On Mars any initial colony will be power limited. Power sources can be brought from Earth, but any power generation capacity would reduce the mass available for other, more useful cargo.

Alternatively, you could produce solar panels locally. It feels like it should be doable locally. The primary materials required by mass are silicon, a frame and supports, and some form of conductor. Martian regolith includes all three. In this situation you wouldn't be attempting high efficiency cells, but a minimal viable product. As long as the marginal import mass of each watt was less than 100g, you're still beating the best fission power we're likely to send (kilopower - amazing NASA project).

Possible process: Wash reolith with water to remove perchlorates and water soluble salts. Reclaim the water via distillation.

Smelt the result with carbon monoxide, to remove both the iron oxides and impurities less reactive than Carbon.

Continue heating the remainder and use electrolysis on the melted liquid to produce Aluminium metal, liquid silicon and oxygen.

It sounds perfectly plausible, especially if you import the material for doping the silicon - you only need trace amounts.

So it sounds perfectly feasible. The materials are mostly available. Question is - given that all the above is true, what's the minimum practical mass required to set up a minimal product production facility?

Or would we be better off importing rolls of very thin film solar panel, mm thick, and producing frames and wires locally?

  • 2
    $\begingroup$ How 'initial' is the colony? First week? Second year? Are we allowed to send active robots and cargo and other capital goods years ahead of the colonists? Or is this a Farmer In The Sky hardscrabble scenario? $\endgroup$
    – user535733
    Sep 20 '19 at 17:12
  • $\begingroup$ Mars Direct style, first 2-4 manned landings. So perhaps 1 year of manned presence, might well have not been continuous. Certainly the first manned mission didn't stay! $\endgroup$ Sep 20 '19 at 17:27
  • 1
    $\begingroup$ "In this situation you wouldn't be attempting high efficiency cells" solar power on Mars is inconvenient enough as it is; you'll be wanting as efficient as possible. You don't want to be burning up all your power and time and effort just to make a teeny tiny bit more power. Leave the difficult fabrication to facilities elsewhere; so long as you have a decent logistics chain it will be the best option until your colony is very well established. $\endgroup$ Sep 20 '19 at 17:33
  • 2
    $\begingroup$ That's only true if the cells are durable enough to have EROEI greater than 1. At 5% efficiency, with the more limited insolation of Mars, a dust storm might cost more energy to clean and repair your solar field than the field generates in a week. $\endgroup$
    – jdunlop
    Sep 20 '19 at 18:31
  • 1
    $\begingroup$ Alright, looked up a few things, because I was curious. On earth, the estimated energy invested for photovoltaics is ~585kWh/m^2. The insolation of Mars (in the most ideal latitudes) is 590 W/m^2. At 5%, that would be ~29 W/m^2, or about 0.36 kWh/martian day. So 5% efficient solar cells would, conservatively, have a payback period of four and a half earth years before you break even, and they have to not require any repairs (incurring more energy costs) during that time, nor be occluded by martian weather. 5% is definitely marginal utility. $\endgroup$
    – jdunlop
    Sep 20 '19 at 18:38

There are many uncertain variables with your question. One being scale, how many solar cells are needed per year and at what efficiency? It is reasonable to presume that existing technology could be miniaturised sufficiently to enable small scale production of solar cells with fairly modest means.

The technology is mature and well understood. Processes such as the reduction of silicon dioxide, zone refining to remove impurities, cutting and shaping, surface preparation, doping and assembly could all be carried out on a small scale with some specialist furnaces, cutting equipment and automated handling equipment. Assuming a relatively pure source of silicon dioxide, at small enough scale and with sufficient development the process should be possible with less than 1 ton of equipment.

However large amounts of energy would be needed to reduce the silicon dioxide to silicon, to zone refine the impure silicon and then to cut and shape the ingots into wafers and assemble into panels. At very low volume the process would also be very inefficient.

Additional energy would also be required to provide a relatively pure silicon dioxide feed stock. Sites containing high purity silicon dioxide are unlikely to be a high priority for early missions so a suitable source would very likely be at some distance and even then would be unlikely to be as pure as needed leading to a number of purification steps involving a range of processes more energy and possibly consumables that would need to be recovered or replaced.

The exact nature of the refining processes required would depend on the crude input material and this is unknown in detail. Particle size distribution, the quantity and nature of impurities and their distribution (pure silica particles mixed with particles of impurity or silica particles containing impurities or more likely a mixture of the two). So which physical and or chemical processes would be best to use are not currently known and would require considerable process development post Mars landing.

In summary it would be possible to make solar cells on Mars given reasonably pure silica even in some of the early missions; however it would not be remotely practical to do so. The amount of crew time and electrical energy required would be out of all proportion to the end result. The best method for purification of silica on Mars is currently unknown as the nature of the raw material at or near the base site is unknown in sufficient detail. Realistically the prospects for practical solar cell production on Mars will not arrive until decades after the initial landings.


There is a problem with current solar panels: even here, on Earth they at most produce less energy for there lifespan (not in $$, but in kW*h) than it was total used to produce them (including extraction and transportation). Top solar panels are now at "1-to-1" zone or nearing it (if placed in California, not in Helsinki). But Mars gets times less solar energy!

It means that it never would be efficient to produce solar panels on Mars. Colonist can bring it from Earth but only as "conservated energy" for startup phase.

Wind generartion has more energy potential. It has about 2-to-1 ratio and require same technology as the rest of basebuilding.

Or, if you are for solar power, you can utilise mirros focusing and heat engine to produce power. It would easer to produce, setup and maintain.

  • $\begingroup$ If you could substantiate the numbers, as an answer "it's never efficient" would plausibly answer the question. (You're wrong about wind though, atmospheric density is less than 1%. It's not actually a vacuum, but...) $\endgroup$ Sep 24 '19 at 6:13
  • 1
    $\begingroup$ Yes, "windmils" would provide little power, but they would last long, easy to produce and (if designed right) would require little to nothing maintaince (unlike any solar power - dust is an issue). It means that net efficiency could be greater. $\endgroup$
    – ksbes
    Sep 24 '19 at 7:19

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .