Grid or Generator failure - Why cant you stay online?

[Bignose]

In the case of a power grid or generator failure, the former which chernobyl was testing for and the later which occurred at Fukushima - why can't you stay online, use the electricity your producing to power the pumps and sink the excess electricity into the ground? I have no engineering backround and am quite baffled by this.

 

[Astronuc]

In the case of a power grid or generator failure, the former which chernobyl was testing for and the later which occurred at Fukushima - why can't you stay online, use the electricity your producing to power the pumps and sink the excess electricity into the ground? I have no engineering backround and am quite baffled by this.

Single nuclear units produce about 1000 MWe, while the plant uses about 25 MWe to run the primary cooling pumps. The turbine-generator isn't designed for 2.5% of production. Normally a turbine-generator set is rolled and synchronized with the grid at about 10-15% of rated power.

When there is loss of off-site power, i.e., loss of grid connectivity, basically the procedure is to shutdown the plant, and use auxiliary systems to remove decay heat from the reactor, which is few % of the thermal output. It doesn't require 25 MWe to run the systems to remove decay heat.

Some background - but it doesn't address power requirements.
http://www.iaea.org/About/Policy/GC/GC53/GC53InfDocuments/English/gc53inf-3-att5_en.pdf

http://www-pub.iaea.org/mtcd/publications/pdf/trs1/trs224_web.pdf


Operating Experience Assessment – Effects of Grid Events on Nuclear Power Plant Performance
http://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/sr1784/sr1784.pdf

 

[Bignose]

I understand all of that but is there any reason why you couldn't just keep it running and just send the surplus electricity into the ground? I think of it as important because the more options there are, the more redundant the system is, the lesser the risk of a meltdown.

 

[rmattila]

In most European plants, there is a house load operation mode, which is attempted in case of failures in the external grid connection. In this "island operation mode", the plant's own generator remains connected to the internal distribution, reactor power is rapidly dropped to around 30 %, and the excess power is managed by dumping most of the steam directly to the condenser through the steam bypass.

House-load operation is not a requirement by the safety authorities but rather the grid companies. It is not without safety concerns, since operation without the stabilizing effect of the national grid makes the plant vulnerable to e.g. voltage peaks caused by disturbances in the generator voltage control. There have been some incidents where safety-classified equipment has been lost due to failures in the non-safety voltage regulators, and this has led into some plant modifications regarding situations where house-load will be attempted.

Another consequence of the house-load capability is the large capacity of the steam dumps. This is not without risks either (at least in BWRs), as spurious opening of the (non-safety) dump valves at low reactor power will lead into rapid rise of the water level due to RPV depressurization, and rapid actions are needed to prevent water from entering the steam lines. Steam lines would probably not withstand a water slug traveling at 50 m/s, and the MSIVs may only have a second or two to close before the water level rises to the level of the steam lines.

 

[wizwom]

All nuclear plants include a bypass line, to stop feeding steam to the turbines in case of grid trip. What you are talking about would be using just one section of the turbine for powering local needs and idling the rest; from an engineering standpoint, its very possible - each section (High pressure, mid pressure, low pressure) around a third of the energy in the steam. But that's 300 MW shaft power, even so.
The problem is the NSSS doesn't make "low energy steam" if its running at low power; It just doesn't make enough heat to make steam at the operating pressure. Nuclear plants are not run in the superheat region, and so can't really scale very efficiently.

 

[rmattila]

All nuclear plants include a bypass line, to stop feeding steam to the turbines in case of grid trip. What you are talking about would be using just one section of the turbine for powering local needs and idling the rest; from an engineering standpoint, its very possible - each section (High pressure, mid pressure, low pressure) around a third of the energy in the steam. But that's 300 MW shaft power, even so.

That's not the way it's done in the BWR plants I know of. At the time of the load drop, the reactor power is rapidly cut to 30 % by tripping the main circulation pumps and scramming some pre-selected CR groups. The turbine pressure control is switched from the normal mode (where the turbine regulator valves control the reactor dome pressure) to a hybrid mode, where the steam dump valves control the pressure, and the turbine regulator valves first close to prevent turbine overspeed, and then start regulating the turbine rpm instead of pressure.

No changes are made in the steam routing within the turbines, i.e. all sections are in their normal line-up. Below 10 % of rated power, the low pressure turbine tends to heat up, and therefore plants equipped with house-load capabilities usually have some kind of special LP steam cooling system/arrangement for these conditions. Besides that, the turbines can be run at a couple % of rated power, as has been demonstrated by numerous transients ending up at house load. The reactor is producing about 30 %, but most of it is dumped directly through bypass, and only 10 % or so of the steam flow (which is 30 % of nominal) goes to run the turbines, as controlled by the turbine regulator in the rpm regulation mode.

 

[jim hardy]

PWR's are likewise equipped with turbine bypass valves, our plant called them"steam dump to condenser".

In theory one could open the bypass valves and reduce reactor power to the amount required by house load , and that's called "islanding" the plant.

That rapid initial heatup from power being mismatched, followed by a rapid cooldown from full power temperature to low power temperature causes water in reactor to shrink and causes bubbles in steam generators to shrink. That results in low water levels both places and usually one of them reaches a trip setpoint during the transient.
Pressurizer and steam generators are both tall skinny volumes so density change causes substantial level change.

IF everything goes just right you could handle loss of grid

but huge machines just aren't nimble.

 

[Bignose]

I seem to understand that dumping energy via steam release may be problematic. Back to my original question - is it possible to only throttle back as far as what would still be considered normal operation, and simply ground the excess. Or is there simply too much resistance in the ground, or it would energize water tables or something crazy like that?

 

[rmattila]

Normal operation would be something between 20 % and 50 %, i.e. hundreds of megawatts. Two megawatts boil a kilogram of water per second, meaning that any stationary object absorbing the energy will heat up quickly.

The biggest problem with the islanding based on dumping the excess steam is not in the final state, but in the transient, where the load suddenly falls and the turbine accelerates rapidly to overspeed. In order to keep the turbine operable, you have to rapidly close the turbine valves and simultaneously open the bypass dump valves to keep the steam pressure constant. A BWR plant will have a reactor trip at a few bar overpressure as well as at too rapid drop of the pressure, and synchronizing the turbine and bypass dump valves so that the pressure stays within that band during the initial transient - while simultaneously preventing turbine tripping from overspeed - is quite challenging.

The PRA models assume a success probability of some 50 %, and this order of magnitude has been verified by experience.

 

[jim hardy]

...simply ground the excess. Or is there simply too much resistance in the ground, or it would energize water tables or something crazy like that?

well, yes...
the electric company goes to a lot of trouble to keep power out of the ground because when it gets into the ground(which i prefer to call 'earth') it disrupts telephone lines and can electrocute people and livestock walking nearby. Imagine if that much power got into the underground pipeline bringing water to the power plant - it'd really light up the municipal water plant at other end of the pipe.

Many people think that Earth is some kind of infinite sink for electricity. You often hear the phrase "electricity wants to get to ground". I think we are imprinted with that idea as children watching lightning storms. But that idea is a myth, pure superstition. Earth (often called 'ground') is just another wire that happens to go most everyplace.

old jim

 

[wizwom]

Thanks for the input, guys. It fills out the stuff we covered in power systems a lot.