January 25, 2010
Michael Kanellos On Nuclear Power
Michael Kanellos, who writes lots of articles about solar power and other renewables for Greentech Media, has written a piece arguing that nuclear power, currently supplying 20% of US electric power, looks hard to avoid if the goal is to stop carbon dioxide emissions from electric power generation.
And, despite all of the rooftops covered in solar panels you see today, solar right now only accounts for around 0.03 percent of power in the U.S. (That's three hundredths of a percent if you don't feel like counting the zeros.
Although pro nuke factoids might sound a little weird coming from someone who works at a research firm dedicated to green technologies, it is difficult to look at America's energy needs for a long time without warming to nuclear. Simply put, nuclear remains one of the most feasible ways right now to produce large amounts power consistently without generating carbon emissions. Constructing nuclear plants generates emissions, but once erected, the plants produce carbon-free power for decades.
When I read people argue that we need to stop global warming from melting Antarctica and Greenland one of the indicators I look at to gauge their practicality is their position on nuclear power. Nukes deliver baseload nuclear power just like most coal electric power plants. Coal electric accounts for one third of total US carbon dioxide emissions. I do not believe wind can displace all coal electric power. But nukes could.
Please click thru and read the full article before raising points of disagreement in the comments. He addresses a number of arguments about nukes versus renewables like wind, geothermal and solar. One important argument: The renewables do not each work everywhere. What works in Texas or North Dakota doesn't work in Maine. Insolation levels, wind levels, and availability of geothermal vary quite a bit by geography.
The geography problem is especially big in densely populated areas. India has over 10 times the population density of the United States and is on course to hit 14 or 15 times current US population density. Nuclear power plants take up small footprints. So they work well for dense populations and in areas where the sun doesn't shine or the wind doesn't blow.
Speaking of India, does anyone know why they aren't pursuing LFTR? They've certainly got the thorium and LFTR's footprint and waste is a factor of ten lower than their current thorium reactor promises to be.
Perhaps I'm missing something - I don't see anything new here. The one data point for solar is glaringly wrong: the EIA only includes commercial generation in their solar data (it comes from the Energy Information Administration, Form EIA-906, "Power Plant Report), so it basically includes only CSP. All distributed, customer-located PV is excluded.
Energy density is a common criticism of wind and solar, when in fact wind consumes very little land - perhaps 1/2 acre per 1.6-3MW wind turbine - much less than other forms of generation, when you include fuel mining and the overall footprint of generating plants (nuclear plants can take up more than a square mile). Rooftop solar doesn't consume any land.
Each form of generation has its own problems - a diverse portfolio is an excellent idea. I think we'll use nuclear, and I think it will work ok. I just can't get excited about it, because of it's link to weapons proliferation, something that Kanellos acknowledges:
"Expanding nuclear power means educating more engineers and technicians on how to build and operate plants. In turn, that means more people that could be susceptible to bribes and blackmail from less democracy friendly nations. This can't be dismissed lightly. If the Physical Dynamics Research Laboratory in the Netherlands hadn't hired and trained a newly minted PhD named A.Q. Khan back in 1972, Pakistan and North Korea may not have missiles today.
On the same day that Areva announced its Fresno plans, an intelligence report from the International Atomic Energy Agency stated that Iran cut a secret deal to obtain uranium from Kazakhstan."
18.5 acres per MW!
The Clinton Power Station is located near Clinton, Illinois, USA. The nuclear power station has a General Electric boiling water reactor on a 14,300 acres (57.9 km2) site with an adjacent 5,000 acres (20.2 km2) cooling reservoir, Clinton Lake. Due to inflation and cost overruns, Clinton's final construction cost exceeded $2.6 billion, leading the plant to produce some of the most expensive power in the Midwest. The power station began service on April 24, 1987 and is currently capable of generating 1,043 MW.
The challenge I see with nuclear is assigning a cost to waste management.
In Washington state we have the hanford nuclear reservation. It has multiple plumes of groundwater driven high level contaminants drifting toward the Columbia river. That may be due to gross mismanagement during and shortly after world war II, but many of the same agencies (mis)managing the cleanup are the ones to managing all domestic nuclear waste.
Vitrification appears to be working, albeit more slowly and costly than expected. Making sure waste is managed properly can be costly, perhaps we can further improve techniques, but most of our current methods are from the 50s, or from the french. Our regulatory structure is also a very real challenge.
This cost is passed on to taxpayers, rather than ratepayers. Perhaps that's better than the even more difficult to determine costs of environmental carbon, but I find choosing my poison difficult.
Nuclear waste falls into two categories; "spent" nuclear fuel (SNF) which is contaminated by the short lived (and therefore quite hot) fission byproducts; and the low level stuff which is often buried, much of it medical waste. We were going to recycle the SNL, but Carter shut that down just before the plant went operational.
When molten salt reactors (MSRs) are ready for deployment, the SNL can be used as "ignition" charges to start them up, any unneeded for this purpose can still be consumed in MSRs. Deep-burn cycle reactors will even take care of the daughter product that make SNL so hot. We're reasonably close to having the technology to make SNL management a non-issue from a technical perspective.
Nick, a wind turbine may only stand on 1/2 acre, but you can't pack them that dense, and no one wants to get perticularly close to a working wind farm either. So the land use is larger than that figure implies. And the capacity factor of nuclear is more than three times that of wind, so a wind farm that actually produces the same average output as a nuclear power plant hase to have more than three times the faceplate capacity.
a wind turbine may only stand on 1/2 acre, but you can't pack them that dense, and no one wants to get perticularly close to a working wind farm either. So the land use is larger than that figure implies.
The land between isn't "consumed" - it can be used for other things. This is clearest for farmland - the land in between can be planted quite nicely. Farmers love wind power (it brings in a lot more money than food crops), and in the US there is an enormous wind resource in farm areas. A nuclear plant, OTOH, encloses it's land for security reasons, so it's really unavailable for other uses.
And the capacity factor of nuclear is more than three times that of wind, so a wind farm that actually produces the same average output as a nuclear power plant hase to have more than three times the faceplate capacity.
The Clinton nuclear plant uses 20 acres per average MW, and a 3MW wind turbine on a 1/2 acre uses 1/2 acre per average MW. Coal and nuclear also have to add in the space used for mining - there are 70,000 existing and old coal mines in the US.
Half an acre = 150 feet by 150 feet.
The Enercon E126 diameter of the rotor is 126 meters (413 feet).
"The "footprint," which is typically between 0.25 and 0.50 acres per turbine, does not include the 5-10 turbine diameters of spacing required between wind turbines. "
In fact, the E126 will need 2000-4000 feet between each turbine.
4000 feet x 4000 feet = 367 acres.
The Enercon E126 is a 7MW turbine, which is rather larger than the 1.6-3MW turbines that are now standard. Right now 60 acres per turbine is pretty standard (probably 1.6MW), for 37.5 acres/MW. 5-10 turbine diameters of spacing means between 92 and 367 acres, for an average of 230, and 33 acres/MW. So, spacing needs are constant or falling.
Farmers have often gotten about $4K per 1.6MW turbine, which meant about $40K on a 640 acre farm. 10 turbines means they only lose 5 acres of productive farmland (less than 1%), and perhaps double their net income. That's huge money for a farmer.
Land owned per megawatt varies wildly for nuclear plants. Land needed is different.
Nuclear plants sites can be rather small or immense. In general the cost of land determines how much will be acquired for the plant. And operators know they will less political opposition if the plant is well isolated - out of sight, out of mind so to speak.
Thus Palo Verde, with the largest capacity in the US, sits on 5,000 acres of relatively inexpensive land in Arizona. Wolf Creek in Kansas has roughly 1/3 the power of Palo Verde but the site is over twice as large at 11,800 acres.
San Onofre in California, with more power than Wolf Creek is on only 84 acres. Granted SO has the advantage of using sea water for cooling. So it is possible to argue San Onofre somehow occupies the entire Pacific Ocean.
A lot of land or water eases cooling costs. You can choose to spend more for acreage in order to spend less on the cooling facilities.
I suspect that rural prisons have more land per inmate than those in urban areas.
Solar has some definite limits for collector area. Only so much sunlight will fall on a given acre. But if all the solar generation is put on roofs we can say solar takes no land at all. And extend that clever argument to wind power.
3MW turbines are 90m in diameter, about 300 feet. 3000 feet spacing is still over 200 acres.
And fromt he picture I posted, that spacing may not be enough.
The reference says 5-10 diameters - you shouldn't always use the maximum. And, I think that 60 acre figure is probably for 1.6MW turbines.
The picture is certainly evocative, but the shadowing in question is unusual - the wind has to be just right for that.
Just right for what? Just right to form visible clouds? Or just right to reduce usable wind in the wake of windmills?
I gather the latter happens even in the absence of the former.
Both. The wind has to line up with the turbine grid, and the grid isn't going to be lined up with the prevailing winds.
Anyone know more about how often this happens?
That depends on how deep the grid goes. From Bruce's picture of turbine wakes, if the grid is 6 towers deep or more, the towers 5 or more in from the edge are in the wake of another tower 100% of the possible wind directions.
The towers 4 in from the edge are in the wake of another tower 90% of the possible wind directions.
The towers 3 in from the edge are in the wake of another tower 65% of the possible wind directions.
The towers 2 in from the edge are in the wake of another tower 55% of the possible wind directions.
The towers 1 in from the edge are in the wake of another tower 33% of the possible wind directions.
Sure, but the farther towers will be less affected by the towers on the edge. If you look at the study, you see that wind direction matters, and that the shadowing effect stabilizes deeper into the wind farm.
This whole topic is obviously mostly old information for people developing wind farms. I imagine that the optimum spacing is a complex balance between shadowing and other costs such as wiring & lease acquisition, maximization of wind resources, etc.
That is a really cool picture of offshore wind farms with cloud cones stretching out from them.
Regards the towers in from the edge: I wonder what the economic calculation is for using taller towers. You'd think height would be more valuable for the inner towers.
A friend spent almost $100,000 last year on a solar-PV system for his house.
It was only supposed to cost $50K, but by the time installation was completed and the amps were flowing, the cost had ballooned.
And here in VA, not the sunniest of places, the PV system only makes 10,000 KW-hr AC at the meter per year total.
This system generates the same electric power PER YEAR that a nuclear plant can make for $130.
$100 grand for less than 50 cents/day (wholesale) of power.
And thanks to green religionists, the federal government will pay $30,000 of this boondoggle in a tax credit in April.
And the State of Virginia, currently firing teachers and closing interstate restrooms due to lack of money, will give him another $17,000 taken from taxpayers.
And electric ratepayers will pay for the power company (forced by law) to buy his solar renewable energy credits (SREC) for almost $6000/yr.
Those SREC's go for over 50 cents per KW-Hr, compared to nuclear power at 1.3 cents/kw-hr.
ALL for less than 50 cents/day (wholesale) of electricity.
Insanity... and economic suicide for the nation if followed larger scale.
We need to build nuclear plants immediately.
Burning coal for power is nuts from a heavy metal release perspective alone.
But "green" alternatives will bankrupt the nation.