Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
Solar Revolution? Don’t Hold Your Breath.
The $2.2B Ivanpah Solar Power Facility is in the news and in danger of closure for underperforming on its contracted power production. If you are a Ricochet reader, you are probably not surprised, but early results from Ivanpah and two other large scale solar power installations — the Desert Sunlight and Topaz plants, which are also located in Southern California — do not bode well for the future of solar as a replacement for fossil fuels or nuclear power. (Ivanpah’s predilection to burn up birds has been covered on Ricochet previously so I’ll skip it).
Solar has offered the promise of nearly limitless carbon-free energy for nearly 40 years, with a few caveats. First, solar power is expensive but prices per kilowatt are falling rapidly and approaching that of traditional plants. Second, size matters, which means that solar will need to be scaled-up from boutique rooftop installations to larger plants that can produce power at a cost on par with other forms of power generation. This is all done, of course, with the advantages offered by federal loan guarantees specifically designed to prove-out the scalability of these concepts so that the private sector can take over.
In the words of Peter Davidson, solar panels “… existed before as a technology, but that technology hadn’t been deployed at a large scale. Once we’ve done that, the government steps aside let the private markets take over.” However, results from these demonstrator plants appear to show that they are not cheaper than even nuclear plants — which are very expensive — even when scaled-up to large size. The line of companies waiting to take over without federal loan guarantees will be a short one.
(Before beginning a technical discussion, a brief explanation for the uninitiated, a watt is a measure of power, while a watt hour is a measure of energy consumption; they are, roughly, analogous to speed and distance travelled. A megawatt is a million watts and a gigawatt is a billion watts.)
The 392 megawatt (MW) Ivanpah plant is unusual in that it focuses sunlight from over 170,000 mirrors onto three 450′ tall towers (see picture above). The towers contain boilers that create steam, which drives a turbine. The Ivanpah plant’s ability to concentrate solar power was specifically conceived to address the scale-up issue. It opened in 2013 and has underperformed on its expected 1,000 gigawatt-hours (GWh) per year of power ever since.
How badly? First, let’s consider Ivanpah’s capacity utilization, which is the amount of power produced divided by the nameplate power (i.e., its gross capacity). If Ivanpah ran around-the-clock for a year at nameplate power production rate of 392 MW, it would generate 3,433 GWh of energy and its capacity utilization would be 100 percent. This is impossible for several reasons, and its utilization was predicted to be 29 percent (or about, again, 1,000 GWh). That proved to be wildly optimistic: in reality, Ivanpah produced 419 GWh of power in 2014 and 652 GWh last year, or about 12% and 19%, respectively, of its capacity. One can easily see why power contract cancellations loom over the plant.
Two other large solar plants have also been completed in southern California during the same time as Ivanpah. The Topaz Solar Farm (who could hate a solar farm?) in San Luis Obispo County and the Desert Sunlight Solar Plant in Riverside County each have a nameplate power output of 550 MW and both began operation in 2013, so we have two full years of operating experience to look at. These two plants use photovoltaic (PV) panels like those you see on houses. There is not a huge economy of scale in scaling-up PV, because they are just like a rooftop solar plant a million times over.
Although these plants still do not have the nameplate power of a world-class power plant fueled by nuclear, coal, or natural gas, they are roughly ten times more powerful than what was available a decade ago. And as Mr. Davidson said above, this scale-up was specifically meant to prove that solar can be scaled up to commercially viable size. After two full years of operation the annual capacity utilization of each plant maxed out at about 27% in 2015 (this was much better in each case than 2014, which I guess was a bad year for sunlight all around). Better than Ivanpah, surely, but how well do these state-of-the-art solar plants to other plants?
In the chart below, I compare the capital expenditures, size, and capacity utilization of these solar plants with a pair of nuclear plants which are under construction. I picked nuclear because they’re expensive, (essentially) carbon-free, and because — like solar — most of the costs are upfront (in contrast, fossil-fuel power plants have relatively low sunk costs, and much of the cost over the plant’s life cycle is associated with buying the coal or natural gas fuel). As such, I picked the two new reactors at the Vogtle plant in Georgia. These reactors will use newer Generation III pressurized water reactor (PWR) technology which incorporates, among other things, passive safety features to avert a meltdown and much less design complexity which saves cost and improves reliability as when compared to older reactors. Vogtle’s new reactors are projected to run several billion dollars over budget and are not yet complete. In order to give solar the best advantage, I have used the cost estimate with overruns here below:
Plant | Type | Capex ($B) | GW (gross) | Cap. Util. | $/GW | $/GWh |
Topaz | Solar (PV) | $2.40 | 0.55 | 27% | $4.36 | $16.16 |
Desert Sunlight | Solar (PV) | $2.30 | 0.55 | 27% | $4.18 | $15.66 |
Invapah | Solar Tower | $2.20 | 0.39 | 19% | $5.61 | $29.56 |
Vogtle 3 & 4 | Nuclear (PWR) | $16.40 | 2.50 | 88% | $6.56 | $7.45 |
So you can see that when a simple $/GW of nameplate capacity metric is used, solar compares well with nuclear. The Desert Sunlight solar plant cost $4.2B/GW of capacity which is substantially less than Vogtle 3 & 4, which together are $6.56B/GW. Not surprisingly, these numbers are frequently used by solar boosters. But as anyone who’s ever stepped outside and has a brain can tell you, the Sun provides its power unevenly throughout the day and throughout the year. When you normalize the cost of the plant by the power actually produced, (i.e. divide capex by GW hours actually produced in a year), you get a very different result: the nuclear reactors cost about half as much to operate as the PV plans, which cost about half as much again as Invapah.
And remember, these solar plants are in nearly ideal locations, having high elevations, clear skies, and relatively southerly latitudes; identical plants in Ohio or New Jersey would perform far below this already-disappointing production. (For the record, the $B/GWh numbers should not be confused with the actual cost to produce power, though that measure has similar units. The numbers I am showing would be the payment required to pay back the capital expenditure in a single year, which is only used to normalize costs by capacity utilization).
So, you can easily see you can’t compare the costs of solar and nuclear without some complexity: Ultimately, the amount of power produced is important when capacity utilization numbers are as low as they are for solar power, and the grid is going to have to depend on other plants to back up the solar power when the sun isn’t shining and those plants have to be paid for also. The cheapest way to do this, is to pay fossil fuel plants to be on standby for when they are needed to fill in. If you’re concerned with carbon emissions, that should be a deal breaker, especially if you’re operating where coal is the fossil fuel of choice.
The upshot here is that several large-scale solar demonstrator plants have been built and operated for years now. The results show that there continues to be large gaps that need to be filled before solar can compete with nuclear plants (and larger gaps yet if it is to compete with fossil fuel plants). Scaled-up photovoltaic plants are easily twice as expensive as nuclear plants when judged by the power provided and the solar tower plants appear to be worse than photovoltaic in every way.
As a policy prescription, the United States should probably re-think offering loan guarantees for large solar power ventures (at least 17 of which have been funded in this way). Instead, the government would likely get a much more for its money by funding research into improving photovoltaic panels or Generation IV nuclear technology demonstrator plants which could be cheaper, safer, and even proliferation proof than today’s nuclear plants. But more about that another time.
But don’t hold your breath.
Published in General
Thank you for putting this together, Ross.
Is the cost of water to keep the reflectors/ PV surfaces clean factored into the cost per Gw/Gwh?
HAHAHAHAHA
I have a former friend who profited nicely from state grants and loans for a solar venture with a University of Toledo colleague.
It folded. There’s so much more that I won’t go into in a public forum. This man is disturbingly far, far left.
Interesting. But the belief that government research dollars leads to results is how we got in this mess in the first place.
Solar cannot compete. And, for gridded power, it never will. Unlimited cheap fracked gas and oil will consistently win.
Well done. Thanks.
I don’t see the power plant life cycle costs factor into the solar/nuclear comparison. PV have a limited life. The arrays on the Spacecraft in low earth orbit tend to put out about 1/2 their BOL capability after 15 years. Given most of the nuclear plants built in the US were suppose to have maybe a 30 life. Yet many have been extended decades beyond their initial designs with mostly a “digital refreshing overhauls” of their control systems.
My understand is we could further reduce the cost of the nuclear if were allowed the reprocessing of the fuel since the first pass thru the core only pulls 5% of the energy available from the uranium material after the current five year life of the rods. (we can thank the only nuclear engineer (Carter) to make it to the Oval office for that bit of schadenfreude). I mean if the French can do fuel reprocessing surely we can.
I suspect that your being very generous with this comparison between solar and nuclear power sources.
I just dislike them because they ruin the desert. Deserts are an important part to the ecosystem,no?
If they close them will they pay people to dismantle them so the Mojave can go on being the Mojave?
It is not. Although I suspect water usage is not substantial for that purpose. A bigger issue is the replacement of inverters which convert the dc power produced by the panels to ac power for the grid. These are expensive and not particularly long lived.
Hi Trink your post on Ricochet was one of the things that got me interested in this some time ago.
I think that is the lie that must be confronted. Solar like nuclear fusion is always 20-50 years away. These plant represent the state of the art on how not to compete.
I make my living in the natural gas transport and I am happy to see the emergence of massive new gas fired plants, but I think if government has a role to play, it is more in basic research.
On the nose, I am being generous in the comparison and still it is not even close. A rigorous approach would take out parasitic losses (for pumping or cooling for example) to other plant systems which is probably low in PV but is probably pretty high in concentrated solar.
I used to make my living in gas fired generation plants years ago and parasitic losses are killers. I suspect if we were to track these solar plants over time, with degradation of the panel efficiencies you mention and inverter replacements I mentioned above and this will look much worse than it does here.
And I will mention again, this is in high desert with perfect conditions and it is still not good.
What is missing in the calculations is who pays for the removal and reclamation efforts when the site fails. Read the fine print. Ratepayers and taxpayers are nothing more that dairy cows in these scemes.
It’s the desert. A wasteland. What’s there to reclaim? (I’m being facetious.)
Nice work. Thanks.
And then there’s the matter of the sun going quiet. Our planetary science son works with a team operating a cosmic ray detector orbiting the moon. The increased cosmic rays getting into our solar system as the sun grows quiet, strike the moon. It’s helping their research. I’m no scientist, but I believe those cosmic rays seed clouds in earth’s atmosphere. Look at our current solar cycle. Going down. We’re going to need cheap and abundant energy.
Trink: “sun going quite”? do you mean quiet?
Thanks MLH . .shouldn’t post after taking Ambien ; )
IIRC, solar also loses efficiency because plants are located much farther from population centers, and energy is lost in transmission. Do you know if this is accurate?
Son of Spengler
IIRC, solar also loses efficiency because plants are located much farther from population centers, and energy is lost in transmission. Do you know if this is accurate?
GLDIII says:
Anytime you are “pushing electrons” thru a wire there are losses. We do “tricks” to reduce these losses like boosting to very high voltages and AC current, but TANSAAFL (There Ain’s No Such Thing As A Free Lunch) in this universe. So yes the further away you create the “potential energy” the more you lose in the transmissions.
GLDIII is spot on here although I don’t know what the losses are as a percentage.
About 8% between transmission and distribution for nominal grids. Long distances losses are counterintuitively low because you stay at the really high voltages, it is in the stepping down and distribution that hit’s the efficiency.
Transmission loss and availability timing are why I’ve always figured that any actual (as opposed to subsidized or mandated) role for solar would be at user premises, using some of the more advanced (and not yet generally available) designs that would allow close integration with building materials. Get something that can be installed with original or re-roofing projects, and that delivers output that’s timed with demand, specifically A/C in hot summer areas. That allows you to displace peak load demand and keeps the production as close to the demand as possible.
I forgot to mention that greens acknowledge that green power will require reconstruction of the electric grid to allow for wind power in places like Texas, Kansas and So. Dakota to replace solar power that is not being generated at night. They view that, however, as a feature of green power rather than a bug. Sort of like a $1 trillion dollar investment opportunity except it is to replace something that already works.
This is a great article!
The comparison to nuclear is even more generous, because the high cost of nuclear is in part caused by anti-nuclear greens, not because of the intrinsic cost of the technology. Nuclear requires a lot of up-front capital, so delays in construction become extremely expensive.
That’s been the game by anti-nuclear activists for decades – keep delaying plants by slapping them with lawsuits and demands for further environmental impact statements and such, drive the price up, then use the high price as an excuse for not using nuclear power.
France and China build nuclear plants much faster, and for a fraction of what they cost in the U.S. New small scale reactors can be made to a standardized design, achieving quantities of scale that could dramatically lower prices and time to market, if they can get through the regulatory hurdles.
I think you’re also being more than fair in that you aren’t adding in storage costs for solar power, or if it’s just being tied into the grid you’re not accounting for the problems that causes, or the fact that you need to maintain your baseline power capacity, so you just have to run your other power sources at a lower duty cycle when solar is online, lowering efficiency of the whole system.
But the solar power nuts never learn. And they think they can just make this stuff happen with the stroke of a pen. Here in Alberta, we already suffer from lower solar flux because of our northerly location. And when we need power the most (the winter time), we only get a few hours of sun per day. This has not stopped our government from mandating that all coal plants be shut down by 2030 and replaced largely with renewable energy. They have a target of 30% renewable energy within ten years. And how will that happen? Why, they passed a law! Their work is done…
Have a look at the incredible disaster that Ontario faces after they tried to accelerate a push to renewables. Their energy prices have skyrocketed, and the government still has to subsidize the industry to the tune of billions of dollars. Ontario is now the largest non-state debtor in the world. It now owes more money than California does, despite having a much, much smaller population. The high cost of energy is hurting their manufacturing industry and putting people out of jobs. And in the end, they’re going to wind up buying grid power or re-deploying gas turbines or something, because their renewables are not working out so well.
Excellent job.
The New York Times printed an op-ed piece today about “keeping fossil fuels in the ground.”
The author purchased oil and gas leases in Utah with no intention of developing the resources (bracketed comments are mine):
“Moral imagination.” I call it “moral vanity.” If only we could build plants to harness energy from that.
Thanks to the editors for the promote and the edits. It is quite a bit easier to read now.
I know what you mean. I almost went into the concept of dispatchable and non-dispatchable power in detail, but I thought it was already too long.
This article is why Ricochet is so cool. I learn so much.
What makes you think water usage is not an issue? The places where these things are usually put are not clean. Dust accumulates quite quickly, and any moisture (condensation, mist or light rain) makes the dust adhere stubbornly to the surfaces.
http://lasvegassun.com/news/2009/sep/18/dirty-detail-solar-panels-need-water/