View Full Version : How does a nuclear reactor's energy scale up?

Philippe Lemay
2012-Apr-13, 03:50 AM
I remember discussing with a friend that in the future things like fission (or possibly fusion) reactors will be easier to build and control as we learn to build mega-structures. Reactors and installations much larger than the ones we see today. Part of the discussion was about how spacecrafts (or at least space stations) could utilize nuclear power more easily because they wouldn't be constrained by space like we are here on Earth. Our reasoning was that for a given source of heat (like a fusion plasma or a fissioning uranium source) the more surface area you have to bleed the heat away, the more efficient your system can be. A reactor 2-meters across would have to be made of tougher stuff than a reactor 5 meters across, or use a fuel that doesn't run as hot. And a 50 meter reactor could hold a much much hotter fuel, because it would have all that extra surface to bleed away the heat. And I remember reading that nuclear reactors can be made more efficient when they get very hot.


I recall talking to a nuclear engineer who mentioned to me that the reactors in nuclear submarines actually operate less efficiently than those in land-based facilities (despite having more efficient, more enriched fuels) because they are so compact and so small. It goes from 33% on land to something like 25% in subs, from what I recall. So this at least proves some of what me and my friend were theorizing. I think it has to do with heat-flux of a surface, and the amount of energy the material that surface is made of is able to withstand. But I recently read an article about fusion power that has me a bit confused. It states that as a reactor scales up, the surface area of the container squares, but the volume within cubes. So a larger more spacious reactor actually holds more plasma, more heat, and would thus be harder to contain.

Both these lines of reasoning seem to make sense, but they seem self-contradictory. Would a reaction get easier to maintain as it's container got larger (and the only reason we don't build them to be so huge is engineering limitations), or do they get more energetic as they get larger, and it actually becomes fundamentally more difficult to contain the energy?

2012-Apr-13, 05:16 AM
Well first off draw a sharp line between fission and fusion. One of the critical factors apart from heat flux with fission reactors is that neutrons escape. More core volume = fewer wasted neutrons = easier chain reaction. Whereas for fusion it is often down to the fine control of the magnetic field. Creating and sustaining the large magnetic fields (either to contain or induce the pinch) in the face of larger and larger volumes of plasma gets to be a problem.

Ara Pacis
2012-Apr-13, 08:21 AM
Another issue may be internal pressure. An engineer or physicist should may be able to elaborate, but I seem to recall that as a structural envelope gets larger, the internal pressure acting upon it increases. Or perhaps that the point of the scaling comment, that you need to increase the thickness of the structural envelope to compensate. Of course, this can vary depending on other things, such as the working fluid. Water and steam may behave differently than liquid sodium or helium.

2012-Apr-13, 09:02 AM
The efficiency of commercial nuclear reactors (33%) has more to do with the steam turbine, not the reactor, I believe. There are also substantial issues with reactor physics as you scale up, even from the size of a submarine core to a commercial core. Has to do with neutron travel distances and such.

2012-Apr-13, 01:59 PM
Shortly after a lecture on the dangers of scaling structures, the Ferry Bridge UK street of cooling towers suffered a major collapse. This was because the aerodynamic loads, ignored on smaller versions, were sufficient to break the walls. Rather similar to Tacoma Narrows USA bridge disaster. BigDon's question about scaling up cooking recipes going badly wrong also relevant. Scaling up is much more complex than just working out the loads, some effects scale as high and sometimes unknown
powers of the linear dimensions. Just to cite one nuclear example the neutron embrittlement of steels is highly non linear and is a case where future replacement is assumed or limits the life of the whole plant. It would be a brave step to scale up without the necessary experiments.

Philippe Lemay
2012-Apr-21, 12:12 AM
So... if I'm understanding this correctly - for fission at least - larger reactors would be more efficient once they are built. But the design, construction, and maintenance of them would be more difficult than smaller-scale modern stuff. So ultimately it is an engineering limitation?

Is that right?

2012-Apr-21, 05:13 AM
Engineering and practicality - in order to keep the reactor power output fairly constant over its life they play fun games with the fuel. This requires that when you refuel you swap around a lot of the fuel assemblies. A monster of a reactor with either huge fuel arrays or many, many of them would be really awkward to refuel (requiring huge lifting gear or a long time).

The main arguments I would put forwards against huge single reactors would be:
Refuelling practicality
Single points of failure (if anything breaks down you lose a lot of energy production)
Engineering issues as mentioned
Amount of waste produced at a single site
Safety - consequences of a single accident are larger

2012-Apr-21, 09:38 AM
So... if I'm understanding this correctly - for fission at least - larger reactors would be more efficient once they are built. But the design, construction, and maintenance of them would be more difficult than smaller-scale modern stuff. So ultimately it is an engineering limitation?

Is that right?

I'm going to have to say no, at least not because of anything presented here.

You seem to be under the impression that larger reactors would necessarily run hotter, due to inherent self-insulating properties, and that would make them more efficient. As I alluded to above, the efficiency of nuclear power plants is chiefly a function of the turbine, not the reactor. Also, I don't think you can assume larger reactors would run hotter anyway. As I also alluded to, the physics of the fission process would dictate optimum core temperature. It might turn out to be lower for larger reactors.

2012-Apr-22, 04:32 AM
In 1994 I worked with a Los Alamos physicist and one day we were discussing the future of fusion reactors. His learned opinion was that in order for a fusion reactor to be practical as a power plant it would have to be scaled-up to where one reactor would power the entire US or even the entire North American continent.

If that turns out to be true, then the biggest engineering problem will be power distribution from the central location.