Mechanical Nuclear coolant pumps built on turbine to generator shaft?

[David Christo]

Hello, I'm new to this board. I have a engineering question about the coolant pumps used to maintain reactors.

I have been studying the basics of a PWR.

I notice the pumps are fed mostly by electricity.

While I see the logic and efficiency of that..I am left wondering why the space on and between the turbine to generator shaft is not used to provide also failsafe mechanical pumps?

In the event of a electrical power loss to the pumps such as Fukushima, or a pump failure..all you would have to do is open a valve and let the reactor regulate and cool itself down mechanically driven by its own turbine to generator shaft.

Is this practical or worth pursuing further?

Thanks,

David

 

[Astronuc]

Fukushima units are BWRs, not PWRs.

PWR coolant pump are electrically driven since they are inside containment while the steam turbine is outside of containment in the auxilliary/mechanical building. The primary cooling system is closed. The steam generator separates the primary and secondary cooling systems. Some plants have steam driven feedwater pumps, but I'm not sure about PWRs. The feedwater is part of the secondary cooling system and goes to the steam generators, which do act as a heat sink.

 

[rmattila]

Also, in Fukushima the main turbines tripped and the shaft stopped right at the earthquake because the acceleration exceeded the turbine trip level.

 

[David Christo]

Thanks for the explanations about that. If the turbines stopped mechanical pumps wouldn't be much help I see.
Just a thought.

 

[QuantumPion]

PWR's have steam-driven aux feedwater pumps called terry turbines.

 

[Hiddencamper]

Hello, I'm new to this board. I have a engineering question about the coolant pumps used to maintain reactors.

I have been studying the basics of a PWR.

I notice the pumps are fed mostly by electricity.

While I see the logic and efficiency of that..I am left wondering why the space on and between the turbine to generator shaft is not used to provide also failsafe mechanical pumps?

In the event of a electrical power loss to the pumps such as Fukushima, or a pump failure..all you would have to do is open a valve and let the reactor regulate and cool itself down mechanically driven by its own turbine to generator shaft.

Is this practical or worth pursuing further?

Thanks,

David

So first off, I believe what you are talking about are injection pumps, things which put water into either a steam generator (PWR) or reactor (BWR).

For PWRs, the auxiliary feed system does this. Aux feed consists of pumps which are either motor driven, electrically driven, steam driven, or some combination of the above. Typically there are anywhere from 2-4 aux feed pumps per PWR. These are low capacity high pressure pumps which are capable of injecting sufficient water to the steam generators with the reactor shut down or at low power, even with the steam generators at high pressure.

For BWRs like Fukushima, almost all Boiling reactors have either RCIC and/or HPCI. RCIC is a steam driven pump which uses reactor steam to pump water in. HPCI is a large steam driven pump. (RCIC = Reactor Core Isolation Cooling, a 600ish gpm pump. HPCI = High Pressure Coolant Injection, a ~3500-5000 gpm pump, which was replaced with the motor driven High Pressure Core Spray in later models of the BWR. Many BWR plants have both RCIC and HPCI).

These steam-turbine driven pumps are all Terry turbines. Terry is a company that made steam turbine pumps. They were bought out by Dresser-Rand, who now manages maintenance of Terry turbines throughout the world.

Some overview on RCIC/HPCI turbine driven pumps. The pumps can draw water either from the condensate storage tank (CST), OR from the suppression pool. The CST stores reactor grade water and is reserve water for the condenser or reactor. The pumps will auto or manually transfer from the tank to the suppression pool if the tank level is too low, or the pool level is too high. The turbines on these pumps need to exhaust their steam somewhere. They exhaust it to the suppression pool (which is where the reactor safety valves also exhaust their steam). The turbines are self-cooled by their own water supply, which during an accident is coming from the suppression pool. In other words, as long as my suppression pool has adequate water to quench steam, and supply water, and the water is not boiling, my RCIC/HPCI turbine will function. Over time, if I don't cool the suppression pool through some means, RCIC/HPCI will fail, and is what we saw at unit 2 and 3.

At Fukushima Daiichi, unit 1 was equipped with a HPCI (and a passive steam generator called IC), units 2 and 3 both had RCIC and HPCI. These systems require DC battery power to start up. Once they are running, DC battery power is used to control their flow rates. In the event DC power is lost, the pump will continue to run "As-Is", and may over-fill itself with water. Interestingly enough, without DC power, unit 2's RCIC managed to maintain a sort of steady state until its failure. RCIC at unit 2 maintained cooling for 70 hours. At unit 3, RCIC and HPCI together lasted for around 32-36 hours. RCIC failed, HPCI was manually shut down and could not be restored. Unit 1's HPCI never initiated, DC power was lost prior to being able to start it up, and it was not already in use for cooling prior to the loss of DC power.

Now you ask, why can't we drive cooling pumps off the main turbine. There are a few reasons. For a boiling water reactor (The type that Fukushima is), my main turbine is outside of the containment, and the largest pipes going from the reactor to outside of containment are the main steam lines, which supply steam to the turbine. During an accident, I need these main steam lines sealed, to prevent water loss from the reactor, and to ensure that no line break can cause a direct leak of radioactive material from the core to the environment. This is the most direct answer to your question, if I were to just have a turbine driven pump, that means I could be porting reactor steam containing fission products to outside of the containment. Even if I did have this option, where is the water coming from to pump into the reactor, and how will I condense the steam? My condenser is already lost, so I can't use the condenser, I would have to vent it to atmosphere (Very bad idea). As for water supply, the condenser hotwell generally has <70,000 gallons of capacity (and is usually kept 1/3rd full), so it's not going to buy a lot of time for injection. I could have this extra pump hooked up to the condensate storage tank like RCIC, but then I'm in no better situation than RCIC, when it runs out I have no water source. RCIC/HPCI can at least draw water from the suppression pool.

Now some other reasons you don't want a pump driven by your turbine. First up, the turbine is extremely sensitive and finicky, if you don't have power for all its auxiliary systems you will damage or destroy it in short order. Secondly, the reactor cannot supply enough steam through decay heat to drive enough turbine load. After a reactor scram, my turbine (I'm at an 1150 MWe BWR) will trip off in 5-7 minutes on its own, as there is not enough decay steam to run it, and the generator will reverse power trip. Third off, the turbine is controlled by instrument air and hydraulic systems, without those systems, the turbine cannot function and will shut itself down. So now I would need to find ways to ensure those systems are available post accident. Fourth off, none of the equipment in the turbine system is designed or credited for post-accident. It's all assumed to be destroyed post accident. Another consideration, large power generation turbines require extremely precise tolerances for operation. Post-earthquake, a large power generation turbine would be unable to safely function (at least without engineering and maintenance performed on it), which would have made it useless in Japan's case.

So ultimately, for reactor safety, we want to rely on simple, safety grade equipment, that is located in the reactor containment or reactor building. Relying on things outside the containment is dangerous as it creates radiological release paths, and also is very unlikely to function. All plants have some form of non-electric cooling for loss of normal heat removal (IC/RCIC) situations.

If you have any questions let me know. I know there's a lot of info here, but I hope its a decent crash course in some of these plants ECCS design.