Questions about reactor cooling systems

[TheVat]

This from an AP News report prompted me to wonder about the apparent need for an external electricity source in many nuclear plants to run the coolant pumps...

Quote

One major concern, raised by Ukraine’s state nuclear regulator, is that if fighting interrupts power supply to the nuclear plant, it would be forced to use less-reliable diesel generators to provide emergency power to operating cooling systems. A failure of those systems could lead to a disaster similar to that of Japan’s Fukushima plant, when a massive earthquake and tsunami in 2011 destroyed cooling systems, triggering meltdowns in three reactors.

So, with these designs, if you lose grid electricity, you go to diesel backup, and if that fails, you better have batteries.

What I have never fully understood is why designers, from the start, wouldn't look at this system producing massive steam and think, for safety's sake, we will put in some emergency shunt to a steam turbine that could provide emergency coolant pumping.  So the pumps run directly off the plant's heat.  As it cools, there is less steam, and there is less need of the steam to run the emergency pumps. 

I'm not suggesting the plant run off its own power, and understand the problems with that.  But the ability to do so temporarily, in an emergency, would seem useful.  Diesel generators can get flooded with seawater, broken by crazy or poorly trained armies with artillery, etc.

Edited March 4, 2022 by TheVat
Fixd

[swansont]

Money is probably a factor. You need to buy these items and maintain these systems, both of which cost money. In a pressurized water reactor a turbine is in a second loop. You have more opportunity for leaks with each new penetration. Plus the fun of either potentially having a thermal shock if the loop is cold and all of the sudden you fire it up, or if you keep it hot you are wasting some of the generated heat, making the plant less efficient. 

It's risk/reward, which is skewed by the nature of the beast: you want to make money, so a certain amount of value engineering goes on, cutting back on costs that are deemed unnecessary. Are four layers of redundancy required, or can you get away with three?

And if you are having problems with the reactor, perhaps it's best not to rely on the reactor itself.

Some reactors can use natural convection and not rely on pumps, but it might not work on a commercial scale plant. Decay heat can be something like 7% of full power, so if you have a 1 GW plant, that's 70 MW you need to remove - that's a lot of water that needs to be moved, and you aren't pulling the energy out by driving a steam turbine.

I'm not sure how much "broken by an invading force" gets considered in the design of a plant.

 

[TheVat]

Thanks.  Had not thought much about the decay heat, especially in the first hour or two.  That's a pretty significant burn with the short lived isotopes.  

And money does seem like a factor, for sure.   Putting that cost into high quality batteries might make more sense.  

[OldChemE]

Having worked on Nuclear plant design in the past I agree the steam idea would be neat.  Diesel generators are used because regulatory agencies require protection against multiple failures.  Typically this means two or three separate systems separated so that no single accident can prevent all from working (unless, of course the designers totally underestimate the potential of a tsunami).  I think steam would be great for one of the systems-- provide the steam is not radioactive.  I do not know if the Ukraine plant has radioactive steam.  Typically, Pressurized water reactors have clean steam-- at the expense of the complex pressurization system that has its own drawbacks.  Boiling water reactors, being simpler, derive their steam directly from the reactor-- at the expense of having radioactive steam.