APR: your source for nuclear news and analysis since April 16, 2010

Tuesday, April 12, 2011

Boiling Water Reactors... some further details.

Now that we're seeing a further media explosion about the Fukushima Daiichi accident being raised to an INES Level 7 event, there are sure to be a lot more questions... and a lot of implications to US plants and plants everywhere else. It's time to take a little closer look at some of the features of the plants there at Fukushima as best we can ascertain from here.

Readers of this blog already know a lot about the setup of the relationships that have to be arranged when a nuclear plant is to be built. If you did not see the article on that, then please click here and read this first. Remember that different architect-engineers will design buildings and structures for a given job differently, and specs provided by the owner-operator can vary. Further, plants built over a span of time have to comply with newer and newer regulations and some of the new requirements are also fitted to older plants when necessary.

We've also talked a bit in the past about containment designs, and shown the evolution of BWR containment. If you are not familiar with this, then click here for a good primer on this, and then click here to see some more information. Readers might also want to revisit our most popular online article on the BWR DRY WELL which will cover the function and construction of this feature.

With that background behind us, let's stop a moment and note that while there are many mentions of the containment designs relative to Fukushima, and elsewhere (these are the MkI, the MkII and MkIII Boiling Water Reactor Containment) there is only scattered mention of the actual reactor design. Actually, this has gotten a bit more coverage than one might think because every single JAIF status chart from very early on until now has listed the core type of the plants at Fukushima Daiichi, as separate from the containment types. Fukushima Daiichi No. 1 is a BWR/3 reactor core design; Nos. 2 through 5 are BWR/4 reactors while No. 6 is a BWR/5 reactor.

These are generational designations created by GE to describe, generally, design features of a reactor core, core vessel and some supporting systems. There were a variety of power ratings for each style of core, using more or less fuel elements so that simply seeing a "BWR/4" designator doesn't necessarily imply the actual core thermal power rating.

There are many differences between the actual reactor models, although after a time some parts are standardized (especially concerning control rod drive.) However, from BWR/3 on, jet pump core recirculation was used and since all involved plants there are this model or newer, let's begin right there with a look at the actual core flow, and in-vessel design in normal operation at one of these reactors. We'll use a simplified WASH-1250 schematic for core flow in a jet pump BWR reactor.

Following the path of water flow beginning at the feedwater line, we see that this water enters the pressure vessel though the feed nozzle (a term we've seen here a lot, since it's one of the few working temperature indicators at the plants) and moves down around the outside of the core barrel into an annular plenum space. Some of the water goes all the way down and is taken into the suctions of the recirculation pumps, where it's pressurized and blasted down through a nozzle pointing into the throat of the jet pump inlet. This is where the rest of the feedwater goes; this jet pump puts a strong suction on the throat and sucks in feedwater from the plenum, which then enters the lower area of the reactor vessel below the core.

As the water then rises up through the core, it is heated as it passes through the core assembly. Some of the water begins to boil, and more and more of it is boiling as the mixture reaches the top of the core. This violent mixture of steam and water is separated by several series (not shown in detail in this drawing) of moisture separators that use either a swirling, centrifugal action or else baffles to make the water droplets fall out of the steam. The now fairly pure steam exits the top of the moisture separators and exits the vessel through the steam line, while the saturated water falls back to be mixed with the incoming feed water in the outer plenum, helping pre-heat this feedwater.

Now, this is a very simple drawing. Let's take a look at another one, from WASH-1082 that shows a BWR of the vintage we're concerned with at Fukushima Daiichi to see how this theoretical design is actually manufactured.

We can see the feedwater inlet on the right, high up, and the steam outlet on the left of this drawing, also high up. The jet pumps can be made out standing in the space between the core barrel and the pressure vessel; see the paired venturi type jet pumps on the left. The recirculation pumps are of course electrical, and are right near the reactor vessel, inside the dry well.

Let's point out one technical fact here; One of the aspects of this design worth noting is the fact that if anything in the recirculating line breaks, allowing water to exit through this line, then this can only drop water level IN THE CORE as low as the top end of the jet pump inlet nozzles. This height is about 2/3 of the way up the active core region. As we will see, there are quite a number of other pipes that tap off the recirculating water piping. This could well explain some of the problems with core cooling at Fukushima Daiichi's reactor plants... if any plant has any leakage or failure in this piping, it could leave up to 1/3 of the core upper end not submerged in water. We already know that the upper portions of the core on all three plants are damaged, and in fact No. 1 plant is said to have over 2/3 of the core damaged.

Also, in the above drawing, look for the core plate. This is a heavy plate lower in the reactor vessel, with round holes in it. The label is on the right. This is the plate on which the fuel assemblies sit. This plate is pivotal in core melt scenarios; one major question being "Does the core melt cause failure of the core plate allowing core material to drop to the bottom of the pressure vessel or not?" If a large amount of the core fuel assemblies melt and drop onto the plate, it could conceivably fail. Further, it's also obvious that melting and demolishing of parts of the core could easily lead to blockage or restriction of water flow through the core itself.

Now look at the very base of the drawing, showing the heavy support on which the reactor sits; inside this are the control rod drive mechanisms. Early on, flooding of the drywells was discussed a great deal (this could take up to 6 hours or so to accomplish just up to the reactor vessel support skirt) and here we see one thing that hurts the ability of this action to remove heat from the reactor; air would be trapped in this area as the dry well filled, limiting the height to which water would enter the area of the vessel support skirt on this design.

Let's take a look at yet another picture of a containment building, again a Mk I containment housing a BWR/4 reactor. This is probably the closest illustration we've yet put up to the actual structures at most of the Fukushima Daiichi plants.

This is from WASH-1082. The building foot is 14 feet below ground. The refueling floor, commonly mentioned in articles about Fukushima here and elsewhere and clearly marked on this drawing is 149 feet above the building foot level.

In our next installment we'll look at some of the actual fluid systems involved to get a better idea of what TEPCO is trying to restore access to, and use of.

3:00 PM Eastern Tuesday 4/12


  1. Very helpful and very informative.
    The schematics make the possible negative consequences of using salt water to cool the reactor stunningly obvious.
    There was an earlier estimate that as much as 50,000 pounds of salt might have precipitated out in each reactor, equal to several hundred cubic feet.
    Is there any modeling on the consequences for the cooling flows and is there any likelihood the current fresh water injections can alleviate this situation?

  2. There is a lot of modeling that's been done, and it sure seems clear that either salt deposition or core demolition is inhibiting core cooling capability. The fresh water is not likely to be able to dissolve the salt, at the kinds of pump pressures and velocities and temperatures being used now.