Since my conclusions are diametrically opposed to solar power orthodoxy, this article is broken into four discrete sections:
- Part-I offers a high-level overview of hourly and seasonal power demand in the US;
- Part-II integrates 400,000 MW of hypothetical PV solar into the national system;
- Part-III uses standard metrics to analyze changes in total electric power costs; and
- Part-IV discusses risks for stock market investors.
This is a blog article, not an exhaustive academic study. The essential facts are derived from independent sources that I consider credible; but I'll be happy to consider additional facts from other comparably credible sources. A copy of the Excel workbook I used in the development of this article has been uploaded to my Dropbox (click here). It is my sincere hope that thoughtful readers will help clarify the issues with fact-centric commentary.
Part-I - Overview of US Power Demand
The US demands several hundred thousand megawatt hours of electricity per hour; 24 hours a day, 365 days a year. My first graph is based on hourly electric system operating data from the US Energy Information Administration, or EIA. For the sake of simplicity, the graph uses an average of the first day and the middle day of each calendar quarter as typical for the season and shows how hourly power demand changes from season to season.
In this graph, peak power demand for the year is 636,300 MW at 6 PM in the summer quarter. If you're responsible for planning national electric power availability, you'll want to make sure that your generating fleet can satisfy that peak annual demand with a 10% to 15% margin of safety. For my Excel model, I've used a 10% margin of safety, which results in a total national capacity requirement of 700,000 MW.
When you total the hourly power consumption data for each of the four seasons, the daily totals range from 9.2 million MWh in the spring to 12.8 million MWh in the summer, and average 10.4 million MWh for all four seasons. Since 700,000 MW of capacity can generate 16.8 million MWh in 24 hours, the capacity utilization rate in my basic model is 64.2%.
Part-II - Integrate 400,000 MW of Hypothetical PV Solar
My next graph modifies the Part-I demand data by integrating 100,000 MW of hypothetical PV solar in each of four time zones and subtracting the hypothetical solar power production from the actual hourly demand data. The Part II model represents a best-case rather than a likely-case scenario because it does not make allowances for weather-related variability or seasonal fluctuations in power output.
The first critical takeaway in this graph is that the peak conventional power requirement of 602,000 MW at 8 pm in the summer quarter is not significantly lower than the Part-I peak power requirement of 636,300 MW at 6 pm in the summer quarter. While 400,000 MW of new PV solar capacity can slash the power required from conventional plants during the middle of the work day, it can't reduce power required from conventional plants during late afternoon and evening hours when sunlight is weakening and societal power demand is increasing.
In my view the Part-I and Part-II graphs represent conclusive proof of the proposition that "PV solar is not an "alternative" to anything because regardless of penetration rates, industrialized societies need a complete fleet of conventional power plants to satisfy peak annual demand."
In all four seasons, the daily solar power contribution is 2.9 MW, which works out to a 30% capacity utilization rate for the PV solar component. When you total the hourly conventional power consumption figures for each of the four seasons, the daily totals range from 6.3 million MWh in the spring to 9.9 million MWh in the summer and average 7.9 million MWh for all four seasons. When you add 400,000 MW of PV solar to 700,000 of conventional capacity and run the numbers, the capacity utilization rate for conventional power plants falls to 47.0%.
Part-III - Conduct Cost Comparisons of Part-I and Part-II Scenarios
This is where the analysis gets fascinating and my conclusions diverge sharply from solar power orthodoxy. To minimize complexity, I've made four "desert island" assumptions.
- First, I assume that combined cycle natural gas, today's cheapest conventional power alternative, is a reasonable proxy for all conventional power technologies.
- Second, I assume that utility-scale thin film PV solar, today's cheapest renewable power alternative, is a reasonable proxy for all renewable power technologies.
- Third, I assume that capital and operating costs for existing power plants are fairly comparable with expected capital and operating costs for new power plants.
- Fourth, I assume that fixed costs increase proportionally as capacity utilization falls.
I believe each of these desert island assumptions is reasonable and to the extent that the assumptions introduce analytical errors, the errors favor PV solar.
"Levelized cost of electricity," or LCOE, is a standard metric that's frequently used to compare the costs of different electric generation technologies with similar duty cycles and operating characteristics. LCOE is usually expressed in dollars per MWh and represents the net present value of the average unit cost of electricity generated during a power plant's useful life. Data inputs for LCOE calculations include capital costs, fuel costs, fixed and variable operation and maintenance costs, financing costs and an assumed utilization rate.
Grid parity is typically defined as the point where the LCOE of renewable power is less than or equal to the LCOE of conventional power.
Every fall, Lazard publishes a "Levelized Cost of Energy Analysis," an in-depth comparison of alternative energy and conventional energy costs that's made available to the public for general informational purposes. Version 9.0, published in November 2015, includes granular data on 11 alternative energy technologies, seven conventional technologies and energy efficiency.
The following table summarizes the key LCOE components for combined cycle natural gas and utility-scale thin film PV solar. The first two columns are the values reported by Lazard in its LCOE analysis. I calculated values for the third and fourth columns to adjust Lazard's LCOE numbers to the capacity utilization rates in my Part-I and Part-II models.
If one only considers LCOE values, utility-scale thin film PV solar with a 30% utilization rate is 8% cheaper than combined cycle gas with a 64.2% utilization rate. The problem is that PV solar can't generate power for more than 6 to 8 hours a day while combined cycle gas can generate power 24 hours a day. Since the two technologies can never have "similar duty cycles and operating characteristics," a simple LCOE comparison leads to a wildly inaccurate conclusion.
As an industrialized society, our only choices are (1) conventional power with renewables or (2) conventional power without renewables. There is no scenario where we can eliminate the conventional power and rely solely on renewables.
When a renewable technology cannot completely displace a conventional technology, the total societal cost of renewables equals the LCOE of the renewable technology plus the unavoidable costs of the sidelined conventional technology. Using numbers from the table, the total societal cost of PV Solar is $52.22 per MWh for the solar facility plus $30.22 per MWh in unavoidable fixed costs for a sidelined combined cycle natural gas plant, or a total of $82.44 per MWh.
My next table uses the daily power production data from Parts I and II and the calculated LCOE values set forth above to compute daily societal energy costs under the Part-I scenario where all power production comes from combined cycle natural gas and the Part-II scenario where PV solar contributes as much as it can and combined cycle natural gas does the rest.
I suspect that most readers will be shocked to see that adding a duplicative layer of PV solar to a robust conventional generating infrastructure could increase electricity costs by 17% across the board and increase the national power bill by $38 billion a year, but those are the numbers.
Overall, I think today's analytical exercise demonstrates:
- Since building a conventional power infrastructure and a duplicative renewable power overlay will cost two or three times more, the capital inefficiencies are staggering.
- Since renewables integration shifts power production from conventional assets to renewable assets without reducing the need for conventional assets, it reduces the capacity utilization rates of the conventional assets and increases their LCOE by spreading fixed costs over fewer units of production.
- While renewables integration saves fuel to the extent that renewable power displaces conventional power, it does not reduce the essential capital, financing and other fixed costs of the conventional power infrastructure.
- Even if PV solar hardware was free, the economics would be marginal in industrialized economies. As long as the PV solar hardware has a cost, Part-II style integration of PV solar and conventional power generation is a sucker's bet that increases everybody's power costs.
- The inefficiencies and diseconomies discussed in this article apply with equal force to wind power and scale ratably as renewable power capacity is added to a conventional core, beginning with the first MW of renewable capacity.
While a 17% across-the-board increase in electricity costs may not seem all that onerous for readers who are worried about CO2 emissions, it should be noted that the 17% increase results from transitioning a mere 26.7% of power production to renewables.
Part-IV - Risks for Stock Market Investors
So far, I've focused exclusively on utility-scale thin film PV solar, the most cost-effective solar technology. The Lazard LCOE analysis actually includes data for the six solar technologies I've summarized in the following table.
Can anybody honestly consider these standalone LCOE values, allow for the unavoidable fixed costs of sidelined conventional power plants and tell me how solar is a good deal for society?
In the long run, I'm convinced the dismal economics of solar power will doom the technology to the scrap bin of history. In the short run, however, I believe politics and public perception will almost certainly outweigh the vulgar exigencies of objective truth. Heck, there's even a chance that solar advocates and political shills will convince the masses that small reductions in CO2 emissions from power plants justify huge increases in the national electric bill. I may not agree with that conclusion, but I'm the only guy I know who cherishes my opinion.
The following table identifies four publicly held players in the PV solar market and provides product class, market capitalization, book value and price-to-book ratios for each of them.
You'll note that First Solar (OTCPK:FLSR), SunPower (NASDAQ:SPWR) and Canadian Solar (NASDAQ:CSIQ) actually make solar cells and trade at modest discounts to book value while SolarCity (SCTY), a low-tech packager and installer of products made by others, trades at a 2.2x multiple of book. I believe these price-to-book disparities suggest that SolarCity is overvalued compared to its peers.
While I'm not a fan of Tesla Motors (NASDAQ:TSLA), I believe the planned merger of SolarCity into Tesla could be catastrophic for Tesla's stockholders because a significant ownership stake in Tesla will be issued to SolarCity stockholders who contribute no real value to the combined companies. The flip side of that argument is that there's no harm in exchanging one over-valued stock for another. For what it's worth, the LCOE values in this article all come from Lazard, the financial advisory firm that wrote SolarCity's fairness opinion for the Tesla merger.
I'm increasingly convinced that solar power promoters are the snake oil salesmen of the new millennium. With the chutzpah of P. T. Barnum, they pontificate on the sustainability and alleged environmental benefits of PV solar and other renewables without acknowledging the technical limitations, economic costs and ancillary societal impacts. The result is a misinformed public that supports and even subsidizes the incremental degradation of the power grid, a critical resource that makes our very way of life possible - a common resource that's no less important than clean air and water.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.