"How high did we have to pump water of the same volume to store the same amount of energy as in our thermal storage system?" Leitner compares Baumgartner's 39 km jump to energy storage

By Arnold Leitner

During the past years the concentrating solar power (CSP) industry has had to contend with a number of formidable challenges. The collapse of the global economy after the bankruptcy of the investment bank Lehman Brothers did much to slow the momentum that CSP had built in Spain since 2005. However, the real trouble for CSP has been a change in direction, not the loss of momentum. Beginning about five years ago Chinese-made photovoltaic (PV) solar panels started on a year-on-year price decline which soon allowed large-scale PV solar projects to match and then underbid CSP plants for the cost of generating solar energy. Today, large-scale PV plants are bidding into the California market at energy prices as low as 9 cents/kWh with no annual price escalators.

The first victims of the energy price matchup were oil-based parabolic trough (PT) plants which are no longer competitive in the global market against photovoltaic projects on a cost of energy basis. With standard single-axis tracking for PV plants the output profile for PV and PT plants is about the same while the energy cost for PT is at its best 12.5 cents/kWh. Oil-based PT plants provide a somewhat smoother output due to thermal inertia in the oil and the option of molten salt storage, but these benefits are either not significant enough or too costly to obtain, respectively, to overcome the significant cost of energy gap between PT and PV in a competitive solicitation.

The main, if not only, selling point of CSP over PV today is the ability to store energy. Here the oil-based parabolic trough is challenged, because the temperature differential between the hot oil returning from the field and the “cold” oil after the power cycle is only about 100 degree Celsius. Ultimately to tap the inherent large-scale energy storage capabilities of CSP plants one requires a higher temperature gap, also referred to as the “delta”, between the heat transfer fluid (HTF) exiting the solar receiver and the HTF returning from the “cold” storage tank. This is because the amount of thermal energy a liquid or solid can store (in the temperature range of 275 to 550 degree Celsius) is linearly proportional to the temperature. (We ignore any materials that change their phase in this temperature range, e.g. go from solid to a liquid). Thus the higher the temperature gap the smaller the storage volume can be.

Although Spain has seen storage with oil-based parabolic trough plants and the oil-based parabolic trough plant Solana in Arizona has storage on a massive scale, storage only makes sense for CSP plants with an energy delta of 275 degrees Celsius as this allows the energy storage to become nearly three times smaller and correspondingly cheaper for the same amount of energy stored than a storage system for an oil-based PT with a temperature delta of 100 degrees Celsius.

As it turns out for the most part the heat capacity, that is a material’s ability to absorb heat for a change in temperature of a certain mass of solid or liquid, is about the same (give it a factor of two) for the various practical solids and liquids that can function in heat storage. For commercial purposes, only a eutectic mix of sodium nitrate (NaNO3) and potassium nitrate (KNO3) at the ratio 60% and 40%, respectively, also known as “Solar Salt” is used as a molten salt storage medium. Therefore, the following discussion, using the example of molten salt, of how good a “battery” thermal storage is, is generally useful.

So, how “good” a battery is a molten salt system? We will let the following examples define what we mean by “good”.

One interesting question, for example, is to ask how much energy can be stored in a volume of molten salt when the delta between its lower (e.g. 275 degrees Celsius) and upper temperature (e.g. 550 degrees Celsius) is 275 degrees Celsius. (This is a typical temperature difference for molten salt storage in a power tower.)

As turns for that delta and temperature range a cubic meter of molten salt holds 677 Mega Joule of thermal energy. If we assume a round-trip efficiency of 95% for the storage and a thermal-to-electric energy conversion of 40% then this is equivalent to storing 75 kilowatt-hours of electricity. A cubic meter is about the volume underneath your desk and given a per capita average daily energy consumption of 37 kWh in the U.S. (2010) this volume could store twice the electric energy average Americans use in a day (for everything from making cars to lighting homes). This seems pretty “good”.

Now, let’s assume that we opted to use pump hydro to store this same amount of electric energy and let's further assume that the pump hydro plant has a round trip efficiency of 80%. How high would we have to pump water (of the same volume as the molten salt) to store the same amount of electric energy? Before we give you the answer so that you have some time to work your own answer —which will likely challenge your intuition—let’s compare the aforementioned molten salt storage system not to pumped hydro but to an electric battery.

For this, we take the example of a lithium-ion battery from one of my old Macintosh laptops (from the time when Mac laptops still had removable batteries). This battery, when new, holds 60 Watt-hours of electric energy and measures 220 cubic centimeters in volume. Lithium-ion batteries pack more electrical energy per volume than pretty much any other battery. Thus compared to such a state-of-the-art battery, how much volume would a molten salt system need? As it turns out the thermos bottle on my desk holds 1.1 quarts or about 1 liter. If this thermos bottle were a miniature molten salt storage system it would hold in its volume 78 Watt-hours of electric energy. In this simplistic comparison, my molten salt system only needs about 4 times more volume than some of the world’s best batteries. Let’s make it a 1:5 volume ratio, because there is a lot of packaging for such a small laptop battery and we will need some plant equipment to turn the molten salt back into electricity. Nonetheless, even on the metric of energy density compared to a lithium-ion battery a molten salt system is pretty good again.

Now that same Macintosh laptop battery that I used for this example barely holds any charge today. I treated it well and did all the full discharge-and-recharge restorative cycles Apple recommended, but it is pretty much exhausted after 5 years and likely less than 1,000 charge-and-discharge cycles. Compared to that what would the degradation of a thermal storage system be after 1,000 cycles? The answer is none. This is because we simply put heat in and take heat out. It is for the same reason that grandpa’s thermos bottle still makes it perfectly good for day of fishing. Thermal storage for all practical purpose will last forever.

So coming back to the earlier question. How high did we have to pump water of the same volume to store the same amount of energy as in our thermal storage system? As it turns out to that number is about 33 kilometers. Only Felix Baumgartner rises higher (by 6 kilometers).

It is self-evident that water is precious in a desert, while there are plenty of rocks and lots of sand. Indeed, it as been proposed to use a reversible lift as an alternative to pumped hydro for storing solar energy in a desert environment. However, given what we learned about lifting water extends to rocks or sand (albeit they are heavier than water which helps) and thus even against that “dry” gravitational storage solution a molten salt storage system seems to be the better solution.

In summary, thermal energy is a formidable storage medium for energy. With some exaggeration and simplification we can say that its cycle efficiency is nearly 100% (primarily because it only stores and extracts heat), it is only 5 x more voluminous than electric batteries, and it essentially lasts forever. That's all very good.