Preble (2000): Prospects for Space Solar Power

Robert Zubrin of the Mars Society has written several fascinating and energizing books on space exploration, with a perhaps natural emphasis on human travel to Mars.

Darel Preble of the Space Solar Power Workshop - - was prompted by Zubrin's Entering Space to examine the numbers used there for solar power satellites, which Zubrin dismisses as unlikely, and provides us with the following critique and more optimistic comparison of space solar power with terrestrial alternatives, originally written in early 2000.

From page 71 of "Entering Space": The anticipated cost for space transportation with Space Solar Power System construction traffic in GSO would be neither $40,000/kg nor $17/kg. At the last Space 1998 meeting, Hu Davis calculated this at $97/kg, which the SSPW conservatively and conveniently rounds to $100/kg for our rough planning estimates. Just twenty-one pages earlier, page 50, is a quote of transportation studies from 1995 with smaller traffic volumes that projected costs of $400/kg or $2000/kg for sub orbital package delivery flights. Of course, if priority package delivery and SSPW traffic models were superimposed on other probable traffic models, space transportation costs could be driven low enough to enable many new space systems opportunities.

Also from page 71: Current efficiency for space photovoltaic panels is 27% not 15%. This is available from Tecstar, ( for example. Twenty five percent space qualified PV is now typical of Tecstar production. It is reasonable, in fact, to expect a slightly higher efficiency with significant SSP prospects, say to 30%. Laboratory demonstrations above that figure have already been done. For example Hughes Spectrolab has announced 32.2% efficient cells.

The weight of silicon photovoltaic fabric would not be 4 kg per square meter, but far lower for the currently planned and/or the several available thin film designs. (see Many current PV semiconductor thicknesses are typically only a few microns thick, (IEEE Spectrum, "Photovoltaics", page 37, Sept. 1999).

According to the most recent NASA HQ numbers I have seen (November/December 1998), a 5 GW Space Solar Power Satellite is expected to have a mass of about 42,000 metric tons in space and double that payload mass launched. That Concept Definition Study estimates SSP wholesale costs around 5.4 to 5.7 cents per Kwh, which is comparable to wholesale costs in most of the world, excepting US wholesale costs, which are about 2 cents per Kwh, nationally, or Japan at much higher numbers. Improvements in NASA's design engineering from such numbers is underway.

From page 73: The claim is that the advantage of SSP over terrestrial PV is about 50% if sun tracking is included. In actuality, studies have shown it is close to a factor of 10 for typical US sites compared to GSO: Using the most current NREL standard tables as listed at

Solar Radiation by City for Flat-Plate Collector Facing South at Fixed Tilt=Latitude
MAJOR US CITIES Average kWh/m2/day
US Average 4.78

The total US Avg = 1746. kWh per year for Flat-Plate Collector At GSO Solar Radiation is 1.367 kW/m2 , so for the year we get 11975. kWh per year Therefore solar radiation is 6.86 times higher at GSO. If sun tracking is added to any solar array location, the collected power increases by only 20%, (not by 100% as claimed on page 73), moving the US average to 2095 kWh per year, which would reduce the expected advantage of GSO to 5.7, but substantially increase the operational costs. Very few installed PV systems track the sun; as it is not cost effective.

Unfortunately, based on the best studies extant, by Texas Utilities's Electric Park, the expected rating for their PV system was predicted to be 140,000 kWh but actual production was closer to 100,000 kWh. The cause was attributed to summer heat, smog and haze. Based on their studies the typical US average advantage of SSP over terrestrial PV is 6.86 * (100/140) = 9.6 and probably higher since Texas (and TU Electric Park) is a better PV site than average US sites. Even more telling, from their actual system studies was the fact that peak PV performance coincided with early spring or late fall when electric system power use was at annual lows. See EPRI TR-106409 or RS-106409 for additional details.

The most serious disadvantage for ground based solar however, is the huge cost of storing vast quantities of energy overnight. There are no known near economic ways to accomplish this. Many schemes have been tried, from pumping water up a hill at night (when power demand is low) to a high reservoir and storing for recycling the next day, to compressing air in giant underground salt domes. None has approached the goal of "only" doubling the original cost of the power.

SSP suffers from none of these problems, in fact its brief 72 minute outages occur during the equinoxes at local midnight, which are optimal timing for power outages. No baseload generating station known can match their low rate of scheduled down time - about 1%.

From page 80 and 88: Here and other places it is stated that Helium 3 is "nonexistent" in the inner solar system and would be almost priceless for fusion energy, e.g. $1,000,000/kg). But it is not necessary to visit the Moon to obtain Helium 3. It has been available at least since 1998 in quantity through many suppliers, e.g.

The natural abundance on earth of the Helium-3 isotope is 1.38 x 10-6. Contrary to one of fusion researcher's popular myths, however, fusion power's severe engineering difficulties have little connection to Helium 3 availability. Although I wrote a thesis in nuclear structural analysis, I contend that there are no experts in fusion power. IF there were, those experts could predict nuclear energy excitation states, (eigenvalues), which they cannot - not even to one significant figure.

The rich bouquet of problems presented by deuterium tritium fusion or He3 fusion begins with our dismal understanding of the strong interaction (i.e the nuclear force). Work in this field stands in stark contrast to quantum electrodynamics (QED), which is the most successful theory man has ever constructed. QED can reliably predict experimental results to more than twelve significant digits! And so, plasma physics would perhaps be sufficient if we were dealing with Hydrogen - Hydrogen fusion (single unbound protons), but the Lawson parameter n-tau_e (density X energy confinement time) required for this most common fusion is unearthly, as you point out. Therefore the plasma state equations should also include strong interaction terms to handle the ionized muclei - such as tritium or He3. Here our math falls apart because our theoretical nuclear physics falls apart. We cannot get even ONE reliable significant predictive digit from nuclear theory! For making bombs this is not such a problem, for sustaining a reaction this is a severe disadvantage.

As Richard Feynman contended just before his untimely death - the mathematics is far more difficult than we can handle. (Feynman directed the Mathematics section during the Manhattan project and had a highly respected opinion in these areas.) This is no reason to give up, of course, but simply a stern warning that pouring a few more billion into fusion is unlikely to quickly solve this intractable challenge. Steel is no substitute for skull work.

More importantly, if you are trying to plan for global strategic energy continuity in the face of the global energy market and environmental stresses very likely during our childrens (or our) lifetimes, it would be highly presumptious to depend on commercial fusion power being there. A more severe fossil fuel energy crisis during that time is far more likely:

Fortunately, we have a large fusion furnace already conveniently located nearby, if we can gather its power at reasonable cost. Unlike fusion power engineering, SSP engineering is relatively simple. Yet funding for SSP work has been nonexistent from 1980 until 1997 and only in 1998 achieved one hundredth the level of fusion power's generous budget, (which is also far below previous levels). Still SSP is making faster progress toward commercial reality.

The powerful energy lobby seems to prefer energy research moneys being expended on a virtually impossible problem (fusion) or peripheral peaking power energy alternatives such as terrestrial PV or wind, rather than one (SSP) which could "eat their lunch" within twenty years. SSP is inherently baseload, and baseload electric power is the ultimate trillion dollar market goal.

(In truth, the prospect of new power generation schemes should not be perceived as threatening to the owners of major energy facilities, merely another rising energy alternative to add to their already busy toolbox. But like IBM giving away the market that Microsoft now dominates, the Swedish proverb "Too soon old, too late smart" still applies.)

Since power generation is the most capital intensive industry extent, the prospect of plant replacement is never pleasant. But just as half of America's nuclear reactors are expected to be decommisioned by 2015, for example - which is about 10 percent of America's current baseload electric power generation - all these plants will one day be decommissioned.

In a curious coincidence, if you believe in coincidences, just as He3 became available to researchers last year, our nation's strategic stockpile of helium was being sold off and will be closed by 2015. No doubt commercial interests can make a few bucks marketing helium, but they will not undertake or substitute for the terribly crucial role of strategic stockpiler that Uncle Sam has performed since the earth's only natural helium (and gas) fields were found around Texas and Kansas. America's effective monopoly on this critical material seems to be being discarded just as He3 is being aggressively studied for fusion potential and helium is becoming more essential for important cryogenic purposes. . .

The best of many general SSP overviews available is "Solar Power Satellites" by Peter Glaser, F.P. Davidson and K. Csigi from John Wiley-Praxis publishers 1998 ISBN 0-471-96817-X.

Created: 2005-02-10 05:38:45 by Arthur Smith
Modified: 2005-02-10 05:43:57 by Arthur Smith