Sequence Networks 6

[birddogger]

What exactly are sequence networks? I have taken courses on power systems analysis and read other material on the subject, and the most I can gather is that it’s simply a different domain, and that voltages and currents can be mapped into this domain and manipulated easier than in the time domain, much like with Fourier or LaPlace transforms. But it still doesn’t really tell me what positive, negative, and zero sequences actually represent. I understand the whole a-b-c and a-c-b phasor rotation thing that defines positive and negative sequence, but it still doesn’t really tell me what they “are”.

Close as I can figure, when you are talking about a fault on a system, there are three components to the fault current – the excess current that is flowing in the direction it would under normal conditions (positive), the excess current that is flowing in the opposite direction than it normally would (negative), and the steady-state current which is equally balanced among all three phases (zero). Am I on the right track? How do the sequence networks vary from line-to-line faults vs. line-to-ground faults or bolted three-phase faults?

Just trying to figure how this applies in the real world. I know that protective relaying makes heavy use of sequence networks, and I was just wondering how it all comes together under such applications.

 

[ScottyUK]

Birddogger,

There is a good explanation of sequence groups in the Areva Network Protection & Automation Guide, in an early chapter on background maths. It's free to download from their website, or the hardcopy costs £70 in the UK and is one of the best protection books you could hope to buy. Suggest you download the guide at work unless you have broadband - kinda big!




-----------------------------------

Start each new day with a smile.

Get it over with.

 

[dpc]

Check this out:

http://www.selinc.com/techpprs/6066.pdf

You'll need a good book on symmetrical components and understanding of phasors to make a lot of sense out of the sequence components.

Basically, it is a mathematical tool that allows analysis of unbalanced conditions using balanced networks.

Good luck.

 

[cuky2000]

Fallow DPC advise. Consider using the following responses as an introductory motivation to this topic.

QUESTION 1:What exactly are sequence networks?
For current and voltage in a 3-phase power systems could be represented by a set of 3 phasors on the time domain (phase A, B &C). This phasors could be discomposed by 3 sets of balance phasors that are the symmetrical components
The advantage of this modeling technique is to converts a 3 phase unbalanced network to a set of three balanced networks that each can be represented by a single-phase equivalent circuit. For additional graphic details see the enclose site:

http://xnet.rrc.mb.ca/janaj/1_2_1_symmetrical_components.htm

QUESTION 2:I understand the whole a-b-c and a-c-b phasor rotation thing that defines positive and negative sequence, but it still doesn’t really tell me what they “are”.
The symmetrical component are “n” arbitrarily decomposition of the “n” polyphase system mapping the equivalent wave from the time domain to the frequency domain.
_{Please note that for a polyphase system of “n” phase the resultant symmetrical component are equal to “n”. Ex. For 3 phase: Sym Comp j= 0 (zero), 1(positive), 2 (negative) (total 3) angle. For 6-phase system: 0, 1,2,3,4 &5 (total 6). The rotating phasor sequence angle areas follow:
Sequence angle = k.360/n
Were:
.n = Number of system phase.
k= 0 (zero seq), 1(positive seq) & 2 (negative seq.) for 3-phase systems.
.k= 0,1,2,3,4,5 for 6-phase system etc.}

QUESTION 3: Close as I can figure, when you are talking about a fault on a system, there are three components to the fault current – the excess current that is flowing in the direction it would under normal conditions (positive), the excess current that is flowing in the opposite direction than it normally would (negative), and the steady-state current which is equally balanced among all three phases (zero). Am I on the right track?
ANSWER 3: Unfortunately not, there are few clarifications that need to be made.
a- The zero sequence is associated with the neutral return path.
b- In steady state, the system could be close to balance or imbalance depending on the single-phase load connection.

QUESTION 4: How do the sequences networks vary from line-to-line faults vs. line-to-ground faults or bolted three-phase faults?
These circuits are then connected in a way that simulates the conditions on the actual unbalanced system and allows calculation of the currents and voltages during a fault
as follow:
a- SLG: Series connection of the 3 sequence impedances.
b- L-L: Series connections of the positive and negative. (zero sequence are not included)
c- 3-phase bolted connection: only the positive impedance is considered.

 

[jbartos]

birddogger (Electrical) Apr 28, 2004
What exactly are sequence networks? I have taken courses on power systems analysis and read other material on the subject, and the most I can gather is that it’s simply a different domain, and that voltages and currents can be mapped into this domain and manipulated easier than in the time domain, much like with Fourier or LaPlace transforms. But it still doesn’t really tell me what positive, negative, and zero sequences actually represent. I understand the whole a-b-c and a-c-b phasor rotation thing that defines positive and negative sequence, but it still doesn’t really tell me what they “are”.
///Symmetrical components that include positive, negative and zero phase sequence components are mathematical description useful to deal with multi-phase systems that have symmetrical or nonsymetrical (meaning unbalanced) power supply source and/or loads.\\Close as I can figure, when you are talking about a fault on a system, there are three components to the fault current – the excess current that is flowing in the direction it would under normal conditions (positive), the excess current that is flowing in the opposite direction than it normally would (negative), and the steady-state current
///the steady state would be better called a common-mode current since it flows in all phase conductors in the same direction toward the neutral, i.e. wye connection with neutral or grounded neutral is a mandatory prerequisite. The delta connection has a very negligible zero sequence current through the delta winding leakages to the ground.\\ which is equally balanced among all three phases (zero).
///Yes, correct.\\ Am I on the right track? How do the sequence networks vary from line-to-line faults vs. line-to-ground faults or bolted three-phase faults?
///It vary widely. This is usually explain in the electrical power system textbooks, e.g. Stevenson "Elements of Power Systems Analysis."\\

 

[birddogger]

To dpc - I have that exact handout from SEL labs, but it only goes up to 3/4 of page 8. It's obviously been apended since then. Thanks for the updated version!

 

[busbar]

Dr Mack Grady’s {University of Texas} PDF notes are online:

http://www.ece.utexas.edu/~grady/PNG.html

 

[DiscoP]

To briefly summarise sequence components, if the three phases are prefectly balanced, it is all positive sequence current and voltage.

If there is an imbalance, there will be negative sequence current. There will be an unbalance during L-L, L-G or L-L-G faults.

If there is a ground fault, there will also be zero sequnce current.

Relays measure or calculate the zero sequence components, and operate if above the setting. Because during healthy conditions there will be no zero sequence current, it can be set more sensitive.

Some modern relays can also calculate the negative sequence component. Like zero sequence, presence of this means a fault, and it can also be set a little more sensitive.

Note that in theory there shouldn't be any negative or zero sequence components, but in reality there is a little bit of each during normal conditions, so the settings need to be above these actual values.

Read some of the other references mentioned for details on how they are determined.

 

[jbartos]

Suggestions to the previous posting marked ///\\
If there is an imbalance, there will be negative sequence current.
///and possibly the zero sequence phase components, depending on the load connection, wye in particular.\\ There will be an unbalance during L-L, L-G or L-L-G faults.

If there is a ground fault, there will also be zero sequnce current.
///It depends, the symmetrical three phase to phase to ground fault will not produce the ground fault current.\\
Relays measure or calculate the zero sequence components, and operate if above the setting. Because during healthy conditions there will be no zero sequence current, it can be set more sensitive.
///Not exactly. The unbalanced load and wye connection with neutral will cause the zero sequence current flowing.\\
Some modern relays can also calculate the negative sequence component. Like zero sequence, presence of this means a fault, and it can also be set a little more sensitive.
///Not necessarily, when it comes to unbalanced loads.\\
Note that in theory there shouldn't be any negative or zero sequence components, but in reality there is a little bit
///Or quite a bit depending on the unbalance of power supply and/or loads\\ of each during normal conditions, so the settings need to be above these actual values.

 

[rsherry]

Sequence network analysis is just a handy mathematical tool (the dq axis analysis for rotating machines is another common example in power systems). These were developed because they reduce the required calculations to a manageable task. Sometimes the model increases understanding of the phenomenon (such as the dq machine model) and sometimes it does not because sometimes the model does not reflect anything in the physical system it is just convenient (e.g. sequence modelling).
The real power system only has positively rotating phase voltages. They are not always balanced (i.e stably 120 degrees apart and of equal magnitude) and this makes the calculations tricky. The sequence model uses a mathematical proof that such a system can be exactly represented by the summation of three sets of balanced rotating voltages - and this reduces the calculations required. With increasing brute force computing power such reductions in computing requirements are becoming less necessary, and just as full 3 phase power flow calculations are now common, full 3 phase fault modelling will not be far behind.

 

[jbartos]

Comment: I would just like to continue with the previous posting.
The handy mathematical tool is actually "very handy" since the transformation can be programmed on computer and easily followed; especially, if they are in some abbreviated, e.g. matrix format.
Once the power supplies voltages, power distribution impedances, load impedances are converter into symmetrical components, then it is relatively simple to process these symmetrical components. If needed, the symmetrical component result can be transformed back to regular phase variables and hardware parameters.

 

[rsherry]

Re : jbartos's comments above.
I agree that the sequence component method is probably the best we have for quick calculations - but i don't think the 3 phase fault caluclation will be inherently much more complicated than the sequence network method.
Would you agree that instead of measuring component impedances to negative and zero sequence voltgaes we will simply have to measure self and mutual phase impedances (or derive these backwards from the sequence measurements).
This can still be easily expressed in matrix format. The 3 phase fault calculation will then need a 3 phase loadflow algorithm to derive the fault currents (the trickier bit computationally). I think the 3 phase calculation will actually be much more understandable as the result will give the actual currents on the network directly - and can automatically include the effects of loads, shunt elements etc that have to be specially treated using sequence components.

 

[MichaelKelly]

Birddogger,

What are the key words that should be used to find the Network Protection & Automation Guide of which you speak in this thread? I have used all the combinations I can think of and the Areva site just returns "no information available."

Thanks,
kellymw@bv.com

 

[birddogger]

You'll have to ask ScottyUK, he's the one that posted that note. I think I remember also having trouble finding that one a couple weeks ago.

 

[umrpwr]

Listed below is the link to download the PRAG guide.

http://www.areva-td.com/servlet/Con...7&lid=en&tab=Chapters&id=1056536208254#spare2

It is also still available in hard copy as well. I believe it can be purchased from Areva T&D which used to be Alstom.

 

[davidbeach]

The Network Protection & Automation Guide is available at

http://www.areva-td.com/servlet/Con...8589697&lid=en&tab=Chapters&id=1056536208254.

It would be nice if it had the balance between IEC and ANSI closer to the point of both being covered equally, rather than mostly IEC (but that probably makes up for a lot of other works that are predominantly ANSI). But this is a great book in that it reasonably up to date and fairly thorough.

 

[davidbeach]

It looks like umrpwr and I answered about the same time. For some reason the URL in my post includes the beginning of the next sentence, so use the URL in umrpwr's post.

 

[ScottyUK]

Davidbeach,

I think the IEC bias probably stems from the days when GEC Relays of England originally wrote the book, long before they were swallowed along with the rest of GEC by the french company.

Umpower,

The book can be purchased direct from Areva, although the copy I received bore the Alstom name rather than Areva's. I got myself an updated copy while feeling generous with the company's money.



-----------------------------------

Start each new day with a smile.

Get it over with.

 

[jbartos]

Comment on rsherry (Electrical) May 5, 2004 marked ///\\Re : jbartos's comments above.
I agree that the sequence component method is probably the best we have for quick calculations - but i don't think the 3 phase fault caluclation will be inherently much more complicated than the sequence network method.
///Reference: IEEE Std 141-1993 Red Book Section 4.4.5 Practical Impedance Network Synthesis:
"""For unbalanced short-circuit problems, the concept of symmetrical components is used for solutions. This concept discloses that any conceivable condition of unbalanced loading can be correctly synthesized by the use of appropriate magnitudes and phasing of several systems of symmetrical loading. In a three-phase system, with a normal phase separation of 120°, there are just three possible symmetrical patterns. ... These loading patterns satisfy the restraints demanded by the analytical method to be used."""\\The three phase fault calculations
Would you agree that instead of measuring component impedances to negative and zero sequence voltages we will simply have to measure self and mutual phase impedances (or derive these backwards from the sequence measurements).
///Measuring is a different thing from theoretical analysis, or conceptual engineering where there is nothing to be measured.\\This can still be easily expressed in matrix format. The 3 phase fault calculation will then need a 3 phase loadflow algorithm to derive the fault currents (the trickier bit computationally).
///Perhaps in symmetrical systems. When it comes to unsymmetrical systems, the symmetrical components reveal their strength.\\ I think the 3 phase calculation will actually be much more understandable as the result will give the actual currents on the network directly - and can automatically include the effects of loads, shunt elements etc that have to be specially treated using sequence components.
///Yes, this is often being done in symmetrical systems.\\