"Difficulty shows what men are. Therefore when a difficulty falls upon you, remember that God, like a trainer of wrestlers, has matched you with a rough young man. Why? So that you may become an Olympic conqueror; but it is not accomplished without sweat." - Epictetus
I thought it might be of interest to discuss some fuel and engineering differences between nuclear-powered ships and nuclear power plants. Reactors for ships and power plants have very different needs. A ship needs enormous flexibility in power output, the ability to frequently shut down when the ship is is in port, and also requires very long periods between refueling - thermal efficiency is not high on the list of priorities, but compact size is.
On the other hand, a reactor for a power station needs the ability to generate incredible amounts of heat at the highest possible efficiency, for months on end, and it will be sized accordingly - they are quite large compared to naval reactors. A power plant will attempt to stay at 100% power output and maintain maximum power and thermal efficiency until shut down for the next scheduled refueling outage.
I'll do a brief overview of civilian nuclear power plants first.
One of the major design constraints in the civilian world is the limitation on fuel enrichment. Enrichment of the U-235 fraction in civilian power plants is typically 3-5%. 20% or greater enrichment is considered "highly enriched", and 20% is the point at which a very crude and heavy nuclear weapon could be made, if enough of it could be gathered together. Even though 3-5% enrichment is "low enrichment", it would still make excellent feedstock for a nuclear weapons enrichment facility, so even low enrichment fuel is carefully tracked and monitored. High enrichment in the civilian world is a no-no, so it's all low-enrichment, with lots of refueling. That works out OK though, because the rest of the power plant needs periodic maintenance too.
All reactors need quite a bit more fuel than what is required to achieve criticality - a self-sustaining nuclear reaction - and then up to full power. Reactors also require quite a bit of excess fuel to compensate for fuel burn-up, and to compensate for the accumulation of reactor poisons. How nuclear engineers deal with the excess reactivity and poison build-up in each type of reactor is the subject of this post.
Typically during a refueling outage in a utility-size reactor, about 1/3 of the fuel is replaced. The fuel at the center of the reactor is removed - it will be very high burn-up fuel, which we will discuss a bit later. The fuel at the perimeter of the reactor core (which has lower burn-up due to being on the edge of the reactor core), is moved to the center. Fresh fuel replaces the fuel around the perimeter, so that the fresh fuel is not exposed to the highest neutron flux at the very center of the core. This fuel management scheme is done to ensure the maximum possible fuel burn, while also not placing fresh fuel into a high neutron flux and causing localized overheating.
In a Pressurized Water Reactor (PWR) power plant, the control rods are completely removed from the reactor during operation, and reactivity is controlled by using a "chemical shim". In smaller reactors, control rods absorb neutrons, and are either "shimmed in" to reduce reactor power, or "shimmed out" to raise reactor power. "Shimming in" a control rod absorbs more neutrons to reduce power, and "shimming out" exposes more fuel to neutrons, allowing reactor power to rise.
One issue that a partially inserted control rod causes is "flux tilting". Flux tilting is anything that causes neutron density to not be consistent throughout the reactor core. When the control rods are partially in the core controlling reactivity and power level, the neutron flux near them is almost zero. So the fuel at those points is not producing any heat. For a given power level, fuel in other regions of the core has to run hotter, and that increases the chance of damaging the hotter fuel. Ideally, heat is produced by the entire core so that there are not hot zones or cold zones. This is why dropping a single control rod is baaaaad. The flux tilting is huge, and the best thing to do is shut down the reactor immediately.
A certain amount of flux tilting is inevitable. Neutrons will be more numerous in the fuel than in the fuel cladding or water channels. Neutron population also won't be consistent from top to bottom in a reactor, nor will it be on a side-to-side cross-section. There will be more neutrons at the center than the edge. That's just how the physics works in a non-homogenous, non-infinite reactor core.
Some flux tilting we have to live with, but for the control rods, there is a work-around, and that is a soluble reactor poison, also called a "chemical shim".
Below: Flux tilting. Neutron population is depressed in the fuel adjacent to a partially inserted control rod (solid lines). On the solid line at points "A", the neutron population and heat generation will be really high to compensate. With the control rod completely removed (using a soluble reactor poison to suppress reactivity), neutron flux is more evenly distributed, so fission is more evenly distributed, and so there is less chance of small areas with localized overheating.
Additionally, if all of the control rods are halfway down in the core, then only the bottom half of the core will generate any heat. Since you are only getting heat out of half the heat transfer surface of the reactor core, the bottom half will have to run much hotter than if the entire core were making power.
Below: Top to bottom neutron flux (and heat generation) inside a reactor core with partially inserted control rods. The lower half of the fuel is making all the heat, while the upper half is doing nothing.
The "chemical shim" used by large utility-scale reactors to completely remove the control rods is dilute boric acid. Boron has a large neutron absorption cross-section, and is slowly added to the reactor coolant as the control rods are slowly removed from the core. The boron atoms dampen the reaction even as more and more fuel is exposed to neutrons, allowing the entire core to generate heat. This configuration utilizes the entire heat transfer surface of all the fuel elements. As the fuel ages, the concentration of boric acid is reduced, allowing the reactor to remain at full power.
A utility size reactor can operate at 100% power for 1-1/2 to 2 years before refueling is required. Here is an interesting fact though: Only 3-4% of the fissionable U-235 atoms are burned up via fission during this period, meaning that 96-97% of the fuel is still available, but the reactor must be refueled at this point. Why can a utility reactor only burn 3-4% of the fissionable part (U-235) of the fuel before it needs a fresh load? Read on!
There is some funky stuff that goes on with nuclear fuel over time in a high neutron flux, and not all of it is fission. Since this fuel just 3-5 percent enriched, the remaining 95-97% is U-238. U-238 is pretty useless in a thermal reactor. It frequently absorbs neutrons as they are in the process of being slowed down, or moderated. This loss of neutrons is called resonance absorption, and it removes a lot of precious neutrons from the reactor. Then the U-238 (now U-239, since it has absorbed a neutron) decays to Neptunium 239, and then decays again to Plutonium 239, which is fissionable.
Toward the end of a fuel cycle, a significant amount of fission in the core will be due to the elevated levels of Plutonium 239. So we have isotopic changes occurring in the fuel. U-235 is depleting, and Pu-239 is increasing, but not by as much. Some of the Uranium ends up becoming Amerecium, which is a useless neutron absorber.
We also have neutron poisons. There are a number of fission products that are voracious neutron absorbers, which have been mentioned here before. Xenon-135 is the most significant neutron absorber, but several others combine to add up to a significant quantity of neutron poison. Neutron poisons are slightly different than other poisons I will address in the next section. Neutron poisons will tend to decay once the reactor shuts down - into some other isotope that will not absorb a neutron. With the reactor shut down, these neutron poisons will clear up over time.
Reactor poisons, (also known as reactor slag) unlike neutron poisons, are those fission products - neutron absorbers - that accumulate in the fuel that are stable isotopes. These build up, and never decay away, permanently reducing the available reactivity of the fuel in the core.
So to recap: A utility reactor that is intended to run at full power for 2 years will eventually accumulate enough neutron poison and reactor poison that it will not be able to run at full power, even though it has only fissioned 3-4% of the available U-235. It's a sorry situation, but that's how the nature makes the physics happen.
The Navy:
Nuclear-powered ships manage to get 15-30 years of operation out of a much smaller load of fuel. How do they manage to pull that off? The secret is two-fold. First of all, they don't run at 100% power all the time - there are no military vessels that need to operate at ahead flank at all times. Second of all, the fuel is highly enriched to ~93% fissionable U-235. This massive pre-load of reactivity compensates for the inevitable build-up of reactor poisons, delaying the need for difficult and expensive refueling.
It should be obvious that highly enriched fuel will tend to be really hot from a reactor control standpoint. With a fresh load of fuel, removing the control rods just a little would tend make the reactor critical. This means that only the bottom few inches of the core would be making any heat - see the second image above for an example. Localized heating would place the lowest part of the fuel elements and place them under excessively high thermal stress.
It is not practical to add boric acid to the coolant, as they do in utility-scale reactors, and remove the control rods all the way from the core. This reactor must be able to respond rapidly to all kinds of propulsion needs, from 10% to 100% power in a matter of a minute or so, and it would not be possible to control the reactivity of the fuel with circulating boron.
Instead of using soluble boron circulating in the coolant, we can get the control rods further out of the core by loading a burnable poison in with the fuel. This dampens the excess reactivity caused by the higher U-235 enrichment. The burnable poison will absorb a neutron and then be gone, and this happens as the reactor poisons build up. This keeps a core full of fresh fuel from being quite so "hot", allowing the control rods to be pulled further out of the core and provide more fuel heat transfer surface. It also flattens the reactivity of the highly enriched core over time, by burning out as the poisons accumulate.
Below: All the stuff that accumulates in the reactor core that combines to stop the nuclear reaction over time. You have to load a lot of extra fuel to compensate, but that makes the reactor run hot.
One favorite burnable poison for a reactor that needs operational flexibility is Gadolinium. It has several isotopes that can absorb a neutron, each with a different cross-section for neutron absorption. With the various isotopes of Gadolinium burning out at different rates, it has the nice effect that if you load the correct quantity, it burns up at about the same rate that reactor poisons accumulate.
You can load Gadolinium a bit heavier at the bottom of the core to compensate for the fact that the bottom of a fresh core will initially be doing all the work, so you can pull the control rods a bit further out to expose more of the core than if the Gadolinium were evenly distributed. You can also load it heavier in the center fuel elements of the core, so that neutron flux will be a bit lower at the center than it would be otherwise - so there is less flux tilting radially too.
Below: Blending solid Gadolinium in with the fuel will cause the reactivity of the core to be suppressed at the beginning of core life, and will disappear near the end, just as the poisons above accumulate. It's a very effective way to balance out the fuel's excess reactivity over core life.
One of the design considerations of any reactor is the "hot channel". The hot channel is an assumed hot-spot in the reactor core, where all of the worst acceptable manufacturing limits have managed to all land in one theoretical place. They assume the "hot channel" has the highest heat flux, smallest acceptable coolant channel (and flow), the fuel has the worst allowable fuel/cladding bonding, etc. Basically the worst-case possible that could still pass inspection.
The "hot channel" has to not have a fuel element failure under a wide range of accident scenarios, and so the hot channel sets the limits of how hard we can run our reactor. This is why flux tilting is such a big deal, and one reason why burnable poisons are added to a fresh reactor core.
1 comment:
Interesting
Post a Comment