When Will Solar Achieve Grid Parity? We're Already There!

Includes: CSIQ, FSLR, STP, TSL
by: Jack Yetiv

I always chuckle when I read an article in the popular press — or a comment on this site — stating something like "Solar is too expensive and will never be a significant source of electric power." Invariably, these articles or comments are never documented to explain how the author reached this faulty opinion.

I believe that I will prove in this article that UNSUBSIDIZED solar is ALREADY at grid parity today against UNSUBSIDIZED "conventional" power sources. The word "parity" implies "equality," and therefore, the only fair comparison is one where ALL costs are taken into account.

As we all know, the primary sources of electric power in the US are coal, natural gas and nuclear, and it is to these three sources of power that I compare solar. We do obtain some power via hydroelectric sources and some biomass, wind and solar, but these are relatively small (but increasing quickly as to biomass, wind and solar). Here are some general comments about each of these power sources:


PROS — coal is widely available in the US and will continue to be available for decades to come; relatively inexpensive, although has doubled in price in the past year

CONS — the dirtiest source of power we have, with release of mercury, sulfur and nitrogen oxides, and a large amount of CO2 (double that of natural gas)


PROS — also very abundant in the US and Canada, used to be very cheap, but has also doubled in price in the past couple of years, much cleaner than coal, releases ½ as much CO2 as coal

CONS -- still puts out a lot of CO2, price likely to rise substantially going forward


PROS — no CO2 production, uranium fuel is relatively inexpensive (even though it has gone up ten-fold in the past few years, it constitutes a relatively low percentage of the cost of operating a nuclear power plant)

CONS — very high capital cost to build, long lead times to build (close to a decade versus 1-2 years for coal or nat gas power plants), although plants are reputed (and probably are) much safer than they were 10-20 years ago, nuclear plants nevertheless present some risk of nuclear contamination (whether through natural causes or terrorism; nothing is fail-safe) that other power sources do not; and the disposal problem of spent nuclear fuel has not been solved

With the foregoing in mind, let me now make the case that if you were deciding today on which kind of "power plant" (the reason for the quotes will become obvious below) you were going to build tomorrow—and you were willing to consider ALL COSTS and NO SUBSIDIES — you would build a solar "power plant" (NOTE: I am aware that wind may be even cheaper than solar, but that would be the topic for a complete other article that someone should write; my goal here is solely to compare solar to the other three power sources noted above; I will say, very briefly and generally, that the choice of solar versus wind is probably a function of the meteorological conditions at the site of the solar panel or wind turbine, and the cost advantage will depend on how much sun versus how much wind the location gets).

To begin the comparison, let us look at what unsubsidized solar costs today. Last quarter, First Solar (NASDAQ:FSLR) announced that its production cost was $1.12 per watt. I will guess that by the third or fourth quarter of this year (the soonest FSLR would provide panels for the imaginary "power plant" we're discussing), FSLR's cost would be well under a dollar. I believe that within a year or two, as polysilicon prices approach $50/kg and efficiencies routinely exceed 20%, other manufacturers will also achieve panel costs approaching $1/watt; obviously, disruptive technologies such as the one Nanosolar is rumored to have may drive the cost substantially under $1/watt). Many people estimate that balance-of-system costs would add another dollar, plus another dollar for installation and another for profit, for a total unsubsidized installed cost of $4/watt, or $4 million per megawatt. Let's add another dollar just to be conservative, and thus, a modest 275 megawatt solar "power plant" would cost a cool $1.4 billion to build.

$1.4 billion sounds like a big number, doesn't it?

Well, it does until you consider "FutureGen," an "almost-zero-emission" coal-fired power plant that was to be built in Illinois. The plant was designed to eliminate essentially all harmful emissions from the plant, including the capture and sequestration of the carbon dioxide. The plant was cancelled when estimated costs for a 275-megawatt plant increased to $1.8 billion. The background of the plant is interesting because the idea for this plant got off the ground ONLY because ¾ of the cost was going to be subsidized by the federal government. In other words, the private sector was only willing to put up $450 million to build the plant, while the taxpayers would foot the other $1.35 billion (it is true that foreign financial support was to offset a portion of DOE's cost-share, but that, of course, does not alter my thesis here which relates to total unsubsidized cost).

On Jan. 29, 2008, DOE pulled funding for the plant, essentially killing the project. Although some people argue that the funding was pulled for political reasons, it seems more logical that the cost of $6.55/watt was the real reason that this project was pulled.

The FutureGen project is clear proof that if we are going to compare apples-to-apples (ie, no harmful emissions from the coal-fired plant as is the case with solar power), coal-fired power plants are MORE expensive to build than solar is TODAY.

Four things make the above case even more compelling. First, it is not clear to me if 30 years of coal were included in the $1.8 billion price tag. I can assure you that I have included 30 years of "sunshine" in the solar project cost of $5/watt. If coal was not included, that would make the coal power plant even less competitive. Even if coal were included in the price tag, I would be interested to know what they estimated the price of coal would be in 10 years, 20 years or 30 years. I can easily tell you what the cost of the sunshine will be in the future (unless the government figures out how to tax us on "sunshine." LOL)

Second, the FutureGen project was a "demonstration project"—ie, we don't know if they could have even achieved the goals of near-zero-emissions for the $1.8 billion price tag. And of course, I don't know what "almost" zero emissions would actually be.

Third, it is likely that the cost of building coal, gas and nuclear power plants will INCREASE going forward, whereas it is almost guaranteed that the cost of building solar power "plants" will decrease.

Finally, it's not clear to me if distribution of the electricity was included in the $1.8 billion price tag. I suspect it was not. If it was not, you can add another hundred million (or even more) for distribution, depending of course on whether rights-of-way are available and what their cost is, distribution distance and power losses.

But for solar, in many cases, there will be very little if any "distribution" cost because solar power "plants" can be decentralized (thus, they are not really "power plants" at all—hence, the quotes in this article). Look at the 250-megawatt "power plant" that was announced on March 27, 2008, by Southern California Edison. SCE will lease 65 million square feet of roof area from owners of commercial buildings and install 1 megawatt per week (250 megawats over 5 years) of photovoltaic panels on those roofs at no cost to the property owners (on the contrary, these owners will collect rent for something they did not think had value—their roofs). The beauty of this plan is that it ALREADY uses existing "distribution" facilities by using the existing wires (that usually bring IN electricity) in reverse, sending power TO the grid when the solar panels are producing power.

Commenting on this project, Governor Schwarzenegger said: "These are the kinds of big ideas we need to meet California's long-term energy and climate change goals. I urge others to follow in their footsteps. If commercial buildings statewide partnered with utilities to put this solar technology on their rooftops, it would set off a huge wave of renewable energy growth."

John E. Bryson, Edison International chairman and CEO, said: "This project will turn two square miles of unused commercial rooftops into advanced solar generating stations. We hope to have the first solar rooftops in service by August. The sunlight power will be available to meet our largest challenge – peak load demands on the hottest days."

As it noted in its press release, SCE's renewable energy project "was prompted by recent advances in solar technology that reduce the cost of installed photovoltaic generation. When combined with the size of SCE's investment, the resulting costs per unit are projected to be half that of common photovoltaic installations in California."

Interestingly, SCE estimates the total project cost will be $875 million (in today's dollars), working out to $3.50/watt. This number is probably lower than my $5/watt estimate because I was trying to be conservative in my estimate (remember the extra dollar I added) and because anticipated price reductions during the 5-year life of this project will probably decrease the cost of the PV installation by 30-50%.

Much of what I said above regarding coal applies to natural gas-fired generation. Carbon capture and sequestration would probably push natural-gas power plant costs to the billion-dollar range, again making gas-fired plants less competitive than solar.

As to nuclear, capital and insurance costs are so high that none of these plants are ever built without significant governmental subsidies, and this is especially true as to spent-fuel-disposal costs (how much has Yucca Mountain costs taxpayers so far? How many more centuries will it take to complete? ANSWER: It will probably never be completed). In addition, lead times to build nuclear power plants are so long that by the time a nuclear plant that is planned today is actually producing electricity, solar-generated electricity will cost half of what it does today, putting the nail in the coffin of nuclear-powered electricity.

Finally, there is a lot of NIMBY (Not In My Back Yard) feeling these days vis-à-vis conventional power plants. Coal plants are almost impossible to get approved, and who wants to have a nuclear plant in their backyard? As the point of this article is better appreciated—that "alternative" energy sources should no longer be considered as "alternative"—NIMBYism against coal, gas and nuclear power plants will only grow.

So there you have it: Behind Door #1, you can pay $1.8 billion for a demonstration, coal-fired power plant that will "almost" eliminate all emissions (maybe, if it all works right), while behind Door #2, you can pay $875 million (UNsubsidized) for 250 megawatts of 100% clean solar power.

I know which door I would pick.

I dunno, but it sure seems to me like we have hit grid parity.

Given that this is an investment site, I have written two other articles that build on what I have said here. My second article ("Solar Power Will Be Transformational In The Next Decade") discusses what I think this means for the solar space in general, and in my third article ("TSL Is the Best Value In The Solar Space"), I have compared 11 solars and concluded that TSL is the best bargain today in that space.

DISCLOSURE: I own a large position in TSL, and a decent-sized position in STP. My CSIQ shares will be called away from me over this weekend since I sold $25 calls against my position about 3 weeks ago, but I may buy more CSIQ on a dip. I am not short or long any other stock.