This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Desert Sunlight, Another 550MW Solar Farm From First Solar, Now Fully Operational

desert solar project operations

The title of world’s largest solar farm is fleeting with California adding solar capacity in half-gigawatt chunks.

Yet another half-gigawatt solar power project is coming on-line in California.

Late last year, the 550-megawatt capacity Topaz Solar project achieved full commercial operation and claimed the title of largest solar plant on-line in the world.

And now Topaz has to share the crown with First Solar’s 550-megawatt Desert Sunlight project in Riverside, California, which went all-on this month, according to the California Independent System Operator (CAISO) website.

The projects use approximately 9 million solar panels each.

Desert Sunlight is co-owned by NextEra Energy Resources, GE Energy Financial Services, and Sumitomo Corporation of America, and is constructed on land managed by the federal Bureau of Land Management. In 2011, the U.S. DOE issued loan guarantees of $1.46 million for Desert Sunlight.

Topaz has 9 million solar panels across 7.3 square miles in San Luis Obispo County on California’s Carrizo Plain. The solar farm was not the recipient of a loan guarantee and is built on disturbed farm land miles away from sensitive areas in the Carrizo Plain National Monument. PG&E will purchase the electricity from the Topaz project under a power-purchase agreement. Followers of ancient history will recall that this project was originated by OptiSolar and that some of the Topaz real estate was once intended for an Ausra CSP solar power plant.

But these two projects from First Solar will soon yield their glory to SunPower’s 579-megawatt solar project in Antelope Valley, Calif., which is scheduled to go fully operational in the first half of this year and claim the title of the largest operational solar project on the planet.

Manufacturers of the world’s most efficient solar panels (SunPower) and some of the world’s less efficient panels (First Solar) are still able to make large solar projects work, revealing that panel efficiency is less important than project economics and execution in 2015.

Information on these interconnections comes courtesy of the WECC website, CAISO’s Master Control Area Generating Capability List and the intrepid sleuthing of GTM solar analyst Cory Honeyman.

Here’s a chart showing the top three U.S. PV power plants under development in the U.S. 

Project Name Developer Capacity (MWac) Capacity On-Line State Offtaker Owner
Topaz Solar Farm First Solar 550 550 CA PG&E MidAmerican Energy Holdings
Desert Sunlight First Solar 550 550 CA PG&E, SCE NextEra Energy Resources, GE Energy Financial Services, Sumitomo
Solar Star  SunPower  579  412.5  CA  SCE MidAmerican Energy Holdings

Source:  GTM Research’s Utility PV Market Tracker

Check out the GTM Research Utility PV Market Tracker for much more information on utility-scale solar deployment in the U.S.

Photo Credit: New Operational Solar Farm/shutterstock

greentech mediaGreentech Media (GTM) produces industry-leading news, research, and conferences in the business-to-business greentech market. Our coverage areas include solar, smart grid, energy efficiency, wind, and other non-incumbent energy markets. For more information, visit: , follow us on twitter: @greentechmedia, or like us on Facebook:

Eric Wesoff's picture

Thank Eric for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


Bob Meinetz's picture
Bob Meinetz on Jan 16, 2015 5:59 pm GMT

Eric, Topaz cost $2 billion and annually generates 1,100 GWh of intermittent electricity, for a capex of $1.8 million/GWh.

San Onofre Nuclear Generating Station cost $9 billion and generated 8,857 GWh of baseload electricity, for a capex of $1 million/GWh.

California will have to build 6 more of these plants, at a total cost of $16 billion, just to replace San Onofre – and they increase California’s reliance on fossil fuels for backup.

Why are we celebrating again?

Joe Deely's picture
Joe Deely on Jan 16, 2015 9:31 pm GMT


Aren’t the Generation numbers you cite for San Onofre low? Are you counting both units?

Bob Meinetz's picture
Bob Meinetz on Jan 16, 2015 10:22 pm GMT

Thank you Joe, you’re absolutely right.

In 2011, the last full year of operation, Unit 2 generated 9,817 GWh, Unit 3 generated 8,290 GWh for a total annual generation of 18,107 GWh.

Capex is less than half what I quoted earlier; it would cost $34 billion to build enough solar to compensate. That’s assuming, of course, that we could afford enough gas peaking plants to back it up.

Bob Meinetz's picture
Bob Meinetz on Jan 16, 2015 10:25 pm GMT

Brett, you’re correct. My math was way off, see correction above.

Solar is less than one-half as viable as I thought it was – to replace San Onofre would cost $34 billion, compared to which $500 million for decommissioning is a drop in the bucket.

There’s also zero chance of a Chernobyl with SONGS, but whatever. What was the yearly cost of “disposing” of the nuclear waste from San Onofre? Do you know?

Nathan Wilson's picture
Nathan Wilson on Jan 17, 2015 3:23 am GMT

Solar can play a pretty sensible role in an ideal clean energy grid for California.  They have a sunny climate.  As these graphs from JuneMarch, and January show, their demand peaks in the summer (35 GW June, and 25-30 GW  in Jan), with a night-time low around 20 GW year-around.

So I would think they would do well to have 20 GW of baseload power (nuclear plus their roughly 3 GW of geothermal and hydro) and 10 GW of solar, nameplate power (about at 88:12 energy ratio).  They seem to have about 5 GW of solar already, so this addition won’t push them over the top.

With today’s prices for utility solar ($1.9/Watt according to SEIA), it’s a good way to add summer peaking to a nuclear-dominated grid.

Solar power (at least at utility scale) is not the problem, it’s the anti-nuclear ideology.

Hops Gegangen's picture
Hops Gegangen on Jan 17, 2015 1:15 pm GMT


Wikipedia says $4.4B to decommision, for whatever it’s worth.

Apparently the last attempt to refurbish SONG was a fiasco. They shut it down, replaced the steam generators, and a couple years later found premature dangerous wear. There was also some discovery of severe human error, even suspicion of sabotage.

The really nice thing about solar is that it is idiot proof, and it’s hard to put a price on that.




Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 4:39 pm GMT

Hops, “fiasco”, “dangerous”, “severe”, “sabotage”. This unsupportable hyperbole generated by antinuclear activism is unworthy of a response – although I’ve been tempted to respond more times than I can remember.

Solar has a time and a place, but its chief disadvantage is an almost complete lack of resistance to idiots. The price on that in California is $34 billion.

Clayton Handleman's picture
Clayton Handleman on Jan 17, 2015 4:48 pm GMT


Thanks for this objective assessment.

I think that their is plenty of positional “ideology” to go around .  I think the “problem” is the nasty / patronizing tone taken by some who are very entrenched in their viewpoint and who turn discussion into argument. 



Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 4:51 pm GMT

Nathan, in California as everywhere else solar does nothing to get us away from gas generation, it has a huge geographical footprint, it’s non-dispatchable, but most importantly it’s expensive. Whatever SEIA says: utility solar, by the numbers, costs four times as much as nuclear in California.

That these facilities can generate almost 2 GWhs of clean energy would be wonderful if we didn’t  have to pay for it, and if we didn’t have a climate catastrophe staring us in the face.

Thomas Garven's picture
Thomas Garven on Jan 17, 2015 5:09 pm GMT

Hi Bob:

I worked at SONGS for about 20 years or so.  Took the plant from bare ground through full power operation and then observed the retirement of the older SONGS Unit 1.  

Some readers may not be familiar with how spent fuel during operation is typically handled.  It is stored in fuel pools to cool before moving to casks for dry storage since we don’t yet have a long term storage repository.  Other costs during operation which are significantly different than a solar plant are the required operational staffs.  At SONGS we had about 500-700 full time employees when the plant was operational. When we shutdown the plant for maintenance or refueling, that number could increase by another 200-400 workers depending on the scheduled work.  And certainly the Steam Generator replacement disaster can not go unnoticed.  I was lucky and received an early retirement package from SONGS in ‘96.

A typical solar plant on the other hand of lets say 500 MWe [½ the size of one unit] might have from 5-10 full time workers and most of those workers are used for maintenance activities not operations. Operation of a solar plant is normally controlled by a dispatch or operations center and is usually as simple as issuing a computer command or activating a piece of switch gear.  There is no nuclear power plant startup, hot standby, accidental or unplanned trips or outages, cool down or broken turbines, pumps, valves or leaking tanks to deal with or I might add; operator errors.      

There is also a significant difference in the salary scales between nuclear workers and solar workers. Let us not forget that solar is more or less automatic.  It turns on when the sun rises all by itself and then at the end of the day it shuts itself down.  No human is required to do anything and the operational errors that were made by humans at Three Mile island, in Russia and our failure to design for an unrealized failure mode in Japan has had the effect of slowing new nuclear development.

I have seen a few cost estimates for the dismantling of the SONGS and the movement of materials off site to storage locations and certainly that will take many years to complete.  After that another few hundred years of either monitoring the spent fuel and nuclear materials or the reprocessing if that is the plan.  I would never stick my neck out and put some dollar cost value on the next phase of retiring our aging nuclear power plant fleet.   

I generally support nuclear power development in America.  I DO NOT support the further development of pressurized water reactors [PWR] for two reasons.  One is the unavoidable errors made by humans and the other is cost. My preference for power plant applications would be the development of small modular High Tempeature Gas Cooled [HTGR] reactors. For other applications, different reactor designs would be appropriate.

Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 5:08 pm GMT

Clayton, since it’s clear you’re referring to me, l’m sorry you take my tone to be nasty or patronizing. That’s not the intent.

I make no apologies, however, for my frustration with widespread ignorance of practical matters with real implications for the future. The nuclear industry has historically overindulged this ignorance to its own detriment, but In the war of public opinion and through no fault of its own, it’s losing. That’s going to change – the gloves are coming off.

Clayton Handleman's picture
Clayton Handleman on Jan 17, 2015 6:22 pm GMT


LCOE of utiliy scale solar is lower than most nuclear.  Suggesting that solar is 4x more expensive is a pretty bold assertion to go without backup.  It would be great if you would walk your talk and post credible references. 

You have a pretty high bar to show that Lazzards is 2x biased, let alone 4x biased.  And certainly you have demanded that others meet that standard.  A sure sign of a zeolot is that they put out wildly inflated position statements and justify them with righteousness rather than facts.

The topic of the article was utility scale solar.  Utility scale solar cost is currently at the low end of of nuclear in LCOE and still dropping.  Your numbers are inflated even for residential solar which presumably is what you are referring to and which is not on topic. 


Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 5:23 pm GMT

Thomas, thanks for your input. I’ve toured SONGS, and you confirm most of the impressions I had when I was there. One question: how does the turbine replacement qualify as a disaster? It’s my understanding it was spilling  70 gallons/day of secondary loop, lightly radioactive water on the plant floor, and could have been repaired, under warranty, in six months.

Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 8:24 pm GMT

Clayton, a demand to source claims is never a harsh one.

Topaz Solar Farm

Cost: “(more than) 2.0 bilion” (using $2 billion).

Annual generation: 1,100 GWh

Capex per annual GWh = $2,000,000,000/1100 = $1,818,000/GWh

San Onofre Nuclear Generating Station (SONGS)

Cost: $8.968 billion

Annual generation (2011): 8290 GWh (Unit 3) + 9817 GWh (Unit 2) = 18,107 GWh

Capex per annual GWh = $8,968,000,000/18,107 = $495,278/GWh

1,818,000/495,278 = ~3.67:1

Thomas points out below that solar has a tiny fraction of the maintenance and security expenses of nuclear, which is relevant. Those costs, fuel costs, and decommissioning costs are dwarfed by the cost of replacing 148,000,000 solar panels every twenty-five years:

18107/1100 (SONGS annual generation / Topaz annual generation) = 16.46 (SONGS generated the equivalent of 16 Topaz solar farms). Topaz uses 9,000,000 solar panels.

16.46 x 9,000,000 = 148,140,000 solar panels. Assuming it costs half as much to replace all panels in 2040 as to build the facility in 2014, and including the generous assumption that the panels will generate 100% of capacity throughout their useful life, replacement would cost $16 billion.

SONGS employee expenses (estimate average $60,000/year salary) x 600 = $36,000,000 annually x 25 years = $900 million.

Fuel costs ~$80 million per refueling cycle for two reactors, every eighteen months.

25/1.5 = 16.66 cycles x $80,000,000 = $1.332 billion.

SONGS decommissioning cost $4.4 billion, unnecessary except in case of a catastrophic failure or plant is closed prematurely.

These are real world costs, not “estimates” by a for-profit research group which very possibly has skin in the game.

Lazard’s disclaimer:

Other factors would also have a potentially significant effect on the results contained herein, but have not been examined in the scope of this current analysis. These additional factors, among others, could include: capacity value vs. energy value; network upgrade or congestion costs; integration costs; costs of adding emissions controls (e.g., selective catalytic reductions systems, etc.) to existing fossil power plants; and transmission costs. The analysis also does not address the potential stranded cost aspects of distributed generation solutions in respect of existing electric utility systems, nor does it account for the social costs or other externalities of the rate consequences for those who cannot afford distributed generation solutions.

Substantiated corrections are welcome.

Thomas Garven's picture
Thomas Garven on Jan 17, 2015 6:51 pm GMT

Hi Bob:

There is never a short answer with me, LOL.  Let me try and answer your question.

I assume [I love that word] you are referring to this statement; “And certainly the Steam Generator replacement disaster can not go unnoticed.”. The replacement Steam Generators were purchased and installed after I retired. However since I spent most of my working career there I still try to follow what is happening.

Without turning this posting into a book, here is a little history.   When the plant was built it was theorized that “X” number of Heat Exchanger tubes would fail and this proved to be true. Steam Generators are designed to tolerate a certain level of tube failures before they are considered no longer servicable.  Its the old bathtub reliability curve in play. Early failures followed by a long term operational phase followed by wear out. Just about what you would expect from any mechanical system or even your car.

However after many years of excellent service the original Steam Generators reached the end of their life and needed to be replaced. New Steam Generators were purchased and installed in the plant which necessitated making a hole in the side of the containment building as well as other modifications which were a very expensive operation. Most people do not appreciate just how difficult and complex an operation like this can be. Just the reinforcing steel alone that needed to be cut and removed and then replaced was #18 bar in many locations. That #18 bar in layman terms is 2 ¼” inches in  diameter. Not something you can buy at your local Home Deport, LOL. We are talking thousands of man-hours of labor. There are lots of news articles you can Google if you are interested in the projected costs.

But in any case from a utility perspective, there were still many years of licensed operation available to recover these costs so the work went forward. In the beginning, the new Steam Generators exhibited the same operating characteristics as the original steam generators which were higher levels of water induced vibration mode failures. This of course was expected and design allowances were made during fabrication of the new Steam Generators. However, the original high failure rate never seemed to decrease like the original Steam Generators. Every time the plant was taken to 100% power levels and the required amount of water was feed into the Steam Generators the vibration took it toll and cause large numbers of tube failures.

The plant was therefor left with only a couple of choices. Operate the plant at a reduced power level which would reduce revenue making cost recovery unacceptabe or replace the new Steam Generators with another set of Steam Generators of a different design. Neither alternative was deemed to be financially sound.

The short answer to your question would have been to just say – the plant was no longer financially cost competitive.  Sorry for the long winded response.  

Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 7:02 pm GMT

Thomas, no apology is necessary and I appreciate getting this information from the source.

It pretty much confirms my understanding, however I thought Mitsubishi Heavy Industries was picking up the tab for all or part of that under an equipment warranty? And what part did activist litigation/politics play  in postponing recertification and running costs up for SoCalEd?

Joe Deely's picture
Joe Deely on Jan 17, 2015 7:07 pm GMT
Good to see the improved nuclear numbers.
Now, here is a different way of looking at cost of solar to replace lost generation.
First assumption – it will take 8GW of solar to replace the 2GW of lost capacity @ San Onofre.
Second assumption – this solar will be built out over 10 years. We will start with 300MW of solar built in 2015 and then increase build over the years until we have a total of 8GW built 2015-2024.
Third assumption – Cost of installed solar installed in 2015 will be $1.9/Watt as Nathan mentioned with his 2014 Q3 SEIA source. A more recent announcement shows a solar project in Dubai with a cost of $1.65/Watt. But let’s be more conservative and stick with $1.9/Watt
Fourth assumption – Solar will continue to drop in price. However, instead of using the 20+% reduction in pricing that solar has seen over the last few years we will be more conservative again and use an annual 5% reduction in cost.
Here are the costs for this scenario:
2015  300MW x $1.90/Watt = $570M
2016  400MW x $1.81/Watt = $722M
2017  500MW x $1.71/Watt = $857M
2018  600MW x $1.63/Watt = $977M
2019  700MW x $1.55/Watt = $1,083M
2020  800MW x $1.47/Watt = $1,176M
2021  900MW x $1.40/Watt = $1,257M
2022 1000MW x $1.33/Watt = $1,326M
2023 1100MW x $1.26/Watt = $1,386M
2024 1700MW x $1.20/Watt = $2,035M
Total 8,000MW @ Cost of $11.4B  with an average cost over 10 years of $1.42/Watt.
By the way – the goal for government’s SunShot Initiative is an installed cost of $1/Watt by 2020.  Obviously if this occurs all my numbers above would be way too high.
Joe Deely's picture
Joe Deely on Jan 17, 2015 7:08 pm GMT


Moved my reply up top.


Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 7:46 pm GMT

Joe, regarding first assumption, mine is correct if even a bit conservative. Solar capacity is irrelevant – it’s only worth how much energy it can deliver. The Topaz Solar Farm delivers 22% of its capacity (capacity factor); San Onofre delivered >95%. Based on my knowledge of California weather it’s likely Riverside will be more productive, however.

Second assumption – I’m not understanding your point. We can amortize nuclear over 25 years as well.

Third assumption – predictions are wonderful, most of them are wrong. My post below is to demonstrate what solar did in fact cost in 2014, and how Lazard’s and Deutsche Bank are investor services which often invest in the very technologies they’re evaluating.

This article in PV Magazine forecasts that SunShot 2020 objectives won’t even be met by 2030. Although the research is done by a company which may be investing in new nuclear, for all I know.

Joe Deely's picture
Joe Deely on Jan 17, 2015 9:32 pm GMT


I like your thinking on this.

Some things to consider:

  – CAISO data is a great source and serves as a pretty good representation of state as a whole but it only represents about 80% of total state market  The in-state municipal utilities including LADWP are generally not part of system. See FERC site for more info on this. 

  – by end of 2015 solar in CA – if you include distributed solar – will have already passed your 10GW number.

  – CA imports about 1/3 of their electricity from out of state. So there is a LOT of inherent flexibility including transmission lines from Northwest, Wyoming/Utah, Nevada/Arizona and more recently Mexico.

My thinking is that within the next decade CA will be able to get away with about 10GW of Baseload power – maybe less.  Any improvements in Storage could lower this further. Plenty of work going into Battery storage so that could improve things substantially.  Also there are projects underway for further Pumped Storage –

1) Iowa Hill  – 400 MW in Sacramento area

2) San Vicente – 500 MW in San Diego area 

Thomas Garven's picture
Thomas Garven on Jan 17, 2015 8:15 pm GMT

Here is a Google search link you might enjoy.

It will be many years before the total costs can be accounted for and lawyers can be a very creative bunch when it comes to billing hours, LOL.  

Have a great day.  I must move on. 

Joe Deely's picture
Joe Deely on Jan 17, 2015 8:23 pm GMT


I think on most points we are close to agreement…except for one really important factor.

– As we both say it will take about 4x Solar capacity to replace 1x of nuclear capacity. So in this case 8GW of solar to replace 2GW of nuclear. Not to quibble too much, but if you want true capacity factors for San Onofre check out this site from World Nuclear Association.

– my second assumption was just used to define a possible “scenario” for building out the replacement solar. I used 10 years because that is about how long it takes to build a new nuclear plant. 

Our main disagreement is on cost of utility Solar. The SEIA link mentioned by Nathan derives the costs by looking at the projects that were actually built in Q3 of 2014.  The Topaz project you mention was started in 2011. Things are changing fast! 

Here is some data:

There is absolutely no way that you can use costs from a 2011 project in pricing out a 2015 project and even more so for work that that will occur in 2020 and later.

You may think that continued price drops are going to somehow magically stop occurring but I don’t see any possible way of justifying this. There are already many improvements in the pipeline that are lined up for at least the next decade.

By the way, I am not saying that this new 8GW solar is as equally valuable to the energy system in CA as the 2GW iof nuclear it replaces. I just want to show that the costs are pretty close.

Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 8:53 pm GMT

Joe, your point about average capacity factor is well taken, and there certainly is more of a possibility for long-term downtime with nuclear. But it should be noted that nuclear plants’  “efficiencies” (capacity factors) have historically improved over time; whereas (once it’s installed) a solar panel’s efficiency deteriorates. Hard to quantify or compare these two factors, but intuitively it seems to be pretty much a wash. I make this point often in discussions with people who point out that small diesel-powered cars can generate less carbon than electric ones. That’s true, in some areas, with new cars. Over the lifetime of the car, that dynamic changes considerably.

I don’t think costs are going to stop dropping on solar PV for a while, but I do think price improvements are going to be harder to come by. You’ll notice that non-module costs declined a fraction of a cent/kwh in 2012-2013.

Solar and wind have historically relied on overhyped promises and predictions to survive. In many cases the industry has exceeded predictions (module cost). In the most important cases, like percent of generation, they’ve proven dismally disappointing. Jimmy Carter predicted in 1979 that 20% of our energy would be provided by solar by 2000. He was off by two orders of magnitude.

Do we trust an issue as important as climate change to possible “new developments in the pipeline”  or knuckle down and go with what works?

Clayton Handleman's picture
Clayton Handleman on Jan 17, 2015 9:32 pm GMT

He is replacing the 2GW nuclear with 8GW solar.  That takes CF into account.  You could split hairs over 22% vs 25% but given conservative cost assumptions, that really is a distraction from the discussion.

“PV Magazine forecasts that SunShot 2020 objectives won’t even be met by 2030.” 
These are not the numbers he used.  He used much more conservative numbers.  To quote SunShot is to misrepresent his analysis.

20 – 25 years is the lifetime typically used for designing PPAs and for doing forcasts.  However field experience has shown the lifetime to be considerably longer.  I worked in the solar industry for years and have personally observed PV systems that were installed in the early 80’s operating well past the 25 year anniversery.  Materials and processes have improved, tier 1 modules can safely be assumed to last considerably past their warranty.  In my travels I have worked with the production folks at module mfgs.  Off the record they said that the internal numbers that they expected for module lifetimes were on the order of 35 to 50 years but that the did not release those numbers out of fear that they would be accused of inflating numbers.  Its not just the nuclear guys who get things thrown back in their faces.  Since nuclear has operational costs longer lifetime PV arrays work in the favor of PV in terms of cost. 

I don’t know what the cost of labor is for nuclear power plants but I would be surprised if $60,000 is an accurate when taking into account benefits and other associated costs.  I think those workers are required to have a good deal of training and other specialized costs on top of the usual benifits and costs of labor.  Add to that that they are a limited commodity and it seems very unlikely that you get that package for $60k.  Thomas Garven, can you shed any light on this.

Thats all I got for today.  Have other obligations.



Bob Meinetz's picture
Bob Meinetz on Jan 17, 2015 9:01 pm GMT

Thank you!

Robert Bernal's picture
Robert Bernal on Jan 18, 2015 5:09 am GMT

Even though I want reliable, 24/7 power, I intuitively like solar, too. However, such farms will most probably cause local heat islands as they are dark in color. Too bad they can’t be only slightly darker than their efficiency. The picture looks like the large farms at Lucerne Valley (by Big Bear).

The continued expansion of solar collection must expand with proportionately less subsidy. It must also entail rapid machine automation of batteries, to store close to the inverse of their capacity factors, in the absence of advanced nuclear.


Also, if the machine automation can make solar for whatever the costs are today minus (most of the) subsidies, then there should be a continued exponential growth curve. (I think) it grew by a factor of about 1.25 from 2012 to 2013. If projected at just half that (at 1.12) then by 2035, they’ll be 10x the capacity and by 2050, 50x the capacity – which is NOT good enough to save the biosphere from fossil fuels. However, if continued at the 1.25 rate, 85x by 2030 and 1,500x today’s capacity by 2050 – which is probably more than the future world could use!

So, the subsidies must be shifted towards paving the path to machine automation of ALL the processes involved!

Nathan Wilson's picture
Nathan Wilson on Jan 18, 2015 4:01 am GMT

Joe, a huge difference between San Onofre’s 2 GW and 8 GW of hypothetical solar PV is that the PV would contribute to fossil fuel lock-in, but San Onofre could be part of a zero-carbon grid for California.

As I’ve said, using state of the art technology, California would minimize their fossil fuel use with 20 GW of baseload power and 10 GW of solar.  California (the CAISO portion) already has 5 GW of solar; so building above 5 GW of additional solar effectively puts a floor on how low their fossil fuel use can go (it’s actually restricted already by their 5.8 GW of wind, but that isn’t really a long-term investement).

Of course there could be breakthroughs in the cost of energy storage which would reduce the fossil fuel lock-in which is caused by variable renewables, but I’m skeptical.   Consider than supplying CAISO’s 20 GW of night time load with storage, assuming lithium-ion batteries with 150 Wh/kg (midrange from this article):  20 GW * 15 hours/(80% depth of discharge)/(150 Wh/kg)= 2.5 billion kg of batteries.  If the batteries last 2000 cycles (about 7 years), then each year, California would have to manufacture and dispose of 357,000 tons of batteries.

But even such a large battery system is just a fair weather resource.  You still need lots of capacity over-build and summer curtailment, or else it’s fossil fuel all during the rainy season.

Hops Gegangen's picture
Hops Gegangen on Jan 18, 2015 12:58 pm GMT


Much ink gets spilled here over dealing with the variability of renewables, so I thought I would point this out: ARPA-e has an interesting project underway with Foro Energy to equip drill bits with high powered lasers that could make it economical to drill through the hard rock in which the best geothermal resources are located.

In parts of the country, that could tap a huge resource of baseload power. And in some places, it would match well with solar because you could allow heat to build up in the reservoir on sunny days.

Joe Deely's picture
Joe Deely on Jan 18, 2015 5:51 pm GMT


As I said in another comment:

“By the way, I am not saying that this new 8GW solar is as equally valuable to the energy system in CA as the 2GW of nuclear it replaces. I just want to show that the costs are pretty close.”

I actually believe the if we had 5-6GW of nuclear we might be able to get really close to zero carbon. But that is not going to happen in next 20 years. Instead CA will move to 50-60% carbon free in that time period first and then we’ll see where things stand.

As for the battery storage comments. We are really early days. Things are changing fast.  For example, the battery article you cite – which isn’t even 3 years old has a cost for lithium ion @600/kWh.

Way too high. I don’t think we can even say that lithium ion will be the only or primary battery storage technology. Any speculation on energy storage density and cycles is just that – speculation.

Let’s get the first 2GW of storage out there and see what is working,evaluate performance and costs.

Finally CA already has some zero-carbon baseload (nuclear,geothermal,bio and hydro) which you seem to be ignoring. Look here to see the actual breakdown of sources for current CAISO workloads.

You are also ignoring pumped storage which CA already has and as I mentioned in other comments there are other projects in various stages.


Clayton Handleman's picture
Clayton Handleman on Jan 18, 2015 6:07 pm GMT


I tend to agree with you.  The article Nathan posted uses $600 / kwhr for Li-ion which yeilds a lifetime performance of $0.40 / kwhr.  Here is a blog post that aggregates a number of sources that point to about 1/5th of that.  So it is pretty safe to use between $0.05 /kwhr to $0.10 / kwhr in the 5 – 10 year timeframe.

I hope to get the time to write an article that explains how most of the benefits of storage can come from load shifting during the charging process.  If that is done then much of the storage benefit of batteries can be had for free by charging when demand is low and / when generation is most available. 


Joe Deely's picture
Joe Deely on Jan 18, 2015 6:16 pm GMT

“Do we trust an issue as important as climate change to possible “new developments in the pipeline”  or knuckle down and go with what works?”


I couldn’t agree with your comment more. Let’s go with what works. In CA that is primarily Solar supported by a variety of other sources and technologies- Hydro,wind,Geo and storage. Currently Nuclear is just not a viable solution in CA.  

The Energy department has a $12.5B loan program  for new nuclear kicking off with the first round of proposals due in March. Just wondering – how many of these proposals do you think will be submitted for projects in CA?  My guess – ZERO.

The only area of the country that will be implementing any new nuclear over the next twenty years is the SouthEast. A remote possibility for more nuclear out west might be in AZ – and I think even that is highly unlikely.Let’s get another 4-5 projects going in Florida, Georgia, LA etc.. – hopefully with newer, more nimble technology.

So, as you say let’s knuckle down. Get CA to 50-70% carbon free over the next twenty years and hopefully then we can finish the job with some nuclear that has been proven out elsewhere.

Clayton Handleman's picture
Clayton Handleman on Jan 18, 2015 6:23 pm GMT


Why FF during the less sunny season (rainy?).  Demand is reduced and renewable production tracks it reasonably well.  I am not seeing major mismatch problems in the forseable future.


Nathan Wilson's picture
Nathan Wilson on Jan 18, 2015 7:54 pm GMT

Clayton, desert solar power with batteries for time-shifting is a fairly reliable power source in the summer.  Furthermore, batteries are most economical when tied to solar power and limited to 6 (and to a lesser extent 14) hours of storage.

On the other hand, no plausible amount of batteries will ever make wind power (and for that matter winter solar) reliable – it’s just too chaotic over multi-day timeframes (I note that you chose a graph of monthly average wind power – I’m sure as an engineer, you well know that mean+-3*standard_deviations is a better metric, and with <24 hours of storage, <1 day granulatity is required!).

That’s why am baffled as to why wind-power advocates are not more supportive of ammonia fuel (which does allow the much needed monthly and even seasonal energy storage).  Perhaps because it would require that them to admit it is needed (or maybe they just don’t like the ammonia smell).

Bob Meinetz's picture
Bob Meinetz on Jan 18, 2015 8:06 pm GMT

Joe, with regards to

Let’s get the first 2GW of storage out there and see what is working, evaluate performance and costs.

Let’s get the first 550MW of panel capacity out there and see whether 9.5 square miles and $2 billion worth of solar panels are delivering as promised.

During February, 2014, when Topaz supposedly had 300MW of functional capacity, it generated 36,044 MWh, for a capacity factor of 17.88%.

Nathan Wilson's picture
Nathan Wilson on Jan 18, 2015 8:15 pm GMT

I don’t believe the battery industry is changing fast.  I can tell you that li-ion was used in cell phones and laptop computers 20 years ago, and while they have replaced NiCads and NiMHyd in several applications, they’ve had no luck displacing lead-acid or pumped-hydro anywhere (i.e. the cost of grid energy storage has not come down at all!).  I’m not saying the (speculative) costs that Clayton mentioned (5-10 ¢/kWh for per-cycle battery costs) are unachievable, but that this is not low enough to make variable renewables a cost effective replacement for nuclear or fossil fuels for most of the world’s population.

I also don’t dispute the value of California’s existing clean baseload.  The problem is that it is not growing.  My big complaint is the variable renewables (when deployed without storage) tilt the balance to favor flexible fossil fuel over baseload (thus producing fossil fuel lock-in).  The EIA says geothermal power is one of the cheapest sources of electricity, easily beating coal!  But it is not growing in the US, perhaps because of wind power. [I consider using biomass to make electricity to be unacceptable environmentally].

I agree that we should pursue storage and renewables, but when they are used as tools to slow down nuclear power, they cause much more harm than good.


Clayton Handleman's picture
Clayton Handleman on Jan 18, 2015 8:18 pm GMT


Thanks, I was not actually trying to address every imaginable contingency.  I was addressing specific comments you made about things happening on a seasonal scale. 

“You still need lots of capacity over-build and summer curtailment, or else it’s fossil fuel all during the rainy season.”

Lets get that handled and then move on.  If you want to make a point on a finer resolution then grab some CAISO data and do it rather than handwaiving. If you look at the data even on a daily basis you will see that during the peak summer months the wind and solar play very nicely together.  If gas peaking plants have to be turned on on rare occaisions it has not real impact on climate change. 

Can we please agree that Li-Ion batteries at $600 are long in the rear view mirror and stop using that number for looking at the world 5 – 10 years from now.  Maybe the economics still don’t work but when looking at what could be in a 5 – 10 year timeframe lets use the mainstream numbers for Li-Ion battery pricing such as the up to date McKinsey, Musk and Navigant numbers.  They are stating Li-Ion battery prices in the $100 – $150 range within 5 – 10 years.  If you disagree with that then please post some credible, current backup. 





Clayton Handleman's picture
Clayton Handleman on Jan 18, 2015 8:54 pm GMT

“I don’t believe the battery industry is changing fast.”

Hmmmm, looks pretty fast to me.  This is the most up to date I could find without paying $5k for an analyst report but it seems to be in line with the press.  What I have read is that Tesla is paying $200 – $250 .

And the Giga factory will have all of the latest and greatest along with the benefits of scale.  The McKinsey and Navigant numbers were from before the giga factory was confirmed. 

Joe Deely's picture
Joe Deely on Jan 18, 2015 11:07 pm GMT


Thanks for the link. I didn’t realize you could use the EIA data browser to view output from a single plant. Nice.

Other than that – not sure what your point is??

It looks like the final 40MW for this plant went live in Nov 2014. Therefore in Oct of 2014 it had 510MW and it generated 113,417MWh for a capacity factor of 25.4%.

So what’s the point? Looks like it is performing pretty close to what was expected, right?

Joe Deely's picture
Joe Deely on Jan 19, 2015 12:00 am GMT
“I don’t believe the battery industry is changing fast.”

Wow.. I guess we have been spoiled by Moore’s law with computers over the last 30 years.
Check out this video. Its the keynote of JB Straubel  – Tesla CTO – at the Energy Storage Symposium in Oct 2014.
You can skip to about 5:30 to see Energy density discussion.
His take – Energy density has doubled over the last 10 years and will continue to double every ten years for at least the next 10-20 years. 
That seems pretty fast to me.
Robert Bernal's picture
Robert Bernal on Jan 19, 2015 12:45 am GMT

I thought that molten salt reactors could be somewhat load following. If that’s the case, then only very little NG (or battery) be needed for precision. Renewables could be seen as “tools to advanced nuclear” if only the environmentalists would promote advanced nuclear, too.

Clayton Handleman's picture
Clayton Handleman on Jan 19, 2015 1:06 am GMT

Yes, at MIT they are looking at Nuclear reactors in which they offtake the molten salt and store it for later use.  This enables it to play nicely with renewables and takes away the traditional constraint of near constant power generation.  Forsberg is one of the researchers working on it.  They have a readable piece on it but I could not track it down.  Probably on this site  if you want to look into it more.

Mark Heslep's picture
Mark Heslep on Jan 19, 2015 2:49 am GMT

Nathan –

Consider than supplying CAISO’s 20 GW of night time load with storage, ….  20 GW * 15 hours/(80% depth … 2.5 billion kg of batteries.  If the batteries last 2000 cycles (about 7 years), then each year, California would have to manufacture and dispose of 357,000 tons of batteries.”

What then did you have in mind for the early evening power gap?  Batteries?

A power gap between your proposed 20 GW baseload nuclear and solar occurs in the early evening window (and to a lesser degree morning).  In the winter, CA peak load occurs aounnd 7pm and in the summer load is still 32 GW to 34 GW at 7pm. The gap then is a peak power of 15 GW, and averaging perhaps 8 GW * 5 hours = 40 GWh.  

In the particular CAISO summer chart you referenced wind happened to (almost) make up the difference in the evening, but even in the good summer wind season Pacific coast wind generation is not that reliable.  In the winter CAISO chart, wind generation is near non-existant.  The “evening” gap still requires, using your battery parameters and a guess of 40 GWh stored,  0.3 million tons of batteries with 47 thousand tons replaced every year.  

If instead of batteries one chooses, say, 15-20 GWe of gas peaking plants then the use of solar is again locking in fossil fuels along side the solar, incurring the cost then of the solar and a rather expensive fleet of low utilization gas plants and low utilization transmission. 

Bob Meinetz's picture
Bob Meinetz on Jan 19, 2015 7:46 am GMT


What source do you have showing 510MW for all of October? From what I can tell, the final phase of construction began in January 2014 and they put capacity on line as it was added (by February 2014 there should have been at least 300MW online).

It’s going to be hard to know until we have a year’s worth of data, but I have yet to see an underestimation among predictions for solar capacity factor.

Nathan Wilson's picture
Nathan Wilson on Jan 19, 2015 5:44 am GMT

Ok, I was off the mark.  The battery industry has changed a lot (not compared to computers of course), and in particular the prospects of EVs is approaching a knee in the curve.  

In his lecture, Straubel was certainly excited about grid energy storage.  He assumed that we would choose a 100% renewable+storage grid in the future; he did not make a case that it would the most economical choice.  

So I think it’s still too early to say renewable-output smoothing will be affordable.

Bob Meinetz's picture
Bob Meinetz on Jan 19, 2015 7:43 am GMT

Clayton, near constant power generation needn’t be viewed as a constraint. A molten-salt reactor could be operated constantly at full capacity by first meeting electricity demand and diverting excess energy to synfuel production. By storing it in hydrocarbon compounds (gasoline, alcohols, methane, etc.) using ambient CO2 and water as feedstocks, we could create a carbon-neutral chemical “battery” suitable for powering transportation, heating, or any purpose currently served by fossil fuels.

See slide 66 of Forsberg’s 2012 lecture at Ohio State, “Alternative Nuclear Energy Futures: Liquid Fuels can be Made From Air“.

Robert Bernal's picture
Robert Bernal on Jan 19, 2015 7:52 am GMT

Thanks, I’ll probably find it. Many say that “their” reactor design is load following but do not list specifics. I know that the French can do it even with their old water reactors – to some rather slow degree.

Hops Gegangen's picture
Hops Gegangen on Jan 19, 2015 10:36 am GMT


I read that the Air Force is working to make jet fuel using the nuclear reactor on air craft carriers so they don’t have dependencies on supply ships. As I recall, CO2 is more concentrated in sea water, so they were going to get the carbon from the water.

This is why I see potential for harnessing nuclear at sea (or on islands) and transporting the energy in the form of fuels. I would think that the regulatory costs would be vastly reduced and some expensive redundancy removed if the reactors were not near population centers. 


Clayton Handleman's picture
Clayton Handleman on Jan 19, 2015 1:39 pm GMT

Link takes me to a dead page.

Bob Meinetz's picture
Bob Meinetz on Jan 19, 2015 4:56 pm GMT
Clayton Handleman's picture
Clayton Handleman on Jan 19, 2015 6:01 pm GMT

Worked that time.

I had the opportunity to sit down with Forsberg and enjoyed the conversation.  I like what he brings to the table.  The idea of coupling nuclear with various symbiotic storage approaches, to move it away from its baseload only paradigm is interesting and promising. 

As you point out though, even with a straight shooter like Forsberg, it is important to look at motivations of the source.  His job is to find ways to do nuclear and he is not immune from taking some liberties with the data to cast a more favorable light.  In his presentation he compares to wind alone and solar alone.  But a quick glance at the graph below (from his presentation) shows that that biases the conclusions considerably.  Wind and solar play very nicely.  When wind drops out, solar is strong and vice versa.  Taken together the storage issue is much less nasty than taken apart.  Given that their implimentation trajectory is tandem, it is a disappointing oops that he does not show them taken together also.  They would still need considerable storaage but the numbers would not be nearly as dire.



Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »