There are a number of proposals provide for the Earth’s energy demands by using solar energy satellites (SES). The most common version is to use satellites in geosynchronous orbits. The satellites would convert the suns energy to electricity and use the electricity to generate a beam. The beam will be transmitted back to the Earth’s surface where it will be converted back to electricity. This is presented as having great benefits over ground based solar power options. Many of those proposing and supporting technology of this type suggest that it could provide power more cheaply than fossil fuels or existing renewable technologies. I believe that this viewpoint is based on several misconceptions. I will introduce the issues below and then try to explain in detail what I mean.
In regards to the cost of SES compared to fossil fuels is a misconception held for many renewable power sources. The fuel for renewable power is “free” so the power generated by it must be cheaper. The misconception here is that fuel costs are actually a small part of the overall cost of power generation. A majority of the cost of power generation is split between initial construction costs and ongoing non-fuel operation and maintenance costs. This is the reason that all renewable energy except hydro are only now approaching or achieving grid parity as the cost of construction per KW delivered has been significantly higher. This would be even more true for a similar installation in Earth’s orbit.
In comparing it to existing renewable technologies it is usually argued that it is more consistent and efficient than ground based options. On the surface this is true but ignores the fact that this is only advantage if it is cheaper than a ground based option that would cover the gaps in generation such as other power generation method or power storage methods. An orbital installation’s construction cost is likely to make it far more expensive than any ground based option including grid scale battery systems.
The advantages that an SES system has over non-renewable options is almost identical to ground based solar power. Ground based solar power solutions are just beginning to be competitive with non-renewable option. As that is the case if an SES system is not able to compete with a ground based solar power system it will definitely not be able to compete with non-renewable options. In addition solar thermal option continue to be significantly more expensive than photovoltaic (PV) options. For that reason from this point forward I will focus on comparing SES to ground based PV.
The two main advantages of a GEO SES is that it gets more hours of sunlight and the power density of the sunlight is greater without the atmosphere filtering it. If the Earth based system uses similar sun tracking technology as would be required to an SES system and the location were chosen moderately well it could achieve and average of 9 hours of full sunlight. This would provide a 2 and 2/3rd advantage to the orbital system over the earth based one. The Earth’s atmosphere absorbs about 35% of the overall but as more of this loss is in the UV and infrared range not used by PV the loss is closer 21% in this comparison. Combined with hours of sunlight this results in an overall advantage of almost 3.38 times the power available.
The Show Stopper
I will start with the single biggest obstacle to the feasibility of a SES system, launch costs. The single most important factor here is weight. I checked several sites to identify what would be the weight of the solar panels to provide 1 KW of capacity. The best numbers I came up with was about 120 lbs./KW. With the number above we could determine that we get the same bang out of only 35.5 lbs. of solar panels. When it is launched this summer the Falcon Heavy will likely be the cheapest way to get to orbit per pound. Details listed on the Space X website give pricing of $772/lbs. to LEO and $1,930/lbs. to GEO. Even if a way is found to transport the panels from LEO to GEO for free this would still mean that it would cost $27,406 just to transport the equivalent of 1 KW of ground based capacity to orbit. The more reasonable cost to take the panels to GEO would be $68,515. Agua Caliente, a fairly recent solar plant project costing $1.8 Billion with a capacity of 290 MW that works out to just $6,000/KW. The worst part of this is that the solar panels are only part of the weight that such an orbital project would require. Some of the things I have not tried to estimate the weight of is the framework that the panels would be mounted on, the motors needed to keep the panels pointed at the sun, the engines needed for station keeping, the construction equipment to build it in orbit or the transmission antenna for beaming the power back to Earth.
Often with show figures like these the reply will be, “the next launch technology will significantly reduce launch costs.” The problem with this is that if launch costs dropped by an order of magnitude and the rest of the project was free it would still not be competitive with existing ground based solar power or almost any other power source. And the bad news just keeps coming. Launch cost are not the only prohibitive costs and in future posts will discuss these and options for orbital beamed power I believe are more likely to work.
- Article referencing pricing for grid scale batteries – Big Batteries Don’t Come Cheap
- Article from U.S. Energy Information Administration – Capital Cost for Electricity Plants
- Post on ‘Do the Math’ blog – Space Based Solar Power
- Space X launch costs – Capabilities
- Wikipedia – Agua Caliente Solar Project