Colonize Mercury

[moonguy]

Colonizing the planet Mercury will allow us greater access to the Solar System's diverse bodies. It will also allow us to conduct space-based astronomy on a more economical basis and, ultimately, provide a means of drastically reducing the costs associated with colonizing other planets. This can be done at a cost level at or below what is currently accepted (by NASA) for establishing bases on the Moon and Mars. Mercury, however, enjoys a number of important attributes that give it an advantage over other venues. Mercury has the promise of becoming a highly productive venue for science and exploration.

 

There are major challenges. Mercury is difficult to reach. The environment there is not people-friendly. Radiation is a particular concern. Mercury does not enjoy the 'Earthlike' imge of Mars and therefore had relatively little attention by comparison. We still have a lot to learn about the planet.

 

I am hoping for a friendly dialog to expand on these points. As a helpful hint, a lot of time and copy space will be saved if postings ask specific questions. There are a lot oif surprizes ahead. . . !

[moonguy]

Space-based astronomy is about to turn an important corner. The recent problems with the Kepler observatory point up the major flaw with orbital telescopes: unless you are willing to spend a billion dollars to send people to fix inoperable parts of an otherwise functional telescope, you have to get used to losing them for things like bad gyros. OR. . . You can put them someplace where they can be maintained more cheaply. If they happened to be on a solid surface , they would be only the observing instruments with no complex propulsion or orientation equipment needed. If the location could provide construction materials for the instrument's mountings, the mass to be launched from Earth for a given instrument would be minimal - much less than the mass of an orbital version of the same instrument. Ideally, the location would have a slow rotation, be airless, be reasonably accessible and provide ample energy and area for operating truly large arrays of instruments if desired. Such a place would be employed by astronomers for decades. In this Solar System, there are only two reasonable choices for such a location. The Moon and Mercury. Of these, Mercury is the better of the two.

 

As an astronomical platform, Mercury can provide 88 straight days of full-sky observing, such as for infrared mapping projects. For projects involving observation of specific targets, a given target would be above Mercury's nigh horizon for 44 days. This is three times longer than what the Moon offers. The Moon's far side only allows 'full-sky' observing for two weeks as it is facing the Sun at two week intervals. Mercury, by contrast, rotates such that the entire sky is viewed, in the dark, during the course of one Mercury night of 88 days - the length of it's year. A manned observatory on Mercury could host instruments covering every part of the spectrum. Large units could be built from resources available on Mercury. Technicians stationed at the observatory could repair or upgrade instruments quickly and at low cost compared to those in Earth orbit. What would this mean to something like the $8 Billion James Webb Space Telescope?

[SaxonViolence]

Well, assuming that we've advanced enough to travel to Mercury and set up an observatory...

 

By then, we'll have even better sensors than we have today—presumably.

 

God alone knows what we might learn in a year's time, focusing our new (20XX whatever the date will be) Sensors on the Sun 24/7...

 

Probably enough raw data to keep Theoretical Physicists busy for a decade or more...

 

Assuming that we haven't already deduced most of it by then.

 

Yeah, sign me up—but we can't even get a rotating space platform or a Moon Colony built...

 

And truthfully, we could start work on them tomorrow.

 

I don't think we have the wherewithal to put a base on Mercury right now.

 

 

Saxon Violence

[moonguy]

Well, assuming that we've advanced enough to travel to Mercury and set up an observatory...

 

By then, we'll have even better sensors than we have today—presumably.

 

God alone knows what we might learn in a year's time, focusing our new (20XX whatever the date will be) Sensors on the Sun 24/7...

 

Probably enough raw data to keep Theoretical Physicists busy for a decade or more...

 

Assuming that we haven't already deduced most of it by then.

 

Yeah, sign me up—but we can't even get a rotating space platform or a Moon Colony built...

 

And truthfully, we could start work on them tomorrow.

 

I don't think we have the wherewithal to put a base on Mercury right now.

 

 

Saxon Violence

Thanks for your interest!!

Why do people assume we need some new technology (Nuclear Thermal? VASIMIR?) to get to Mercury? The J-2X engine being developed as we speak has the exhaust velocity needed for a manned mission to Mercury. Hohmann transfers to Mercury are less than half as long as flights to Mars, so the payloads are significantly lighter. What matters is mission cost. You end up spending about the same for either a Lunar Base, Mars or Mercury because propellant, masses required are actually about the same. Of course this assumes we are wise enough to design a crew module/payload for Mercury missions and not try to adapt a Mars mission module for it.

Of course, there is a lot more to tell. . .

[moonguy]

Settlements on Mercury are better placed to support - at reasonable costs - aggressive programs of exploration to all of the bodies in the Solar System. From Mercury, it is possible to launch probes to Venus twice a year. Launches to Venus from are possible only every 1.6 years. Mercury could send three or four probes to Venus for every one Earth could launch. For Main Belt asteroids, the ratio is three or four per year to any given asteroid. Jupiter and all other outer planets have four launch opportunities per year from Mercury. Using solar sails, Mercury could launch a program of exploration that would dwarf what has been accomplished so far from Earth.

 

More frequent launch opportunities means a given probe can be more specialized as the capabilities of one large probe can be divided amongst a series of probes, all launched within a few months of each other. If launched from Mercury, the smaller probes require much smaller rocket vehicles for launch form Mercury's surface and no rocket after that at all of they use solar sails to get to the destination planet. If that planet is airless, they would only need a small lander stage, because all of the planets beyond Venus that have solid surfaces - with the one exception of Mars - have gravities less than the Moon's in strength. This would be important for places like Jupiter and Saturn where each have several, large satellites.

 

Individual probes could be partially built on Mercury by having the more high-tech instrument packages sent form Earth and installed on structural frames made on Mercury. This reduces Earth-launch weight as it enables a smaller rocket launched from Earth (with several instrument packages) to support several different missions. Normally, that smaller rocket (Falcon 9? Atlas V?) would only support one mission.

[CraigD]

I’m far from convinced that landing humans on, let alone colonizing, Mercury, could be funded by the US’s NASA or a private US-based company, or even by a cash-rich, prestige-seeking national space program like China’s. I doubt that anybody, even teams of the most experienced and visionary administrators, accountants, and entrepreneurs, can accurately calculate to within factor of 4 the cost of any major construction project, because project cost overruns and other mishaps are common in such enterprises.

 

But these are business questions, and by the power vested in me by my administrator account and paying the hosting bills, I decree that Hypography’s Space forum is more for space science and engineering than business! ;)

 

So let’s get to the engineering!

 

The main problem I (and as best I can research, most others – see this Wikipedia section for a summary) can see with landing anything on Mercury is shedding velocity. Unlike Earth and Mars, which have atmospheres that can be used for aerodynamic braking and controlled descents, Mercury, like Luna, is for practical purposes surrounded by vacuum, so transferring from orbit to landing on either – or on any large, airless body – requires a large change in velocity. Assuming a circular orbit of 110 km altitude (what the Apollo moon missions used), this final delta-V is abou 1679 m/s for Luna, about 3005 for Mercury.

 

Getting from Earth (where all the humans are now) to Mercury is much more difficult than getting to Luna, but for now let’s consider this problem somehow solved, and focus on the “last step” problem.

 

Breaking the mechanics down, the 110 km orbit to landing on Mercury involves a small delta-V (about 5 m/s) to change orbit to an ellipse that reaches its surface, and a large one (about 3010) to change from orbiting speed to the practically 0 speed of the surface. In addition, since the final change can’t be instantaneous, thrust must be spent to prevent hitting the surface before matching speed with it. In the case of the Apollo, including a margin of safety to allow finding a safe landing site, the landers had 1.4 times the delta-V needed for an ideal, instantaneous change maneuver, a reasonable rule-of-experience factor for estimations. This scheme is why nearly 60% of the mass of an Apollo lunar lander was propellant used for descent, while the “Earth lander” part of an Apollo used practically no propellant. The mechanics are worse (by about 40%) for Mercury, and can’t be solved with low-thrust systems like Solar Sails or high-specific impulse, low-thrust rocket motors, because the system must have at least enough thrust to equal the surface gravity acceleration of 3.7 m/s.

 

As best I can imagine it, the only way to avoid all this lander rocketry and propellant is to use something akin to the arrester system on an aircraft carrier, except rather than an about 100 m landing runway to stop a craft moving about 67 m/s, the speed is 3010 m/s, so the runway must be much longer. The math’s pretty simple,

[math]d = \frac{v^2}{2a}[/math]

 

Using the acceleration a = 25 m/s/s of an aircraft arrestor system, gives a length of about 181 km. Properly couched and packed humans and cargo can safely endure 2 to 4 times, this, though, so the runway can be shortened to something between 90 and 45 km.

 

As a bonus, such a system could be designed to both remove or add speed, so could also be a nearly (the craft would still need that little 5 m/s delta-V on top to circularize its orbit rather than looping back to ground level) propellant-less launch system.

 

So the first construction project I’d propose for a Mercury base (or a Luna base, though given the much lower delta-V and cost of delivering propellant to lunar orbit, such the need for it is less critical there) would be this 45 to 90 km “space runway”.

 

For the actual mechanics of the system, it’s intuitively obvious that scaling up the steel cables and big hydraulic pistons and cylinders of an aircraft carrier arrestor isn’t an option, nor is anything that involves direct contact of mechanical parts, a conventional runway surface, or even something unconventional like a giant trough full or bouncy balls. Magnetic levitation and propulsion seems the only option to me. Getting much deeper in details without a lot more detailed engineering estimation is premature, but my guess favors a system where the levitation force component of the system is passively levitated-body induced, essentially the Inductrack system. That NASA and related organizations have been playing with this system for launching spaceplanes from Earth is a plus.

 

A key high-level design decision is whether to put the “carriage” part of the system – analogous to the landing gear on an airplane, or the car on a maglev transport system – on the spacecraft, or keep it on near the surface, matching velocity and attaching to the spacecraft at apoapsis. If the latter, spacecraft mass is reduced, and the carriage can be more robust, but additional runway length it needed to accelerate it to around 3010 m/s before meeting the landing spacecraft, and a point-of-failure involving this precise and timely rendezvous is introduced.

 

Another family of key design decision is how much of this, or any other system, to manufacture “onsite” on Mercury vs. offsite and delivered, and how to “bootstrap” onsite mining and manufacturing. Mercury is believed (no direct analysis of its surface composition has been made) to have about the same surface composition as Luna, so mining for materials that can be used to make things to mine and make more is no easier than on Luna, but “seed” machines and materials are a lot more difficult to deliver to Mercury than to Luna. It’s worth noting that it took US$450,000,000 and some very clever orbital maneuvering just to put the 485 kg MESSENGER into a 200x15000 km altitude Mercury orbit using conventional rockets – about the same as the $480,000,000 to land the 350 kg Phoenix on Mars.

 

That nice little 5 m/s initial delta-V I calculated for transfer from 110 km circular Mercury orbit to its surface is a big 548 m/s for MESSANGER’s highly elliptical orbit, and can’t be avoided with any landing system trickery.

 

Landing even a few 100 kg of anything on Mercury with current technology looks terribly expensive. This strongly persuades me that, to put stuff softly on Mercury

• solar sailcraft will be needed for the Earth-to-circular Mercury orbit step
  
• a landing/launch system like I describe in this post will be needed for the circular orbit-to-surface step
  

It’s tempting to discard the landing/launch system, because it’s so big, and because, once the technology is successful, solar sails could be scaled up to carry lots of propellant for the orbit-to-surface step, but this begs the question of where will all this propellant come from? Could a few very expensive landings using a few 10,000s of kg of propellant from Earth deliver a factory that could make propellant and fuel that could be used to launch propellant from surface to Mercury orbit?

 

Unlike proposed missions like Mars Direct, a factory on airless Mercury can’t make Methane and Oxygen from CO₂. There appears to be water ice, and some organic molecules, in deep polar shadow on Mercury – MESSENGER confirmed this old, Earth-base radar observatory-raised suspicion late last year – but how much, and how easily exploitable, I believe remains very uncertain.

[moonguy]

While I applaud your acumen in orbital mechanics, I found most of this posting unnecessary. You took great pains to elaborate on the 'difficulty' of the easiest part of the mission! Boosting hundred-ton spacecraft away from Earth so they can project themselves into an orbit over Mercury. . . THAT is the difficult part of the mission!! Making the ~3.1 km/sec. descent for a vehicle with payload well below five tons is easy by comparison. For an engine like the RL10-A-3 series,(Isp 465 sec.) using LO2/LH2 at 5.88:1, you need a mass ratio of 2 for such a mission. So what is this bizarre notion about aircraft carrier arresting gear?

 

As for the 'business' issues, what is space transport technology all about if not cost reduction? That is a discussion I hope we have when I've completes a few more posts. . .

[moonguy]

Mars is the big plum in NASA's manned space program. Unlike other planets, when the discussion shifts to Mars, it is not about 'exploration'. It is about 'colonization'. Quite a few environments on Earth are unoccupied even though they are easier and less costly to reach and survive in. No matter. NASA wants to go to Mars, so we are going to Mars. How can a settlement on Mercury help such an effort?

 

Skyscrapers, football stadiums, aircraft carriers. . . They all are designed to support the habitation of hundreds or thousands of people. If only on a temporary basis. Colonies on other planets are likewise designed to support large numbers of people - but on a permanent basis. A Mars colony could easily be compared to a skyscraper - only one with very large gardens, some factories and a rocket propellant plant nearby. Think Chicago's Sears Tower and O'Hare airport combined. The point here is that for any group of people put anywhere permanently, it is absolutely certain there will be more mass required by a factor of ten or more then the person themselves weighs.

 

Right now, the largest rocket NASA has plans for could put something close to 25 tons of actual payload on Mars' surface. It would be at a cost of about $60,000.00/kg. NASA thinks this is a reasonable proposition. Ok. But there is a problem.

 

That largest rocket can be launched two at a time. At best - if we care about safety - we could launch two every 20 days and build an entire fleet of Mars-bound payloads - and their stages - in earth orbit. If we except the Earth-Mars synodic period of ~780 days as a working figure, divide by 2 launches every 20 days, we ultimate have 78 payloads ready to go when the launch window opens!!! At a combined cost of $117 BILLION dollars. It gets worse. If you send people, you have to subtract those flights because they only bring people, not construction material. Those are just launch costs, by the way. They do not include the operation of a station(s) in orbit that can sustain that rate of deployment. You need those stations because the departure stages have to wait ('loiter') up to 26 months to fly and they are only designed to loiter for 3, maybe 4, months.

 

Take a breath. Pour another cup of coffee. I'm only just getting started. . .

 

With just one exception, no in-space transportation technology known to man will help the above scenario because the problem is not the Earth-Mars transfer. It's getting the construction material off Earth in the first place. Getting the construction materials from the Moon would help, somewhat. If you built the rockets from lunar materials and fuelled them from lunar water would not help because you have to build the base first creating yet more need for Earth-launched materials.

 

Frustrating, but how does Mercury make any difference? Mercury can do the same things as the Moon to provide construction materials. It's just there are some huge differences. . .

 

Mercury has no launch window bottleneck. With three launch opportunities per year to Mars, Mercury would need to produce and deliver only one-seventh the mass at each opportunity to keep pace with Earth capability.

 

Since solar sails would be the transporter-of-choice from Mercury, the demand for chemical propellants there would be very modest compared to any Earth/Moon launch scenario. Carefully note I have not specified either a payload mass or sail size. Depending on exact costs for NASA's SLS, it may prove cost-effective to use smaller launchers at Earth to deploy payloads to Mercury with solar sails.

 

Mercury's 20-fold THERMAL energy advantage over Mars makes utilizing poor-grade regolith materials a practical proposition without resort to (billion-dollar) nuclear power systems. If we did not know how to safely utilize heat like that, we would not have steel mills.

 

Earth could use solar sails very effectively to send payloads to Mars. This would not resolve the bottleneck problem. The only way Earth can compensate for that is to somehow put seven times the payload mass onto a sail that will be inherently less efficient than its Mercury-launched counterpart because it is starting flight from 1 AU. Or deploy seven times as many sails to carry reduced payloads. Or deploy MUCH larger sails. Either approach results in greater cost-per-kilogram delivered than the Mercury alternative.

 

Overall, construction material for a colony anywhere will make up 90% or more of the mass to be shipped. Air and water may be the greater share of the final mass, but they are both available on both Mars and Mercury.

 

Mercury is better able to produce and ship the 10,000 tons of a Mars colony's deadweight mass. Mercury can do this without relying on SLS elements indefinitely. Mars really does not have that option.

Edited September 23, 2013 by moonguy

[moonguy]

How would crews get to Mercury? Is there any way we can beat the high delta-V penalties for classic style ballistic flights? Yes, there is: Cyclers.

Mercury’s orbital period is 87.97 days. Earth’s is 365.25 days, or 4.15 Mercury ‘years’. A cycler deployed to a 351.5 day orbit, with Earth’s orbit as the aphelion and Mercury’s orbit as the perihelion, will encounter Mercury every time it (the cycler) reaches perihelion.

Crews departing Earth would still have to generate a high delta-V. In fact, cycler missions would require a delta-V around 9.5 km/sec. while some standard (‘Hohmann’) transfers can be done for 6-7 km/sec. The difference is that a cycler mission puts most of th payload mass required for the ~176 day transfer time on the cycler. This drastically lowers the payload mass injected into the transfer. In a classic Hohmann transfer scheme, a manned payload would be injected with an upper stage able to effect the Mercury rendezvous and orbit insertion. For a 10 ton payload and a stage using a J2-X engine, a mass ratio of at least 4.1 would be needed for even the most favorable MOI delta-V, which is about 6.3 km/sec. The resulting stage would be about 60 tons. Pushing this into a transfer orbit from Earth would require a stage which also has a 4.0 mass ratio – and hence masses over 100 tons by itself.

The cycler reduces the requirement to launching the crew in their Earth-return capsule. This could be an Orion, A DragonRider derivative or something similar. If it had the same 10-ton mass as the first example, it would not need to be boosted with a second stage. Instead, propellant for the maneuvers at Mercury would be derived from water stored on the cycler. Food and living accommodations would also be on the cycler as well. A 15 ton cycler could easily store enough food and consumables for several mission cycles. These supplies would be replenished to the cycler periodically by the same solar sail that deploys the cycler to the 351 orbit. The water used for propellant would also be supplied using solar sails. Initially this would be from Earth, but it would eventually come from Mercury.

With an orbit of 351 days, the cycler would encounter Earth every third orbit of the cycler. This is due to the synodic period of Earth and Mercury is 115.9 days. Multiplied by three yields 347.7 days. There is just over a 4 days discrepancy between an exact encounter., However, launch windows to Mercury are open for about 20 days, so it could be assumed a delta-V penalty would be incurred to make up for the 4-day difference.

The cycler mission requires much less propellant be available in Earth orbit. Only one stage is used. This is refueled at the cycler for the MOI burns and again for the return to the cycler for the return trip. If the Earth-entry interface velocity is too great for the return module’s heat shield, the crewed stage could refuel a third time to execute a burn into Earth orbit. This architecture enables a crew to launch to Mercury every 347 days. Twice as frequently as flights to Mars.

[JMJones0424]

Unless I'm completely misunderstanding you, the delta v requirement for an Earth bound vessel to reach and dock with the Earth-Mercury cycler is about the same as the delta v required to reach Mercury from Earth in the same orbit as the cycler. There's no such thing as a free lunch.

 

To say it another way, if you have your cycler with an orbit that is tuned to meet Earth at aphelion and Mercury at perihilion, if you send a vessel to dock with the cycler, it must have at least the delta v required to match that orbit. You aren't gaining anything in the process.

Edited November 17, 2013 by JMJones0424

[moonguy]

You are correct. The delta-V for the crew vehicle must be the same for the Cycler. The difference, compared to classic concepts, is that there is only the crew module to be boosted as payload. In traditional approaches, the payload boosted away from Earth is the crew module AND a stage for insertion into orbit over Mercury. In the Cycler concept, the 'free lunch' is not free, it is just paid up in the form of solar sail deliveries of propellant to the Cycler prior to the crew's departure from Earth. These require no propellant at all for Earth departure or rendezvous with the Cycler. A square sail 820 meters to a side weighing just 4000 kilograms can deliver 10-ton cargo masses (propellant, supplies etc.) to a Cycler in flights of about a year from Earth. A typical insertion to Mercury orbit might require 50 tons of propellant. This allows a fully supplied Cycler to support several crews over about five years of missions.

[tkron31]

As far as I'm concerned, if you can solve all of the obvious problems involved with colonizing Mercury (i.e. no atmosphere, EXTREMELY hot on the sun side and paradoxically cold on the night side, and the intense radiation that comes with being so close to the sun), you'll have solved all the problems that the news rags talk about when it comes to colonizing other planets. Otherwise, why not just mine it for the metals?

[CraigD]

Welcome to hypography, tkron31! :) You seem a spaceflight enthusiast – I hope we can have some good conversations on the subject.

 

... the obvious problems involved with colonizing Mercury (i.e. no atmosphere, EXTREMELY hot on the sun side and paradoxically cold on the night side ...

It’s important to note that Mercury isn’t tidally locked (1:1 rotation:orbit resonance) to the Sun, so doesn’t have a fixed day and night side. It has a 3:2 rotation:orbit resonance, resulting in each point on its equator experiencing exactly 1 2111.28 hr-long Mercury year of day, followed by the same length night.

 

Due to Mercury’s great orbital eccentricity (aphelion about 52% greater than perihelion) the rate of the sun’s progress across its sky varies extremely during each day, fastest at sunrise and sunset, and actually stopping and moving slightly in the opposite direction around mid-day.

 

This 3:2 resonance was one of the big astronomy surprises of the 1960s, when radar observations revealed it

 

... and the intense radiation that comes with being so close to the sun) ...

Mercury is, obviously, closer to the Sun than earth, so its daylight is stronger than on Earth (from [math]\frac1{0.4667^2} \dot= 4.6[/math] at aphelion to [math]\frac1{0.3871^2} \dot= 6.7[/math] times), but like earth, it has sufficiently strong magnetic field to deflect “hard” solar wind radiation. It’s got no atmosphere to speak of, so no protection from UV and other electromagnetic (light) radiation absorbed by one.

 

Night on Mercury, though, is about as dark and low-radiation as on Earth. At night, the surface of Mercury would arguable be about as human-friendly as the Moon’s.

 

As far as I'm concerned, if you can solve all of the obvious problems involved with colonizing Mercury ... you'll have solved all the problems that the news rags talk about when it comes to colonizing other planets.

I hope the info I’ve given above illuminates how unique Mercury is in the Solar System. While many of the engineering solutions for allowing people to live there, especially spaceflight systems, would be more than needed for other planets (Earth-to-Mercury transfer orbit delta-V is 7533 + 9612 m/s, making it several time more momentum costly than Earth-to-Mars (2945 + 2649) and even slightly more than Earth-to- Jupiter (8794 + 5643)), many others would be unique to Mercury.

[Murga]

As a long time Mercury watcher, I can add some points to the discussion.

Yes, a 180 day or so effective (not sidereal, IIRC) 'day'. The best location for any astronomical base, which the threadmaker desires, is of course at either pole. The seasons on Mercury for axis tilt are the least of any known large body in the solar system, I recall. The low tilt, long 'day' and low temperatures (at the poles) makes astronomy pointing and other issues easy compared to elsewhere. However, there is the odd apparent retrograde motion of celestial bodies with all locations at the surface. It has to do with the resonance orbit and eccentricity I think. Moon Guy might have a better way to put it.

Surprisingly, the temperature of the regolith is quite high at the equator (180 C?) even in the last part of night where 90 days of darkness would be expected to act as a heat sink. The soil has an inch or so of very, very fluffy material which acts as a great insulator, since there is no atmosphere to speak of to do that job. Think of super down feathers. Obviously, this is no place for an astronomical observatory if at all possible.  The heat contractions could soon bend the apparatus to weird contortions, and man power would be at a premium far beyond that of LEO. 

<p>-- To be continued, getting sleepy --</p>

[Murga]

 

Due to Mercury’s great orbital eccentricity (aphelion about 52% greater than perihelion) the rate of the sun’s progress across its sky varies extremely during each day, fastest at sunrise and sunset, and actually stopping and moving slightly in the opposite direction around mid-day.

  That is what I meant of the eccentricity.  Yes, somewhat difficult for an astronomical device, but compensated by being very slow motion. 

[moonguy]

This presents an interesting question: How much propellant is needed to get a crewed payload to Mercury if it is one-third the mass of a crewed payload used for a Mars missions?  As the post notes, delta-V to Mercury is roughly three times that for Mars.  A crewed payload, however, would be much lighter, for several reasons: 1) Typical ballistic flights to Mercury are about 43% as long as those to Mars.  This reduces food/consumable needs accordingly.  2) The reduction in consumable mass also means a reduction in vehicle volume - hence structural mass - for the crew module while maintaining the same volume ratio per crew member.  3) Flights to Mercury are always closer to the Sun than 1 AU, requiring less mass for a (assumed) photovoltaic system compared to a Mars vehicle.  4) Crew size matters.  We are used to thinking of Mars missions with four or more crewmembers.  Credible Mercury missions can be flown with crews of three, though, of course this would not be a straight up, al-things-equal comparison.  This reduces consumables even further for a total mission payload mass that reaches about one-third the Mars mission payload value.  This issue here is not whether Mercury is 'less expensive' than Mars.  It probably is not.  The issue is whether Mercury can be reached for costs less than the three-fold (compared with Mars) delta-v figures would seem to indicate.        

[TomKalbfus]

You see the latest News from Mercury, more water than we thought.

#https://www.space.com/38274-mercury-has-surprisingly-icy-north-pole.html

Maybe Mercury can sustain a human presence indefinitely, the hard part is getting humans there. Once buried under the north pole, they should be well protected from whatever the Sun puts out. I am thinking about a linear mass driver built at the North Pole. the planet rotates slowly, so when the mass driver is pointed in the right direction, you can fling the payload into an orbit that will intercept Earth. Since Mercury is particularly dense for its size, it may contain a higher abundance of precious metals such as platinum than Earth, and there are less environmental restrictions against strip mining on Mercury. We might want to send something to Earth to pay for the colony, and ingot of platinum ought to survive atmospheric entry into Earth quite handily, as the metal has a high melting point.