Solar power after sunset

[dragon]

Nice how they heat the salt to continue to generate full power for up to 6 hours after sunset.

solana generating station can supply power for up to 6 hours after dark
Phoenix – Unlike other solar-powered electrical plants, the new Solana Generating Station keeps the sun’s energy working after dark to produce electricity for APS customers across Arizona.

Solana represents an important technological advance in solar energy production compared to the more-common photovoltaic technology, which needs direct sunlight in order to produce electricity. The three-square-mile facility near Gila Bend uses concentrated solar power (CSP) technology and thermal energy storage to capture the sun’s heat to generate clean, renewable electricity.

This technology enables Solana to produce electricity at full capacity for up to six hours after sunset, including the early evening hours when customer demand for power typically peaks in Arizona.

“Solana is a monumental step forward in solar energy production,” said Don Brandt, APS President and Chief Executive Officer. “Solana delivers important value to APS customers by generating power when the sun isn’t shining. It also increases our solar energy portfolio by nearly 50 percent. This provides a huge boost toward our goal to make Arizona the solar capital of America.”

Solana is one of the largest power plants of its kind in the world with a capacity of 280 megawatts, enough to provide energy for 70,000 Arizona homes – a city the size of Yuma. APS is purchasing 100 percent of the output from Solana, which was constructed and is owned by Abengoa Solar.

The plant’s CSP technology produces electricity by collecting the sun’s heat to create steam that turns conventional turbines. The process begins with 2,700 parabolic trough mirrors, which follow the sun to focus its heat on a pipe containing a heat transfer fluid. This fluid, a synthetic oil, can reach a temperature of 735 degrees Fahrenheit. The heat transfer fluid then flows to steam boilers, where it heats water to create steam. The steam drives two 140-megawatt turbines to produce electricity, much like a traditional power plant.

What separates Solana from other solar power plants is the ability to store the heat from the sun up to six hours for electrical production at night. In addition to creating steam, the heat transfer fluid is used to heat molten salt in tanks adjacent to the steam boilers. The thermal energy storage system includes six pairs of hot and cold tanks with a capacity of 125,000 metric tons of salt, and the molten salt is kept at a minimum temperature of 530 degrees Fahrenheit.
When the sun goes down, the heat transfer fluid can be heated by the molten salt to create steam by running it through the tanks instead of the field of parabolic mirrors.

“With Solana’s substantial thermal heat storage capacity, we can manage electrical output from the plant much more effectively than from other solar power sources,” said Pat Dinkel, APS Vice President of Resource Management. “With photovoltaic technology, generated electricity needs to be used immediately or it’s lost. Solana’s technology extends the use of solar energy to produce power whenever our customers need it most, including evenings.”

With the addition of Solana, APS will have 750 megawatts of solar power on its system by the end of 2013, enough to serve 185,000 Arizona customers.

APS, Arizona’s largest and longest-serving electricity utility, serves more than 1.1 million customers in 11 of the state’s 15 counties. With headquarters in Phoenix, APS is the principal subsidiary of Pinnacle West Capital Corp. (NYSE: PNW).​

link

[Borgholio]

Adding salt to the existing solar / thermal setup is a nice touch. It's a good thing that it works so well. I wonder, however, how long it takes to heat the salt in the morning? If it takes 6 hours or so to heat it fully, then it might be that this sort of plant is best used to help with power usage after dark, as opposed to in the morning when it's still heating up.

[krakonfour]

Simple calculation tells me that 125000 metric tons of molten salt at 549K takes 37.4*10^12 J to get to 673K, with 2400J/kg.K specific heat capacity.
Since the station produces 280MW, it takes 37 hours to reach this temperature. Unlikely.

Now, if we assume that the station only heats up the mixture to provide 6hours of 280MW production, then the temperature during the day is 0.006K higher than at night.

I think the 125kT of molten salt figure is wrong.

[Simon_Jester]

Simple calculation tells me that 125000 metric tons of molten salt at 549K takes 37.4*10^12 J to get to 673K, with 2400J/kg.K specific heat capacity.
Since the station produces 280MW, it takes 37 hours to reach this temperature. Unlikely.

Now, if we assume that the station only heats up the mixture to provide 6hours of 280MW production, then the temperature during the day is 0.006K higher than at night.

I think the 125kT of molten salt figure is wrong.

Several possibilities you're missing:
1) The plant produces 280 MW of electrical output; this does not translate into saying that it applies 280 MW of power to the process of heating the molten salt. Those numbers are only vaguely related.
2) 125 thousand tons is the capacity of the tanks; it is not a given that all the tanks are being used to full capacity at all times.
3) I don't know where you got your 673 K figure from, if you got it from "735 degrees Fahrenheit," then you've forgotten that that is the maximum operating temperature of the heat transfer fluid, not necessarily any temperature ever reached by the molten salt.

[krakonfour]

Several possibilities you're missing:
1) The plant produces 280 MW of electrical output; this does not translate into saying that it applies 280 MW of power to the process of heating the molten salt. Those numbers are only vaguely related.
2) 125 thousand tons is the capacity of the tanks; it is not a given that all the tanks are being used to full capacity at all times.
3) I don't know where you got your 673 K figure from, if you got it from "735 degrees Fahrenheit," then you've forgotten that that is the maximum operating temperature of the heat transfer fluid, not necessarily any temperature ever reached by the molten salt.

Most solar arrays function in the 100-200W/m2 range. 280MW requires 16km squared already. If it is going to produce 280MW on top of heating the salt, just how big is this plant?

125kT is massive. If they are not using it all, and will never have enough output to heat it up within a day, then what's the point of having so much?!

Thermal efficiency rises with an increased temperature difference, because the heat transfer rate is higher. It is only natural to assume that the operators would want to have the hottest salt possible to increase their production efficiency.

If I had 125kT of molten salt, and could only heat it up by 0.001K by hour, then I'd get rid of most of it and keep only the amount I can realistically heat up to the maximal operating temperature.

[energiewende]

Solar concentrator is a lot more sensible than PV and molten salt is probably about the least-bad storage option. This is probably about the best set-up possible with current technology. I would be very interested to see how the levelised cost of electricity compares to conventional (fossil and nuclear) plants. 125kT of salt (heavier than a supercarrier) isn't going to come cheap.

If I had 125kT of molten salt, and could only heat it up by 0.001K by hour, then I'd get rid of most of it and keep only the amount I can realistically heat up to the maximal operating temperature.

This is amount is also determined by engineering considerations as well as physical limits. With such a huge amount of salt most of it is unlikely to be directly heated. They will have to wait as heat diffuses into the tank, or alternatively have a much greater theoretical storage capacity than will ever be used so that the outermost surfaces don't exceed the operating limits.

[energiewende]

Also, I don't agree with your numbers. 280MW gives 2.80E8 W * 3.6E3 s = 1E12 J/hour.

The salt, using your figures, stores 1.25E8 kg * 2.4E3 J/K.kg = 3E11 J/K. So the salt's temperature increases by about 3K per hour, which sounds sensible.

I think you forgot to multiply by 3,600s.

[krakonfour]

My mistake, energiewende. I hadn't considered the system's operating limits, but the point still stands: higher temperatures = higher efficiencies, and anything less than the maximum acceptable is going to cost them.
3.6K per hour is still only 40 or so Kelvin over the course of a day. They could have gone with boiling water at those temperature ranges and massively saved on infrastructure and insulation.

[Magis]

You guys aren't thinking about this correctly.

The mass of the salt is irrelevant to the energy balance. The reason why more salt mass is always better is because a large mass lowers the variation of temperature of the salt during the 24-hour cycle, which is better for the boiling hydraulics and is also necessary to prevent phase changes in the salt.

Anyway, 280MW is the electrical output of the plant. At a 35% thermodynamic efficiency, the thermal power of the plant will be around 800MW input to the turbine and even more than that input to the salt, let's call it an even 1000MW of solar input.

The mass of salt is so large because it's desirable for its temperature to remain as constant as possible during a 24-hour cycle. When the article quotes a 'minimum' temperature, that is merely the temperature necessary to keep the salt in the liquid phase; it won't be designed to ever reach this temperature during operation. But given a max. temp of 673 K and a min. temp of 550 K, then during the night a max. of 123 K can be removed from the salt.

The specific heat of the salt depends on its chemistry; the article doesn't specify which salt will be used and it isn't exactly table salt. Assuming it's potassium nitrate, with a specific heat of 0.95 J/g/C, then the salt's temperature will decrease by 24.25 K per hour if it produces at full 800MW-th capacity during the night (which it wouldn't).

In reality, the electrical output would vary through the day and night based on demand on the grid and the temperature of the salt, and would be controlled by the operating staff in anticipation of near-term production requirements and the safe operating limits of the facility. The quoted 280 MW-e capacity is simply the capacity of the coupled steam turbines/generator; there's no reason to think the plant would ever deliver this much to the grid except perhaps on exceptionally sunny days when the salt is near its maximum operating temperature.

[Irbis]

The mass of the salt is irrelevant to the energy balance. The reason why more salt mass is always better is because a large mass lowers the variation of temperature of the salt during the 24-hour cycle, which is better for the boiling hydraulics and is also necessary to prevent phase changes in the salt.

In fact, yes, keeping it very hot is a must, otherwise bane of all solar plants, event known as 'some clouds' would cause the salt to fossilize and permanently wreck the shit out of the whole installation.

Also, I still love how they combined desert, colossal installation, probably the most cloud-less sunny landscape in US, and the combined effort still only produces 280 megawatts. Hell, their entire in state solar portfolio powers less than my small city of residence

For comparison, ancient Soviet lead salt-cooled power plant installed in Lira-class ships could output 155 megawatts, in 1960s. Not from dozens of square kilometres, from cube less than 7 metres on each side. Progress!