December 07, 2007
Pebble Bed And Other Gen IV Nuclear Reactor Designs

An article in Popular Mechanics examines generation IV nuclear reactor designs now under development and reports that pebble bed reactor designs look likely to get built before other Gen IV designs.

Kevan Weaver, like most of the lab's 3500 employees, works in a sprawling group of campus-like buildings on the outskirts of Idaho Falls. Standing in his third-floor office, the fresh-faced nuclear engineer holds what could be the future of nuclear power in his hand: a smooth graphite sphere about the size of a tennis ball. It could take years to weigh the pros and cons of all six Gen IV designs, Weaver says, but Congress can't wait that long. In addition to replacing the aging fleet of Generation II reactors, the government wants to make progress on another front: the production of hydrogen, to fuel the dream of exhaust-free cars running independent of foreign oil.

As a result, the frontrunner for the initial $1.25 billion demonstration plant in Idaho is a helium-cooled, graphite-moderated reactor whose extremely high outlet temperature (1650 to 1830 F) would be ideal for efficiently producing hydrogen. There are a couple of designs that could run that hot, but the “pebble bed,” so named for the fuel pebble that Weaver holds, is attracting particularly intense interest.

A typical pebble-bed reactor would function somewhat like a giant gumball machine. The design calls for a core filled with about 360,000 of these fuel pebbles--"kernels" of uranium oxide wrapped in two layers of silicon carbide and one layer of pyrolytic carbon, and embedded in a graphite shell. Each day about 3000 pebbles are removed from the bottom as fuel becomes spent. Fresh pebbles are added to the top, eliminating the need to shut down the reactor for refueling. Helium gas flows through the spaces between the spheres, carrying away the heat of the reacting fuel. This hot gas--which is inert, so a leak wouldn't be radioactive--can then be used to spin a turbine to generate electricity, or serve more exotic uses such as produce hydrogen, refine shale oil or desalinate water.

The ability to make hydrogen more efficiently will only matter if and when we find better ways to store hydrogen. Since Gen IV reactor designs are easily a decade away from initial use in commercial reactor construction better methods for storing hydrogen will become available by then.

The biggest promise of pebble bed is much more rapid construction. A substantial part of the cost of nuclear power is the interest cost of reactors when they are only partially constructed. If a reactor takes 5 years to build then the portion of the cost spent in the first year doesn't start earning back on its investment for over 4 years. That period during which the capital equipment is sitting idle while the rest of the plant gets constructed makes nuclear power far more expensive.

The article claims a new demonstration reactor build decision won't be taken until 2014. So the development of new nuclear reactor designs seems really slow. Does it have to take that long?

Share |      Randall Parker, 2007 December 07 12:00 AM  Energy Nuclear

Wolf-Dog said at December 7, 2007 5:05 AM:

The real problem is that every year, ridiculously small amounts of money are being allocated to nuclear electricity research. Just compare these funds with the astronomical sums that are being wasted in Iraq and Afghanistan. If $50 billion per year were spent on alternative energy since 2001, we would now have a solution.

Steve said at December 7, 2007 7:33 AM:

Once machine intelligences can design our machines, design time will decrease. Remember--the regulatory load is just one more mega-ton of constraints that designers must take into account.

All of the constraints in nuclear design are killers--there is no quick fix. It takes a sophisticated computational machine to balance all of the constraints in a reasonable time period.

To check the design work of a machine intelligence, will require another machine intelligence.

Brock said at December 7, 2007 9:03 AM:

Even if we had the designs proven today, it takes years to build a nuclear reactor, and as Randall said, we don't have the storage mechanism or batteries to use nuke-produced electricity or hydrogen efficiently. In the interim we need the oil.

I recall reading somewhere (probably here on Futurepundit, though I can't find it) that pebble bed reactors are the least efficient fuel usage of all Gen IV designs. Has that been addressed. I would hate to replace "Peak Oil" with "Peak Uranium" or what have you ...

Wolf-Dog said at December 7, 2007 9:19 AM:

If the US government had allocated just $25 billion per year to battery research and development since 2001, the new batteries would have been ready now. Already there are some electric sport utility vehicles with ranger over miles, but these are expensive batteries that are not in mass production. But had a national effort started on September 12, 2001 for both batteries and alternative energy, then electric cars would have been much closer to mass production.

It is true that Pebble Bed Reactors are very inefficient for uranium fuel usage (I was among the ones who said at Futurepundit that Pebble Bed is inefficient for fuel usage). But on the other hand, since the construction of the modular Pebble Bed Reactors is very fast, probably less than a year because the pieces would be manufactured in a standardized factory and then shipped to be assembled very fast, there are advantages to Pebble Bed reactors. Additionally, reprocessing plants would simply extract the plutonium and also unused uranium, as well as other long term fissile materials from nuclear waste and then add these to nuclear fuel to be burned in reactors once again.

The fast neutron reactors are nearly 100 times more uranium efficient than regular reactors because these are breeder reactors that can use even the non-fissile uranium to make plutonium to be burned in the process, in addition to the fissile uranium.

But seriously, if the US pulls out of Iraq next year, there will be some $100 billion per year to spend on energy research and development. Say $50 billion per year for nuclear reactor research, $25 billion per year for battery research, and another $25 billion for alternative energy including solar and biomass.

K said at December 7, 2007 12:02 PM:

Does it have to be this slow?

No. Expense expands to the limits of funding. Completion is never pondered. Those phrases, stolen from Parkinson? and mangled by me, bear upon our problems.

Today these programs are not funded to achieve anything but to decide if the proposal is perfect. Adequate is never adequate since some improvement or risk can be always be conjectured.

Every detail is to be studied and theoretically refined for years or decades. Then funding and attention will be shifted to a pleasingly new concept which has made those old ideas obsolete anyway. Motives, funding, and directions change with each Congress, and Presidency. Quite frankly there is seldom anyone involved who gains from finishing.

This pattern is hard for engineers and scientists to see, especially in their own work. They are trained to look for solutions and improvements, not to assess how infrastructure and economics and management affect society. To grasp it, look at city transportation systems, aka rapid transit.

In the 19th and 20th centuries our major cities routinely built elevated rail and subways. Some are still in use today with little change. Bus service was regular and reliable. Five years was a very long project, ten almost unknown.

Today most of our cities spend billions, much of it federal grants, and take decades to produce a few miles of light rail. It is happening now in greater Phoenix where I live. The reasons are not technical but because few benefit from completion. And many - managers, politicans, contractors, planners, and the workers - benefit when overruns and delays extend the work. The odds on a big publicly funded activity being cancelled for poor performance are lower than your risk of being hit by lightning.

Randall Parker said at December 7, 2007 6:39 PM:


One of the purposes of pebble bed is faster construction. Nuclear would become a whole lot more appealing if a utility company could go from decision to full operation in a couple of years. Costs and financial risks would go way down as many components could get manufactured on an assembly line and shipped to sites

Efficiency: Not sure.

What I really want to know: Is there any obstacle to using a mix of thorium and uranium in a pebble bed reactor?

aa2 said at December 8, 2007 12:24 AM:

K - great post.. Our society is at a point where even what was routinely built before we are having trouble building. And of course its not lack of technical ability or workers with the skills, its the endless bureaucracy. It seems all nations go through phases. I look at us like the Soviet Union in the late 70's maybe.

Meanwhile other nations are at different phases of the cycle. Like China can just smash in solutions when it faces challenges(while at the same time working on bringing the modern world to 100's of millions of persons). They have something like a big building built and in use, while we are still in the approval stages.

Wolf-Dog said at December 8, 2007 7:27 AM:

It turns out that operating expenses constitute most of the cost of nuclear electricity, and uranium fuel is only a small fraction of the cost of electricity. Now if Pebble Bed reactors become widespread, the money that is saved, can easily finance the construction of a few reprocessing plants to make more fuel from spent fuel, dramatically increasing the net fuel efficiency.

Yes, thorium can be used in Pebble Bed Reactors, but more research and development is needed to make it happen within a few years. If only $50 billion per year were allocated after September 12, 2001, then these Pebble Bed reactors would have already been ready for mass construction.

Here is a list of countries with a lot of thorium. The United States is number 2 on the list,
and India is working very hard on thorium reactor research, since it has a lot of thorium:

Paul F. Dietz said at December 9, 2007 9:33 PM:

The ability to make hydrogen more efficiently will only matter if and when we find better ways to store hydrogen.

Funny, millions of tons of hydrogen are already made each year, without there being 'better ways to store hydrogen'. Many uses of hydrogen are continuous industrial processes that do not require a great deal of storage. If we find a better way it could be plugged into the existing hydrogen infrastructure, including the gulf hydrogen pipeline operated by Praxair.

It turns out that operating expenses constitute most of the cost of nuclear electricity, and uranium fuel is only a small fraction of the cost of electricity.

No, capital costs (including cost of financing), not operating expenses, are the dominant part of the cost of nuclear electricity.

Randall Parker said at December 9, 2007 10:03 PM:


The demand for hydrogen in industrial processes probably is not enough to incentivize the development of nuclear reactors that operate at much higher temperatures for hydrogen production.


Paul is correct about capital costs and nuclear power. Capital costs are much more than half of total nuclear electric costs. That's why nuclear only makes sense for base load demand. Once you build a nuclear power plant you pretty much want it to produce max power continuously.

Curiously, a nuclear power plant that could produce hydrogen would make nuclear power usable for non-baseload. Make the hydrogen at night. Make electricity during the day when demand is higher.

Wolf-Dog said at December 10, 2007 4:57 AM:

When electric cars become widespread, then the demand for electricity will go up sufficiently to make nuclear energy be superior to coal. In fact, when cheaper and more efficient reactor designs become available in the future, then the capital costs and operating costs of nuclear reactors will not matter any more. For the immediate future, Pebble Bed reactor is supposed to be a major step forward for reducing the capital and operating expenses. And in the more distant future the molten salt reactors and the Integral Fast reactor, would be even more efficient, but government money is needed to develop these.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©