transitory stability

[bunda222]

Hi,

I am performing a simple work to evaluate the transitory stability of the system shown attached when a three-phase fault happens in the substation S. Osorio. Due to this fault three-lines shown in the picture have been cut off from the system. Now the system lost its synchronism, I need to find a way to make it come back to synchronism again. I was trying to cut off the generation that is attached to the S. Osorio bus, but according to the simulation I'm doing that is not the solution of the problem because the machine's angle grows forever to different directions. I would like to know if you have some tip on what to do to keep this system operating without the 3 line transmissions.

Thanks.

 

[Mbrooke]

More info is needed such as the impedances of the lines, their ratings, the Kv involved, the size of the generator and the load connected to the bus.

But from your question I gather the problem is that when the breakers trip on those 3 lines the resulting phase angle then becomes to great to reclose? Or the generator looses rotor angle stability? Is this what you mean by loss of syncronism? When you say 3 phase fault in the S Osorio, this would be a bus fault, correct? If so a bus fault (assuming a straight bus) would clear more than just those 3 lines. Thus I am assuming the issue is the loss of the two lines between S Osioro and C Mourao and the line between S Osioro and Areia?


Apologies if I am misunderstanding.

 

[QBplanner]

You may try different generation shedding capacity or shedding times at the generation station connecting to S.osorio

 

[bunda222]

The loss of synchronism I mean when the generator loses its rotor angle stability. And it is a bus fault in the bus S. Osorio

I was thinking of that generation shedding due to the loss of that 3 lines. That'll be an over generation in the station Osorio and I need to reduce that generation or find a way to drain out that energy. I was thinking of putting a bank of resistors there. But I don't know if it is something usual to do.

I'm sending a picture of the generation rotor angle for different machines of the benchmark when I put a bank of resistors. Its capacity was chosen to be the line fluxes sum before the breakers trip.

 

[Mbrooke]

The easiest thing would be to convert S Osorio to breaker-and-a-half, ring bus, or double breaker double bus. That way a fault on any bus section or segment at most removes only one line from service instead of 3. These configurations can also mitigate similar consequences arising from breaker failure. This concept is routinely done in the US when a bus fault or breaker failure could produce undesirable results.

Another possibility is to just live with the risk. In a well maintained substation bus faults tend to be fairly rare; and there are many, many cases around the world (even the US) where loosing load and/or generation is considered acceptable on the account its a low probability event. If you can continue supplying critical loads via the remaining bus section it becomes even more tolerable.


And while I am not an expert in RAS (remedial action schemes) and GRB (generator run back), someone else on here who is might be able to determine if thats a valid option.


But I am curious- In your case once the 3 lines and generator trip, what is the per unit voltage on the remaining bus? Does the (faulted) bus section in question also supply load?

 

[bunda222]

The generation does not trip in my case. It is a possible solution to loss of synchronism. In my case when I run the load flow with no fault, the line fluxes in the two lines of C. Mourao - S. Osorio is 109.28 MW and from S.Osorio to Areia is 105.089 MW. Then I think what I need to do is to reduce the generation by the sum of this amount (109.28+109.28+105.089 = 323.663MW). Althougt I don't know if this can be done in a real substation (I am using a simulation software).

 

[crshears]

What MBrooke just said, plus my admittedly quick take on this would be a connectivtiy- or topology-based generation rejection scheme coupled to a load rejection scheme. Due to loops the balance of the grid could be expected to survive.

CR

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

 

[Mbrooke]

But- if you loose those 3 lines and the generator: 1. Does the voltage on the unfaulted bus remain within limits with load staying below the parabola of the power transfer curve? (Ie, the voltage might drop to 0.90pu, but the loads adjusts and fairs well without voltage collapse) 2. Do the remaining lines stay within thermal and sag limits? If the answer is yes to both, loosing the generator may not even be a concern.

 

[QBplanner]

If the generators are Hydro units, then just shed the units until the remaining generations are within synchronism with the rest of eth system. Generation run back can not do much to the rotor angle stability it takes at least 5-10 seconds to act which the machines are already gone. From the curve you sent I assumed you applied a fault at 1 second and it lost in the first swing.

 

[QBplanner]

When u get into a N-3 scenario, any capital investment to improve the system is not necessary. Under N-3, only a few things you can do, RAS to trip the units and shed load, system separation to avoid cascading etc. I have not seen any utilities spending money to fix N-3 scenarios.

 

[bunda222]

I've come up with two solutions:

First: As QBplanner pointed out, if the generators are Hydro Units I can reduce the generation in substation S.Osorio until there is a match between load and generation.

Second: This is more complex. In this scenario the generator unit may not be dispatchable, so I need to cut it off from the system when I lose those three lines. The voltage from the remaining unfaulted buses drops to 0.7 p.u (P. Branco, Xanxere, C. Mourao, Maringa and Apucar.) So I need to cut off those loads too, I guess.

The graph attached shows the voltage at different substations when the generator is unplugged from the system when the fault is cleared 150 ms after the fault. Also the loads are unplugged!

I would like to know whether those solutions are doable.

 

[bunda222]

Here is the rotor angle of the remaining generators after the fault is cleared and the generator connected to the S.Osorio is also unplugged. The rotor angle keeps growing but synchronism is not lost though. The angle difference remains the same.

 

[QBplanner]

I assume you used absolute angle other than relative angle for each machine. You may want to use relative to the system one.
I did not see the 0.7p.u. voltage profiles in the drawing. The voltage profile seems too high at two buses all beyond 1.1pu.
Hydro units is easy to shed within cycles without causing damages.

 

[bunda222]

Those voltages beyond 1.1 pu happen when I cut off loads from the S. Osorio, P.Branco and Xanxere Station that's why there's a voltage swell. The 0.7 pu happens only when no action is taken after lost of three lines.

 

[QBplanner]

Got it which means you over shed the load.

 

[Mbrooke]

Second: This is more complex. In this scenario the generator unit may not be dispatchable, so I need to cut it off from the system when I lose those three lines. The voltage from the remaining unfaulted buses drops to 0.7 p.u (P. Branco, Xanxere, C. Mourao, Maringa and Apucar.) So I need to cut off those loads too, I guess.

Under voltage load shedding on none critical distribution/sub-transmission feeders to bring the voltage back up to stable limits before voltage collapse can take place. Most SEL "utility" class relays have under voltage trip options.

Just be careful how you set it though, to much load being dropped through out the entire system for major regional disturbances can back fire.

The graph attached shows the voltage at different substations when the generator is unplugged from the system when the fault is cleared 150 ms after the fault. Also the loads are unplugged!

What do you mean by the loads being "unplugged"?

 

[Mbrooke]

When u get into a N-3 scenario, any capital investment to improve the system is not necessary. Under N-3, only a few things you can do, RAS to trip the units and shed load, system separation to avoid cascading etc. I have not seen any utilities spending money to fix N-3 scenarios.

I know I am nit picking, and technically you are right by which way you look at it, but I politely beg to differ. Utilties frequently invest into preventing a system from going into N-3 and over. Typical example I can think would be separating a triple circuit tower into 3 single circuits, eliminating circuit cross overs, upgrading straight buss substations to breaker-and-a-half, replicating station relaying, and the like. Yes its often not cost practical for most utilities to design a system to lose 3 elements without dropping any load, but in this case a bus fault would count as a single individual element contingency.

 

[QBplanner]

@Mbrooke
No problem, even rudely beg to differ is fine for me. What I imply is for an "existing system", if there is a bus fault causing three elements out, typical utility way nowadays is to save money by either trip generation or load other than re-configure the system like what you proposed. But for a new system to be designed, your proposal flies.

There is no exception here in North America, I agree with you if one fault tripping out three elements, then in the old NERC TPL standard, it is treated a multiple contingency and allow load shedding. However, in the new NERC TPL standard001-04, NERC keeps silent on the number of elements out but more focus on single contingency or multiple contingencies. The scenario you described belongs to P2 -bus section fault taking out three elements, if the bus section is EHV system then non-consequential load loss is not allowed and utility has to spend money to fix the problem such as what you proposed.
Cheers!

 

[QBplanner]

"One exception"

 

[bunda222]

@MBrooke
you said about too many load drops can back fire the whole system, I think for this N-3 scenario there's not much to be done tbough. The two unfaulted substations (P. Branco and Xanxere) are fed by the Generator tied to S.Sorio bus, so if I don't drop those loads the voltage level will drop too, the same happens to substation C. Mourao, the voltage drops to 0.7 pu due to generator disconnection. I know there are other options to restore the voltage level such as a bank of capacitors there, but they do not exist in the substation when fault happens, so I cannot account with them, maybe the substation does not have physical space to shelter those bank of capacitors.