The energy potential of the sun’s light is obvious to anyone who has ever seen a magnifying glass used in the sun to burn things like ants, paper, or themselves. Direct sunshine delivers a lot of energy to what it strikes. By focusing even a small area of it down to a small point, a magnifying glass can create enough heat to burn through many materials.
To produce electric power from heat, we generally use a steam turbine. To run a steam turbine, you need something that gets hot (like a nuclear reactor core), and something that is used as cooling (like water from a lake or river). The difference between the hot and cold temperatures defines how efficient the heat to electricity conversion can be. The majority of the world’s electricity is produced by steam turbines. The difference between forms of power is generally where the heat comes from. Here we are talking about using sunlight as our heat source.
What is so special or unique?
Solar thermal power generation has some properties that make it very special, perhaps even unique among all the forms of renewable energy. Keep in mind that solar thermal is very different from solar photovoltaics. Solar photovoltaics are what people are generally referring to when they say ‘solar panels’. Solar thermal plants look and function very differently from photovoltaics. Wikipedia has some nice images if you are interested. We will go into the physical construction of these plants later in this piece.
Very High Temperature
With a steam turbine, the higher temperature you can reach, the more efficient your conversion of heat energy into electricity can be. Humans have also built turbines that use working fluids other than water, and some even use gases. These more advanced turbines are capable of reaching higher efficiencies because the steam turbine begins to lose its effectiveness compared to them when the heat source is above 600⁰C.
Some forms of solar thermal power are capable of reaching temperatures around 800⁰C. This leads to increased efficiencies and to other advantages which we will now look into.
Heat storage is easier than electricity storage
There are no good solutions yet to the problem of storing electricity. These solar thermal systems can get around that problem by storing heat instead. Storing heat energy is more efficient and cost-effective than any currently used method for storing electricity.
The sun’s light is used to heat up a working fluid of some kind. This fluid is then kept in an insulated tank until it is needed for power generation. It is then pumped out to where it can be utilized by the turbine to produce electricity. Afterwards, it has lost a lot of its energy and is much cooler. It is then reintroduced to the heating system to be heated back up by the sun’s rays.
Can produce dispatchable baseload energy
Being able to store energy is extremely important. Having energy there when we need it is the foundation of a reliable grid. The most valuable sources of energy we use today are those that are:
- Baseload. Which means they generally produce the same amount of power all the time.
- Dispatchable. Which means they can be turned on at will, and can power up in a short amount of time.
- Both. Systems that are both dispatchable and can fulfill baseload generation duties are incredibly valuable.
Solar thermal generation combined with heat storage has the potential to be both baseload and dispatchable. This is the first directly solar technology to have these properties. Of all renewable energy options, only one has been broadly developed so far that can fulfill both of these roles, and that is reservoir-based hydropower. Most good reservoir hydroelectric spots in the world have been developed due to this fact.
Being able to build systems that have similar properties to hydro, but in different places, is a profound step forward for renewable energy. It appears to us that this may be one of the key technologies in the energy revolution of our times.
For further reading, you may be interested in our article on making dispatchable power from renewable energy sources.
Backup power easy and cheap to implement
In the case of a cloudy spell lasting hours or days, the power output of these systems would drop. It is advantageous to have a power source that can be relied upon very firmly, so sometimes these systems paired with fossil fuels for backup1 . This can be done quite cheaply since all of the turbines are there already. The only detail is that one has to plan a method by which another energy source such as natural gas can be turned on to produce heat when the sun isn’t getting through.
Water use can be minimized
Since some of these systems are capable of reaching such high temperatures, it may be possible to build them in places that do not have much available water. This is advantageous since solar energy is great in deserts, but there isn’t much water to cool your power plant with.
It is possible to use air cooling for power plants to save on water usage. For example, the steam that comes out of the plant superheated can be condensed and cooled using the air, with most of the water being recaptured. This hot water would then be fed back into the cooling loop. The problem is that cooling in this manner is not as effective as using colder water. That is, the temperature of the cold side of the turbine is much colder when using water from a river or lake than it would be if it had just been condensed from steam using air cooling. The efficiency of a turbine is related to the difference between the hot and cold temperatures. So air cooling in this manner leads to slightly less efficient electricity production and some additional expense.
With much higher temperatures available for the hot side of the turbine however, the efficiency can remain reasonably high even using just air cooling. This is not as effective of a solution as using direct water cooling is, but it frees us from a design restriction. Instead of having to build beside large water sources, we can build almost anywhere there is lots of sun. Thus the deserts of the world can become our energy sources.
To see an example of this in planning, visit the Wikipedia article about the Ivanpah Solar Power Facility.
Cheaper than large-scale photovoltaics
Solar thermal systems can create more electricity than photovoltaics (PVs) in same amount of space. They are also cheaper. Estimates for solar thermal power generation run as low as 10¢/kWh for currently planned projects. See our section on cost for more details. This is cheaper than any solar PV project we know of. This is however not to say that PVs are useless in comparison. PVs have a lot of abilities that solar thermal systems do not have. For instance, they can be deployed at any scale, while solar thermal has certain sizes that are more advantageous. Also, solar thermal requires considerable construction at a power generating site while PV requires only very minimal construction. Lastly, PV uses far less water than solar thermal. The small amount of water required for PV is used for acquiring the materials, manufacturing the panels, and keeping them clean during their lifetime of use.
Cheaper turbines if used for baseload
If for example, we want to produce 50MW for 24 hours (using energy storage) rather than 150MW for 8 hours, we would be producing the same amount of total energy. Energy is power multiplied by time, so in either case we have 24 hours multiplied by 50 MW equals 1200MWh (Megawatt Hours), which is a measure of energy related to the kilowatt-hour. This is important because in the case where we produce 50 MW for 24 hours, we only need a turbine system capable of producing 50 MW. This would be much cheaper than the turbine that could produce 150 MW.
Using turbines in this way for baseload would also lead to less turbine wear-and-tear since they can operate at a relatively steady power output rather than being turned on and off. Powering up and down a turbine is often a major factor in aging it. It is possible to use cheap, long-lived turbines for baseload production.
Of course the storage system does cost some money. Estimates place it at around $50 per kwh to install2 . With a big plant such as Andasol, this came out to be around 5% of the total cost of the plant. Therefore there is certainly a trade-off in cost with a smaller turbine being cheaper, but storage costing some money.
It is worth noting that these systems could load-match (or peak-match). They could be dispatchable power to some extent, and that would require that they have a more powerful turbine. The turbine would have to be able to produce more power than the energy storage would be able to run 24/7. This is because with dispatchable sources used for peak-matching, we are looking for power for a few minutes or hours. Optimizing for maximum cost effectiveness in this realm would mean a very different setup than one for baseload most likely. Turbines designed for peak-matching are also different than those designed for baseload. They are meant to be turned up and down as well as powered up in a short amount of time.
How do we build these?
This type of power generation is still being actively developed. There are a number of different approaches being pursued. Each has its own advantages. This list is by no means exhaustive, it just includes the basic types that we are interested in mentioning right now. If you want more details about what exists and is being developed, check out the high quality Wikipedia pages on Solar Thermal Energy and Concentrated Solar Power.
A parabolic mirror will reflect all of the sun’s parallel light rays to the same spot. An excellent image is available at Wikipedia of the focusing ability of parabolic mirrors. The mirror is angled during the day to follow the path of the sun across the sky. A tube in the focal point of the mirror carries the working fluid. The working fluid has the sun’s rays focused onto it, so it heats up. The fluid is pumped through sections of mirrors, then back to a central location for use.
Parabolic solar power plants have existed for decades. This is a well-known technology that has continued to pique the interest of energy developers worldwide.
Power Tower (Heliostat)
The power tower is a large mass on top of a tower. See an image of one at Wikipedia. A field of mirrors are angled so that they reflect the sun’s rays up to hit near the top of the tower. The top of the tower is thus heated up by the reflected light. A working fluid such as a molten salt is used to move the heat downward towards a turbine system and possibly a heat storage system.
The mirrors are controlled automatically by computer systems. These systems must keep track of variations in the suns position during the day, and also season to season. The sun’s position in the sky changes during the year, so these systems correct for that effect. This is very easy to do with today’s computer technology. The only major maintenance issue with the mirrors is the motors that perform the movement of the mirrors. There are also sensor systems that exist that can tell when a motor is failing to function correctly or when the mirror is not pointed in the correct direction. These systems are quite affordable and do not add significantly to the cost of the power plant.
The distinct advantage of the power tower is that it can reach much higher temperatures than the parabolic trough design. Higher temperatures means a higher efficiency of converting heat into electricity, assuming that the cooling used is of the same temperature. Power towers with heat to electricity conversion systems of 38-41% efficiency currently exist, and ones with 50% may be feasible within the next decade3 . We have the knowledge to build these towers. One can hope that policy makers and energy companies take note.
In 2003 the National Renewable Energy Laboratory (NREL) looked at Parabolic Trough Solar Thermal Generation. According to the NREL, the costs of parabolic trough solar generation were around 10-12.6 ¢/kWh in 2003, and were projected to drop to very competitive levels within a few years assuming that investment in them was aggressive. The numbers they cite for 2020 are incredibly low: around 3.5 to 6.2 ¢/kWh. At this point, parabolic trough solar would be competing with even the cheapest compliant coal generation units. During the same study they also looked at power towers. Their estimate for 2020 was 5.47 ¢/kWh. The NREL is quoted in this Scientific American article as saying that electricity from solar thermal power plants, with or without storage will cost around 13 ¢/kWh. There are also good reasons to believe that the price of solar thermal is edging its way closer to 10¢/kWh as they are scaled up.
Conclusion: Great Potential
Solar thermal power generation presents a unique opportunity among renewable technologies today.
- It is well-understood. Prototypes and commercial power plants of this sort exist.
- It has capabilities for both baseload and peak-matching power generation.
- It is affordable in locations with lots of sunlight. Costs are in the range 10-15¢/kWh currently, with great potential for the future.
- Extremely low C02 footprint.
- It can be coupled with fossil fuel based backup systems for even better reliability at a low cost.
- Low water-use designs exist for areas without access to a lot of cooling water.
- Possibilities for using steam turbines and other infrastructure from existing thermal power plants.
These properties make solar thermal a veritable superstar among the up-and-coming renewable power sources. Most renewable power sources are intermittent, not dispatchable, and relatively expensive compared to their conventional competition such as coal and nuclear. Solar thermal seems to be capable of being as desirable as hydro, perhaps more so. With less land use than hydro and the flexibility to be used in any sunny area of the world, this may be one of the key energy technologies of the future.
- Solar Thermal Power Plants. Ferrostaal AG. 2010. [↩]
- How to use solar energy at night. Scientific American (2009). Retrieved Sept 17th, 2010. [↩]
- Forsberg, C., Peterson, P., Zhao, H. (2007) High-Temperature Liquid-Fluoride-Salt Closed-Brayton-Cycle Solar Power Towers. Journal of Solar Energy Engineering, Vol 129, 141. [↩]