In a message dated 9/10/2011 10:05:43 A.M. Eastern Daylight Time,
edmiston@bluffton.edu writes:
Bob Zannelli responded to my questions about nuclear power plants. Thanks
Bob. Could you please respond to a few more questions.
My reactor experience comes from research reactors. I believe the largest
of these that I worked at was the 40 MW high flux beam reactor at
Brookhaven
National Lab, which has now been shut down. Obviously, dealing with 40 MW
of power is a lot different from dealing with 3000 MW. Also, the research
reactors are not putting power onto the the grid, and therefore don't need
to worry about producing a lot of power for a load that could suddenly
disappear. Therefore I realize it’s a whole different ball game when
electric power reactors are compared to research reactors.
Here are some questions for which the answers would help me better
understand the electric power reactors.
[1] If a reactor is generating 3000 MW, of which 1000 MW is appearing on
the
grid, how many turbines are there? Is it a single 1000 MW turbine, or
four
250 MW units, or ten 100 MW units, or what?
)))))))))))))))))))))))))))))))))))
BZ
In the United States and Europe the rule is single turbine generators.
Turbines generators as large as 1300 MWe or even larger exist. In the
former Soviet Union the infamous RBMK design used two turbine generators. In
principle you can design plants with any number of turbine generators. I am
not sure why the Soviet design used two turbine generators , perhaps they
has technological issue building such large turbine generators.
[2] If it is not the case, why couldn't it be the case that there is a 10
MW
or 50 MW or 100 MW turbine (whatever would be the appropriate size) that
could keep running in order to power the nuclear controls and cooling
system
in the event of grid loss?
))))))))))))))))))))))))))))))))
BZ
I am not sure I under stand your question. I think you mean run a small
turbine generator from reactor power. First when the plant suffers some
kind of failure you want to shut the reactor down, terminate the chain
reaction , not perform a slow ramp down in power. Things happen fast when
something goes wrong and the maximum rate of controlled power change is a large
nuclear reactor is far too slow to respond to any grid upsets some other kind
of major failure this way. In the case of loss of offsite power you have
a reactor -turbine system pumping over 1000MWe onto the grid one second and
no load the next. The only safe thing to do is quickly trip the steam
valves to the turbine, trip the generator breakers and scram the reactor.
))))))))))))))))))))))))))))))))
[3] Depending on where the reactor was in its fuel cycle, if a SCRAM
occurred, wouldn't there be enough residual heat to run a 100 MW or
smaller
turbine/generator? If this is true, why aren't power plants designed this
way?
)))))))))))))))))))))))))))))))))))))))))))))))))
BZ
They are to a point, but sooner or later you need electrically powered
systems.Boiling water reactors use the residual heat of the reactor to maintain
water level , a steam turbine drives a kind of emergency steam pump. Even
though I did start up engineering many years ago for the Clinton station in
Illinois , a BWR , I mostly worked on distribution and Instrument control
systems and I am rusty on the nuclear steam supply systems for a BWR, but
this is basically correct off the top of my head.
For PWR there are also stream driven pumps to provide Steam Generator
feed. There are also emergency electric pump backup. On loss of off site power,
, normally there are three diesel generators that can supply vital system
power. All systems are still in a safe condition by a conservative design
criteria if one diesel fails to start, though in reality a single diesel
can protect the core absent some other major failure.
Third generations nuclear designs exist which dispense altogether with the
need for any diesels, they are able to use passive cooling to keep the
nuclear core within safe limits. The Europeans have a pressurized water reactor
that still needs diesels, but these plants have far more redundancy,
they can suffer the loss of several diesels based on conservative design
criteria.
)))))))))))))))))))))))))))))))))))))
.
[4] If a 3000 MW reactor system has the capability of getting rid of the
2000 MW of power that is not going to the grid, then it certainly has the
ability to get rid of 1500 MW of thermal energy (without using the grid)
if
the reactor were shut down to 50% (or less) with only sufficient turbine
load to generate local power. So why shut it down all the way? Couldn't
it
provide local power at reduced fission rate, and thereby be in sort of a
standby mode to come back online more quickly as the grid begins to
recover?
)))))))))))))))))))))))))))))))))))))))))
BZ
You just can't ramp power down this quickly and maintain control. When the
Grid goes black the turbine will rapidly spin up and the steam plant
rapidly increases its pressure and temperature. One second your pumping 1.5
million horse power to the grid , the next second there is no where for the
power to go. You have to shut down things quickly to prevent damage to
equipment. The ability of any nuclear plant to respond to changes in heat load is
determined by the sizing of the steam generators or the size of the reactor
vessel in the case of BWR. I suppose you could design a nuclear reactor as
a 30,000 MWth unit and limit its power to 3000 MWth but that's not
economically viable. And there would likely be other problems with such an
absurd design. BTW naval reactors are actually small enough to deal with load
changes pretty well. But this is a question of scale.
)))))))))))))))))))))))))))))
[5] If most large nuclear power plants have only one very large
turbine/generator, I can see there would be little opportunity to operate
at
reduced capacity. I can also see that it would be difficult to
synchronize
this one large turbine/generator onto a grid that was not nearly fully
functional. So if there is only one turbine... why? Couldn't you more
easily synchronize and load four (or ten) turbines one at a time?
)))))))))))))))))))))))
BZ
It's economics. I don't think the small gain in stability would offset the
extra costs of multiple generators in the light water reactor designs.
However, some third generation plants being proposed use smaller reactor cores
with multiple reactor plants. The reason for this is to enhance passive
cooling, these type designs might use multiple TGs. None have been build as
far as I know.
))))))))))))))))))))))))
[6] There already has to be the capability of getting rid of the full
thermal output of the reactor in the event that the (all) turbine(s) are
shut down, so why would the reactor have to SCRAM if there were a
satisfactory way to maintain the cooling system even if there were no grid
power?
)))))))))))))))))))
I think I answered this question above, hopefully adequately. The question
of whether or not we should expand or even retain nuclear power in the mix
is complex. It depends on a lot of factors. Obviously if we can produce
enough power with renewable energy we should foreclose the nuclear option. But
I don't we can be certain this possible , though it certainly may be. But
energy available and quality of life are intimately linked we should avoid
basing our energy decisions on ideology, but rather sound science. Since
there seems to be some interest in this topic I'll post about one interesting
possibility.
Bob Zannelli
[7] Bob may answer some of these questions by repeating his comment that
current nuclear power reactors were designed as base load suppliers of
electric power. Okay, but is that the only way they could be designed?
Michael D. Edmiston, Ph.D.
Professor of Chemistry and Physics
Chair, Division of Natural and Applied Sciences
Bluffton University
1 University Drive
Bluffton, OH 45817
419.358.3270 (office)
edmiston@bluffton.edu
-----Original Message-----
From: Spinozalens@aol.com
Sent: Saturday, September 10, 2011 6:40 AM
To: phys-l@carnot.physics.buffalo.edu
Subject: Re: [Phys-l] Nuclear Power and the Grid
In a message dated 9/9/2011 6:38:41 P.M. Eastern Daylight Time,
edmiston@bluffton.edu writes:
Under the "all-too-predictable blackout" thread, John Denker mentioned
that all power plants should have "black-start" capability. He also
mentioned
that nuclear plants will generally shut down during a cascading blackout,
and then be the last to come back online.
I strongly agree that all plants should have black-start capability, and I
have been saying that for years.
I also know that nuclear plants tend to SCRAM when there is trouble on the
grid, and I wonder if someone can explain that to me. If grid power is
lost, and the reactor SCRAMs, and there is no black-start capability, the
reactor probably cannot be restarted for a pretty ling time. That means
reactor cooling and spent-fuel cooling has to be provided by diesel
generators,
or something similar... and this might be required for a time longer than
there is sufficient fuel to run the backup power.
)))))))))))))))))))))))))))))))))))
BZ
Commercial Nuclear reactors are never started in a black out condition.
These plants are designed to feed power into a live grid. It's very
difficult
if not impossible to maintain frequency and voltage stability with a large
plant feeding into a power grid with virtually no load.
)))))))))))))))))))))))))))))))))))))))))))))))))
It's not obvious to me why the reactor has to SCRAM in the first place.
Although grid trouble might indicate some sort of nasty thing coming (like
an earthquake), why not wait to SCRAM until you know there is something
nasty coming. It's difficult to imagine that waiting for a few minutes
after a
grid anomaly before initiating the shutdown would make much safety
difference.
))))))))))))))))))))))))))))))))))
BZ
A typical nuclear reactor produced over 3000 MW of thermal power, which
becomes about 1000 MW sent out on the grid. If the ability to send power
to
the grid is lost, the reactor still producing 3000 MW of thermal power will
rapidly heat up and raise the water pressure ( in light water reactor
designs) to dangerous level absent a scram. You just can't swing power fast
enough to prevent this from happening. Nuclear power plants are used for
base load, that is they aren't designed to rapidly swing power levels,
basically the nature of nuclear power plants makes this impossible.
))))))))))))))))))))))))))))))))))))))))))))
Anyway, if a SCRAM eventually does occur, and if the nuclear plant has
black-start capability, once it is ascertained that nothing at the nuclear
plant is damaged, the reactor could be restarted, the generators could be
brought back up (because black-start is provided) and the nuclear plant
can
provide its own energy for cooling itself and the spent fuel pools.
)))))))))))))))))
BZ
This is never done, see above.
)))))))))))))))))))))))))))))))))))
So what's the point of a reactor shutdown when the external grid power is
lost? Indeed, what in the world would happen in a system where *all*
power
is provided by nuclear plants, and there is a grid problem? They all
shutdown, and you can never get any of them running again if external grid
power is a condition of restart. Sounds like you'd have the proverbial
"catch-22."
)))))))))))))))))
BZ
Grids are very big, they can tolerate a certain percentage of units
tripping off line and still maintain stable operation. But of course if
something
goes wrong on the grid, it's quite possible you will get a cascade
effect
with all the power units tripping off the line over a large area.
Engineers attempt to design power grids to minimize this possibility but
in
engineering shit happens.
Bob Zannelli
Retired Nuclear Engineer
)))))))))))))))))))))))))
Michael D. Edmiston, PhD.
Professor of Chemistry and Physics
Chair, Division of Natural and Applied Sciences
Bluffton University
Bluffton, OH 45817
Office 419-358-3270
Cell 419-230-9657