Hypothetical question in regards to system design and protection 6

[Mbrooke]

Would it be possible to design an entire power system where all 500kv, 345kv, 230kv and 115kv lines are only protected via step distance? No POTT, DUTT, DCB, or any communication between relays- of course very short lines would have differential. Perhaps the real question is- what can theoretically be done to increase the critical clearing time of a large power system? Ie GSU impedance, conductor size, generator inertia... I know this question is off the wall (for North America), but rather in regards to a developing country. Can the laws of physics even allow for this?

 

[QBplanner]

" increase the critical clearing time "
I thought all the system design and P&C schemes is to reduce the critical fault clearing time not to increase it. Am I missing anything?

 

[HamburgerHelper]

Lowering impedance to the rest of the system. I have seen this with an autotransformer that came in to connect a generator hanging on the 138 kV to 345 kV. The utility that got a 345 kV autotransformer got it as part their ISO's transmission investment plan. Every utility had to be given something and they got an autotransformer to help their generator.

Force the generators to need to supply more vars. The more excited their rotor is, the stronger they are coupled to the grid. Running the grid high, I suppose, would help too.

Conductor size does help but the spacing of the conductors in relationship to each other makes a bigger difference. There is a vendor that sells a transmission tower with curved bars that pulls the 3 phases closer together than standard transmission towers to reduce the impedance and the need for right of way space. I can't remember their name but Austin Power has them.

Series capacitance to reduce the impedance between the generator and the system. I think you can compensate up to like 70% before you have issues.

Increase generator or system inertia. Dams have a lot of inertia. Wind generator controls for type 4 turbines can be programmed to provide fake inertia.

Reduce the power output so you are operating at a lower power angle. The closer you are to the critical angle, the less time you have before the generator becomes unstable.

You can reduce the acceleration of a generator by reducing the mechanical energy acting on the prime mover. Up by Idaho, there is or was a long section of transmission that was used like a heating element to slow down an accelerating generator. I think it is like 10 miles of transmission used as a heating element. One of the utilities that Warren Buffet purchased, installed a fast acting valve to blow out steam so that they could operate at a higher power angle.

I suppose if you wanted to, you could have all your instantaneous reaches over reach so could have instantaneous tripping at the expense of security. You would have a lot of false tripping but your fault would always be cleared fast.



QBPlanner,

Doesn't planning dictate critical clearing time with their stability studies? We ,P&C, just try to undershoot that time and that usually means we try to make sure breaker failure can do its thing in that time period. I work at a utility that has a very strong system so this issue for us only comes up occasionally.

 

[Mbrooke]

@QBPlanner: By increasing the critical clearing time the goal is to design the system itself (not the protective relaying) to be able to hold a fault for a longer period of time before the system becomes unstable.

@HamburgerHelper: Well written! Are there any down falls to running a grid high? Also- I always assumed the opposite that having a smaller impedance increases CCC in that if you have many transmission lines with thin conductors as apposed to a few with large conductors a fault on any one circuit draws less short circuit current- but you are rather referring to having a low impedance from all parts of the system into the faulted portion?

 

[HamburgerHelper]

Mbrooke,

By running the grid high, I meant above nominal voltage. If you look at the classical equal area stability curve, you have

Electrical_Real_Power_Given_to_the_System =V1*V2*sin(power angle)/Z.

http://nptel.ac.in/courses/Webcourse-contents/IIT-KANPUR/power-system/chapter_9/9_4.html

If you can run V1 and V2 a few percent above nominal, it will lower your pre-contigency power angle and buy you some more time. I don't know if this is practical. Low voltage does lend itself to more stability issues, though.

When a 3 phase fault (the worst) happens, the generator starts not putting real power into the system. When this happens, the imbalance in real mechanical power fed into the generator vs electrical real power fed into the grid causes the generator to accelerate. When the fault is cleared, the generator starts dumping the extra mechanical energy it has into the system and starts deaccelerating. If it is a stable swing, the generator will not swing so far as to slip a pole (power angle exceeding 90 deg). If its swing exceeds 90 deg, it be unstable and will swing all over the place, alternating between generating and motoring until something breaks.

I think you have to look at it as the the thevenin equivalent between the generator and its loads. If you have a generator in the middle of nowhere, you would like the connection between it and what is using the power ,maybe a city, to be of low impedance. The more impedance there is, the higher the power angle the generator will be resting at during normal generation and the lower the amount of time you have to clear the fault before the generator swings too far out.

 

[QBplanner]

@HamburgerHelper
System planners typically perform the various rotor angle stability studies to define the required fault clearing time and pass the requirement to P&C. If the machine loss of synchronism after faults no matter how fast the fault clearing time can do 4 cycles the fastest one I can think, System planners will propose various alternatives to improve the system stability such as what you already pointed out:

Fast excitation with high ceiling
High voltage level, bundled conductors
series compensation.
Larger inertia of generator
braking resistor
fast Valve for steam turbine
etc.
I once required a 3 cycles fault clearing time because the system is weak via a 300+ km long line with a 300MW + generator at the end of the line.

@Mbrooke
Now I got your point. Sorry I was confused with the reserved logic to relax the P&C requirements.
See above alternatives provided by Hamburgerhelper
Some non-typical ones such as high resistance lines connecting the generators or tripping the units if they are hydro ones or wind.
Keep in mind any system design changes trying to increase fault clearing time will be expensive than the P&C schemes because P&C schemes do not cost much it is the telecom system which is still a smaller portion comparing to the modifications of system design such as installing series caps, braking resistors, Fast valves if works, requiring larger inertia generators.
Hope it helps.

 

[QBplanner]

A new system design is more complicated depending on lot of factors distances between the loads center and generations, size of the loads and generators, required design criteria, network topology you want to achieve from long term perspectives, type of the generators, types of the loads, how to interconnect to the existing network etc.
To answer your questions yes it is allowed but costly,

 

[Mbrooke]

@QBplanner: Its ok, I would have been confused as well now that I think about it- the wording could have been better on my part. My apologies here

@:HamburgerHelper: I need more explanation in regards to why a generator accelerates during a fault. I've always assumed that as load increased, even for a fault, the rotor slowed down. I know from diesel backup generators (25-2000Kw) that a short circuit on the secondary can actually cause the generator to lurch to one side and tip over if not bolted down due to the generator all of a sudden acting as a "break" on the engine. Why does a 500MW steam generator accelerate while a 50kw diesel decelerates?

In regards to slipping a pole and motoring, this would be called "out-of-step" correct? Where the motoring during the pole slip can cause zone 2 distance elements to pick up if reaching through the GSU? Or would I be mistaken?


But from everything I am gather thus far is this: If I was to build a new system having many dozens of 500MW generating stations scattered around a dozen load pockets 20 miles away- or really any system with lots of load a generation evenly scattered about- it is absolutely in my best interest (in terms of CCT) to make sure all the lines linking everything together are of low impedance? Ie, this would involve selecting 500kv instead of 345kv, 230kv instead of 115kv, having many more lines linking things together (say 4 230kv lines where two would suffice) and having each circuit with a higher rating (ie, 2x 1590 ACSR instead of single 795 ACSR)? Would the GSU impedance also factor in here?


FWIW I've read that having many smaller generators increases the CCT instead of smaller but larger MW units- but I believe that has to do with the inertia in the generator itself.



Really eye opening and educational responses thus far.

 

[davidbeach]

Any generator accelerates during a fault. There's very little real power in a fault; power delivery to a voltage of zero is zero. When the electric power out drops during the fault the mechanical power in doesn't change immediately to match. Since that mechanical power can't be converted to electric power it does the only other thing it can do and accelerates the generator and the prime mover.

Consider a single generator feeding a single load. The generator's real power out is consumed by the R component of the connected "load" and the reactive power out by the X component. Most of the R will be in the load served and some in the circuit between generator and load while most of the X will be in the circuit. Now place a short across the load terminals. The bulk of the R is gone and most of the X remains. The impedance is much lower, the current is much higher, but the power factor is much lower. So all that increased current carries much less real power.

 

[bacon4life]

Solar PV doesn't accelerate during a fault. Seems like a build grid fully based on power electronics sources wouldn't have a CCT driven by inertia, but would be instead driven by the overload ratings of the inverters. Of course solar PV also doesn't have inertia to help reduce the rate frequency of decline after a generator trip
How do wind turbines behave during a fault? Since they have a wide range of operating speed, seems like they could be allowed a very large amount of acceleration while having the power electronics stay synced to the grid. I don't know how fast can wind turbines can adjust pitch to reduce mechanical power input.

Here is a video of the BPA Chief Joseph braking resistor in action.

https://www.facebook.com/bonnevillepower/videos/10154155355534374/

 

[Mbrooke]

Any generator accelerates during a fault. There's very little real power in a fault; power delivery to a voltage of zero is zero. When the electric power out drops during the fault the mechanical power in doesn't change immediately to match. Since that mechanical power can't be converted to electric power it does the only other thing it can do and accelerates the generator and the prime mover.

Consider a single generator feeding a single load. The generator's real power out is consumed by the R component of the connected "load" and the reactive power out by the X component. Most of the R will be in the load served and some in the circuit between generator and load while most of the X will be in the circuit. Now place a short across the load terminals. The bulk of the R is gone and most of the X remains. The impedance is much lower, the current is much higher, but the power factor is much lower. So all that increased current carries much less real power.

But what confuses me is that the circuit itself has R as well. Even very close to the generator, the conductors between it and the fault are not super conductors- there is I2R losses and those conductors simply behave as heaters in the process. Basically a really big load placed across the generator. Unless you are saying that under short circuit conditions there is far more X then R in the conduit? Am I correct on this?

Solar PV doesn't accelerate during a fault. Seems like a build grid fully based on power electronics sources wouldn't have a CCT driven by inertia, but would be instead driven by the overload ratings of the inverters. Of course solar PV also doesn't have inertia to help reduce the rate frequency of decline after a generator trip sad
How do wind turbines behave during a fault? Since they have a wide range of operating speed, seems like they could be allowed a very large amount of acceleration while having the power electronics stay synced to the grid. I don't know how fast can wind turbines can adjust pitch to reduce mechanical power input.

Here is a video of the BPA Chief Joseph braking resistor in action.

https://www.facebook.com/bonnevillepower/videos/10...

Neat video

I agree with you on PV- and to be honest my understanding is that the current system in use is not meant for 100% renewables like solar and wind. There has been talk of having DC and semi-conductor breakers taking over from here.

 

[wroggent]

R is very small and so I2R is as well.

 

[QBplanner]

"But from everything I am gather thus far is this: If I was to build a new system having many dozens of 500MW generating stations scattered around a dozen load pockets 20 miles away- or really any system with lots of load a generation evenly scattered about- it is absolutely in my best interest (in terms of CCT) to make sure all the lines linking everything together are of low impedance? Ie, this would involve selecting 500kv instead of 345kv, 230kv instead of 115kv, having many more lines linking things together (say 4 230kv lines where two would suffice) and having each circuit with a higher rating (ie, 2x 1590 ACSR instead of single 795 ACSR)? Would the GSU impedance also factor in here?"

Your case is a system design issue. depending on the area you are in, Let's say you have 10X 500MW CCGT, and 4000MW load 20 miles away.
You can go with 500kV alternatives: let's say 3 lines
345kV :5 lines
230kV :7 lines

At 20 miles distance, my guess is you will not likely encounter rotor angle stability issues but only thermal over loading after loss of 1 line or 2. But you need to run the study to make sure under fault conditions, the generators will survive.
The rest will be losses, cost and NPV analysis unless you want to go to the next level which is network topology design for future then it will be more complicated.

"But what confuses me is that the circuit itself has R as well. Even very close to the generator, the conductors between it and the fault are not super conductors- there is I2R losses and those conductors simply behave as heaters in the process. Basically a really big load placed across the generator. Unless you are saying that under short circuit conditions there is far more X then R in the conduit? Am I correct on this?
"
You are correct, the line will act like a braking resistor I had this case before, a generator connects to grid via a 300kV 138kV line and during the fault within the grid, the generators survive the fault, I checked the generator cluster output during the fault, because of the high line losses on the 138kV line over 300km, generators provided almost 35% of the their nominal capacity to the line. The line resistance act like the braking resistor in this case. Does it answer your question?

Again, you will have heavy line losses during normal conditions, you need to compare the losses among the alternatives and select the best voltage level suitable for your system.

 

[Mbrooke]

Getting clearer. As I understand now the breaking affect is after the fault is cleared?


Also, what exactly do you mean by this: "unless you want to go to the next level which is network topology design for future then it will be more complicated" By network topology you mean anticipation of future load growth or designing the system as a mesh?

 

[HamburgerHelper]

Anything that uses real power brakes the generator. If your line is long enough and has enough R, it will help brake the generator during a fault. It will also brake the generator during normal operations ,too, since it is consuming real power. The braking resistor that Qplanner is talking about is switched in when the generator starts accelerating. You are trying to chew up as much real power as you can so that the generator doesn't accelerating enough to create an unstable swing.

 

[HamburgerHelper]

I can't help but post this.

So, at least as I understand it some trains have electric motors at each wheel now. In this video, I think you can see where each individual resistive brake is as the train passes by. The things that are glowing red I think are the resistive brakes.

https://youtu.be/2PNQ2hrb3Rw?t=52

Here is a dynamic braking bank on top of a train that has problems and is sparking.

https://www.youtube.com/watch?v=5OwMzI9P79c

It blow my mind that this works. An incredible amount of energy has to be dumped into these banks to slow an electric train down. Does anyone know anything about trains?

 

[QBplanner]

No the Braking effect is during the fault to allow the generator to continue to output the electrical power and have less kinetic energy stored in the machine therefore less accelerating power after the fault is cleared and the generator will not swing beyond certain angle say 90 degree C (just a text book value, I saw some machines swing up more than 120 degree) some utilities in North America will switching in the Braking resistors. In my case, the transmission line acts like a braking resistor.

"By network topology you mean anticipation of future load growth or designing the system as a mesh? "
Yes, if you have the existing transmission system and try to mesh the new system in you may want to do more studies and analysis the related issues with different topologies. Or you start with a brand new system you may want to refer to the existing system in some utilities' such as Hydro Quebec's 735kv system. But if your system is a short distance one, then you may want to refer to Japan or Europe system.

In the country I came from back to 1990s we started to build up our bulk system 500kV and 765kV system and higher and also in the larger cities we started to develop network to provide supply to cities. We visited Japan , Europe like France learn the Paris network and BPA system and trying to learn how to develop the system. We pay a lot of attention to the topology design and make sure it will not have any major impacts to the system under extreme conditions to avoid here in North America in 2006 the black out happened in First energy. The big loop system is a trouble maker under certain conditions with the reversed flows.
Best wish to your new system.
Not every planners have the chances to design a brand new system. Some in their whole career do not even have one chance to plan anything new like the people around me now.

 

[QBplanner]

Again, you may require a lot of experts when designing a new EHV system. The reason I refer to Hydro Quebec is because it has a strong transmission planning team back then and has IREQ to provide technical supports for any type of issues relate to their 735kV system. I worked there for a few years very impressive place to work maybe the last utility in North America who still keep research institute within the utility.

 

[Mbrooke]

This system is being evaluated as a mesh, at least the goal is to have the 500kv, 345 and 230kv as a strong tight nit mesh or in the least a series of loops. Power can flow in or out in all directions.

 

[crshears]

Are there any down falls to running a grid high?

Not that many unmanageable ones; beyond the associated damage that can occur at sustained high voltages, and, in specific circumstances, sometimes even despite them, both our IESO and we as the transmission operators for my utility definitely prefer to run our grid voltages, across the board, at the upper end of their acceptable range for stability purposes. Static shunt capacitors deliver the most reactive power that way, power transfer capabilities are maximized, etc.

There are more downfalls to allowing a grid to run low than high; indeed, allowing the voltage to sag too far can be problematic. During periods of increasing load, if the system voltage gets too low, the incumbent caps' output will drop due to their output varying as the square of the voltage, and it can become a toss-up whether we get the remaining available caps in service in time to correct the sag, and yes, falling behind the curve has happened; not pretty.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]