Control.com - Nerds in Control

Hi everyone,

Relatively new here and have had some really useful insights from a lot of guys here treating many posts since joining, though this is my own very first post (cries of help).

We operate a series of GE 9E in SC configuration, single fuel DLN-1 in Africa, supplying power to the grid. Problem is not being able to optimise the units (500MW combined plant capacity) due grid swings. Mostly have to go to pre-select load once the grid goes beyond certain frequency. Hopeless to say units rated at 126MW ISO being operated at pre-select loads of 70MW because of grid. Even if we produce our own fuel! What can be done to keep our units in baseload control even at relatively higher/lower grid frequencies? I have heard of "increasing the deadband", if anyone knows what this means. I will suppose this means incresing the droop deadband nominally set at between 49.5hz to 50.5hz to something wider?

If this is done, what will be the implication on the turbine and the generators?

Thanks

The droop curve is usually set at 4% (speed/load). Yours may be set too low even below 4%

Increase the droop above 4, Increase to 6, 8 or 10 and see what happens.

The response will be a lot less.

Hello Mashll,

First of all, I don't think that your turbine being operated at preselect 70MW load has anything to do with the software, nor can be solved modifying the software. If you have to operate it at such loads it has to be due to the fact that there is not enough demand in the grid.

Then, frequencies higher than 50Hz are normally due to few factors:

1. Sudden shutdown of big consumers

2. Bad grid regulation with individual power plants not participating (or not being able to participate) to frequency regulation

3. Bad "grid clock" - each grid has one or more generators that set the grid frequency.

If you have a normal 9E, your turbine is set to stay synchronized as long as the grid frequency is anywhere between 47.5 and 52.5%. Outside these values, your generator is no longer safely operated.

Now, when you say you have to go to preselect from Base Load when above certain frequency, you mean the machine is going automatically out of Base Load? I have to admit, I don't have too much experience with unstable grids. But I don't really see how this would happen automatically. I checked a Frame9e software and found nothing, but I might have missed something. I would rather expect that you are asked for lower loads by the grid operators instead of a result of the control system software (well... I am sure CSA will have something to say about this, something new to learn for me as well)

Mashll,

Your situation is "hopeless", per your description. The situation is this: As the grid frequency goes up and down, so will the speed of your turbines--because speed is directly proportional to frequency. There is no change that can be made to alter the direct relationship between speed and frequency. None. Full stop. Period.

As the speed of a single-shaft gas turbine increases and decreases, the air flow through the machine will change. As the frequency goes up, the speed will increase and the air flow will increase. And, if the unit is operating at Base Load, the power output will increase--because exhaust temperature control (i.e., Base Load) doesn't care about frequency, it just wants to make as much power as possible by keeping the exhaust temperature as high as possible for the ambient- and machine conditions. If the frequency goes down when the unit is operating at Base Load, then the air flow will decrease and the power output will go down. Again, Base Load says keep the exhaust temperature as high as possible for the current operating conditions, and when the air flow goes down then amount of fuel that can be burned goes down and that means the power decreases.

If you are operating at Part Load, say 70 MW to use your example, and you are NOT operating with Pre-Selected Load Control enabled and active then as grid frequency changes the load will change--per Droop Speed Control governing action. That's what's supposed to happen--because it's intended that units with reserve capacity (those NOT operating at full output) are supposed to change their load to try to help maintain grid frequency. If the frequency goes down, then load is supposed to increase. If the frequency goes up, then load is supposed to decrease. Droop speed control is designed and intended to support maintaining grid frequency by having machines vary their load as grid frequency changes.

Now, if a large plant (such as yours) unilaterally decides, "NO! We don't want our load to change when frequency changes--we insist that our load remain stable regardless of that darned grid frequency!!!" and decides to operate their plant at Part Load with Pre-Selected Load Control enabled and active, then that large plant is actually making the grid frequency fluctuations worse. Yes, WORSE!

By not allowing their turbines to respond to grid frequency disturbances as nature, and Droop Speed Control intended, they are exacerbating the problem and making the situation worse.

Let's say you are riding on a bicycle with several other riders and each rider has a set of pedals on a crankset, and all the cranksets are connected together with chains and fixed gears, and to the rear drive wheel by a chain. Now, the orders for this group of riders are that they are to transport some packages in baskets on the bicycle from one location to another at a constant rate of speed--meaning that the riders all have to work together to maintain the speed (no one rider can provide enough torque to keep the bicycle and riders and packages moving at the desired speed; several riders will have to work together to maintain the speed).

Now, one rider suffers an injury to his legs and can no longer pedal the bicycle. To keep the same speed, the remaining riders are going to have to pedal harder (apply more torque to the pedals). If one rider applies more torque than is necessary to keep the bicycle moving at the desired speed then the bicycle will speed up unless another rider or riders decrease torque. It takes some coordination to keep all the riders working together to maintain the desired speed.

Another rider now suffers an injury and is no longer able to pedal the bicycle. Now, the remaining riders will have to pedal as hard as they can just to maintain desired speed.

Let's say the bicycle passes a station and another package is added to one of the baskets, increasing the load, which causes the speed to decrease. If all of the riders are pedaling as hard as they can, then the speed will not be maintained. If one of the riders that was injured can begin pedaling again then the speed can be increased to desired.

Now, let's say that one or two of the riders just decide without any consultation with any of the other riders that they are only going to provide 50% of the torque they are capable of providing and no more. And, let's say that that's enough to keep the bicycle moving at desired speed as long as the other riders are pedaling as hard as they can. Some more packages are added to the bicycle and now the speed starts to decrease. But, the majority of riders are already pedaling as hard as they can, and those two riders are only producing 50% of what they could produce, even though the speed of the bicycle is decreasing. They could increase their output, but they don't. They don't want to pedal any harder even though the bicycle speed is less than desired.

Those two bicycle riders are making the speed problem worse. They could increase their output and help maintain speed, but they won't. They just want to keep producing the same amount of torque, regardless of bicycle speed. So, the bicycle slows down instead of maintaining speed.

That's how Droop Speed Control works. As load on a grid increases the frequency will start to decrease. As the frequency starts to decrease those units with reserve capacity--and which are not operating in Pre-Selected Load Control--will increase their output to try help maintain frequency. If there are enough units increasing their output then the grid frequency won't decrease by very much.

Everywhere I've ever worked in the world, when a plant signs an agreement to connect to a grid they also tacitly, at least, if not contractually, agree to help support grid frequency through Droop Speed Control governing action. In fact, most contracts require the power generator to specify the amount of Droop Speed Control, 4% o 5% depending on the machine and the application and the location. Which means that for a 1% change in frequency the load of a machine with 4% droop will change by 25%--up to rated output. If the unit was at essentially zero load and in a new and clean condition and ambient conditions were at rated, then if the frequency decreased by 4% the power output of the turbine would increase by 100%. (Now that's no 100% true, because, again, as speed goes down so does air flow, and power output decreases as air flow goes down, but the 25% output change for each 1% frequency change is the key to Droop Speed Control.)

When your generator is synchronized to a grid with other generators the frequency of your machine is controlled by the grid. Just as on the bicycle where all the cranksets are connected by chains and the speed of all the cranksets is the same at any bicycle speed (in essence, they are "synchronized"). In the generator, it's magnetic forces that keep the generators locked in synchronism. On the bicycle, it's the chain that has a fixed number of links (usually the same number of links on gears with the same number of teeth) connected by a final drive chain to the rear wheel, thereby keeping the cranksets rotating at the same speed relative to bicycle speed.

But, if everyone isn't "sharing" in maintaining the bicycle speed, then the bicycle speed isn't going to be maintained very well. And, in the case of combustion turbines when air flow through the machine changes as speed changes, then power output will also change (at any load). A properly regulated and controlled grid requires responsible power generators who understand how their machines operate, how a grid operates, and how their machines contribute to a properly operating grid.

It's as simple as that. Full stop. Period.

And there's no setting that will allow your machines to run at different frequencies (speed) than all the other machines on an AC grid. Full stop. Period.

So it is when you are connected to a poorly regulated grid. Magnetics and physics dictate that your speed, and the amount of power being produced, is fated by the grid frequency and how well the regulators are controlling generation to match load. If generators are prone to tripping excessively and there isn't sufficient reserve capacity to make up for the lost generation--or there are some plants which just arbitrarily decide not to participate in Droop Speed Control schemes--then the grid frequency isn't going to be maintained very well.

And, don't be fooled because your Speedtronic panels are in Droop Speed Control when Pre-Selected Load Control is enabled and active. Because Pre-Selected Load Control over-rides Droop Speed Control when it's enabled and active. So, if you are operating your machines at a constant output as grid frequency is changing, you are part of the problem--not part of the solution.

Sorry, but them's the facts.

No matter what you want them to be, them's the facts.

I hope this helps!

Can anyone let me know how generators on a grid remain synchronized? I mean why they don't simply lose synchronization.

I can understand what happens when they are first synchronized with all the phase angle and speed and voltage matching but what keeps these parameters synchronized thereafter?!

Thanks

Magnetism.

GenSet_Expert,

The whole purpose of matching speed and phase angle is to make sure that when the generator breaker is closed that the rotor magnetic poles are properly aligned with stator magnetic poles so that when the breaker is closed rotor North is near stator South and rotor South is near stator North.

Because if the rotor South pole is near the stator South pole and the rotor North pole is near the stator North pole, REALLY BAD things can happen. You know what happens when you try to force two magnets together when the South Poles (or the North Poles) are near the point of contact--it doesn't happen. You can push and push and push, and while you might eventually get the two magnets to touch each other, when you release them they fly apart.

And you know what happens when you put two magnets together with opposite poles near each other--the two magnets fly together without much assistance. And require force to separate them.

This is exactly what is happening in a synchronous generator. The magnetic forces of attraction are VERY strong, and keep the rotor from spinning faster than the "rotating" magnetic fields of the stator. The rotor North pole follows the "rotating" stator South pole, and the rotor South pole follows the "rotating" stator North pole--with very great forces of magnetic attraction. The magnetic forces of attraction are so great that the prime mover cannot break them to spin the rotor faster than the speed proportional the frequency of the AC flowing in the stator.

This is why the speed of a synchronous generator is directly proportional to the frequency of the grid with which it is connected. The forces of magnetic attraction keep the generator rotor running at synchronous speed--the speed that is proportional to frequency--which is what is called synchronized.

If the phase angle and speed weren't matched during the process of closing the generator breaker, the magnetic forces of repulsion--which are just as strong as those of magnetic attraction--which try to spin the generator rotor extremely fast in one direction or the other (possibly against rotation!) to attract the proper rotor poles. And, that can cause great mechanical damage to the generator, the load coupling, and the prime mover.

And, once the process of synchronization is done, the generator remains "synchronized", running at a speed that is proportional to frequency REGARDLESS OF THE LOAD OR THE FUEL OR THE STEAM BEING ADMITTED TO THE PRIME MOVER DRIVING THE GENERATOR.

It's simple magnetism. Period. Full stop. Nothing more. Nothing less.

EVERY synchronous generator synchronized to an electric grid with other synchronous generators is running at synchronous speed--the speed that is proportional to the AC grid frequency. And, they are all locked into synchronous speed by the forces of magnetic attraction at work inside the synchronous generator. No generator synchronized to an electric grid can spin faster or slower than its synchronous speed (which is a function of the number of poles of the generator rotors.

This is the part that most people miss about AC electric power systems and the part that helps people understand Droop speed control: Once the generator breaker is closed on a well-regulated electric grid, the speed of the generator (and of the prime mover driving the generator) DOES NOT CHANGE. Sure, when the prime mover is being started and accelerated, the speed changes as fuel or steam is admitted to the prime mover. But, once that generator breaker closes the speed of the generator and of the prime mover driving the generator is fixed by the frequency of the grid. Adding more fuel or steam to the prime mover DOES NOT increase the speed of the prime mover or the generator.

Most people think the speed changes as fuel or steam flow is changed because that's what happens during starting and acceleration. But, it doesn't. The grid frequency controls the speed when the generator breaker is closed. And it's all because of the forces of magnetic attraction.

Pretty amazing, huh?

Hope this helps!

I probably need to add to the explanation that in an AC machine (motor or generator) the magnetic poles created when current is flowing in the stator appear to "rotate", and they rotate at a speed that is proportional to frequency and the design of the machine (which includes the number of poles of the rotor). On a well-regulated grid of any size the frequency is, or at least it should be, stable and relatively constant, which means the speeds of the electric machines connected to the grid will be constant and relatively stable.

Also, if the generator breaker remains closed when fuel or steam is removed from the prime mover driving the generator the generator will automatically become a motor and will continue to spin at the same speed (proportional to the frequency of the AC grid with which it is synchronized). This is called "reverse power" and is not good for most prime movers (especially steam turbines and reciprocating engines) because the "generator" (which is now a motor) is spinning the prime mover, and there are generally protective relays to ensure the generator breaker opens in a sufficient time to prevent damage to the prime mover.

It's all about magnetism. Simple, and easy. There are lots of youTube videos about motors and generators and "rotating" magnetic stator fields, as well as reference materials all over the World Wide Web.

Remember: There is no difference between a motor and a generator--except the direction of current flowing in the stator, and whether or not torque is being supplied to the electric machine (when it is a generator), or the electric machine is supplying torque to some device like a pump or a fan (when the electric machine is a motor). Electric current is the medium by which torque is transmitted from the electric machine being driven by a prime mover to an electric machine driving a pump or a fan or some other device that requires torque (including those "virtual torque" devices called computers!).

And on a properly regulated AC grid of any size (an "island" with a single or two or three generators, to the European continent) all of the generators that are synchronized together to supply all of the lights and fans and pumps and computers and monitors are spinning at a speed that is directly proportional to the frequency of the grid with which they are connected: synchronous speed. And no single generator can spin faster or slower than synchronous speed when synchronized to a grid with other generators. So, in effect, they are all spinning at exactly the same speed: synchronous speed.

And magnetism is the reason. Pure and simple.

If the excitation being applied to the rotor of a generator drops below a certain amount (which varies with a lot of different parameters) then the rotor can "slip a pole" which means the North pole of the rotor jumps out of "lock" (the forces of magnetic attraction are broken) with the "rotating" South pole of the stator (and the same with the South pole of the rotor and the "rotating" North pole of the stator) and when that the rotor accelerates VERY fast and when the North pole of the rotor approaches the "rotating" North pole of the stator (and at the SAME time the South pole of the rotor is also approaching the South pole of the stator!) the forces of magnetic repulsion can try to stop the rotor or even try to to force it to turn backwards--which can be VERY destructive to the generator, the load coupling, and the prime mover. VERY destructive. VERY destructive.

So, there are protective relay functions (minimum excitation limits and loss of excitation) to open the generator breaker to stop current from flowing in the stator and to eliminate the "rotating" magnetic fields of the stator to prevent the damage which can be caused by slipping a pole.

But, in any case: the answer to your question is one single word.

Magnetism.

It's what makes electric machines (motors and generators) go 'round! And why the speeds of AC electric machines are proportional to the number of poles of the machine AND to the frequency of the applied AC.

Variable frequency AC is used to vary the speed of AC motors--but the magnetic principles are exactly the same, just the frequency is variable. (The number of poles of an electric machines' rotor is not usually variable.) Some very large gas turbine generators actually use the generator, driven by a variable frequency source, as the starting means to accelerate the unit to near rated speed, at which time, the generator becomes a generator again.

But it's still about magnetism and North- and South magnetic poles. And the attraction between. Current flowing in a conductor, and in coils of conductors, creates magnetic poles. Got two magnetic fields, one which "appears to rotate" and the other free to rotate, and you've got yourself the makings of an electric machine. Apply torque to the electric machine and it can become a generator. Apply current to the stator and it can become an electric motor.

But they all require magnetic fields.

Isn't the world a very simple place, after all? At least the things that have been around for a long time, anyway. And they haven't changed, which is kind of nice, also.

What are off peak and peak demand capacity of your grid?

Dear Mashll, I would suggest using the search feature of control.com to look for previous post having to do with grid frequency etc. There have been a few posts that got very complex, and long!!

It is difficult to fully understand your situation but the situation of unstable grid frequency regulation is not new. As stated by other authors a lot has to do with how your area is regulated, the size and configuration of generators in the area, and how generator governors are set. But an unstable grid is typically an issue of inadequate system generation and incorrect relay settings for load shedding.

In a fairly stable grid with adequate generation, if load increases, generators in the area ramp up. Not all can be operating at full (base) load, so that they can provide output for the increase in load. When load drops, generators reduce output to keep frequency stable. If load increases above generation limits, lets say a generator is out of service, then all generators will ramp to full output, but frequency will fall due to lack of generation. In this case transmission relays typically will shed load to protect system frequency.

In your case I am not sure why you are going to preselect load during frequency excursions? If your units were operating at baseload and frequency goes high, they should reduce output to "help" reduce system frequency. It will depend on the magnitude of the swings, the size of your generators VS the total system load of your area, and how long the swings last, but if frequency stays high, theoretically the units should unload as needed.

Now if frequency goes low and your units are operating at baseload, then they will not be able to increase load to support frequency, since they are already operating at the maximum limit. In this case the output of the machine will actually decrease due to lower axial compressor speed and lower mass flow. This is unless you have some of the newer software from GE that allows the unit to overfire slightly to help these situations.

So without more information its hard to provide a real answer for you. One question I have is how are the units controlled? Do the machines get a setpoint from some sort of central power control center? I have encountered some areas where the unit setpoints were recieved from a SCADA system that was not real time. During times of frequency excursions the SCADA setpoint was actually contributing to the problem.

Lastly I would be hesitant to be changing the deadband or settings for droop output without knowing what you are trying to accomplish. Unfortunately in many cases companies are trying to change machine operational programming, instead of trying to fix the real problem.

Know this your not alone, and many here will offer what advice they can to help.

MIKEVI,

When a GE-design heavy duty gas turbine is operating at Base Load without any special "grid code" sequencing (which is optional and not typically provided on most machines) if the grid frequency increases the gas turbine power output will also increase. That's because the air flow increases so the amount of fuel that can be burned also increases, and together, the increased mass flow of air and the additional fuel increase the power output. There is no Droop Speed Control action when a GE-design heavy duty gas turbine is operating at Base Load. In fact, FSRN (Droop Speed Control FSR) is set to a slightly higher value than FSRT (Exhaust Temperature Control FSR) and that differential is maintained as long as Base Load is active. (The "flickering" of the RAISE- and LOWER SPD/LD targets/indicating lamps while the unit is on Base Load is the Speedtronic raising or lowering FSRN to maintain a constant differential above FSRT.)

The opposite happens when a GE-design heavy duty gas turbine is operating at Base Load and grid frequency decreases. The air flow through the machine decreases, the amount of fuel that can be burned decreases, and together they combine to reduce power output.

Now, both of the above results are exactly opposite of what the grid regulators want to happen when there is a grid frequency disturbance! Because when grid frequency increases it means there is an excess of generation relative to load (lights, motors, computers and monitors, etc.) and that translates into a higher than desired frequency. So, even if there are machines operating at Part Load with Droop Speed Control active, the gas turbines operating at Base Load will still be doing exactly the opposite of what the grid regulators, and the grid customers, want to have happen.

It's a dirty little secret of gas turbines--of any manufacture--that this happens at Base Load. It can be particularly disruptive.

I could understand when the grid frequency at Mashll's site goes high the national grid regulators calling to say reduce load to some value below Base Load, perhaps even to 70 MW, though that seems extreme without knowing more about the conditions. But, if the site is unilaterally reducing load when there is any grid frequency disturbance AND putting the units into Pre-Selected Load Control, then they are not contributing to grid stability. While the power output of the turbines might be more stable, the fuel valves are probably going wild trying to keep up with the frequency (axial compressor speed) changes. And that's not helping grid stability either.

As axial compressor speed goes up and down when a GE-design heavy duty gas turbine is operating at Part Load the air flow through the machine is increasing and decreasing. And since gas turbines are primarily mass flow machines, that makes the power output fluctuate in and of itself. The fuel valves, being controlled by a Pre-Selected Load Control Setpoint, are going to be jumping around trying to maintain a constant load depending on the nature of the frequency variations.

Lastly, Mashll said his turbines had DLN-I combustors. YIKES! That's really not good for DLN combustors, having a fluctuating axial compressor speed and mass flow. I can envision the IGVs and the gas valves swinging pretty wildly. The reason they probably go down to 70 MW is to remain in Primary- or Lean-Lean Combustion Mode, because it would probably be VERY hard to stay in Premix Steady-State Combustion Mode, or just Premix Combustion Mode, with grid frequency disturbances above approximately 90-100 MW. The gas fuel valves and IGVs could really be having a go with large and frequent frequency disturbances.

And, though Mashll indicated the units were Simple Cycle, there's still the possibility that they have Inlet Bleed Heating. And that adds dimension when trying to control load and combustion stability during frequency disturbances.

All in all, not a very good situation. Grid regulators need to get a handle on the situation, and power producers need to fully understand how their machines operate, and how they can improve, or de-improve, grid frequency disturbances.

Really, though, it's not a situation that can be solved via a forum like this.

Dear Marshall

your reply is excellent but i would like to ask u one thing. How much frequency will vary or how to calculate if some load got shed ie. say a generator is down?

The dead band +/- 0.5 Hz has already very high. Further increasing it won't help you to solve fundamental problem with the frequency oscillation. At best, it may reduce the number of event related to load oscillation for your plant. At worse, once it happens the load oscillation will become even bigger.

Before you responding to my question, by looking at your plant configuration I'm looking at your grid system peak load demand is 5000MW and above. I might be wrong. Hopefully I'm not wrong since if I'm wrong you might not able to solve your system's frequency oscillation problem for the next 5 years or more.

If your grid peak demand is above 6000MW, such problem can be solved, assuming the biggest per unit capacity of your grid is 300MW. Most likely it has something to do with improper load share scheme via droop setting and/or secondary response overwhelmingly bigger than the primary response (droop response). At least I would that say very remote possibility that it cannot be solved.

Minimum good frequency control and regulation can be achieved if the ratio between the biggest per unit on bars and the peak demand capacity is 0.05 or smaller. If such ratio is greater than 0.15, I would say it is very remote such problem can be solved. The ratio has to go down first before frequency oscillation can be made to reduce.