In this question, Geoffc's answer stated that the ISS solar arrays are about 14% efficient. Higher efficiency panels were available when these arrays were manufactured. Given the high cost per pound of launching stuff to orbit, why were more efficient panels not used? This question estimates the mass of the arrays at 30 tons. Double the efficiency (as seems to be possible with triple-junction panels), and you'd be able to halve that.

  • 7
    $\begingroup$ When putting something into orbit weight is a major concern, but so is reliability. It has to work, so tried and tested technologies are picked over new. $\endgroup$
    – GdD
    Jun 30, 2015 at 9:00
  • 5
    $\begingroup$ One less known fact about ISS solar cells is that they're double-sided (bifacial) to also charge by Earth's albedo illumination. Since this reflected insulation's photon flux is much weaker, this would reflect in ISS SAW (Solar Array Wing) total efficiency. $\endgroup$
    – TildalWave
    Jun 30, 2015 at 12:30
  • 8
    $\begingroup$ Even if the panels were not the most efficient in terms of energy collected per area, they might still be the most efficient by energy per weight. $\endgroup$ Jun 30, 2015 at 13:14
  • 1
    $\begingroup$ Also, recall that the first array was launched in 2000. Which means the design was probably frozen in the mid 90's. So consider only available space rated arrays in the mid to late 1990's. Todays arrays would of course be more efficient. $\endgroup$
    – geoffc
    Jun 30, 2015 at 13:27
  • 2
    $\begingroup$ @TildalWave So I asked it as a separate question. I am betting they did NOT upgrade them. But what do I know? :) $\endgroup$
    – geoffc
    Jun 30, 2015 at 15:45

1 Answer 1


As 2012rcampion noted you are using the wrong metric of cost in your efficiency figures.

Triple junction solar panels are extremely inefficient wrt weight when compared to single junction.

To understand why this is you need to understand what a triple junction PV is.

To do that I need to explain the major inefficiency of standard single junction PV is due to quantum effiency.

Each single junction PV has a semi conductor junction with an associated voltage or colour.

That PV junction will only give you energy for photons of that colour/voltage or bluer. So if you have a PV junction that is tuned to green, that panel would give no power under a red light.

Conversely higher energy photons (bluer) would only give off an electron with that same voltage (the rest of the energy lost as heat).

Triple junction solar panels literally stack a red, green and blue solar panel on top of each other to maximise both the number of photons turned into electricity and the electrical energy gained per photon.

But if you are stacking 3 panels you could just build 3 panels instead and place them side by side so they don't block each other.

So for a naive approach triple junction is 3 times heavier, but not 3 times as powerful.

  • 1
    $\begingroup$ this would be true if whole panel would consist of just p-n junction ... but there is considerable amount passive balast such as substrate ( silicon, glass ... ) contacts, and structural parts (e.g. metalic truss and girders ) if this passive balast is more than 33% of the weight of the whole panel (which I guess it is ... but I'm not sure) than going from 15% to 30% efficiency while would be usefull $\endgroup$ Aug 27, 2015 at 18:39
  • $\begingroup$ The substrate scales with the number of junctions. The structural parts do not (by the way, glass counts as structural). However, you have to remember that solar panels in space do not need good structural support (zero g). I also completely neglected the non-quantum losses involved (silicon is actually not very transparent, extra layering decreases transmittivity). That is why I said that for a naive approach it is 3 times heavier. $\endgroup$
    – Aron
    Aug 28, 2015 at 1:14
  • 2
    $\begingroup$ but thickness of wavers is ~0.1-0.2 mm, while thickness on active junction is often just several microns. By "substrate" I mean that many thin-film solar cells (where juction is even < 1 micron ) are still deposited on some substrate like glass for structural rigidity (ISS panels are also rigid, so I guess they have quite thick substrate/support in comparison to junction thickness). ad "in space do not need good structural support (zero g)" ... I agree I would prefer something like phys.org/news/… ... but this is not the case of ISS SAW $\endgroup$ Aug 28, 2015 at 5:46

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.