Is Solar Cheaper than Nuclear?

by Jigar Shah, Founder SunEdison and CEO of the Carbon War Room
This is an interesting article from Greentech Media. It basically is saying that if we are aiming to decarbonize our electricity grid we cannot do it without Nuclear energy. 
I like to break this down into bite size chunks so that I can understand it better. The Florida public utilities commission recently slowed down the Nuclear plant from Progress Energy so that serves as a useful example: WSJ article here.
The plant seems to cost about $20B for a 2,237 MW plant. So the question is, “For the same $20B over 10 years what could you get?”
– Nuclear: 2,300MWs and 18,000 GWhs of electricity for $20B in ratebase
– Solar: 14,000 MWs and 18,000 GWhs for less than $10B in incentives (with my cost curves the number is actually only $2B, but . . . )
– Energy Efficiency: 6,000 MW and 18,000 GWhs for less than $5B in incentives (structural energy efficiency like HVAC, windows, building automation, etc)
– Ice Storage: shifting peak to off peak to load level the grid could be done for less than $5B.
Whether you agree with my modeling or not, the numbers are in the right ballpark. The challenge here is that the utility industry is not set up to embrace customer empowering technologies because neither they nor their investors understand how these technologies fits with their mode of profit making.
The next stage of this is national security. As has been shown time and time again, distributed, customer empowering solutions are more robust because they are spreadout. Our traditional infrastructure is susceptible to terrorist attacks (or tree branches in the case of Ohio).
Lastly, we have to look at economic security. If the grid goes down with theses central station plants, as happened in Florida in 2008, there are economic losses.
Nuclear power is a wonderful power source that has been serving us ably for many decades. But that isn’t a reason for bypassing the tough economic analysis required for a decision on this magnitude.

9 replies
  1. Anonymous
    Anonymous says:

    I don't understand. 1MW of solar generates 1,500 MWhs a year in Florida, and costs $3 million to build without incentives (ground-mounted). 1MW of nuclear generates about 8,000 MWhs a year, and costs about $9 million (based on your numbers).In other words, nuclear is 3x as expensive to build, but generates 5x as much electricity as solar. What economic analysis can justify installing solar? Solar is a waste. Period. It is simply a scam employed by greedy people exploiting the gullible US taxpayer.

  2. cg
    cg says:

    This sounds interesting, but please briefly explain your assumptions. Focusing just on nuclear and solar:- Where do you get 18,000 GWh? – I assume by incentives you mean to get the PV panels cheap enough that the consumers wants to buy and install on their houses. Is your solar MW peak or estimated real power produced? Regardless, $10B invested in 14,000 MW of PV panels is only giving $0.71 per W. Cost of PV is currently at absolute minimum $4/Wpeak disregarding balance of system http://www.solarbuzz.com/Moduleprices.htm . So you subsidizing at most 18% of the cost which is enough to make it affordable to consumers. Please explain!

  3. Mike Cohn
    Mike Cohn says:

    What the world is not being told is that solar THERMAL (the mirror farms that focus light on a liquid tube that makes steam to spin a turbine) is cheaper than nuclear – not the PV panels on roofs that do a direct conversion of light to electricity.Solar thermal electricity has been dropping in price while nuclear has gone above the old price of solar.This is really only a recently discussed advantage, but one not marketed by the green community for a variety of reasons including the purist's desire to focus on the environmental benefits which only results in preaching to a saturated choir and doesn't motivate the masses.People think incorrectly that the utility bill bears all the cost of nuclear power and only solar is subsidized. The reality is that the Federal government pays for some of the most expensive parts of making nuclear power – they pay for old nuclear plants to be retired when they are too old to operate safely, they pay for the storage of nuclear waste in perpetuity and they pay for liability and catastrophe insurance as well as evacuation and cleanup – NOT the utility.As for getting solar to places where the sun doesn't shine, the smart grid has to be built anyway to prevent regional and local blackouts. That new grid also won't let most of the electricity on it evaporate when transmitted beyond 300 miles like the current grid does. The new smart grid will carry solar and wind from the middle or south to the coasts. And it will let small local producers (like homeowners) push some electricity back into the grid creating even more stability. Then fossil fuel plants can be reduced to only providing fill-in power to the beautifully matrixed outputs of solar which shines during the day and wind which blows harder at night.When the green community purists can get over it's guilt issues and start bragging about it's financial advantage weaponry, the people will naturally reject nuclear for solar just as Wall Street has. The same misconceptions exist with Amtrak who doesn't even own the rails it rides on. Those against Amtrak say it's subsidized, as if paved roads and airports are not subsidized, as if our car payments and airfares pay for them. Rail is many times more construction/fuel/ton/passenger/dollar efficient than paved road and air transit. If the public doesn't stop to think about how much subsidy goes to nuclear vs solar or who owns and subsidizes paved roads and airports, who do they think owns and maintains the railroads? Do they think train builders own them? Or the freight railroad shippers? Let's ask Jay Leno to find out.For now, politicians must promote nuclear because they know the public mistakenly wants it.

  4. Jigar Shah
    Jigar Shah says:

    To CG,Yes exactly. The paradigm shift I am talking about here is that distributed generation is being cost shared by individuals that are interested in putting specific technologies on their load. This could include solar, but also included aggressive energy efficiency, demand response, load control, ice storage, fuels cell, and other technologies. The $0.71/Wdc is what the subsidy has to be for this example. By 2020 when solar and these other technologies won't need any subsidies they will be installed based on high retail electricity rates.

  5. Jigar Shah
    Jigar Shah says:

    On the Anonymous comments. The difference is really the cost of capital. For Nuclear, you have to expend the money over 8-10 years. You also have variable costs of $0.015/kWh and decommissioning costs of an additional $0.01/kWh. With solar you have free fuel and maintenance of about $0.007/kWh. My point was not to say that solar will replace Nuclear. My point was to say that Distributed Resources from solar PV, solar thermal , fuel cells, cogen, aggressive energy efficiency, smart grid, ice storage, etc provides a viable financial alternative to new nuclear. As the deployment of these technologies continue to rise globally the costs are coming down and the delta between Plan B and Nuclear will grow to the benefit of ratepayers.

  6. cg
    cg says:

    Hi Jigar,Thanks for your response. First, please enlighten me on your calculations. You say 2300 MW of installed nuclear power used used over 10 years gives 18,000 GWh. There must be something I am missing, because it should give an order of magnitude more (e.g. 200,000 GWh). Unless the plant is only operating 10% of the time..Similarly, your $0.71 per W of incentives must be a time-averaged number based on your models? e.g. your model must know that today we need $3/W incentive which logarithmically gets lower over the 10 years as PV price gets cheaper, or something like that. Because as a value today, $0.71/W does nothing: PV installed today costs $6-7/Wpeak or around $30/Wreal, which gives the home installer electricity at minimum $0.20/kWh – reducing that by 1-10% doesn't matter."The challenge here is that the utility industry is not set up to embrace customer empowering technologies because neither they nor their investors understand how these technologies fits with their mode of profit making." — I am a very big proponent of solar, but it is important to consider it realistically. The utilities cannot likely profit from customers installing solar PV on their rooftops. At first with low numbers of rooftop PV, perhaps the argument could be made that it helps them manage peak demand. However at high penetration, the utility incurs the cost of managing an intermittent energy supply for the home user (e.g. either storing that energy for the user, or running their other power sources much more intermittently than they currently do, which raises their costs), which they do not make money from. Home solar could still scale up but needs the home user to also install storage at home and not depend on the grid for help. But it may also be more reasonable that solar power plants are still distributed just not installed at homes, e.g. small power stations in nearby fields of each neighborhood. It may be more manageable in the end than each house having PV panels.To Mike: Indeed solar thermal electric is cheaper today. But solar PV may have a greater potential for cost reductions – no moving parts, etc.In any case, I think that solar can supply a big proportion of our energy demands in the future, but some things need to be worked out.

  7. Anonymous
    Anonymous says:

    Cost and generation capabilities aside. The waste from nuclear power will haunt this planet for centuries, the waste from solar doesn't exist. Even if solar thermal isn't as cost effective as nuclear in this authors eyes it is much more favorable for the planet I want my children and their children to inherit.

  8. Anonymous
    Anonymous says:

    When analyzing nuclear power, we need to reconsider MOLTEN SALT reactors, promoted by Alwin Weinberg (Head of ORNL and father of many of today's commercial designs)as a SAFER alternative. More specifically, the Liquid Fluoride Thorium Reactor or LFTR. LFTR uses plentiful Thorium 232. It operates in the thermal spectrum, at near atmospheric pressures; implying significantly lower construction costs (est. 50% less). The LFTR is inherently safe/stable, with a very high negative temperature coefficient of reactivity. It produces 0.1% of the long lived waste of conventional designs. What long lived waste it does yield decays to background rad levels in 300 years, NOT 300,000 years. Its designs are scale-able. Since it does not efficiently yield weaponizeable isotopes, it does NOT present a proliferation problem. It operates with a very simple 2 stage process on-site to separate self generated Uranium-233 fuel from the Thorium-232 blanket, avoiding the expensive and complex reprocessing costs associated with conventional solid fuel reactors. We have enough Thorium-232 on Earth to power the world's population at USA per capita consumption levels for 100's of thousands of years. So, what is the downside…no synergy with the military industrial complex and no solid fuel reprocessing, which is the conventional nuclear industry's bread and butter. See LFTR, on Google TechTalks, for more info.

  9. myth buster
    myth buster says:

    Photovoltaic cells are dependent on nuclear for materials, unless of course you want to use Chinese rare earths. There are only two major sources of rare earths on the planet- Chinese mines and used nuclear fuel.

Leave a Reply

Want to join the discussion?
Feel free to contribute!