“The 2200 million dollars per 120 million watts represents an 18 dollars per watt ($/W) investment. By way of comparison, another nonpolluting source of electricity, nuclear power plant, the Millstone reactor No. 2 in Connecticut, operating at 880 MW since 1975, cost 0.5 $/W, making Ivanpah 36 times more expensive (inflation excluded).”
The newest and largest solar power plant in Mojave Desert has completed its second year of operation. The news is not good–and, in fact, very bad.
This type of power plant generates electricity by concentrating sun rays on a “boiler” making high pressure, high temperature steam that drives a turbine generator. The generator machinery is identical to that common in the fossil-fuel and nuclear plants, except that it has provisions for the every-day shut-down at dusk and restart with the rising sun.
Usually called Concentrating Solar Power (CSP) plants, that term was coined to distinguish them from the Solar Thermal Power (STP) intended for direct heating such as in warming swimming pool water.
Concentrating Solar Power
CSP plants were promoted by the Department of Energy from the time DOE was created under President Carter. Several were built over the decades and none performed adequately. One burned up in 1986 and was rebuilt, enlarged, DOE arguing that the plants have to be big to take advantage of economy of scale.
Following that logic, this latest 392 MW (name-plate) giant was built on 13 km2 of land in Mojave Desert at a cost of 2.2 billion dollars. It generated a disappointing .4 billion kWh thus producing at an average rate of 46 MW the first year.
[Note: It is typical for renewable energy projects to show different units for input, rated output and actual output. This practice makes performance and efficiency comparisons cumbersome, and is therefore not pursued, allowing misinformation to flourish. In the above paragraph, the former value is in “W” but the latter in “Wh.” The author wishes that such reporting use the same unit (W, as with the 392 MW and 46 MW above) or it states, as an example, “…. the plant has been producing 12 % of its name-plate power.”]
Dissecting the Failure
Rather than focusing on the poor performance of the Ivanpah, the following text and numbers show how the planned-for performance, capital and operational expenses and earnings compare with another power plant. The predicted Capacity Factor of 31 percent indicates a 120 MW expected actual average output. That is the number worked with below.
The 2200 million dollars per 120 million watts represents an 18 dollars per watt ($/W) investment. By way of comparison, another nonpolluting source of electricity, nuclear power plant, the Millstone reactor No. 2 in Connecticut, operating at 880 MW since 1975, cost 0.5 $/W, making Ivanpah is thus 36 times more expensive (inflation excluded).
With about 1000 employees receiving salary and benefits, the annual outlay for that alone is roughly $100 million. Selling the annual 3.8 EJ at the projected 0.028 $/MJ yields $106 million. Ouch — only $6 million is left for other expenses, notably for natural gas whose burning produces 8 percent or more of the total output.
For comparison again, the Millstone nuclear plant complex employs also about 1000, and its two reactors have been producing 1870 MW actual electrical output. Assuming the same salaries, benefits, and the electricity selling price, the operating expense is 15 times higher at Ivanpah.
Note that the above two outlays are 35 and 15 times higher, and that this huge discrepancy exists in an industry where a difference of a few percent means the difference between success and bankruptcy. The magnitude of the discrepancy hints also at the reason why the “free” solar electricity is so expensive. 
As for the occupied land comparison, those 120 MW spread over 13 km2 represents 9.2 W/m2. In contrast, ground based nuclear plants produce some 2000 W/m2 thus utilizing the land area some 200 times more effectively. And they can be erected in any climate and in proximity to users.
It should be pointed out again, that these dreadfully unfavorable ratios apply to a plant delivering its planned-for output. In the first two years the plant delivered only 1/3rd of the assumed 100 percent. The already “bad” ratios are actually three times worse than those calculated here.
We must be either very rich or very ignorant to be building power sources of the type that produce power we cannot afford. If the purpose of the CSPs is to cut CO2 emissions, that expectation is unrealistic. The construction, operating, maintaining and eventually dismantling this plant will at best match the amount of CO2 claimed to be saved in non-burning fossil-fuels for that relatively small amount of electricity. And producing intermittent electricity causes CO2 generation elsewhere.
And there are other items to consider with respect to the net output and CO2. It was not apparent from the description whether the electric output was measured at the outlet from the generators (or transformers) or whether it was that output minus the electrical demand on the grid for electricity consumed from dusk to dawn in the plant, such as for lighting, air-conditioning, washing mirrors, water pumping, restarting machinery, etc.
The plant also burns gasoline, diesel fuel and, prodigiously, natural gas. These considerations should be included in the performance data for the net electricity delivered and CO2 saved. Perhaps they will be available someday.
Concerning the DOE covering the invested billions, it was not just tax- and rate-payers who paid. Google, among others, invested millions from its “green” fund, the same Google that abandoned its own PV solar facility and related R&D last year (2014).
 In the summer of 2015 the wholesale price was 2.0 cents per kWh in New England, a region with the highest rate in the US at this time. And those two cents cover also taxes, debt-servicing, dividend payments, etc. and provide profit.
Note: You might be curious to know how efficient the Ivanpah plant is in utilizing solar insolation (irradiation). With the annual average of 180 W/m2 in the sun-rays’ heat in that location spread over the 2.5 km2 area of just the heliostat mirrors, the input power is 450 MW. Based on the above 120 MW output, the efficiency is then almost 27 percent. PV plants by comparison utilize 220 W/m2 in that location and so would theoretically procure more power at the same efficiency level. They also occupy a fraction of the ground area CPS necessitates for equal output because the mirrors must be surrounded by open area on all sides to avoid shade when tilting to follow the sun.