I'd like to know what is the lightest possible solar array per unit power with current or near future technology.

The arrays I have been able to find:

  • Vanguard Space THINS 400-500 W/kg (2014; thin film)
  • SLASR 362 W/kg (2006; concentrator with GaAs cells)
  • SCARLETT 45 W/kg (1998; concentrator with GaAs cells; flown on DS-1)
  • Orbital/ATK UltraFlex 150 W/kg (since 1995-98; triple junction GaAs; flown on Mars Phoenix and Mars Insight)

Thin film solar cells (for example CIGS) can have efficiencies around 10% with a thickness of 1-3 microns (excluding substrate). A 10 micron laminate with a thin film solar cell would weigh 30-50 g/m^2 depending on composition; this would put the output at 2600-4300 W/kg, which might be a reasonable ballpark near-term goal without major technology advances. With a bit of a stretch, a 3-5 micron laminate with a 1-2 micron cell could be over 10 kW/kg. From that perspective, the numbers for existing arrays are a bit low.

What is the state of the art and most promising research for ultra-lightweight solar arrays? What are the major technical challenges to higher power/mass ratios?

  • 4
    $\begingroup$ Current SOA is 1kW+/kg BOL at 1 AU. but this will constantly change and it also largely depends on reliability, efficiency and the environment it would be exposed to. Basically, it would help if you edit to also mention some specific application for it and required durability and output at EOL. Out of the ones you list, SLASR MJC should be most effective for long duration and high radiation environment missions. UltraFlex/MegaFlex have good switching & voltage regulation. SCARLETT has good efficiency at small insolation,... i.e. they're not exactly directly comparable. $\endgroup$
    – TildalWave
    Commented Nov 29, 2015 at 22:46
  • $\begingroup$ @TildalWave: I'm thinking of various solar-electric tug scenarios, for example from outside the van Allen belts to GEO. Tugs that go through the belts would not last as long :) $\endgroup$
    – Alex I
    Commented Nov 29, 2015 at 23:21
  • $\begingroup$ @TildalWave: Could you elaborate on the current SOA? $\endgroup$
    – Alex I
    Commented Nov 29, 2015 at 23:22
  • 1
    $\begingroup$ A flat circular disk has area pi r^2. A sphere has surface area 4 pi r^2. So while removing need to point there's a factor of four hit on insolation per unit surface area. And a sphere would also need structure. $\endgroup$
    – HopDavid
    Commented Nov 30, 2015 at 2:05
  • 2
    $\begingroup$ Don't forget the option of ultra-light mirror array (a'la solar sail) concentrating sunlight over a very small area; with "economy of scale" this has a potential of the least gram per watt, providing the (obligatorily rather heavy) cell proper gets enough watt from the mirror. Plus means of not sailing away... $\endgroup$
    – SF.
    Commented Nov 30, 2015 at 6:17

2 Answers 2


This will vary as we develop new technologies but there are some very lightweight cells. Space Future mentions 4300 W/kg, which should be the current best, but we will surely see improvements and new concepts.

  • $\begingroup$ Thank you! This is the highest number mentioned here so far, and matches my back of the envelope. It's nice they are talking about 17kW/kg as near future capability also :) $\endgroup$
    – Alex I
    Commented Dec 20, 2016 at 6:00
  • $\begingroup$ No problem. Do you have an idea or something you would like to see that would benefit from a very high kw/kg ratio? $\endgroup$ Commented Dec 20, 2016 at 23:52
  • $\begingroup$ I do have something in mind; a drive which has extremely high exhaust velocity, possibly over 300km/s. No free lunch, it uses exactly as much energy as you'd expect. Having a very light power plant is needed to make this useful in practice, otherwise any fuel mass savings are offset by power plant mass. For specific mission types I was thinking LEO to GEO tug, but other options may also be interesting, probably also of the tug variety. $\endgroup$
    – Alex I
    Commented Jan 17, 2017 at 9:35
  • $\begingroup$ Ion thrust, fusion drives they would both need a lot of power. Not a 300km/s exhaust velocity but potentially a much higher one for take off when a lot of thrust is needed. A few days ago they produced metallic hydrogen in a lab if it remains stable when the pressure is removed it stores a lot of energy so it's a potential super fuel for rockets launch phase along with its other potential application for room temperature superconductivity. $\endgroup$ Commented Jan 31, 2017 at 10:29

For large systems the most mass efficient way to collect solar energy could be large and light mirrors concentrating sunlight into somewhat smaller collectors that are more optimized for efficiency.

I say this because it appears that $W/m^2$ and $W/kg$ seem to trade off against each other. Solar tech (e.g. thin film) with very good power/weight is very inefficient, wasting 90% of the solar energy.


(From https://slideplayer.com/slide/3564234/ - note this is terrestrial, not space)

Large wasteful arrays will cause problems as they take up a lot of area, and more mass is then spent on securing that area as the vehicle accelerates or decelerates (this depends on the application, of course).

Another possible advantage is Spectral light management - using mirrors, lenses, diffractive elements etc to split light into a spectrum, so that each wavelength can be converted in a separate cell, hopefully at close to the maximum possible efficiency. It's an area that's undergoing active research at the moment, so I can't say what the performance could eventually be.


  • 1
    $\begingroup$ Including the mass of the structure that keeps the reflector a paraboloid really hurts the W/kg. Unless spinning it is enough to do that... $\endgroup$ Commented Aug 14, 2019 at 20:59
  • $\begingroup$ How much does it hurt though? $\endgroup$
    – Roko Mijic
    Commented Aug 15, 2019 at 13:28
  • $\begingroup$ And you can have a series of linear concentrators hitting a series of narrow linear collectors. The concentrators could be mirrors or thin lenses, and there are options like stretched film. There's a patent on that here: patents.google.com/patent/US6075200A/en $\endgroup$
    – Roko Mijic
    Commented Aug 15, 2019 at 13:32
  • 1
    $\begingroup$ If paraboloid-holding structure weighs 5x as much as the OP's thin film 20 g/m^2, so 100 g/m^2, and increases wattage by more than 5x, then it's worth it. But holding that precise shape is hard... photos of solar sailers show the sails to be far crinklier that what's needed here. $\endgroup$ Commented Aug 15, 2019 at 15:27
  • 1
    $\begingroup$ sure, but the masses are tiny on the scale that we're talking about. I looked at this: ieeexplore.ieee.org/abstract/document/6496892 most of the mass budget they have goes into cooling! $\endgroup$
    – Roko Mijic
    Commented Aug 15, 2019 at 20:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.