Parts of this article previously published at Real Clear Energy.
Solar Energy Is Not Competitive with Fossil Fuels
Norman Rogers
In a typical electrical grid environment, electricity generated by a utility-scale solar farm costs about seven times more than electricity from a natural gas generating plant. Yet many people think solar electricity is a breakthrough. No, it is a wasteful boondoggle.
This article may be criticized because it disagrees with the claims of solar promoters. Promoters get the most press, but there is a large community of solar skeptics. The promoters minimize or ignore solar subsidies. They neglect the need for time-shifting batteries. They make spurious claims implying that fossil fuel subsidies are equivalent to solar subsidies. This article makes the case that solar mandates are a subsidy, something that is ignored by the promoters. The promoters make dubious claims that solar costs will continue to rapidly decline. A powerful fact is that electricity rates are stratospheric in Germany and California, states that are far down the solar road. Plenty of homeowners in California pay over 50 cents per kilowatt hour.
A solar generating station provides intermittent power. The electricity stops if clouds obscure the sun, or if it is night. Solar must be backed up by a reliable source of power that fills in when needed. Commonly, the backup is a plant powered by natural gas.
Solar does not require fuel but the construction cost is very high – about ten times more than a natural gas plant for the same amount of electricity. The high capital cost is why solar electricity is seven times more expensive than natural gas electricity. The high cost is masked by government subsidies for solar. The public pays for both the electricity and the subsidies.
The size of power plants is often expressed by the maximum number of megawatts of power they are capable of delivering, called the nameplate rating. The average power delivery is less. For a natural gas plant it is possible to generate an average power of near 100 percent of the nameplate. The plant can run continuously except for maintenance pauses. But for a solar plant the maximum capacity factor is about 25 percent in sunny areas. The low capacity factor of solar plants, that mainly work in the middle of the day, is the reason solar is so expensive.
A lot of bad energy accounting comes from the misuse of the metric, levelized cost of energy, abbreviated as LCOE. LCOE is an unstable measure. For example, the LCOE for natural gas-generating plants can be completely different for nearly identical plants. The LCOE depends not only on the plant's fundamental properties but also on how the plant is operated. Most writers on solar energy compare the LCOE of fossil fuel electricity to the LCOE of solar electricity. This is a mistake because LCOE for a gas or coal plant is unstable.
LCOE estimates how much a megawatt hour of electricity costs. The cost has three elements, the capital cost of building the plant, the variable or fuel cost, and the operating/maintenance cost. The operating/maintenance cost is minor and is mostly ignored in the discussion here. The capital cost allocated to each megawatt hour can vary wildly. If the plant duty cycle or output is half what the plant is capable of generating, then the capital cost allocated to each megawatt hour will be doubled, because the same capital expense is allocated over half as many megawatt hours.
Solar generating plants, due to power purchase agreements, are operated at full capacity, allowed by the weather, the daily solar cycle, and the ability of the local grid to accept the power. Over time the relation between megawatt hours of electricity generated and capital investment is predictable. LCOE makes more sense for solar plants than for fossil fuel plants, but still requires careful application.
Rate of Return on Capital Investment
The solar developer that provides the capital to build a solar farm expects a rate of return on his investment, often between 8 and 12 percent, depending on perceived risk and current interest rates. The sale of electricity provides the money to pay for operating/maintenance, and the return on investment over the assumed life of the plant.
Estimating the necessary rate of return for a solar farm when there are no government subsidies or mandates is somewhat speculative. Utility-scale solar electricity would hardly exist if there weren’t subsidies and mandates. We estimate a 12 percent rate of return would be needed in the absence of government meddling. Solar would not be remotely competitive in normal circumstances due to the considerable capital cost and the intermittent delivery of power. One can imagine niche applications where it could be competitive. For example, in remote communities that use very expensive medium-scale diesel generation. In such a small-scale, niche application, at least a 12 percent rate of return would be needed.
In utility-scale systems with government mandates that lead to long-term power purchase agreements, an 8 percent rate of return is reasonable. Such a solar farm has bond-like characteristics due to the utility guarantee. Utilities can go bankrupt but they can’t go out of business because the lights can’t go out.
Local Operating Grids
Although there are three large, synchronous electric grids in the United States, grid management occurs in smaller local grids, such as a single state or a regional area. Electricity entering or exiting such a local grid is considered an import or export.
The generating plants available to a local grid operator are optimally dispatched in order of the variable cost of the electricity, lowest variable cost first. The capital cost associated with each available generating plant is a sunk cost and cannot be changed by anything the grid operator does.
If the operator has nuclear, coal, and natural gas plants available, the nuclear plant wlll be dispatched first because nuclear fuel, the variable cost, is very cheap. The capital cost of nuclear is huge. The next plant type dispatched will be coal or gas depending on the cost of the fuel per megawatt hour. Favoring low variable cost generation minimizes money spent to generate the electricity. The capital investment has already been spent and can’t be retroactively changed. Capital cost is ignored when deciding the order in which plants are dispatched.
Role of Solar Added to an Existing Grid
Solar electricity is always added to a preexisting operating grid. Electric grids have been around far longer than utility-scale solar electricity. The economics of solar are such that very few utilities would use solar except that many states mandate the purchase of ever-increasing amounts of renewable electricity. In sunny areas, solar will be a major part of the renewable electricity.
There are many types of renewable electricity, but wind and solar are dominant because they are scalable, mature technologies. The exact definition of renewable energy is confusing and inconsistent, but for practical purposes, it means either wind or solar energy.
State Renewable portfolio laws force utilities to purchase solar or wind electricity. Typically, these laws require that renewable electricity comprise a certain percentage of total generation by a future date. For example, Nevada has a mandate of 50 percent renewable electricity by 2030. It will be mostly solar because Nevada has better sunshine than wind. About 30 states have renewable portfolio laws.
One justification for renewable portfolio laws is the supposition that we are in danger of running out of fossil fuels. But U.S. reserves of coal and natural gas are sufficient to last for hundreds of years. Trying to protect the nation from a hypothetical problem set to surface in hundreds of years is a dubious quest. If the Sierra Club were around when Columbus discovered America, it might have demanded that he produce a plan for interstate horse trails.
Another justification is that fossil fuel plants emit dangerous pollution. But modern natural gas or coal plants are good neighbors. For example, Arkansas's John W. Turk coal generating plant removes 90-99 percent of pollutants. Natural gas plants are remarkably free of pollutants. Atmospheric pollution from fossil fuel generating plants is a scare story.
The usual method of obtaining solar electricity is to have a utility sign a power purchase agreement with an independent solar energy developer. For example, the agreement might specify that the developer will supply 500 megawatts of average power at $40 per megawatt hour for the next 20 years. The solar developer will then build and operate a solar farm to supply the electricity. The real cost of the solar electricity is much higher than $40 per megawatt hour because federal and state governments heavily subsidize the solar developer.
A government mandate to purchase renewable electricity gives wind and solar developers enhanced negotiating power. The utility is forced to buy renewable power, even if the renewable power is exorbitantly expensive. The developers exercise their negotiating power by demanding long-term contracts in the 20- to 25-year range. Since these projects are competitively bid, the solar developers compete with each other on how low the rate of return on their investment will be. Given that there are a limited number of companies with the expertise and resources to build billion-dollar solar farms, the competition may be gentlemanly competition between the members of an oligopoly.
The solar developer has a relatively simple calculation to decide how much per megawatt hour to bid. Solar does not use fuel, and the utility is obligated by contract to purchase all the electricity generated. The lifetime of the solar farm can be taken to match the term of the power purchase agreement, for example, 20 years. Although a solar farm can last more than 20 years, 20 years is safe because the developer is not counting on an extended life beyond 20 years. The construction cost can be obtained from tables published by the National Renewable Energy Laboratory, although it is likely that an experienced solar developer knows the cost better than the laboratory.
The long-term power purchase agreement is a subsidy because it removes risk by guaranteeing a market for 20-years or more. That lowers the return on investment needed and thus lowers the price of electricity. Lowering the rate of return from 12 percent to 8 percent is a one third subsidy. The utility pays for the subsidy because it is giving a guarantee of future purchase of electricity. Such guarantees are nearly as good as cash.
Local grids where solar electricity exceeds about 15 percent of total generation will suffer from solar congestion unless the solar farm is equipped with time-shifting batteries that increase the cost of the farm by about a third. Solar congestion is caused because solar energy has a large midday peak that can exceed the grid's ability to accept the power surge. The solution is shaving the peak with batteries and releasing the stored power later in the day.
The Proper Way to Compare the Cost of Solar Electricity to Natural Gas Electricity
To compare the cost of natural gas generation with solar generation the marginal cost of gas electricity should be compared to the total cost of solar electricity.
When solar electricity is driven into a local grid, natural gas generation must be reduced to make room for the solar electricity. Reducing natural gas generation will save fuel but it will not save capital cost because the capital cost is sunk and cannot be changed by the grid operator. The fuel cost in a modern combined cycle natural gas plant is approximately $20 per megawatt hour. The cost of the solar electricity is set by contract with the solar developer. Without federal and state subsidies the solar cost would be about $150 per megawatt hour. Because there are subsidies, the utility pays about $40 per megawatt hour, and the remaining $110 per megawatt hour is paid by the taxpayers and electricity consumers that finance the subsidies. The difference between the $40 paid by the utility and the $20 savings from reduced natural gas generation requires an additional subsidy that will be paid in the form of increased electricity rates.
This may seem like trickery in the cost comparison because I’m not counting capital cost for the gas electricity, but I am counting it for the solar electricity. It is possible to compute the capital cost of solar per megawatt hour because solar always runs at the maximum output, per the power purchase agreement, limited by weather conditions. Thus, the solar output is proportional to the capital cost excepting random weather variations that average out over a period of time. The solar developer can accurately calculate what he must charge for electricity to recover his investment plus a profit.
The scenario above is real. Every megawatt hour of gas electricity replaced by solar saves $20 worth of gas but would cost, without subsidies, $150 for the solar electricity. Since the gas plant was already present before solar was added, the solar does not change the arrangements for recovering the capital cost of the gas plant. If a regulated utility owns the gas plant as part of its rate base, the depreciation of the gas plant is charged to the electricity customers. Adding solar won’t change that arrangement. The utility does not have to concern itself with the details of estimating the cost of solar. It only has to pay the invoices from the solar developer that operates the solar farm.
The Subsidies
Utility-scale solar is about 85 percent subsidized. State renewable portfolio laws foster long-term power purchase agreements. Those laws are a subsidy because they lower the rate of return on investment that the solar developer needs to be profitable. Because of renewable portfolio mandates, the developer can demand a 20-year contract to sell his electricity. That dramatically reduces business risk and lowers the rate of return needed from about 12 percent to about 8 percent—a one-third subsidy. The renewable portfolio law subsidy reduces the cost of solar electricity from about $150 per megawatt-hour to $100. The utility pays for the subsidy because it is locked into an unfavorable long-term contract.
Two federal subsidies, tax credits and tax equity financing provide an additional subsidy of approximately $60, lowering the cost to about $40 per megawatt hour. The last $20 subsidy, the difference between the $40 paid by the utility and the $20 worth of fuel saved, is paid by increasing the cost of electricity. These subsidies should not be taken as exact numbers but as a rough view of the subsidies.
Ancillary Cost of Solar
Introduction of solar electricity will create a need for new power lines. For example, in Nevada a $4 billion greenlink power line has been made necessary due to the introduction of large amounts of solar. Because solar peaks at midday and does not exist at night power line utilization may be low.
The Utility of Solar for Reducing CO2 Emissions
Substituting expensive solar electricity for cheap gas electricity reduces CO2 emissions, but the cost, based on the subsidy for solar, is about $400 per metric ton of CO2 emissions avoided. If nuclear is substituted for natural gas and solar, the cost of CO2 reduction is about $140 per metric ton, three times less.
Reducing CO2 emissions is an attempt to solve a hypothetical future climate change problem that might occur after all the present-day climate doom promoters are conveniently dead and thus immune to embarrassment.
Conclusion
Utility-scale solar electricity is a waste of money. No more plants should be built.