January 25, 2010
New Nuclear Power Costs Too High?
A November 2009 report by Citibank about nuclear power costs and viability of new nuclear plants in the UK and Europe provides useful information to those (such as myself) interested in the economics of nuclear power. The Citi report claims most of the time new nuclear power would cost more than the the wholesale price of electricity in Britain. See the graph on page 10 (PDF).
Power Price: Nuclear power stations have very high fixed costs and
relatively low variable costs. Their cash flows and profitability are
therefore particularly sensitive to the price that they sell their power.
As we show later, even at the low end of the build cost estimates, we
calculate that a new nuclear station will require €65/MWh
(£58.5/MWh) in real terms year in/year out to hit its breakeven hurdle
rate. As we show in Figure 5, the UK has only seen prices at that level
on a sustained basis for 20 months of the last 115 months. It was a
sudden drop in power prices that drove British Energy to the brink of
bankruptcy in 2003. No nuclear power station has ever been built to
our knowledge where the developer takes the power price risk.
I've come across reports claiming that nuclear power can't compete in Europe without a carbon tax of at least 40 Euro per metric tonne.
The report points to cost and schedule overruns in recent nuclear plant projects and argues that new nuclear plants have considerable construction cost risks. These risks raise the cost of capital (the market wants higher interest rates on bonds) and therefore raise total costs.
Both Westinghouse and Areva claim to be able to construct a new third generation
plant (AP-1000 and EPR, respectively) in 3 years from first pouring
of concrete. However, evidence to date suggests this is not necessarily the
case, as Olkiluoto and Flamanville projects have both suffered delays, while the
first AP-1000 unit under construction, in SanMen China, is running
significantly over its $1,000/KW construction cost target and is expected to be
over $3,500/KW target on current estimates.
The SanMen delay tells us that the Olkiluoto and Flamanville are not outliers.
Note the wide range of cost estimates. This is an indication of uncertainty and uncertainly means risk and higher capital costs.
Georgia Power stated in mid 2008 that two 1100MW reactors would cost up to $14 billion, depending on financing terms. This gives significantly high cost
assumptions of $6,360 per kilowatt.
In November 2008, Tennessee Valley Authority updated its estimates for
Bellefonte units 3 & 4 relating to two AP1000 reactors of 2234MW combined. It
said that overnight capital cost estimates ranged from $2,516 to $4,649/kW for
a combined construction cost of $5.6 to $10.4 billion.
The next few nukes built in the US will have US federal credit guarantees that will lower the cost of capital. If the builders can finish construction in a timely manner those plants will probably turn out to be profitable.
Does anyone know how long the US wind production tax credit lasts on newly installed turbines?
The Energy Bill recently passed by the US Congress recognises such risks and
provides production credits of 1.8 cents per KWh for the first 3 years of
operation, equivalent to the subsidy provided to the wind generation segment.
The US can't cut carbon dioxide emissions without a large nuclear build. Given a large (a few hundred) reactor build the US could pretty much eliminate the one third of total US CO2 emissions that come from coal electric power. We'd also breathe less soot and ozone too.
I'm surprised the SanMen reactor is having that much trouble. Much of the cost for nuclear reactors has been blamed on the stringent safety requirements in the U.S. and Europe. With lower safety and quality standards in China I assumed it would be less costly there. Does anyone know more about the problems?
--I'd use the ERCOT merchant nuke plant ($3,500/kw) as a realistic price
--Georgia Power/SoCo has a sweet relationship with the Georgia regulators...they have been able to squelch merchant gen in their ISO by all kinds of means....don't get me started. Given that they get a guaranteed return on equity, I'd suspect ehre is a LOT of fat in that per kW price
--It's been 5 years since I looked at developing power plants, but at the time, for comparison, we used $550-$650 per kW for combined cycle gas turbines, and $1800/kW for coal. Adds credence to the merchant gen estimate of $3500/kW.
--Of COURSE people want subsidies if they can get them...wouldn't you? My guess is they are trying to slaughter the same AGW hog all the other "watermelons" (Green on the outside..red on the inside) are.
--The USS Nimitiz-class carriers cost $4.5 Billion, displace 90,000 tons and carry a 200-MW reactor. The USS-America (LHA-6) are similar but non-nuclear (powered by LM-2500 GT's, which in a civillian application cost about $1,000/kW) 70MW equivalent...displace 45K tons and cost $3 Billion. Obviously it's inappropriate to compare the two as some EPC costs are the same regardless of displacement, but that hints that it doesn't cost $6 grand/Kw....
The economic case depends on future fossil fuel costs. Those are unknown. And to some extent upon future nuclear fuel costs. Also unknown.
This is sure:
New nukes won't cost less if we don't build them to standard plans. And build repeatedly so that construction crews gain experience and the suppliers have some confidence in future orders.
Study, but don't place great importance on, reports about the first Chinese or Indian standardized constructions. Or from anywhere else. Very soon we will learn what it costs to build the next ones.
Yes, subsidies for solar power in Germany are ridiculous. Angela Merkel is in the process of cutting them. Solar stock prices are dropping as a result.
Germany has been the biggest buyer of solar panels in the world. Never mind that it is so far north with lots of cloudy weather. The subsidies drove the big build-up. The Germans would have been better off spending the money on insulation, hybrids, and nuclear power. Of course the Germans were going to phase out nukes. Angela Merkel is also intent on changing that. Not sure if she's succeeded yet.
I'm fascinated by the cost of nuclear power because so much is riding on it. Cheap nukes would make it cheap to phase out coal for electric power. Combine cheap nukes with cheap and high energy density batteries and we could replace most oil usage in transportation with clean nuclear power.
France is mentioned. Flamanville is in France. Also, the Finnish nuke that is so over budget and schedule is getting built by Areva - the French nuke builder.
France's experience is getting rusty. They built out to 58 nukes (versus the US with 104) and then stopped building for a while until recently.
Randall: I have always supported nuclear for generating power.
As a child I lived in Oak Ridge during WW2 and some of my relatives stayed in the nuclear industry for years. So I picked up my bias at the dinner table and have never felt edgy about the technology.
But I must confess I can't decode the cost figures of today either. Or why projects go years over schedule. (Both costs and schedules routinely go 200% over projections.)
Or why today we seem unable to even match what we built for over thirty years ending in about 1993. Of course we don't want to restart the past anyway, far better designs exist and I wouldn't miss the those old behemoths.
About all the layman can do is watch the news and apply a well tuned BS filter. As economic pressure grows the political obstacles will erode.
And as I said the costs and experiences of other countries will soon reveal a great deal. We have learned from the huge solar initiatives of Germany and Spain. (The Germans and Spanish learned too, perhaps more than they wished to know.)
The UK is now starting some huge off-shore wind farms. Next year I hope to be here to track how that is going.
The reason that nuclear plants can have such cost overruns is because the construction schedule is so extended. A plant built from parts requiring only final assembly on the site could have a construction schedule more like a wind farm. A molten-salt reactor or LMFBR feeding a supercritical CO2 turbine might be built in pieces small enough to be trucked to a site and essentially bolted together. With the actual schedule becoming more like the "overnight" figure of financial numbers, the potential for overruns should be greatly reduced.
One would expect the construction companies to oppose this, but given a world initiative to phase out coal they could make it up on the volume.
Someone is going to have to address the "broken rice bowl" issues of the coal industry (the miners and their families, even moire than the shareholders) and the states where coal is a big part of the economy. And soon. Blowing these issues off means you accept that they'll be nuclear's fearcest opponents, and for good reason. Significant new, additional deployment of nuclear energy will adversly impact the coal industry, in fact that is part of the argument for nuclear energy, that it will displace dirty, polluting coal.
E-P. I agree that the biggest cause of construction overruns is the long schedules which introduce great uncertainty.
And interest on borrowed capital must be paid or accumulate for years before the plants produce zero revenue. So longer schedules produce big costs.
The longer the schedule the more likely a future government will cancel everything. In the US opponents have used the courts to delay projects into oblivion (mixed metaphor alert). Foreign courts usually have less power to intervene.
I try to consider political uncertainty and there still seem to be too many huge errors in both schedules and costs. But only professional study could assess that. And some countries and/or bureaucracies are pretty tight about admitting true costs and outcomes.
Newer and standardized designs, firm political commitments, and volume are the only way to get nuclear going here.
Cost is always a relative issue. If you think new nuclear is expensive, imagine how much it will cost to run out of energy. Try to get a grip on the cost of solar/wind systems that include transmission, storage and back-up. My guess is that Nuclear will appear downright cheap in comparison.
Another consideration is that the evolution of nuclear power has been retarded by the anti-nuclear campaign. Only now are we starting to look seriously at a new generation of reactor designs:
www.gen4forum.org, gif.inel.gov, www.gen-4.org
Real world costs from 2007/2008 for greenfield 1,800MW reactor:
Cost for permit is $150m. Time for permitting process is 12 years, therefor you spend a huge sum of money and 12 years to determine the design and cost of the machine. If you go forward, the all-in cost to build will be ~$6,000 - $8,000 per kW and around 8 years for construction. So, at the end of the day you have spent billions of dollars and 20-years before you start generating revenue. Discount the cash flow and you need greater than $250/MWh for your energy. No way do you build such a power plant under a merchant scenario, only a utility would with full rate recovery guaranteed. The answer is to shorten the permitting process and promoting modular plant design.
Legislation passed in 2005 (2006?) reduced the number of permitting steps and decreased permitting risks. I would not expect 12 years for permitting now. Are you sure of your number?
28 billion has been pissed away on wind since the mid nineties, if that alone had been spent wisely, we would be looking at slipstream shots of failed crap and getting all giddy that we didn't do anything as stupid as the Europeans, instead we are looking at the legacy of sixties idiots now in their seventies.