When asked what energy source will be the most important in a world 15 years from now, most respondents (27%) in a recent survey named solar energy[1].  This is not surprising. The sun is by far the most abundant source of power in the solar system.  Nearly all energy forms used on Earth are, either directly or indirectly, based on solar energy, the most prominent being fossil fuels, biomass, wind, hydro energy, tidal, and solar energy.  The latter can be directly converted into electricity in several ways. So it would appear that direct harnessing of solar power should be one of the biggest hopes as an alternative energy source. However, extrapolating from current growth estimates, one arrives at a much more modest role for solar energy in the future -- probably still in single-digit percentiles in 20 years.  It appears the only way that solar power can live up to its expectations is through some dramatic break-through that will dramatically increase its growth rate.  In this paper, we assume that this breakthrough happens by some break-through in research.  A break-through appears rather likely -- several approaches are currently underway to dramatically increase the solar capture efficiency, most notably concentrators, semiconductor heterostructures, and quantum wires and quantum dots for multiple exciton generation to capture a larger fraction of the solar spectrum.  In this paper, we focus on photovoltaic (PV) power generation (although century-old thermal power generation is seeing remarkable revival in recent years).

Solar cells are commonly used in very complex systems, such as satellites. So  how can we predict the effect of increased efficiency of PV generators--only one of many subsystems---on the overall system configuration? It would be prohibitively difficult to re-optimize the system's configuration. So instead we approach the problem from a broad view-point.  We scale the relevant parameters to determine the 1st order approximation of the system's new configuration.   The most important parameters are the size/mass, power, performance, and cost.  Thus, keeping one of these constant, we can estimate how the other three change.  For example, for system X,  we ask: ``If the power requirement is unchanged, how will the increase in solar cell efficiency affect the mass and cost, and performance of X?''  By keeping one of the parameters of the system constant, we do not explore the whole parameter space, so our prediction to the Scenario Statement is only a first-order approximation. But it will be good enough to understand the fundamental relationships between the solar cell efficiency and the whole system and make rough numerical estimates. 

We will restrict the analysis to the most relevant applications in space and on Earth.  The first includes predictions on mass, cost, and performance, for satellites; atmospheric drag of satellites (or the ISS) in low-earth-orbit; solar ion propulsion technology; and solar-powered deep-space probes.  The second discusses effects on off-the-grid as well as commercial, on-the-grid, power production. 

See full report: SCPaper-englund.pdf