Powering martian civilisation from ebay.

[speedevil]

Bottom line up front - There is no need at all to develop solar power systems for Mars, you can get entirely suitable ones nearly off the shelf. You can return a BFS to earth for $5M/4 tons earth launch cost, and martian power is comparable with earth costs.

Assuming lightly modified commercial products from vendors that are overjoyed by getting a $10M order.
Pricing parts from one off quantities online, rather than assuming bulk discounts, and neglecting interest.

What amount of infrastructure can we get in one BFS - 150 tons?

What does this imply as a $/kWh.
What does it imply as a $/capacity to refuel a BFS each year.
What does it imply as $/capacity to grow astronauts food using solar-PV.

Assuming IAC2016-like costings of $130/kg (this implies reuse of BFS).

Average insolation at the top of the atmosphere is around on average 185W/m^2 (counting nights), so 100W/m^2 counting weather, and the lack of a pivioting system is plausible.

$1/W 50W/kg flexible solar panels.

$250+500 launch/4kg/20W, or $37/W,  on the surface of mars (average). (Assuming efficiency remains at 20%)

100 tons of these provides 0.5MW (av).
Adding 50 tons of tesla model 3 batteries gives you 7.5MWh of storage, or conveniently sized for overnight use at the same power level.
This works out to  $0.3/kWh, assuming 20 years system life.


What sort of power is needed for various tasks?

Returning BFS to earth:
Recent talk gave 50% as their assumed efficiency for propellant production. In order to get back to earth, (150 tons * 44MJ/kg)/2 years = 200kW.

So, as a first cut, one BFS can supply power for refuelling approximately two BFS per synod, or over 20 years, 40.

Or enough power to supply a minimal diet for 150 people. (3000W, for potatos).

Martian power price can be reasonably assumed as several cents a kilowatt, available 24h.
Mass payback for grown potatos over shipped cheese is four years. (considering only power)
Mass payback for returned BFS is about one year.

The last point is especially interesting, because it implies that the cost of returning a cargo BFS to earth is nearly zero, in the steady state. (naively $5M, or 4 tons cargo from earth).

(I have intentionally neglected cost and mass of fuel production and raw materials collection).

[aero]

Assuming lightly modified commercial products from vendors that are overjoyed by getting a $10M order.
Pricing parts from one off quantities online, rather than assuming bulk discounts, and neglecting interest.

Do you think there is a chance that SpaceX wouldn't go with equipment from Tesla Solar? Tesla Solar is one of Elon's companies after all. It is the new name for Solar City.

[speedevil]

Assuming lightly modified commercial products from vendors that are overjoyed by getting a $10M order.
Pricing parts from one off quantities online, rather than assuming bulk discounts, and neglecting interest.

Do you think there is a chance that SpaceX wouldn't go with equipment from Tesla Solar? Tesla Solar is one of Elon's companies after all. It is the new name for Solar City.

I do not have ready prices for them, or masses.

It is clearly possible to optimise, but in many ways this is uninteresting, given the $0.3/kWh implied price. Getting to $0.1/kWh, while it might be possible changing vendors doesn't change much.
The vendor in question was only picked as I'd bought panels from them and noticed that the flexible ones were quite light.

[Robotbeat]

You should know that the kind of flexible/semi-flexible cells you typically buy on ebay are super heavy. If you're actually launching it to space, you can get much lighter ones.

[speedevil]

You should know that the kind of flexible/semi-flexible cells you typically buy on ebay are super heavy. If you're actually launching it to space, you can get much lighter ones.

Aerospace cells are several orders of magnitude more expensive, and purchase cost is already half of the price of launch.


For a minimal total cost mission, you don't want to go much more expensive than launch cost.
(Initial missions where you are costing on the basis of not getting the BFS back may differ, and $40/W at half the weight might be worth it)

[Robotbeat]

They don't have to be triple junction Galium-arsenide ones. Just thin film on the right substrate. Or just silicon.

Typical cells you can buy online are about 11 grams for a 4.5 Watt silicon cell, or about 400 Watts per kilogram at the cell level. But that's for 200 micron thick cells. You can get thinner cells, like 50 micron or so. That's about 1600 Watts per kilogram, although 100 micron thick cells are a lot more common, or about 800 Watts per kilogram still using silicon. Still very cheap to buy. The best I've found on ebay so far is about 145microns thick, and 22% efficient, or about 640W/kg (at Earth) at the cell level (and still much less than $1/Watt even if bought like individually). https://www.ebay.com/itm/SUNPOWER-FLEXIBLE-Solar-Cells-DIY-Solar-Panels-5-x5-3-34-Watts-ea-AWESOME-/181890493881?hash=item2a5984adb9

SpaceX already does cell integration themselves for Dragon. They buy cheap silicon cells from like Sunpower or something (I don't remember), then integrate them into panels. So they could add their own thin encapsulation for those cells, probably no worse than 300W/kg at Earth, or about 150W/kg at mars (not counting night, etc). Much better than the ebay flexible panels.

[Robotbeat]

By the way, the spot price for silicon cells is like 16 cents per Watt. Mono cells are like 20 cents per watt. And the thinner cells may actually be cheaper, as there's less material required. But it may be hard for individuals to acquire thinned cells (made with epitaxial liftoff). SpaceX should have no problem, though.

Thinned cells might be something you always make on Earth even after Mars has an industrial base. You can make like 1 micron thick solar cells and ship them to Mars where you'll assemble them, attach them to substrates, and encapsulate them. That'd be equivalent to like 80000 Watts per kilogram, and a shipping cost of less than 0.2 cents per Watt (i.e. about 1% of the costs of the cells today, though maybe a larger portion in the future when cells are cheaper). 12 Gigawatts of cells per BFS landing...

[Ludus]

By the way, the spot price for silicon cells is like 16 cents per Watt. Mono cells are like 20 cents per watt. And the thinner cells may actually be cheaper, as there's less material required. But it may be hard for individuals to acquire thinned cells (made with epitaxial liftoff). SpaceX should have no problem, though.

Thinned cells might be something you always make on Earth even after Mars has an industrial base. You can make like 1 micron thick solar cells and ship them to Mars where you'll assemble them, attach them to substrates, and encapsulate them. That'd be equivalent to like 80000 Watts per kilogram, and a shipping cost of less than 0.2 cents per Watt (i.e. about 1% of the costs of the cells today, though maybe a larger portion in the future when cells are cheaper). 12 Gigawatts of cells per BFS landing...

What would be needed on Mars to do that? How many BFS loads of equipment? What local resources?

[Robotbeat]

The sorts of things you already want to have for making buildings and things. Glass, epoxy, wire.

[speedevil]

What would be needed on Mars to do that? How many BFS loads of equipment? What local resources?

Quite - part of the reason for this thread is a lot of the 'better' alternatives are somewhat hard to quantify and need at the very least extensive preparation to get working, never mind external industries.
If your solar panels are going to require external coverglass or stiffening, beyond a simple wire frame, that's not a freestanding easily understood solution, and working out the whole cost to ship it to Mars is hard.

I don't disagree that they are better, but at some point 'It works, ship it' is a reasonable solution.

At the least it's useful for ballparking to know that you can get from one BFS enough power 'simply' to refuel a couple of BFS a transit, with no clever engineering at all, no development, and no significant investment.

It's something that you might further optimise now, not a constraint on development.

[Robotbeat]

It’s worth noting that SpaceX has been working on this problem for years (there has been discussion by Musk of inflatable solar arrays from years past for Mars surface power) and have experience with Dragon and Starlink in integrating COTS silicon solar cells into operating space capable arrays. It may be easier for us to imagine things if they’re on eBay, but SpaceX has no problem doing the legitimate engineering on these.

[AC in NC]

This is a great topic and right in speedevil's wheelhouse for informing a subject area through easily quantifiable data. 

I encourage all to engage and enjoy it in the spirit offered rather than pressing the banal variations on the retort that "SpaceX isn't going to power Mars off EBay". 

[DusanC]

IMHO first few BFS that go to Mars for ISRU will be used as fuel storage but surface on their backs would be made of thin flexible solar cells, just like Dragon2. No need for complicated robotic unloading, setup, support structure etc.
BFS would be solar powerplant on Mars, they just have to land in right orientation with their back toward the Sun.

[speedevil]

IMHO first few BFS that go to Mars for ISRU will be used as fuel storage but surface on their backs would be made of thin flexible solar cells, just like Dragon2. No need for complicated robotic unloading, setup, support structure etc.
BFS would be solar powerplant on Mars, they just have to land in right orientation with their back toward the Sun.

This gives around 8m*48m of solar panel area.
Taking the above figure for 20W/m^2 average input equals 10kW.
This is a useful amount of power, and may be enough for 'hotel' loads and light charging of small rovers, but is clearly totally inadequate for ISRU - other than testing on a very small scale.

It also means that the solar cells have to take the radiated heat of reentry and remain working, as well as being able to re-land on earth without damage to the vehicle if an abort happens.

[biosehnsucht]

Assuming the first landed BFS are chompers, and the chomper doors can open and stay open under Martian gravity without any problems...

How about some big folded arrays deployed from the chomper? You could position an array on either side of the chomper door, and still use the middle area (most of the volume) for cargo or inflatable tanks or whatever.

Gets around the heat load issue, and is still a relatively simple technique to get some limited power going (versus rover deployed solar fields). You could still ship rover deployed solar fields on these first flights, but they can be deployed later, semi-autonomously or with local remote control or whatever, but the pop out chomper arrays could give you power practically immediately using relatively simple (as in, relatively known mechanisms and such) systems.

The question then is, based on the size of the volume given from the chomper bay, and the mass of the panels per volume versus necessary structure, how big can you go? Simplest solution is probably a roll out array of some kind versus fold out, but you still need at that size some reasonably robust structure to carry the load, even in Martian gravity.

[Lar]

I think there is a lot of merit in some sort of instant on power that powers the BFS and systems (and maybe can do rover (no, utility vehicle) recharge) that means you're not on just batteries until you get the main load out of more cells and frameworks and etc out and operational.. So if putting cells on the OML, or a popout from the cargo bay works, it is likely that SpaceX will do it.

[Bob Shaw]

I think there is a lot of merit in some sort of instant on power that powers the BFS and systems (and maybe can do rover (no, utility vehicle) recharge) that means you're not on just batteries until you get the main load out of more cells and frameworks and etc out and operational.. So if putting cells on the OML, or a popout from the cargo bay works, it is likely that SpaceX will do it.

You could use flexible folded arrays like on the ISS, and simply extend a couple out of the airlock and allow them to dangle down. I presume that landed vehicles will be anchored by cables anyway, and youy'd do the same with the arrays, just tie them down to stop them flapping about - you could even tie them off at an angle to maximise insolation.

Perhaps there are some ISS engineering models in store which SpaceX could play with...

[speedevil]

Perhaps there are some ISS engineering models in store which SpaceX could play with...

ISS solar panels are of the order of half the output per kilo of the commercial panels mentioned at the beginning of the thread, and not particularly better suited.

ROSA might in principle be better suited.

But to be scalable, it has to be very cheap, or you run into cost issues.
ISS panels were $3500/W, or $100K/kg.

If cargo launch to Mars, for even the first 'disposable' BFSs cost $1.5K/kg, and you bring 2 tons of solar panels, you might as well instead bring a whole nother BFS and just fill it with heavy glass panels as fitted to homes.
For the panels mentioned at the beginning of the thread, the purchase price on earth would be 1/30th of the launch price to Mars with a disposable BFS

That is probably optimised too far in the other direction, and obviously you would want to replace the plastics with more UV stable ones, add some integral lightweight foldable frame, magnetic or easily chainable power connectors, ...

But this gets you to $500/kg or so from a vendor that is overjoyed to get a $10M order.

[ThereIWas3]

I think batteries in a modified Powerwall configuration make more sense than in the Tesla Model 3 car configuration.  The modifications would take into account the much lower gravity and so the cabinetry and support hardware could be made lighter.  But I wonder how they plan to deal with the temperature?  It gets really cold on Mars (average -63C) and batteries hate cold.  Assuming the batteries are kept inside the ship (or later, inside habitats), how do you heat those?  Or insulate the walls enough to prevent heat loss.  Being on the planet means no more free "thermos bottle" insulation from being in a vacuum like all previous space missions.  It is fortunate that the thin atmosphere means practically no "wind chill" effect.

Luckily, solar panel efficiency actually goes up in cold temperatures.

[Robotbeat]

Batteries are liquid cooled. But the batteries can be heated if need be, too. They are for the Model S (etc) in cold weather. But really, keeping a large powerpack-sized unit warm won’t be a problem as long as it is regularly used. A small amount of insulation will do it. The real hard part is the heat rejection. But again, they’re already using liquid, so shouldn’t be too hard to figure out a solution.