Facing the Numbers – at a typical US location:
Why a ground solar plant at Austin, TX needs a collector area 7.5 times larger than a 1 Gigawatt-rated
collector to produce 1 GW of Base Load Power. John K. Strickland 1-1-2010 jkstrick@io.com
Base Load power accounts for most of the US average electrical demand of about 500 Gigawatts, and it needs to be
un-interrupted. Wind and solar power are intermittent. To get base load power from solar or wind takes a huge plant
and site. Austin, Texas is planning to build a ~30 Megawatt photovoltaic plant to cost about $250 million. It would
intermittently produce 28.9 Megawatts ($8.7 million per MW) with an active collector area of 267,000 m2, or about 66
acres. It will use 171,000 Suntech Collector modules mounted on 14,000 single-axis tracking collectors tilted at 20
degrees to the south (very similar to the NREL data for a 15 degree tilted collector). Graph 1 shows the average
number of hours of Effective Full Sun (EFS) per day in Austin, Texas for such a collector system with a tilt of 15
degrees. (EFS means the equivalent hours of steady sunlight at 1 kilowatt/sq. meter. Power from this plant will cost
about 3 times more than current Austin power, (about 16 cents / kwh vs. about 5 cents / kwh). Graph 1
Average EFS Hours by Month
10
8
6
4
2
0
Jan
M ar
M ay
J ul
S e pt
Nov
National Renewable Energy Laboratory data for Austin: 1961-1990
EFS values are the same as kilowatt-hours per square meter per day. These values are related directly to the capacity
factor (the ratio of hours of full power production out of 24) for a solar plant during each month). If the sun were
available directly overhead for 24 hours a day with no clouds, the site would produce about 24 kwh per square meter
per day and thus 24 EFS hours, (a 100% capacity factor). The average kilowatt-hours per square meter per day for
the 30 year period 1961-1990 in Austin is 6.4 kwh, with the average maximum obtained in July with 8.7 kwh and the
average minimum in December with 4.4 kwh. The capacity factor average for this installation in July is 0.36, for
December it is 0.183, and for the year it is 0.266 Thus, in the winter, to supply the 30 Megawatts of power when the
sun is not shining, Austin must rely on gas fired power for an average of 19.6 hours a day, in July for 15.3 hours a day,
for an average of 17.6 hours a day. The combination of intermittent power and low power during winter months is what
creates the low capacity factors for solar collectors in most areas. Austin has frequent week-long cloudy spells during
the fall, winter and spring. Austin is a good site for solar plants only in the summer.
Graph 2 shows typical power outputs for a 1 Gigawatt base load (such as a nuclear) plant and intermittent daytime
only solar plants during clear summer and winter days and cloudy winter days. The total area under each curve
represents the number of Megawatt-hours that would be generated by a plant with a 1 Gigawatt-capacity collector that
day. It is clear that the blue area representing a clear winter day covers only about 25 percent or less of the total area,
indicating why the capacity factor for most solar plants is so low. This idealized curve represents a non-tracking or
fixed collector system such as would be installed on a rooftop. Graph 2
To keep a storage example very simple, assume a city needs an average of 1 Gigawatt of base load power. (This is
comparable the base load amount that Austin, Texas uses). The city will thus need 24 Gigawatt-hours of such power
each day. Assume that it wants to run (as an extreme case of trying to use an environmentally purist methodology
with essentially no use of fossil fuels, which is exactly what many are now advocating), entirely on ground based solar
and/or wind, with no fossil fuel backup. Using a worst-case (winter) scenario, it can be assumed that the capacity
factor for both wind and solar is about 0.16. (Wind power is weakest in the summer). Thus 16 percent of the time
(about 4 hours a day on an average day), the city’s utility needs to gather all of the energy required for those 4 hours
and also for the 20 hours when it will not be gathering any energy. The energy collection system (solar, wind or a
combination) thus needs to be at least 6 times larger (24 / 4) than a system that produces 1 Gigawatt with full solar or
wind energy input, in order to collect all of the 24 Gigawatt hours in just 4 hours during an average winter day.
However, it also needs to store the captured power for night-time use. If the storage system (such as pumped hydro
electric) has a recovery efficiency of 75%,and the city stores exactly 20 Gigawatt-hours, it only gets back 16. (Note
that in the very best solar sites, use of thermal solar collectors and thermal (molten salt) storage could remove the
additional storage requirement since thermal storage efficiency is close to 99%). This means that the city must store
30% more or 26 Gigawatt-hours to get back the 20 it needs, suggesting that the collector for this site must actually be
sized to collect 4 + 20 + 6 = 30 Gigawatt-hours or 7.5 times larger than a collector used only when sunlight is available.
(Shown in Graph 3 below). These values assume average winter conditions, and do not cover the week-long cloudy
periods which are quite frequent. What happens during such a spell? Graph 3
Gigawatt-Hours Needed / day for a 1 Gigawatt Base Load Power System
where: Collector Capacity Factor = 0.16 & Storage System Efficiency = 75%
Units in Gigawatt hours 1 Gwh = 1000 Megawatts for 1 hour
Storage Losses, 6
Daytime Power, 4
Daytime Power
Nightime Power
Storage Losses
Nightime Power, 20
Power collected
4 GWH
20 GWH
6 GWH
30 GWH collected
Power Stored
0 GW hours direct use
20 GW hours night use
6 GW hrs storage losses
26 GW hours stored T
Collector capacity ratio
1 GW
5 GW
1.5 GW
7.5 GW
Note that 7.5 Gigawatts of collector capacity is equal to the entire installed capacity of Texas wind turbines at the end
of 2008. If these were all 2.5 Megawatts in size, they would represent 3000 turbines, covering a wind resource area of
450,000 acres, just to produce 1 Gigawatt of base load power. To cover just the current U.S. car and light truck
demand of about 500 Gigawatts would take 1,500,000 such turbines covering 350,000 sq. miles. Using the cost of the
Austin solar plant ($8.7 million / MW), or $8.7 Billion / GW, as a model of current ground solar costs, it can be shown
that to provide base load power with ground solar and using no fossil fuel (thus requiring a collector 7.5 times larger)
will cost about $65 Billion per Gigawatt, plus the cost of a massive energy storage system ($2-5 Billion for 26 Gigawatthours storage). This is about 30 to 60 times higher than costs for existing base load plants and still does not cover the
multi-day cloudy periods with no sunshine. Power from such a base load plant used without fossil fuel backup would
cost about 25 times more than conventional power rates, and this does not include the cost of building the power
storage facility (intermittent solar plant cost of solar of 16 cents / kwh times 7.5 or about 120 cents / kwh.
The mass of a 1 Gigawatt Solar Power Satellite has been estimated at no less than 2,500 tons. If an Ariane V rocket
payload costs $10 million per ton, and assuming that an Ariane was scaled up to a true Heavy Lift Vehicle size with the
same launch cost per ton, the cost for orbiting a single powersat to Low Earth Orbit is no less than 25 Billion dollars
per Gigawatt, or $25,000 per installed kilowatt. If construction of the ground antenna (to receive the power), the
satellite itself, and its transfer to Geosynchronous Earth Orbit, all costs about 5 Billion, the Space Solar system (which
is still extremely expensive due to high launch costs) is still only about half of the calculated current cost of the model
ground based load solar system. With lower launch costs, Space Solar becomes a viable clean energy option.