How to design a nuclear power plant reactor power control system ?

In summary: SCD) is a measure of how fast the plant is responding to transient changes. There is a controller in the turbine that reads this SCD and uses it to adjust the power output of the turbine. The PWR roughly 157 assemblies (900 MWe), or 193 or 205 assemblies (1300 to 1450 MWe)?The PWR is roughly 157 assemblies (900 MWe).
  • #1
reboothit
11
0
Can you give me some advice of this topic?

Thank you!
 
Engineering news on Phys.org
  • #2
If you tell us what your thoughts are, you will be more likely to get some feedback.
 
  • #3
Are you referring to software or hardware? You'll have to be more specific.
 
  • #4
And the reactor type (BWR, PWR, something else?) has a big influence on the very physical mechanisms used for controlling the reactor power. In a BWR, power control is typically achieved by controlling the recirculation flow, whereas a PWR reactor power may be controlled either by active control rod manoeuvers or through physical feedback mechanisms by controlling the turbine power.
 
  • #5
reboothit said:
Can you give me some advice of this topic?

Thank you!
What does one know about control theory?
 
  • #6
rmattila said:
And the reactor type (BWR, PWR, something else?) has a big influence on the very physical mechanisms used for controlling the reactor power. In a BWR, power control is typically achieved by controlling the recirculation flow, whereas a PWR reactor power may be controlled either by active control rod manoeuvers or through physical feedback mechanisms by controlling the turbine power.
While this is essentially correct, in PWRs core power is often controlled by varying the soluble boron concentration, particularly in base load plants that do not have grey rods for power shaping or power adjustments for load follow or frequency control. Most PWRs in the US do not have power shaping rods. Reducing feedwater temperature or using the turbine is usually performed at EOC.
 
  • #7
Astronuc said:
While this is essentially correct, in PWRs core power is often controlled by varying the soluble boron concentration, particularly in base load plants that do not have grey rods for power shaping or power adjustments for load follow or frequency control. Most PWRs in the US do not have power shaping rods. Reducing feedwater temperature or using the turbine is usually performed at EOC.

That's right, of course. I was thinking more of the load follow -type fast power control, where the dilution/boration control is probably too slow.

Regarding turbine control, BWR:s can be made quite simple: the reactor power is controlled by varying the main circulation pump speed based on feedback from the generator power, and the turbine controller just maintains the steam dome pressure, i.e. "turbine follows reactor". How is it in the referred US PWR plants: are the turbines operated without feedback from the generator power ("following the reactor" similarly to the BWR case), or is the turbine power actively adjusted to meet the desired power output, and the primary inlet temperature will then take care of adjusting the reactor power to match the turbine (="reactor follows turbine")?
 
  • #8
Look for a book called "Nuclear Reactor Engineering" by Glasstone & Sessonske.

I find the older version more readable, a sort of pink front cover with a power plant not the later one with red&yellow graffiti motif.

He has an excellent chapter on the subject.
Book is often on Ebay.

old jim
 
  • #9
gmax137 said:
If you tell us what your thoughts are, you will be more likely to get some feedback.

I want to know how to design a mode "G" reactor control system for PWR .
Thank you!
 
  • #11
for three decades I maintained an analog PWR reactor control and protection system that was designed in late 60's. But i am not familiar with the term "Mode G". I would guess it means something we old guys are used to but not by that name.

Were i in your shoes i'd study what the ancients did in 1960's for starters.
If you set out to re-invent the wheel you'll have to stumble up from bottom of learning curve. Why not start from halfway up?

In early 70's we could load follow and with a little luck survive a somewhat greater than 50% load rejection transient. But over the years increasingly stringent conservatism made us basically an 'all rods out' baseload with chem(Boron) shim.

In our plant the basic automatic control made reactor follow turbine. That way the plant could load follow as directed by system dispatch.

Load on turbine is inferred by measuring steam flow through it which gets shifted(mx+b) into a desired reactor temperature.. Measured temperature is subtracted from desired to produce a temperature error signal. Temperature error becomes a contol rod speed&direction signal that sets the rods into motion. The more temperature error the faster rods move. When measured temperature matches desired there's no error anymore so the rods stop. It really is basically that simple.

For better transient response another difference signal is developed, difference between reactor power and turbine power. Rate of change of this difference signal is added to the temperature error signal. It trims rod speed during transients and can help prevent over/under-shoot.

The reactor would not know whether it is being controlled by an old analog system like mine or by a fancy computer system like i assume you'll build. So your algorithms will probably start from old timey control theory basics.

In my day Bailey Controls had excellent technology, some say the best, and they in turn were owned by Babcock&Wilcox.
You'll probably find some good nuclear plant control system theory in the book "Steam its Generation and Use" published by Babcock & Wilcox.. look for a late 1970's edition ( it's been in print since at least 1920's.)

old jim
 
  • #12
Description of the Mode G control strategy for a VVER 1000 reactor can be found in this article.
 
  • #13
jim hardy said:
for three decades I maintained an analog PWR reactor control and protection system that was designed in late 60's. But i am not familiar with the term "Mode G". I would guess it means something we old guys are used to but not by that name.

Were i in your shoes i'd study what the ancients did in 1960's for starters.
If you set out to re-invent the wheel you'll have to stumble up from bottom of learning curve. Why not start from halfway up?

In early 70's we could load follow and with a little luck survive a somewhat greater than 50% load rejection transient. But over the years increasingly stringent conservatism made us basically an 'all rods out' baseload with chem(Boron) shim.

In our plant the basic automatic control made reactor follow turbine. That way the plant could load follow as directed by system dispatch.

Load on turbine is inferred by measuring steam flow through it which gets shifted(mx+b) into a desired reactor temperature.. Measured temperature is subtracted from desired to produce a temperature error signal. Temperature error becomes a contol rod speed&direction signal that sets the rods into motion. The more temperature error the faster rods move. When measured temperature matches desired there's no error anymore so the rods stop. It really is basically that simple.

For better transient response another difference signal is developed, difference between reactor power and turbine power. Rate of change of this difference signal is added to the temperature error signal. It trims rod speed during transients and can help prevent over/under-shoot.

The reactor would not know whether it is being controlled by an old analog system like mine or by a fancy computer system like i assume you'll build. So your algorithms will probably start from old timey control theory basics.

In my day Bailey Controls had excellent technology, some say the best, and they in turn were owned by Babcock&Wilcox.
You'll probably find some good nuclear plant control system theory in the book "Steam its Generation and Use" published by Babcock & Wilcox.. look for a late 1970's edition ( it's been in print since at least 1920's.)

old jim

Thank you for your explanation about temperature control and mismatch control of PWR and the other information.



MODE G : load follow.
MODE A : basic load.
 
  • #14
who have the description of the Mode G control strategy for a Westinghouse AP1000 reactor?

I need it very much!
 
  • #15
reboothit said:
Can you give me some advice of this topic?

Thank you!

rmattila said:
Description of the Mode G control strategy for a VVER 1000 reactor can be found in this article.

Thank you very much
 
  • #16
jim hardy said:
for three decades I maintained an analog PWR reactor control and protection system that was designed in late 60's. But i am not familiar with the term "Mode G"...
I'm not familiar with that term either.
reboothit said:
MODE G : load follow.
MODE A : basic load.

Thanks.

But it makes me wonder about Modes B, C, D, E, and F. Where are these terms coming from?
 
  • #17
reboothit said:
who have the description of the Mode G control strategy for a Westinghouse AP1000 reactor?

I need it very much!
You may wish to look at their DCD, which is filed with the NRC. The AP1000 grey RCCA (GRCA) uses 4 AIG rodlets (fingers) and 20 Stainless steel rodlets (as filed in the UK).
http://www.nrc.gov/reactors/new-reactors/design-cert/ap1000/dcd/Tier 2/Chapter 4/4-1_r14.pdf


However, in the US design, W indicates 12 AIG rodlets (fingers) and 12 304SS rodlets
AP1000 Design Control Document
CHAPTER 4 REACTOR
4.1 Summary Description
http://pbadupws.nrc.gov/docs/ML0832/ML083230318.pdf

Of course, one needs a core monitoring system and I&C.

See - http://pbadupws.nrc.gov/docs/ML0832/ML083230868.html - for AP1000 DCD Rev. 17
One would look at Chapter 4 (particularly 4.2, 4.3) and 7.

From section 4.2
Gray Rod Cluster Assemblies
The mechanical design of the gray rod cluster assemblies plus the control rod drive mechanism
and the interface with the fuel assemblies and guide thimbles are identical to the rod cluster
control assembly.

As shown in Figure 4.2-11, the gray rod cluster assemblies consist of 24 rodlets fastened at the top
end to a common hub or spider. Geometrically, the gray rod cluster assembly is the same as a rod
cluster control assembly except that 12 of the 24 rodlets are stainless steel while the remaining 12
contain the reduced diameter silver-indium-cadmium [AIG] absorber material clad with stainless steel as
the rod cluster control assemblies.

The gray rod cluster assemblies are used in load follow maneuvering and provide a mechanical
shim to replace the use of changes in the concentration of soluble boron, that is, a chemical shim,
normally used for this purpose. The AP1000 uses 53 rod cluster control assemblies and 16 gray
rod cluster assemblies.
 
Last edited by a moderator:
  • #18
I have the impression that "mode G" would be the French control strategy developed for the 900 MWe plants in the late 70's to enable better load following capabilites by utilizing gray control rods, hence the name "G" for "Gris" or "Grey".

There's one source (in French): http://www.sfen.org/IMG/pdf/ST6-15mars2007.pdf
or translated by Google to "English".

Other modes listed are "X" for Axial imbalance power control (for the N4 generation of PWRs) and "T" for what I guess is the core average temperature control used in EPR.
 
  • #19
rmattila said:
That's right, of course. I was thinking more of the load follow -type fast power control, where the dilution/boration control is probably too slow.

rmattila,

PWR's naturally "load follow" or "follow the turbine". The power of a nuclear power plant with
a PWR is controlled by the turbine throttle valve.

Suppose a factory fires up and there's a big increase in the load on the plant. That would tend to draw more current from the plant, and the additional back-EMF of the generator would tend to slow it down. The generator has to remain phased to or sync'ed to the grid, so a controller on the generator opens the turbine throttle. This increases turbine power and it draws more energy from the primary coolant. The primary coolant goes back to the reactor "cooler", and the moderator temperature feedback forces an increase in reactor power until it balances the additional power demanded by the turbine.

The PWR has a fairly simple controller on the generator that controls the turbine throttle and the reactor just naturally follows.

Greg
 
  • #20
Morbius said:
rmattila,

PWR's naturally "load follow" or "follow the turbine". The power of a nuclear power plant with
a PWR is controlled by the turbine throttle valve.

Suppose a factory fires up and there's a big increase in the load on the plant. That would tend to draw more current from the plant, and the additional back-EMF of the generator would tend to slow it down. The generator has to remain phased to or sync'ed to the grid, so a controller on the generator opens the turbine throttle. This increases turbine power and it draws more energy from the primary coolant. The primary coolant goes back to the reactor "cooler", and the moderator temperature feedback forces an increase in reactor power until it balances the additional power demanded by the turbine.

The PWR has a fairly simple controller on the generator that controls the turbine throttle and the reactor just naturally follows.

Greg

Yes, that is the "control by physical feedback mechanisms" I mentioned as one alternative in my first post:

rmattila said:
In a BWR, power control is typically achieved by controlling the recirculation flow, whereas a PWR reactor power may be controlled either by active control rod manoeuvers or through physical feedback mechanisms by controlling the turbine power.

I know this "reactor follows turbine" by physical feedbacks is the method used in VVER 440 -type PWR:s operating on base load, whereas the EPR is designed to actively alter the reactor power by control rods in order to keep the reactor average temperature constant and thus minimize thermal transients. Reading Astronuc's answer to my post, correcting that the power in US PWRs would be controlled by chemical shim, I wondered if there would be yet some other control strategy in addition to those two. But apparently they use this "reactor follows turbine by physical feedback" approach, then.
 
  • #21
Actually, most LWRs (both B's and P's) in the US use baseload strategy, with high availablity/capacity factors in mind. BWR's are more adept at load following using flow control. If load-following, PWRs would likely use grey/gray rods, or turbine control depending on the magnitude of the swing. Some load-following of PWRs was tried in the 1980s.

The Gen-III+ PWRs are designed for load follow with gray RCCAs.

Most utility grids in the US load-follow with gas-fired (or perhaps oil) peaking units, AFAIK.
 
  • #22
"The primary coolant goes back to the reactor "cooler", and the moderator temperature feedback forces an increase in reactor power until it balances the additional power demanded by the turbine. "

That control scheme would work. It has this disadvantage in a PWR: since at higher load more temperature difference is required to push the heat across the steam generator tubes (see LMTD at http://www.engineeringtoolbox.com/arithmetic-logarithmic-mean-temperature-d_436.html), steam temperature (and hence pressure) would drop as load increases costing you turbine efficiency.

Mr Turbine would be happiest with constant steam presure vs load. To provide required temperature difference across steam generatror tubes as mentioned above, that scheme (constant steam pressure) would require both reactor outlet and inlet temperatures to increase with power.
That has the disadvantage that reactor outlet temperature gets so high it would take unreasonably high pressure to prevent boiling in the reactor.

Mr Reactor would be happiest with constant outlet temperature as Morbius described.
Reactor inlet temperature would decrease with load . As i said, that scheme results in low steam pressure at high load which makes Mr Turbine unhappy.

The PWR i worked on uses a compromise. Reactor inlet temperature is kept nearly constant, reactor outlet temperature increases with load. That let's you operate with turbine steam pressure that drops a little with load but not too much. To that end reactor outlet temperature is programmed to increase with turbine load as described a few posts back. Average of measured inlet and outlet temperatures is used for automatic rod control.
Load follow was intended to be with control rods. But the nuclear fuel cost is so much lower than fossil the unit stays base loaded.

old jim
 
  • #23
rmattila said:
Reading Astronuc's answer to my post, correcting that the power in US PWRs would be controlled by chemical shim,

rmattila,

Actually, the chemical shim is used for long time-scale reactivity control. The concentration of boron is high near the start of cycle, and is diluted as fuel is burned up.

Chemical shim is not used for short time scale reactivity control, such as that contemplated by your feedback control system.

Greg
 
  • #24
jim hardy said:
"The primary coolant goes back to the reactor "cooler", and the moderator temperature feedback forces an increase in reactor power until it balances the additional power demanded by the turbine. "

That control scheme would work. It has this disadvantage in a PWR: since at higher load more temperature difference is required to push the heat across the steam generator tubes (see LMTD at http://www.engineeringtoolbox.com/arithmetic-logarithmic-mean-temperature-d_436.html), steam temperature (and hence pressure) would drop as load increases costing you turbine efficiency.

WRONG - in fact just the OPPOSITE!

The PWR i worked on uses a compromise. Reactor inlet temperature is kept nearly constant, reactor outlet temperature increases with load. That let's you operate with turbine steam pressure that drops a little with load but not too much. To that end reactor outlet temperature is programmed to increase with turbine load as described a few posts back. Average of measured inlet and outlet temperatures is used for automatic rod control.
Load follow was intended to be with control rods. But the nuclear fuel cost is so much lower than fossil the unit stays base loaded.

Jim, even with a base loaded plant, there are small swings in load because the load on the grid is not constant. As load changes on the grid, ALL the plants on the grid experience the change in load; not just the ones that are not-baseloaded.

The baseload plants also feel a change in load. They have to compensate. If the load on the grid goes up, the grid draws more current from all generators on the grid. The increased load doesn't leave the baseload generators alone, and they only generators that feel the load increase are the peaking units.

No - all the plants have to respond. Again, the increase load tends to slow down the generator. The controller responds by opening the turbine throttle. That leads to a "cooler" reactor inlet temperature. The temperature feedback responds by increasing reactor power so that the inlet temperature goes back to equilibrium conditions, and with a higher reactor outlet temperature. ( You said it precisely BACKWARDs above.) The load following of a PWR does precisely what you want it to do - increase reactor power and outlet temperature.

You missed the fact that the inlet temperature goes back to equilibrium conditions, otherwise the reactivity wouldn't be zero at the new higher power.

Greg
 
  • #25
"""The temperature feedback responds by increasing reactor power so that the inlet temperature goes back to equilibrium conditions, and with a higher reactor outlet temperature. ...
The load following of a PWR does precisely what you want it to do - increase reactor power and outlet temperature.

You missed the fact that the inlet temperature goes back to equilibrium conditions, otherwise the reactivity wouldn't be zero at the new higher power.
"""

Indeed the built in negative feedback from moderator coefficient pushes the system back toward equilibrium, in fact most of the way back. Beautiful how we can enlist Mother Nature's aid.
reactivity is a good approach to think it through...
if inlet temperature = same
and outlet temperature = higher
then average temperature in reactor is higher, which removes a smidge of reactiviy via negative moderator temp coefficient ,

i contend that "smidge of reactivity" had to be provided as reactor temperature rose higher.
Else reactor would settle at a new average temperature, one where reactiviy surplus from smidge lower inlet temperature offsets the reactivity shortage from smidge higher outlet temperature.. that's my "most of the way back" mentioned above.

But i think really because of an ambiguity in my post we are debating two different points.
My line, "since at higher load more temperature difference is required to push the heat across the steam generator tubes"
refers to temperature difference between reactor side and turbine side of steam generator tubes, not to reactor's inlet-outlet temperatures.

Similar fugue, different symphonies.

Sorry about the ambiguous wording :)

old jim
 
  • #26
Morbius said:
Chemical shim is not used for short time scale reactivity control, such as that contemplated by your feedback control system.

That being generally true, there are however PWR control strategies that use boration/dilution to control the axial power distribution (=to counter Xenon oscillation in the tall PWR cores) indirectly by causing the control rods (which are automatically moved to keep core temperature at desired values) to move up or down. Not exactly "short scale", but anyway shorter than the normal reactivity compensation.
 

1. What is the purpose of a power control system in a nuclear power plant reactor?

The power control system in a nuclear power plant reactor is responsible for regulating the power output of the reactor. This is important because it helps maintain a stable and safe operating condition for the reactor.

2. How is the power control system designed?

The design of a power control system for a nuclear power plant reactor involves a combination of advanced computer software, control algorithms, and physical components such as sensors, actuators, and control rods. It also requires thorough analysis and testing to ensure its reliability and safety.

3. What factors are considered in designing a power control system for a nuclear power plant reactor?

Some of the key factors that are considered in designing a power control system for a nuclear power plant reactor include the reactor's power output, temperature, pressure, and neutron flux levels. The system must also be able to respond to any changes in these factors quickly and accurately to maintain safe operation.

4. How does the power control system ensure safety in a nuclear power plant reactor?

The power control system continuously monitors and regulates the power output of the reactor to prevent any potential safety hazards. It also has safety mechanisms in place, such as emergency shutdown systems, to quickly and effectively shut down the reactor in case of any abnormalities or malfunctions.

5. How is the power control system maintained and updated?

The power control system in a nuclear power plant reactor is regularly maintained and updated by qualified technicians and engineers. This includes performing routine inspections, tests, and repairs to ensure its proper functioning. Any updates or modifications to the system are also carefully planned and implemented to maintain its safety and efficiency.

Similar threads

  • Nuclear Engineering
2
Replies
41
Views
3K
  • Nuclear Engineering
Replies
5
Views
1K
  • Nuclear Engineering
Replies
1
Views
1K
Replies
46
Views
3K
  • Nuclear Engineering
Replies
30
Views
874
Replies
5
Views
688
Replies
15
Views
798
  • Nuclear Engineering
Replies
1
Views
1K
Replies
5
Views
860
Replies
2
Views
2K
Back
Top