on Terra.

What about the history of Cape Fear?

It would start as a waypoint for the initial Mars expeditions, the first ones that are more than flags and footprints. Stick a payload of a mining drill, filtration and pumping system plus a few habitat modules on a big LOX/LH2 stage, about a couple of hundred tons all up (minus the Earth Departure Stage (EDS) fuel). Arrive at Phobos, berth to the moon, and go to work. Pick a spot near an ice deposit. Use the drill and filtering system to reach it, and use the empty EDS tank as storage. Call it Collins Base, because the people manning it don't get to land on Mars.

One possible configuration of mission is the old Project Apex concept.

Now we wave our hands and postulate the expedition finds MacGuffinite on Mars. Could be just the fuel on Phobos. Could be political reasons. But you need something to justify moving things to and from Mars, i.e., the establishment of a Transport Economy. In such an economy, owning the only gasoline filling station at the half-way point is a gold mine. The refilling station will allow the cargo spacecraft to increase their maximum cargo mass by about four times.

Say on Phobos they find some methane and CO2 as well as water ice. The base is expanded with another payload shipment. More drilling equipment, some laser sintering, and another large empty EDS tank for storage. They cut into one of the big caverns, sinter the walls to seal it, and use that to expand the base so it can act as a waypoint to Mars. Maybe half a cubic kilometer at first. Heh. In this report on page 12 it suggest that Kuck Mosquitoes will remove the good stuff from under the shell. Which will leave nice voids for habitat caves. The question is whether there are significant sized cavities or if the porosity is only in small volumes of less than a cubic meter or so. If the cavities are large enough, use a laser to sinter the dust and rock into an air-tight skin.

Power is a bit of a question. At that distance, solar panels only produce half of what they can crank out around Terra. So a top-of-the-line panel can generate about 125 W/m2. In direct sunlight. So a 15 megawatt solar panel will need a surface area of about 120,000 square meters (about 30 acres), could be a square about 346 meters on an edge.

Ideally you'd want a SNAP or two, but the initial base probably can't afford the expense. When Cape Fear becomes a commercial base then they'll have SNAPs. When the money starts coming in and Cape Fear can prove it is a going concern.

When the basic base is established, sooner or later somebody will realize that with a little extra investment they can turn a profit. And Elon Musk, Planetary Resources et al prick their ears up. "Profit? Did somebody say Profit?"

So Phase 1 is the initial NASA facility. Phase 2 is commercial. NASA's initial base could be by on solar power. They only need a few hundred kilowatts to harvest a few dozen tons a year. They might even be able to get by using those big half-silvered Mylar balloons instead of solar cells, which would be delightfully retro. They are lightweight and fold up real small.

The major problem with solar power in this case is that both Phobos and Deimos are tidally locked in synchronous orbit. This means solar power units on the surface of Phobos get 3.5 hours of sunlight in every 7 hours. This may mean that NASA's Phase 1 facility might need a nuclear SNAP, expensive or no. A SNAP-10a can produce 12 kW, which could split 7.6 metric tons of water ice into hydrogen and oxygen in 100 days.

As it turns out, there are some areas of both Phobos and Deimos that will stay in constant sunshine for the duration of one Martian "season" (about six Terran months) That is, you divide Mars' year (one orbit around Sol) into four seasons based on its equinox and solstices. For the duration of certain seasons Phobos and Deimos have constant sunshine areas. Presumably at other seasons the moons periodically have Sol eclipsed by Mars. Refer to the maps here.

In Phase 1, it would be something like eight to twelve astronauts in each Mars mission, with two staying on Phobos to keep the machinery running. You want to make sure that the Phobos tanks contain enough propellant for Earth Return when each mission arrives, in case an immediate abort is required.

In Phase 2, you are probably have a commercial base with a few megawatts from nuclear reactors, and a population around 600 to 3,000.

About this point Rob figured that Deimos would make more sense than Phobos. So we changed it from Cape Fear to Cape Dread. Instead of Phobians we now have Dreadheads.

Naturally, in Cape Dread, most of the furniture is going to be metal. Because there ain't no trees to supply wood. There are science fiction novels where spacecraft containing wood paneling and furniture is the height of conspicuous consumption, just to show how filthy rich you are. In his game High Frontier, Philip Eklund has his space colonists building items out of "foamed" titanium. He calls it "space wood". Plastic and ceramics also would be expensive (since they are not available locally) so replacement items made of those materials would be brought in the form of powders, 3D blueprints, and a 3D printer. Print on demand.

Actually, now that I think about it, there might be one type of wood available: Bamboo. In Martin Caidin's novel Exit Earth he sings the praises of bamboo in the space environment. As does Allen Steele in his novel Clark County, Space. Blasted stuff grows so fast it looks like you are seeing it under time-lapse photography. Three to one hundred centimeters in 24 hours. And it has a thousand and one uses. They use it in Clarke County because the space habitat needs an inexpensive, renewable supply of building material. It makes nice paper as well. Bamboo will probably grow taller in lower gravity, and since it is cultivated it would require less management than growing actual trees. Trees grow too slow anyway.

Glass might be possible to produce with locally available Demosian materials. There probably is silicon, add oxygen to make silicon dioxide, which can make a sort of slag glass. You need something transparent for solar power bubbles, domes, and whatnot.

Both Phobos and Deimos have similar composition to Carbonaceous chondrite meteors, which contain all sorts of useful elements. Including crude amino acids and other organic compounds. These are valuable as feedstocks for plastics and food. If you are thinking about a long term civilization, an important chemical is phosphorus. Isaac Asimov called it "life's bottleneck."

Asimov noted that some chemical elements are more common in the bodies of living creatures than in the rocks and dirt composing Terra. Compared to the rocks, those elements are concentrated in living things. The higher the concentration factor, the more vital that element is to organisms and the more rare the element is in rocks. By far the element with the highest concentration factor is phosphorus.

What this boils down to is that a planet's supply of living things is limited to its supply of phosphorus. It is the first thing that will run out. The point being that will also apply to a civilization living inside Deimos. If there is not enough native phosphorus it will have to be imported.

It is in Cape Dread's interest to become as completely self-sufficient as possible. Anything that must be imported is a vulnerability, a way that enemies can put the pressure on by restricting supplies. Or a survival issue if, for instance, Terra loses its heavy boost capability due to global thermonuclear war, a dinosaur-killer asteroid strike, or something equally nasty. Space is not as forgiving as Terra, it requires high technology just to breath. Air is not free. And I'm going to pointedly ignore the Three-Generation Rule as a can of worms I'd just as soon not deal with.

If Terran surface-to-orbit services can bring the boost price down to a few thousand dollars US per kilogram, that is low enough for an individual with Bill Gates or Jeff Bezos levels of wealth could create Cape Dread on their own dollar. People were joking about how James Cameron could easily spend a few million dollars with NASA to film his next space movie on location.

(ed note: in a game I was a playtester in called High Trader, each player was a "faction". Factions like "Daimler-Chrysler-Ford Aerospace", "Lockheed-Martin-Mercedes" and "Agencia Espacial Brasileira". One faction was a billionaire spaceflight junkie named after me: Kutai 'Crank' Nariji-Chung)

You'd need a few kilotons of payload from Terra delivered to Deimos. One kiloton delivered at Deimos requires 3.6 kilotons Initial Mass in Low Earth Orbit (IMLEO), assuming the LEO-Deimos delivery is by chemically powered rockets. So roughly $3.5 billion US assuming $1000/kg Terra-to-LEO boost cost. For launch only, the hardware and other payload is extra. Call it $5 billion US.

If reusable boost rockets can bring the boost cost down to $500/kg, that's only $2.5 billion US. That is doable for the super-rich. Elon Musk managed a $1.2 billion investment, and Bill Gates et al are an order of magnitude richer. It would be a multi-billionaire investing almost their entire personal wealth into it. Or a consortium of mulit-billionaires investing part of their personal wealth into it.

Imagine that there was an asteroid scare. A medium sized civilization-destroying asteroid goes whizzing by between Terra and Luna. A large near-miss (Asteroid "Close, But No Cigar"). That would be enough for a couple of billionaires to go "Huh, maybe we should get some people off-planet in case this happens again, we fail to intercept again, and this time it doesn't just miss on its own." You want to get all the eggs out of one basket. So they put down three or four billion dollars in Cape Dread, Mars, and Luna. Now of course, there's not much money in bomb shelters now that the cold war is over, and five or ten years after the near-miss things are largely back to normal, and Cape Dread and their lunar and Mars facilities have basically become a economic bubble and popped.

But what that initial run has done is created infrastructure. This also gives an excuse of why the infrastructure might be a bit overbuild. It's not just the basic facility, it's designed to be expanded quickly if another one comes. It's basically these four or five billionaires who found their company, and then sit down with the engineers to say, "Look, that was too close. Here's a bunch of cash. Build us a base that can be expanded to house a sustainable basic population as a refuge for us and our off-shore accounts...I mean, as insurance for the human race."

In the initial boom, the Deimos corporation could offer subscriptions to billionaires. They will offer a safe place to dine on caviar while having a ring-side seat to watch all the great unwashed masses getting eradicated by the next killer asteroid. And there are other benefits, such as being a tax haven and data haven. The data transmission to and from Terra will have a 15 minute lag time, but nothing is perfect.

At that point, it doesn't matter if their reasons for investing in Cape Dread are rational ones or not. Once the infrastructure is built, the game changes. It wouldn't matter if the original founders went bankrupt. Maybe the apocalypse they were preparing for never showed up and the investors came calling. Then they sell out in bits and pieces. So Cape Dread becomes still mostly corporate owned, but not the same corporates.

if you worked the parts depot on Deimos, and you believe in the much would the founders be selling assets for? Even if it's millions of dollars in spares...they're stuck at the wrong end of a 270 day trip to Terra, and mostly space-specialized. You could buy the whole stock for cheap, pennies on the dollar. Maybe scrap value plus a bit. Certainly in the price range of a worker on Deimos who had been banking all their pay because there is nothing local to spend it on. So that gives you local businessmen. Even if they still end up buying their stock from Lockheed-Boeing or SpaceX, or whatever, they own their stock there and that chunk of facilities. Because they didn't go home when McCrazy's finances crashed and the base was mothballed. They are in for the long haul. After McCrazy leaves, some of the people at Cape Dread are going to want to stay and make the place a going concern.

Don't forget about the NASA base, which was mothballed and abandoned in-place.

And then MacGuffinite is discovered on Mars, after Cape Dread been established for about 25 years now, what with the initial NASA fuel dump days, then McCrazy's multi-billion-dollar buildup. More mature, some long-term inhabitants.

So you have the old-timers vs the new reps from the companies trying to get into the picture now that MacGuffinite's on the scene...

Once you have a significant population at Cape Dread, there will arise a secondary economy.

A source for inspiration about this stage of Cape Dread's history would be Alexis Gilliland Rosinante series. The Revolution From Rosinante, Long Shot For Rosinante, and Pirates Of Rosinante. In the novels the protagonist is an engineer who was contracted to build and L5 colony in the asteroid belt. For a variety of reasons the economy of Terra slips into recession, and the colony is loses its funding. The protagonist is suddenly finds himself transformed from an engineer into the role of governor of the L5 colony. Much like we are postulating for Cape Dread.

What constitutes the MacGuffinite is a problem. It could be something on Mars, could be something on Deimos.

Maybe Dread Bourbon, made from the finest yeast culture, distilled with authentic all-natural vacuum! Imported at great expense from Deimos! The perfect way to impress all of your snob friends! ....nah, that would only be a niche market.

Deimos is a carbonaceous asteroid, unlike Luna. Luna has only trace amounts of carbon in the regolith. So Deimos is a good source of organic chemicals as feedstocks for various things. It also has some silicates, so it might make some microprocessor chips. As a "clean room" to make microchips Deimos' microgravity is an advantage over free fall. Dust dispersed in the clean room's atmosphere will eventually settle out, unlike in a space station. Carbon is also needed for carbon nanotubes.

Deimos could also supply materials to build an L5 colony, with the rest of the materials coming from Luna. I'm not saying that an L5 colony will ever be built. If the reasons for building it are plausible enough, it might be possible to fool enough rich people to invest in the project. Just long as the stupid rich people pour money into Pie-in-the-Sky-L5 to give Cape Dread enough time and resources to become at least semi-self-sufficient. Tell the rich that you are building Elysium, they'll love that. Then even if the L5 project goes bust, Cape Dread will be standing on its own two feet.

As long as Deimos has something that Luna doesn't, there is a reason for Cape Dread.

Once you have some second and third generation population, and Cape Dread is, oh, 80% self-sufficient, that might be enough. It will be self sustaining.

There will be a trickle of rich tourists visiting Cape Dread. "But Charles, everyone who is anyone has visited the Cape Dread Hilton". I'm sure the view of Mars overhead in the sky is spectacular (16.7° or x33 the full moon on Terra). For when a two-week trip to the Lunar Hilton isn't far enough away from those plebeian masses.

Since the Terran launch windows for Deimos Hohmann transfer orbits occur every 26 months (two years and two months), Cape Dread's "off season" will be in between Hohmann synodic periods. Keeping in mind that since spacecraft can refuel at Cape Dread, it is less critical to find the optimal transit. This means that the launch windows will become broader. Typically Terra-Deimos launch windows occur every 26 months and are 3 month's "wide". For example in the year 2020 you could launch from Terra to Deimos into a Hohmann transfer anytime from July 2020 to Sep 2020. Since you can refill propellant at Cape Dread, the launch window will be a few months wider than three.

If there is no cheap propellant at Terra LEO, then you'll see the trip from Terra to Cape Dread rigidly constrained by Hohmann launch windows. Fuel at LEO is either super expensive fuel boosted from Terra, or medium expensive fuel boosted from Cape Dread to LEO. In either case the price of fuel will ensure that launches to Cape Dread will be strictly by Hohmann.

Meanwhile fuel at Cape Dread will be super-cheap, much like purchasing gasoline for your automobile at the oil refinery. No fuel transport cost cost. So trips from Cape Dread to Terra might not be constrained by launch windows at all. Having said that, at the height of the "vacation" season (every 26 months when the flood of ships arrives from Terra on Hohmann trajectories), Cape Dread might raise the price on propellant. Just like any beach resort raises its prices during the vacation season, because they have you over a barrel. Perhaps groups of spacecraft owners might form a co-op in order to get a better price on propellant.

Implication: every few months ships will leave Cape Dread for Terra in order to avoid the seasonal price increase on propellant. However, the ships will gradually accumulate in LEO, since the expensive LEO propellant forces them to only launch for Cape Dread every 26 month Hohmann window. And 270 days later it is the tourist season at Cape Dread as the huge convoy of ships arrive from LEO.

Then Rob and I tried going into the timeline in more detail.

Around 2022 or so, Fobos-Grunt 2 manages to avoid being eaten by the Great Galactic Ghoul and conducts close investigation of Deimos on the way to Phobos, confirming their similar composition (except that Deimos is "fluffier"). Then when it lands on Phobos, radar and sampling confirms significant sub-regolith ice and large cavities within the body. Large amounts of volatiles are discovered.

When NASA plans a crewed mission, they decide to use Phobos as an orbital propellant depot, and an orbital base for telerobotics operators to support the Mars astronauts. The gold rush is on for Phobos' propellant. NASA does one, maybe two landing to set up Collins Base on Phobos around 2030 or 2035, and three or four major missions to staff it and the Mars ground sites.

Some whiney little nation in the UN introduces a resolution to apportion the Phobos ice for all member nations, and is ignored. Much like the current Moon Treaty.

NASA is wrapping up the Mars missions when asteroid "Close, But No Cigar" comes shooting by and scares the poo out of everybody, around 2039.

This lights a fire under a group of billionaires. Call them the Exodus Foundation (est. 2040). Other billionaires start leaning on their owned politicians to increase NASA funding, to "Rescue Humanity" (translation: "Rescue Us Billionaires")

The plan is to make a second base. Not NASA, but the Exodus Foundation. By invitation only. Mostly only if you are rich enough.

So in the next six years the hardware is developed, launched into LEO, moved to Deimos, starts being set up ... and then it goes bankrupt (2046).

The work crew at Deimos is about one hundred or so, though the base is being built to handle a few thousand. In 2046 they are told to pack up and prepare to be returned to Terra. About 15 or 20 refuse to return, they are staying on Deimos. They might even file a lawsuit against the toxic assets at Deimos as compensation for unpaid back leave.

"You owe us 20 workers six years of back wages, plus hazard pay, plus early termination payout, plus of maybe a half mill each, or more. Or you can give us XYZ, theoretical worth of $1.2 billion, but auction price of right around $8 million. Because while a Morris Technologies MT5600 printer or a closed loop-life support system is worth a lot new in box on Earth....used on Deimos, return shipment not covered? Worth...less." (Some might even suggest removing the ellipse between "worth" and "less.")

So then a few years go by, the market recovers from the "slight hiccups" after the asteroid near-miss, and meanwhile the people who stayed at Cape Dread (I think they may be the ones to name it that) are fitting out the facility a bit. Rationing the stored food, expanding the horticulture, discovering the hard way what they forgot to bring, trying to find something they can trade for a "care package" mission from Terra containing whatever they forgot to bring (radio message: "Hey NASA, we have some Deimos geological data. We are willing to trade it in exchange for some stuff from you").

In a way they're lucky. The job they were sent to do was to fit this out to house roughly two orders of magnitude more people, so they have some spare supplies. The supplies will last about two orders of magnitude longer (supplies for a couple of thousand people are being used by 20 people). And stuff like the 3d printer/processor and such means they have fewer "missing pieces", provided they can get it all working and keep it that way. First order of business is to use the printer to clone the printer.

In 2055 someone decides to do some L5 colonies, maybe it's an eco-group trying to prove self-sufficiency, planning to export knowledge workers, maybe its some billionaires, maybe it's...I dunno. Maybe all of the above. This is the start of the Lagrange station boom, taking advantage of the infrastructure and tech left over from the Exodus Foundation bubble of 2040-2045. They need fuel for orbital transfers, they need raw materials, they need a lot of stuff. And though it's 270 days away, Cape Dread is the "nearest" source in terms of delta-V. So they contract Cape Dread to supply some vital raw materials. Giving Cape Dread some vital capital to purchase desperately needed items, and to upgrade their industrial base. And to import labor. 20 people isn't enough of a base staff. Not if they want to buy silicon, fuel, and products by the kiloton. Cape Dread really starts taking off about 2058.

Somewhere along the line it becomes lucrative enough that SpaceX or somebody starts up a LEO-Cape Dread passenger service. Or it may be incidental on the cargo transit. You don't want the cargo ships returning to Cape Dread to be empty, now do we?

So in 2063 is where Cape Dread goes from two main docks (slips?) to three or four, and expands to a population of a few hundred, plus another thousand-odd temp worker. Their children are in their early teens. Cargo to Mars may end up fairly cheap. If a ship is designed to carry 200 tons of fuel and mixed goods back, then an extra couple of tons on the way over is cheap. After a while, somebody wants to, say, stage an expedition the asteroid belt, and they contract Cape Dread for some refueling. Cape Dread will want to create its own gasoline service-station industry, offering maintenance and repair services in addition to just fuel. Eventually Cape Dread will purchase or make its own cargo ships. Then the secondary industries start appearing.

A Fly In The Ointment

Since this page was written, a potential problem arose.

Remember that assumption #2 was "Phobos and Deimos contain lots of ice"? Well, there is a chance that ain't necessarily so.

      Hello Dom, Ulrik, & other playtesters of High Frontier 4th edition:

     If you are in LEO, the closest object to you (in terms of energy) is not Earth, not Luna, but the moon of Mars called Deimos.

     After the Russian Phobos 2 probe (although abortive) returned data on both moons of Mars, widely interpreted as indicating hydrated clays on the surface, similar to Ceres spectrographically, and analogous to Cl and CM meteorites. This suggested water could be mined there for space activities, which has excited space enthusiasts for 20 years, including me. I based Phobos and Deimos in the game on reports of water vapor detected by Phobos 2 presumably from Deimos, and I set up Deimos as the most valuable piece of real estate in the game. If you have a chance, read "The Deimos Water Company" by Kuck, a paper about establishing a profitable water factory on Deimos that became practically the basis for the entire High Frontier concept.

     Old timers may recall that Deimos at one time was Spectral Class D. This is from the report from the ISM instrument onboard Phobos 2, and from the IRTF telescope in Hawaii, that reported that "Deimos resembles the D-class asteroids found mainly in the Trojan asteroid clouds, rather than the C-class asteroids in the main belt which are a close spectral match to Phobos and Deimos in the visible wavelength spectrum."

     The reason this was changed to C-class in High Frontier 3rd edition is the "windfall problem". If a player en-route to Mars serendipitously shoots a raygun, or lands a buggy, successfully Deimos, he gains a powerful lead. This was blunted by changing this from D to C. But in our recent playtest, a player got a serendipitous success on C-class Deimos, and it was still a powerful advantage.

     As it turns out, re-interpretations of the data make it clear that Deimos does not have hydrated clays on the surface, and neither Phobos or Deimos even have hydrated silicates, making them much drier than when I first designed the game. In the words of the paper "It is clear that Deimos is much drier than Orgueil (a hydrated carbonaceous chondrite) (and similar to) Karoonda, a totally anhydrous chondrite."

     There are many ways to spin this. The most straightforward is to make Deimos D-class, once again, but reduce its hydration to 1 or 0. This effectively plummets the value of Deimos, and I think removes any windfall problem.

     But there is another way to spin the data. All D-class worlds appear to be crazy dry. This is unexpected, because D-class are way out there, most are in the Trojans, and supposedly the further you go out, the colder it gets and the more ice you find. The moons of Jupiter and further out are almost entirely icy. The leading theory is that water never "mobilized" on D-class worlds, never heated enough after accretion to melt the ice and create hydrated silicates through the action of groundwater protect from the vacuum of space by a permafrost layer. The words of the paper: "lt is possible that the depth of the 2.9 μm clay absorption bands is anticorrelated with the total abundance of water in the low-albedo asteroids."

From the High Frontier 4 Playtest Mailing List (2019)

      I'd heard bits of the same, plus speculation that their low densities might be attributed neither to convenient volatiles or giant cavities begging to be turned into pressure volumes but by being rubble piles that never consolidated and still have space between the gravel bits.

     Just goes to show the difficulties of speculating settings for hard science fiction in the absence of "ground truth". JAXA is finally going to go get some after the failure of Phobos-Grunt, but we won't really know until 2029 (!) on that mission's schedule. I think Phobos-Grunt may be the mission failure in the last 10-15 years I'm most frustrated by. Spirit and Oppy giving up the ghost was sad, but they succeeded beyond all expectation first. Schiaparelli was disappointing, but had few instruments and ESA are still going back.

     But Phobos-Grunt...a sad computer issue so it never even left LEO, no follow up by the agency that built it, and the next missions with the same target won't fly until 2024, a dozen years later. It'd be cool if Cape Dread turned out to be possible, but I hate not even knowing.

     Even if Phobos and Deimos are dry, I still like the mental image of the colony half built on, half built into, and half docked to the surface of those oft-overlooked little dirt clods. And of the people sent to do a job being unwilling to give up on it when corporate does.

From a twitter thread by Rob Davidoff (2019)

Raw Data

Satellites of Mars
Orbital characteristics
Periapsis9,234.42 km23,455.5 km
Apoapsis9,517.58 km23,470.9 km
9,376 km23,463.2 km
0.3189102 days
(7 h 39.2 min)
1.263 days
(30.312 h)
0.31891 days1.26244 days
2.138 km/s1.3513 km/s
Physical characteristics
Dimensions27 × 22 × 18 km15 × 12.2 × km
Major axis
13.4 km7.5 km
Median axis
11.2667 km6.2 km
Minor axis
9.2 km5.2 km
1,548.3 km2495.1548 km2
Volume5,783.61 km3999.78 km3
Mass1.0659×1016 kg1.4762×1015 kg
1876 kg/m31471 kg/m3
0.0057 m/s
(581.4 μg)
0.003 m/s
(306 μg)
11.39 m/s5.556 m/s
Temperature≈233 K≈233 K

Water is one of the most useful substances available in space. It can be electrolytically split into oxygen and hydrogen for use as chemical rocket fuel or hydrogen for nuclear thermal rocket propellant, astronaut breathing mix or later use in regenerative fuel cells. Water can be used straight (instead of hydrogen) by a nuclear thermal rocket, abet with a performance penalty. It can be used to create hydrogen peroxide, which is a rocket monopropellant. It can be drunk by astronauts, fed to hydroponically grown plants, used as coolant, or used as radiation shielding.

Currently it costs about $5000 US per kilogram to boost something into orbit from Terra. It would be an incredible savings if one could find water or water ice already in space.

As the delta-V for a mission goes up, the amount of propellant required goes up exponentially (or looking at it another way: the amount of payload shrinks exponentially). Large amounts of propellant are expensive, but the higher the mass-ratio the higher the likelihood that the spacecraft will not be resuable. Propellant expense is bad enough, but that's nothing compared to having to build a new spacecraft for each mission. Increasing the mass ratio means things like making the walls of the propellant tanks thin like foil, and shaving down the support members so they are fragile like soda straws. With such flimsy construction it does not take much normal wear-and-tear to turn the spacecraft into junk.

Having a propellant depot at the mid-point of a round-trip mission cuts the required delta-V in half. Instead of the spacecraft having to lug enough propellant to go to Mars then return to Terra, it only carries enough for half the trip but re-fuels (re-propellants, or re-remasses) at the halfway point. And when you are dealing with exponential growth, cutting the delta-V in half cuts the propellant amount much more than half.

So one needs a source of water ice, something to mine the ice, and tankers and lighter to keep the depots filled up. Examples of tankers include Kuck Mosquitos, Zuppero Water Ships, and Zuppero Lunar Water Trucks.

If your orbital propellant depot is located in Low Earth Orbit (LEO), the closest hydrated body in delta V terms is Deimos. It its closer than ice frozen in the Lunar poles (the 3 km/s take off delta V is very expensive), and it is a zillion times closer than Terra. Deimos also does not require high thrust rockets for soft landings and high energy take-offs, as is the case with Luna.

The drawback to Deimos over Luna is that launch windows to Deimos occur only every 26 months (two years and two months), and the transit time is 270 days (nine months).

However, since spacecraft can refuel at Deimos, it is less critical to find the optimal transit. So the launch windows will be wider.

For more details about the economics of Deimosian ice, see The Deimos Water Company by David Kuck. Yes, the same scientist who came up with the Kuck Mosquito concept.

Phobos and Deimos are both thought to have water ice, with Deimos ice being closer to the surface (depth of 100 meters at the equator and at a depth of 20 meters at the poles). Deimos also takes less delta V to reach than Phobos. Both Phobos and Deimos are in synchronous orbit.

delta-V for Transfers from LEO
Locale to LEOLEO to Locale
Localedelta-VTrip Timedelta-VTrip Time
Lunar Base6.2 km/s3 days3.2 km/s3 days
Deimos5.6 km/s270 days1.8 km/s270 days
Mars4.8 km/s270 days5.7 km/s270 days

Phobos and Deimos have spectra, albedoes, and densities similar to C-type or D-type asteroids. They are composed of rock rich with carbonaceous material. They probably contain hydrated (water-containing) minerals, and may have deposits of water ice in their interiors.

Table 2. Hypothetical cash flow for the project using the drill rig vehicle presented in "Exploitation of Space Oases" at Princeton in 1995. This uses nine of the original twelve drill rig vehicles.

Proposed cash flow - 12 initial vehicles
Development & Construction
      Drill & Vehicle @ $12.5mx3
Launches @ $70m Sea Launch
Launches @ $60m Proton
General Expense Estimate$10m$10m$10m$10m$10m$10m$10m$10m
External tank to Mir$10m$10m
Total Hardware Expenses$10m$10m$10m$257.5m$10m$185m$20m$10m
Interest @ 20%$2m$4.4m$7.3m$60.2m$74.3m$128.1m$155.8m
Grand Total Expenses$10m$22m$36.4m$301.2m$371.4m$630.7m$778.9m$944.7m
Product Sales
      100 tonnes Deimos/Phobos @$8k/kg$800m
      200 tonnes Deimos/Phobos @$6k/kg$1,200m
Total Gross Income Estimate$800m$1,200m
Accumulated Net Profit$169.2m$1,055.3m

Deimos versus Phobos
Mars Arrival (2033) plus Earth return (2035) ΔV~2.9 km/s~3.3 km/s
Two-way speed of light lag to nadir point on Mars0.134 s0.040 s
Max visible Mars latitude (with 5° elevation mask)77.6°64.8°
Fraction of Mars surface visible97.5%90.5%
Duration of comm line-of-sight to asset on Mars equator59.6 hrs4.2 hrs
Gap between comm passes to equatorial asset on Mars71.8 hrs6.9 hrs
% time a typical Mars surface site is in view45%38%
Max eclipse duration84 min54 min
Typical nighttime duration15.1 hrs3.8 hrs
Max eclipse % of orbit period4.6%12.0%
Max continuous lighting duration in Northern hemisphere~300 days~140 days
Average eclipse season duration~83 days~228 days
Max continuous lighting duration in Southern hemisphere~225 days~95 days

Advantages of Deimos:

  • Round trip ΔV from Earth to Deimos is about 400 m/s lower than to Phobos
  • Longer communication access to assets on Martian surface
  • Communciation access to higher Martian latitudes
  • Superior line-of-sight to Earth from Deimos due to fewer Martian occultations
  • Twice as much time with constant sunlight and only a third of the eclipse season duration as Phobos

Advantages of Phobo:

  • The gap between comm passes to Martian surface assets is much shorter
  • Phobos is closer to the Martian surface, resulting in higher data rates or smaller antenna & power
  • The maximum possible eclipse duration is 30 minutes shorter on Phobos
  • Phobos appears to be more geologically interesting than Deimos
  • Sample return to Phobos is easier from low latitude Mars sites

The surfaces of Phobos and Deimos are very dark like carbonaceous asteroids, but they lack a detectable absorption feature due to chemically bound water. This does not preclude interstitial water, only chemically combined water, such as in phylosilicates. This fits with the classification of Deimos as a type D body, which may never have been heated to a temperature adequate to hydro-thermally form phylosilicates. Carbonaceous type C chondrites are divided into sub-classes P, D, RD, T, F, G and B. Only in the middle asteroid belt were bodies heated enough after accretion to melt the ice and create hydrated silicates through the action "groundwater" protected from the vacuum of space by a permafrost layer. In the P-class and D-class asteroids, ice is still present and was never mobilized. Deimos contains water as permafrost even though the surface is anhydrous.

Hartman has reported that Phobos, Deimos and some NEAs are class "D" bodies which originated near the orbit of Jupiter at 5.2 AU. Ice is stable at this distance as a solid without transpiration into the vacuum of space.

The surface temperature varies from 400C (3130K) at the equator to -2100C (630K) at the poles. The axial tilt causes large annual temperature swings as a function of latitude. This would cause any surface volatiles to be driven off long ago. With Deimos being a class-D asteroid having no combined water and the baking of the surface, the anhydrous spectra should be expected.

Fanale calculates that ice should exist at a depth of 100 meters at the equator and at a depth of 20 meters at the poles of Deimos. Thus, the drilling equipment proposed in 1995 by Kuck should be able to reach ice at or near the poles, but not near the equator.

Two isolated solar wind disturbances about 5 minutes in duration were detected by the Russian spacecraft Phobos-2 upon its crossing the wake of the Martian moon Deimos about 15,000 kilometers downstream from the moon on 1 February 1989. These plasma events are interpreted as the inboard and outboard crossing of a Mach cone that is formed as a result of an effective interaction of the solar wind with Deimos. Possible mechanisms such as remanent magnetism, cometary type interaction caused by heavy ion or charged dust production or neutral gas emission through water and other volatile loss by Deimos at a rate of about 1023 molecules/sec. Due to the age of Deimos, the later interpretation is favored. This is the equivalent of a geophysical anomaly indicating the presence of water on Deimos.

From THE DEIMOS WATER COMPANY by David Kuck (1997)


Of the elements assumed to compose Phobos, many would be important when processed into water, propellants, and other materials. These materials would then have applications in interplanetary travel, Mars Exploration, base construction, or Earth uses.

For the base on Phobos to be used as a transportation node for inter-planetary travel, the production of water and propellants would be important. Because of the abundance of water on Phobos, a base for water supply could be very valuable and economical. In addition, the water could be processed with electrolysis or thermochemical reactions to yield the propellants, LH2 and LO2. These propellants, however, would only be produced for immediate use since their highly reactive and explosive natures make them difficult to store safely. CH4, Methane, is another propellant which is less reactive and more stable than LH2 or LO2, but it yields a lower specific thrust. Methane could also be considered for fuel production.

For the Phobos base to be economically valuable for Earth supply, silicon semi-conductors could be produced with higher precision and lower cost than on Earth. Indeed, Phobos' abundance of silicon and low gravity make it ideal for the crystal growth and vacuum casting for this application. Also of use on Earth and in space are ceramic magnets (MgFe2O4).

Ceramic magnets have a wide variety of uses in communications for antennae, cassette tapes, deflection transformers in monitor screens, and computer disks.For use in the Phobos base and in other space structures, Phobos has many material capabilities. The production of Iron and Magnesium is feasible and will be discussed later. Other possible building materials are ceramics, glass and fiberglass which are processed from Al203, MgO, SiO2, Na2O, and CaO. With the exception of Na20 and CaO, the other elements are found in abundance on Phobos. Unfortunately, the manufacture of metals and metal alloys is not as feasible because only trace amounts exist of the pure metals. In regolith, most metals eventually become oxidized, therefore it is more difficult and costly to extract from their oxidized forms.Below is a discussion of two of the most feasible materials that could be manufactured. The processes involved are discussed using the resources available on Phobos. These two materials, Iron and Magnesium, could be manufactured from their oxide forms found on Phobos.

Iron Extraction

Ferrous compounds (FeO and FeS) are relatively plentiful on Phobos. Table 2.2 shows how iron cold be extracted from at least on of these compounds (FeO). The compounds should be easily obtained by a magnetic separator which the regolith is run through prior to water extraction. Silicon will reduce FeO into iron at 1300° C according to the equation:

2FeO + Si ⇒ 2Fe + SiO2 (A)

This reaction requires pure silicon which is not present on Phobos. There are as stated before, plentiful quantities of silicon dioxide. Silicon dioxide can be reduced to silicon at 2300° C by the reaction:

SiO2 + 2C ⇒ Si + 2CO (B)

Pure carbon is required for this above reaction, in which Phobos's composition should be approximately 3.62% carbon. However, the simplest method of isolating this carbon would be to reduce one of its gaseous compounds that is released with water vapor in the oven during water extraction. The below reaction demonstrates how carbon monoxide can be reduced to pure carbon:

CO + H2 ⇒ (intermediates) ⇒ C + H2O (C)

Carbon monoxide can be isolated by use of a condenser which takes advantage of carbon monoxide's unique vapor point.

Magnesium Extraction

Phobos should contain an ample amount of magnesium oxide which can be reduced to pure magnesium. The process would involve heating magnesium oxide, silicon, and calcium oxide to 1200° C to produce vaporized magnesium and solid Ca2SiO4. The magnesium vapor is then liquefied by a condenser and then poured into molds to form magnesium ingots. The problem of this method is that the quantity of calcium oxide is relatively scarce on Phobos. Glass production, discussed later, will take all the available calcium oxide. Another method which requires a higher temperature (2300° C) uses the following reaction:

MgO + C ⇒ Mg + CO

The advantage to this method is that carbon is more plentiful than calcium oxide which is required in the first method. This method does require more energy because of its higher temperature, but nuclear power should provide an ample amount of energy so that this will not be a problem. Therefore, this carbon method will be the preferred method.

Other Products

What other production possibilities exist on Phobos? Table 2.2 shows that silicon dioxide should be 23% of Phobos composition. Therefore, glass could be produced since 72% of its composition is silicon dioxide. The other compounds that make up the other 28% of glass are also present on Phobos but not in large quantities. Calcium oxide and sodium oxide make up 1.51% and 0.76% of Phobos respectively. However, some small scale production of glass should be possible using entirely Phobos substance.

The carbon gases (CO, CO2, CH4) that are released during the water extraction process can be processed into ethylene (C2H4). Ethylene is the building block of polymers.

If glass and polymers can be produced then their composite, fiberglass, can also be produced. Fiberglass can be useful as a structure material.

From PROJECT APEX (1992)

The Phobos plant concept is sized to obtain 600 tonnes per year (t/yr) of water from rock and soil. The weak gravity of Phobos presents significant challenges but mining operations may prove more efficient than typical terrestrial ones. The shape and reflectivity of Phobos and Deimos suggest that they may be similar to carbonaceous chondritic asteroids. If so, they could consist of up to 20 percent water. The Phobos propellant plant design assumes a 5 percent water content and is based on a rock-penetrating prototype device that was developed at the Los Alamos National Laboratory. Laboratory and field tests with this prototype indicate that it is effective with most types and conditions of rock and soil. The plant, depicted in figure 2.4.6-4, uses a rock melter configured as a coring device. An impermeable glasslike lining forms in place around the borehole during penetration and seals in the released volatiles so that they do not escape into the surrounding porous rock. The released volatiles will probably contain such impurities as carbon monoxide, carbon dioxide, and hydrogen sulfide. Gross separation occurs when condensing water from the gases is emitted from the borehole. Absorption filters further purify the water which is then dissociated by electrolysis. The resulting oxygen and hydrogen are liquefied and stored. Between boring operations the plant makes short movements to new bore sites. This is accomplished by using legs with end-effectors after raising the plant with hydraulic jacks. The mass of a plant that extracts 600 t/yr of water is estimated at about 86 t with a power requirement of about 1067 kWe. The mass estimates include a self-contained nuclear power supply (20t), radiation shielding, and habitat for crew (20t).


The chosen crater is nearly 1/8 mile in diameter, making it a good target for landing. It sits just forward of the nearly horizontal polar table with respect to the equator and is tipped slightly toward Mars for a good view while providing lateral and aft shielding with its walls. The rotating solar array, addressed below, can be anchored on the smooth plain just behind and above the habitat/lab crater at highest latitude, less than 1/10 mile away.

Landing here, with the low surface gravity and low surface density, will be more of a rendezvous and docking maneuver than a landing. I would not expect that there will be much in the way of exhaust scouring of regolith but some dust may be redistributed by vernier firings. Larger RCS thrusters will most likely not be used to control the descent but instead be reserved for abort. I would expect that a controlled slow fall to the surface, combined with auger screw legs that are counter-rotating at time of touchdown to drill into the 50 meter thick regolith blanket upon contact, would be the most effective method of staying there once you get there. Another method would be to fire harpoons into the regolith from altitude and reel yourself in. In this low gravity field, the surface will not be very dense. It will not behave in ways that normal dirt would be expected to behave in a one g field, as there has been nothing much to compress it. Think of it as deep snow and how efficient you are at moving in that. It may support you after initial compression but gives way under thrust. Moving shielding materials and larger equipment around would most likely be done with small assist devices, perhaps gas thruster powered, though the dust kicked up by them may itself be problematical.

Huge structures and how to erect and support them are not as difficult as they may seem... merely different in how you go about it. For the most part, you can just land them all in one piece. Keeping tall ones from falling over is another story. Guy wires would be largely ineffective. At the low level on the mast to which they would have to be attached, the lever arm of the upper, unsupported section of the mast would have a large advantage on the lower portion and the low density of the regolith would pose little resistance to the anchors pulling out under steady tension. One good hit off center with a meteor, harmonic oscillations build up and the whole thing gyrates, goes out of column and topples over or flies off into space. Think of it as driving a stake into Corn Flakes, Rice Crispies or Grape Nuts.

Given these limitations for conventional stabilization, one is also given rise to a solution, Just as objects can be pulled out of the regolith with ease, so may large objects be thrust into it. The masts for the solar arrays are designed having a long, narrow, sharp, fluted cone base that is thrust fairly deeply into the regolith by firing it at low velocity from above the surface. The panels are assembled to it and unfurl in place. The whole mast, panels and all, turns on a dual full-floating bearing near the base, driven by a heliostat to follow the Sun as Deimos travels around Mars. The array is situated so that only one panel is eclipsed by another at one time, ensuring that power fluctuations will be kept within design limits. Burying the habitat can be done in a number of ways, but it does pose its own problems. Heavy equipment would not be very effective in this low gravity field, if at all—mostly because it wouldn't be heavy. Get a blade full of dirt and your tracks will simply dig out from under you.

It may take a while to move that much mass, but a good ol' HUGE hand shovel might be all that is really needed—but that isn't very glamorous. A beefed-up RMS style arm on top of the habitat that pulls the crater limb in backhoe fashion would be most effective, as the lab is already screwed into the surface. Leverage, rather than brute force. With the scraper removed, the arm serves as a crane for working in close proximity to the lab.

On the other hand, a triad of properly calculated and placed shaped charges would be most efficient to move the dirt where you want it. WHUMP! Create three new craters and do seismic studies all at the same time. It'll take a while for the dust to settle in this low gravity but we'll at least get a show and some new scenery out of the operation.

From WORKING ON DEIMOS by B. E. Johnson (2000)


     A human mission to one of the two moons of Mars has been suggested as an easier precursor before a mission to land on Mars itself. Astronauts would explore the moon in person and teleoperate rovers on the surface of Mars with minimal lag time, returning samples to Earth. Lockheed Martin evaluated such a mission as part of its Stepping Stones sequence of missions in the spirit of the “Flexible Path” approach advocated by the Augustine Committee.

     In this paper, we compare Deimos and Phobos as potential destinations, including trajectory design, communications access to Earth and the Martian surface, solar illumination, expected radiation environment, planetary protection issues, and physical access to and from the Martian surface. While prior mission concepts have tended to focus on Phobos, we conclude that Deimos is the better destination for an early teleoperation mission largely because it is farther from Mars than Phobos. This reduces the required mission ΔV by 400 m/s, provides longer communications access and line of sight to 15 deg higher latitudes on the Martian surface, and reduces the frequency and cumulative duration of eclipses by Mars so that a solar powered mission is easier on Deimos than on Phobos.

     Using a shape model of Deimos, we performed global lighting and communications access analysis and determined that there are two specific regions on Deimos which are the most favorable landing sites. Small areas along the North and South arctic circles on the Mars-facing side of Deimos experience a continuous view of Mars, continuous sunlight for up to ten months during polar summer, and continuous line of sight to Earth during most of the sunlit season. These sites are centered near 60° N 0° W, and 51° S 7° E.

     A timeline for a mission to these two sites is provided for the 2033-2035 opportunity. This is the easiest opportunity during the next few decades because optimum Earth-to-Mars orbital geometry will likely coincide with the phase of the solar activity cycle that provides the most protection from galactic cosmic rays, reducing the effective radiation dose.

     During this mission, the crew would land at the southern hemisphere site first (landing site Alpha), during the middle of the southern summer season. After a four month stay, the crew would depart the surface of Deimos to orbit for 50 days during the equinox and eclipse season, when lighting is unfavorable at any location on the Deimos surface. At the beginning of northern summer, the crew would land at the northern site (landing site Beta) and stays for ten months before returning to Earth. In this way, the crew can explore both hemispheres of Deimos without requiring advanced power systems (meaning it can use off-the-shelf solar power instead of risky nuclear power).

II. Comparison of Deimos and Phobos

     Phobos and Deimos are both small, irregular objects comparable in size to the largest terrestrial mountains. Their origins are debated and their composition uncertain. They may be captured D-type asteroids, or remnant debris ejected from early large impacts on Mars (similar to the formation of Earth's Moon), or material left over from when Mars first accreted. Both moons are tidally synchronized to Mars so that the same side faces the planet at all times. Both moons have nearly circular orbits very close to their parent planet within a few degrees of the equatorial plane. The orbit of Deimos is just beyond Mars's geosynchronous orbit altitude. For comparison, its orbit altitude is similar to the orbit used by Earth's GPS satellites. Phobos is even closer to Mars, with an orbital period only one quarter that of Deimos. The orbit altitudes of these moons determine several of the parameters which are key to this study, including communications access to Mars and solar lighting. Relevant parameters are provided in Table 1 below and discussed in more detail in subsequent sections. For comparison the table also includes data on two other potential mission orbits: a low altitude orbit and Mars geostationary orbit.

Table 1. Comparison of Phobos, Deimos and potential spacecraft orbits
Dimensions (triaxial radius)-13 x 11 x 9 km-8 x 6 x 5 km
Mean orbit radius3797 km9377 km20462 km23460 km
Mean orbit altitude400 km5980 km17065 km20063 km
Orbit mean inclination (relative to equator)Any1.1 deg0.0 deg2.4 deg
Orbit period1.97 hr7.7 hr24.6 hr30.2 hr
Orbital velocity3.37 km/s2.13 km/s1.45 km/s1.35 km/s
Maximum eclipse duration42 min54 min78 min84 min
Max eclipse % of orbit period35%12%5%4.6%
Eclipse season duration-228 days-83 days
Average night duration-3.8 hr-15.1 hr
Max visible latitude on MarsInclination+26.569.8 deg80.4 deg84.1 deg
Max latitude with 5 deg horizon maskInclination+2264.8 deg75.5 deg80.2 deg
Two-way light time to nadir point on Mars3 ms40 ms114 ms134 ms
Duration of line-of-sight to Mars equatorial siteDepends on
4.2 hrsContinuous
(or none)
59.6 hrs
Time between communications passesDepends6.9 hrs-71.8 hrs
Apparent angular size of Mars126.9 deg42.5 deg19.1 deg16.7 deg

A. Communications Access to Mars and to Earth

     If a primary function of a Martian moon mission is for astronauts to teleoperate robots on the surface of Mars then the differences in communication capability from the two moons to the surface are significant. Because of Deimos's higher orbit it moves more slowly than Phobos and an antenna on Deimos can communicate with assets over a larger swath of Mars. Assuming that a communications antenna on the Martian surface may have a 5 degree elevation mask due to terrain on the horizon, then astronauts on Phobos would have line of site communications to a rover up to 64.8 degrees latitude on Mars whereas from Deimos they could control assets up to 80.2 deg latitude. Phobos-based astronauts could directly communicate with most of Mars, but not the polar regions. For example, the landing sites for the Phoenix (68.3 N) and Mars Polar Lander (76 S) missions are only in line of sight from Deimos and not Phobos.

     Because Phobos moves so quickly it has short communications passes of 4 hours to sites on Mars (changing slightly with latitude) compared to more than 2.5 days duration from Deimos. However, the gaps between passess would also be much shorter. Phobos passes over a site on Mars every 11.1 hours, while opportunities from Deimos occur on a 131 hour cycle. The relative merits of short but frequent communications (Phobos) vs long communications passes with long gaps (Deimos) will depend on the concept of operations for surface assets. However, a given site at 30 deg latitude on Mars is in view from Deimos 45% of the time, but only 38% of the time from Phobos, giving Deimos a distinct advantage.

     Though speed of light latency is greater from Deimos than Phobos, it should not be a significant impediment to teleoperations from either moon. Two-way speed of light lag is 40 ms from Phobos and 134 ms from Deimos. On Earth, surgeons perform remote surgery with longer latency. The speed of light lag is short enough that hardware latency may be a larger contributor to total communications latency than the distance to Mars.

     Sites on Deimos also have more frequent direct line of sight communications to Earth than from Phobos, because as viewed from Deimos, Mars does not occult the Earth as frequently. From appropriate locations on Deimos it is possible to have many months of continuous Earth communications.

B. Lighting Conditions and Eclipses

     The moons of Mars have their polar axes aligned within a few degrees of Mars' polar axis, which is tilted 25° to the ecliptic. Phobos and Diemos therefore have distinct seasons and lighting conditions, which coincide with the Martian seasons. Like Earth and Mars, but unlike the Moon, they have a summer season in which the Sun is high in the sky and a winter season when it is low in the sky. In high latitude regions, the Sun can remain visible continuously during summer and may set for many days during winter, as on Earth. Northern hemisphere summer for Martian moons lasts significantly longer than southern hemisphere summer because Mars' orbit is eccentric and apohelion occurs during northern summer. Martian dust storm season generally occurs during southern summer when Mars is closer to the Sun, a scheduling issue relevant for missions which operate assets on the Martian surface.

     Because Phobos and Deimos orbit close Mars they also have eclipse seasons. Eclipses of the Sun by Mars occur repeatedly during the period when the line of intersection between the moon's orbit plane and the ecliptic points toward the sun. Since the orbit plane is roughly the Mars equatorial plane, eclipses occur around the time of Mars' vernal and autumnal equinoxes. Each eclipse season for Phobos lasts about 228 days, whereas the Deimos eclipse seasons are only 83 days long because Deimos orbits much farther from Mars. Maximum eclipse duration on Phobos is only 54 minutes, or 12% of the orbit period. Since eclipse occurs during the middle of the local day for the Mars-facing side of the moon, the combination of eclipse and night time can add up to a maximum of 62% darkness over the orbit period. On Deimos peak eclipse duration is longer, 84 minutes, but constitutes a smaller fraction of the orbit period, and there are many fewer eclipses. It will be difficult to operate a solar powered spacecraft on either moon during the equinoxes at peak of eclipse season, but during the rest of the year Deimos is better illuminated than Phobos.

     For purposes of this paper, we define summer to be the period after the vernal eclipse season ends and before autumnal eclipses begin, rather than the astronomical definition beginning at solistice and ending at equinox. Dates and durations for the summer seasons analyzed in this paper are given in Table 2.

Table 2. Dates and duration of summer sunlight season between eclipse seasons
Southern Hemisphere
Northern Hemisphere
Southern Hemisphere
Northern Hemisphere
11:09 AM
9:32 AM
9:54 AM
4:32 PM
6:27 PM
6:51 AM
1:46 PM
8:14 PM
Duration90 days144 days225 days297 days

D. Additional Considerations

2. Radiation environment

     Early in our investigation we expected that appropriate sites on Phobos might offer a reduced radiation environment compared to Deimos because Mars would fill more of the sky, blocking cosmic rays and solar particles. However, the differences are small. Landing on either moon provides shielding from half the sky due to the bulk of the moon, and perhaps more if the landing site is in a crater or other depression. But, Mars fills only 3.4% of the 4 pi steradian sky as seen from Phobos, vs 0.5% as seen from Deimos. So, differences due to proximity to Mars are likely to be smaller than differences due to local terrain. In either case, using a moon for radiation shielding is beneficial, and can reduce cosmic ray effective dose by on the order of 150-300 mSv compared to staying in high Mars orbit. (This negelects a small but unquantified increase due to albedo neutrons from the surface.)

III. Identifying Landing Sites on Deimos and Phobos

     In order to make a low-cost teleoperation mission feasible, we hoped to find locations on Deimos or Phobos where solar power is readily available and the surface of Mars is visible simultaneously. The following section describes the method and results of the search.

A. Analysis Methods

(ed note: they used a commercial software product: version 9.2.1 of Satellite ToolKit (STK) from Analytical Graphics Inc. Elevation data was from NASA Planetary Data System repository: Planetary Data System, Small Body Shape Models V2.1 by Peter Thomas, retrieved from For more details read the paper)

B. Deimos Results

     Figure 2 shows views of the southern hemisphere of Deimos. There are several regions on the Mars-facing and anti-Mars lobes which are sunlit up to 100% of the time during southern summer, shown in part c of the figure. These could be good landing sites for any solar powered spacecraft. For a mission which will control assets on Mars, the combination of solar power and Mars visibility is desirable. A small region on the Mars-facing lobe combines continuous sunlight and Mars access. It is highlighted in yellow in Figure 2d, and is located at 51° S, 5-10° E. A similar region exists in the northern hemisphere, as shown in Figure 3. It is centered at 60° N and extends several degrees on either side of the 0° longitude line. Coincidentally, this northern region is only a few hundred meters east of where the only high resolution images of Deimos were taken by Viking (such as image 423b62 and 423b63). The Viking images of this area show a smooth surface with muted craters which appear to have been filled in by a deep layer of regolith.

     The southern polar region of Deimos offers an interesting potential storage location for missions with an Earth return stage using cryogenic propellants. The south polar region is a depression between two large lobes which shadow the south pole. During southern summer the south pole receives sunlight during only parts of the day, and no sunlight at all during winter (see Figure 2c). It is also shielded from thermal energy emitted by Mars. The south polar region of Deimos may be one of the coldest places in the Martian system. Furthermore, the average surface gravity on Deimos is roughly 0.004 m/s2. This is similar to the low acceleration used to settle propellant in cryogenic propulsion stages such as Centaur today. Settling the propellant separates warm ullage gas from colder liquid, which simplifies thermal management, venting, and mass measurement. The south polar region of Deimos may be a good location to store a cryogenic return stage during the many months that the crew stays in the Martian neighborhood.

     We provide lighting maps for Deimos northern summer in Figure 4 and southern summer in Figure 5. They indicate the percentage of time that each location is sunlit, not the lighting at a particular time. Black regions recieve no sunlight, and the brightest yellow regions have continuous sunlight during the summer seasons.

     Figure 6 combines the illumination and Mars access data on a single map. In this figure, only the regions with 100% continous sunlight during the respective hemisphere‟s summer season are marked in yellow. The area inside the green boundary can see the entire face of Mars. In the area between the red and green boundaries, Mars would appear on the horizon and only part of its disk would be visible. The two previously identified regions which combine full sunlight and full Mars visbility are quite small. However, larger regions at higher latitudes have continuous sunlight and visibility to part of Mars.

Atomic Rockets notices

This week's featured addition is SPIN POLARIZATION FOR FUSION PROPULSION

This week's featured addition is INsTAR

This week's featured addition is NTR ALTERNATIVES TO LIQUID HYDROGEN

Atomic Rockets

Support Atomic Rockets

Support Atomic Rockets on Patreon