Reply #10: Concrete doesn't contain gases at high pressure (well at least not in the real world). [View All]

"It is also clearly different than any other design in that the concrete is not performing the function of containing gasses"

Once again you are wrong. It is clearly different because it can be passively cooled (with no human operators and no electrical power) however concrete missile shield has never had the job of containing gases that is the job of the steel liner.

Concrete doesn't contain gases, certainly not gases at high pressure, it is porous.

NRC regs requires a containment building handle a variety of stresses:
1) contain overpressure for steam explosion
2) handle accidental or intentional impacts from outside (missiles, aircraft, natural disasters)
3) prevent accidental release of nuclear material in event of a failure inside containment.

As such the containment building consists of multiple layers. Now in the past the design simply put the steel layer and concrete layer together. However nothing in the regs requires that. If concrete could prevent release of gases there would be no need for expensive steel liner to begin with. Just build the whole containment out of concrete.

You make the false assumption that concrete layer is used to prevent escape of gases and that leads you to false conclusion. The concrete layer is designed to handle high speed impacts (missile shield). The steel layer is designed to prevent release of radioactive material. Combined they are provide enough strength to withstand a worst case overpressure (about 10 atmospheres of pressure).

Now the AP100 design has some disadvantages in terms of cost. By putting steel layer and concrete layer together you get improved strength with less material (which is why all previous reactors do). Since AP1000 seperates the layers both now need to be thicker, heavier, higher strength to meet the NUREG-1150 requirements. The thicker concrete missile shield isn't that big of a deal but the AP1000 requires roughly the steel containment to be about 3x as thick. Steel is expensive especially high tensile strength corrosion resistant steel. So why do that? Why spend MORE money on steel when trying to lower cost of nuclear power?

Because existing reactors have one (although highly improbably flaw). In loss of all electrical power fission can still be stopped but decay heat will eventually melt the reactor. By having the steel layer exposed to air it acts as a giant heat sink, this is added by cooling the steal with water (via gravity) and air via natural draft. In an overpressure event water from emergency cooling tank cools outside of steel layer and heat is transferred out of the reactor PASSIVELY. The water tank and passive design can cool decay heat without any human response or even any electricity on site (to run pumps).

You love to bash how new reactors are more expensive. This is an example of why. The newer design is more expensive and It is unlikely this passive cooling will ever be needed however it provides a final line of defense beyond what existing reactors have.

So:
AP1000 w/ hole in steel liner and radioactive material loose in containment = leaks
Gen II PWR w/ hole in steel liner and radioactive material loose in containment = leaks

Of course both those scenarios ignore the fact that containment is not radioactive under normal operations (and not even radioactive under most emergency situations).

So
1) an accident would need to occur
2) fuel assembly would need to breech (likely from overheat)
3) the reactor would need to breech or be intentionally depressurized
4) the containment would need to have an undetected hole in steel liner

If even one of those things is not true then by virtue of defense in depth radiation would note be released from structure.


Funny thing is NRC tests the steel liners of containment structure by over-pressurizing containment building. If there is a hole pressure will drop and the rate of pressure drop helps to estimate the size of the hole. If concrete contained gasses at high pressures well then that test would be kinda useless huh?