3-Phase Transformer Backfeed

[TNPE]

I'm only posing this for consideration by others.  Suppose you have a 3-phase tap that is protected by fuses and you lose a phase due to a line-to-ground fault.  A 3-phase XFMR is down line from the blown fuse, but is still energized on the other two phases.  On this tap, the blown-fused phase feeds a small residential development.  What would you anticipate to happen (neglecting single phasing and ground imbalance conditions that could take this line out entirely either at the fuse bank or further upline at an electronic control or relay)?  

Note:  I'm using a distribution grid as the example system and the winding configuration of the XFMR is not necessarily an overbearing factor

@rg1 @cos90 and others prepping for the exam, this is for y'all.  I am merely using this example for understanding backfeed conditions, or if even back feeding would be incurred on the "blown phase."  Feel free to respond and I'll chime in as needed.

 

[rg1]

I'm only posing this for consideration by others.  Suppose you have a 3-phase tap (Tapped line by single phase Xmer- Disconnect Switch with fuses???) that is protected by fuses and you lose a phase due to a line-to-ground fault.  A 3-phase XFMR is down line from the blown fuse, but is still energized on the other two phases.  On this tap, the blown-fused phase feeds a small residential development ( Tap is before i.e. upline side of the fuses??).  What would you anticipate to happen (neglecting single phasing and ground imbalance conditions that could take this line out entirely either at the fuse bank or further upline at an electronic control or relay)?  

Note:  I'm using a distribution grid as the example system and the winding configuration of the XFMR is not necessarily an overbearing factor

@rg1 @cos90 and others prepping for the exam, this is for y'all.  I am merely using this example for understanding backfeed conditions, or if even back feeding would be incurred on the "blown phase."  Feel free to respond and I'll chime in as needed.

I know it will be a very interesting question and those who have worked on distribution networks(I have not) always get beautiful scenarios for discussion. I have not understood the question fully, can you elaborate the circuit for me. I have some understanding, I marked in red, are they correct? You can write elements of circuit in order- need not draw it and then explain the problem. 

This is not a question of broken neutral wire, I hope??? 

 

[TNPE]

No, not broken neutral.  Let me try to further explain so that it is more clear.

Imagine a main feeder (3-phase), and at some arbitrary junction, a 3-phase tap is pulled off the main feeder.  At this junction, the lateral tap is protected by a bank of fuses (3).  Think of this tap in the same way you would a basic circuit that has a parallel path.  Suppose a single LG fault occurs down line from this fuse bank and takes out, say, C phase.  Also down line from this same fuse bank, a 3-phase XFMR (winding configuration is not of paramount concern) is servicing a commercial/industrial complex (doesn't really matter, just take note there's a 3-phase service that is now only energized from the grid on phases A and B).  Also on this same lateral tap with the blown C phase fuse, a small residential development is serviced by the blown phase (C phase).  What would you expect these residential consumers to see?  Will this XFMR backfeed onto the grid and put load on the "open" phase?  If not, why won't it?  I'm asking more for concepts, not numbers.  

This will most likely give it away, but what's different between a single phase XFMR and a 3 phase XFMR?  No, not that it's 3 phases and not 1, think more along the lines of circuitry, connections and internal design.  How is flux generated in a XFMR? 

Does this make sense?  If not, I'll continue until I make it clear for you.  I've tried to upload pictures on here before but have been unsuccessful. 

 

[knight1fox3]

I've tried to upload pictures on here before but have been unsuccessful. 

Due to forum file size limitations for users, you'd be better off hosting said images using TinyPic or Photobucket, then simply embedding the image in a post using the "Insert other media" feature (select "Insert image from URL" option). :thumbs:

 

[TNPE]

10-4.  That explains my struggles.

 

[PwrEngr]

The open fuse will behave as an open circuit, meaning the C phase is still energized. However, there is the possibility that the customers on the C phase will experience power quality issue like flicker).

The transformer will start to overheat and experience losses due to very high voltage back feeding into it.

(Just a quick explanation based on prior reading)

 

[cos90]

I also don't work on distribution systems. But this is an interesting question that I would have got wrong. I was assuming 1/3rd of the customers would get dropped. Just goes to show don't assume on the exam, always check your references or actually solve the problem schematically or numerically even if the problem doesn't give you numbers.

 

[cos90]

This effect can be observed with only one phase energized, according to this document:

http://www.swedeus.org/SWEDEpdf/2010/Voltage Backfeed on D-Lines - Oncor -TDM.pdf

This is not an intuitive effect.

 

[TNPE]

@cos90 you're correct, it can be seen, though at very low voltage levels (at least with regards to nominal).  If you think of a three phase XFMR, the flux is coupled between the legs of the core simply due to the design (usually either 4 or 5 legged to mitigate heating, as @PwrEngr alluded to).

For modern appliances/heat pumps/etc., most have sensors that limit them from turning on when low voltage is present (usually on the order of 60-70 % of nominal and ranging from a few cycles up to dozens of seconds under these conditions).  A cursory investigation would lead you to consider that only resistive, unsensored circuits will be 'on'.  Further investigation would indicate that these consumers will experience what PwrEngr mentioned, flickering lights and brownouts.   

I don't foresee that you would see this exact scenario on the exam, however, I presented this for pure understanding purposes, and to illustrate how real-world scenarios apply on a power system.  

I'll post more scenarios as time permits and as I think them up.

Here's a fairly simple one, but it may be of use to someone reading this board, what purpose do fuses over a cap bank serve? Seems intuitive and, for the most part, it is; but it is a fundamental question that can trigger someone's thought process when working with capacitors, reactive power, PF correction and the concept of "negative" VARs.

 

[rg1]

Are we discussing about the behavior of transformer on single phasing- What will be Voltage induced on the dead or disconnected phase(HV/LV). Then it will depend on whether the transformer is 3 or 4/5 legged and it will also depend on how the primary is connected, i.e. whether it Y, Y-G or Delta. Am I on right track for understanding the question? If these things are known we can formulate the maths out of it?

 

[TNPE]

From a mathematical, numerical standpoint, that is correct.  However, I was purposely dodging the math for a number of reasons, including core design, flux, reluctance, leakage reactance, hysteresis, load and other factors that would dictate the actual voltage present on the "open" phase.  I was strictly aiming for concepts, not x=y so z=a.  

 

[rg1]

Thanks TNPE. Long back I had discussed a few scenarios of this kind. I thank @cos90for sharing the that ABB document, that is quite exhaustive. Because the forum is for PE I do not know how much it will be useful for all, studying all scenarios, but one or two which are commonly faced in real life like TNPE shared should be understood IMHO.

 

[PwrEngr]

what purpose do fuses over a cap bank serve?

Fuses can help locate failed capacitor units in a capacitor bank. See document attached

The presence of capacitors in a circuit is beneficial for power factor correction due to their leading power factor/neg. vars. However, due to a potential of high charging current in the circuit, fuses may be needed as OCPDs. @TNPE, please let me know if that meets the intent of your question.

View attachment Shunt Capacitor.pdf

 

[TNPE]

I was thinking in terms of the fuses protecting the system from the caps, not the other way around.  Maybe that's what you were saying and I misunderstood. 

Sure, a very simple concept, but often times misconceptualized.  Most think of a fuse bank as protecting down line sections (i.e. caps down line from the fuse bank), but in understanding capacitors and what they do, it is quickly realized that is not the case.  

 

[PwrEngr]

I was thinking in terms of the fuses protecting the system from the caps, not the other way around

Ok. I can see fuses protecting a given circuit against line charging current coming from the caps, if properly sized

 

[rg1]

I was thinking in terms of the fuses protecting the system from the caps, not the other way around.  Maybe that's what you were saying and I misunderstood. 

Sure, a very simple concept, but often times misconceptualized.  Most think of a fuse bank as protecting down line sections (i.e. caps down line from the fuse bank), but in understanding capacitors and what they do, it is quickly realized that is not the case.  

IMHO the fuses always protect the equipment they are meant for. The capacitors are also like any  other loads (Resistors, Inductors, motors) except that they are conventionally considered to be generators of VARs which is taken care by the maths. Rest all is same. Yes capacitors and Inductors unlike resistors have transient behavior which is different from steady state and we have to account for this behaviors too while using them. However most transmission line questions  are steady state. 

 

[TNPE]

Agree, and yes, the fuses are two-fold - they protect and isolate in the event of an internal failure/case fault/etc., but they also protect the system from the caps themselves.  I was simply illustrating that the behavior of capacitors (and inductive loads) are very different than resistive loads.

 

[TNPE]

The primary differences between these loads and resistive loads - they are passive energy storage devices (magnetic field for inductive and electric field for capacitive). Resistors store no energy, they dissipate.  And it follows from this that you get into the mathematics, PF correction, leading/lagging current, etc.  

I try not to explicitly answer questions on this board immediately.  I find it to be detrimental in developing others' reasoning skills.  As you've noticed, I attempt to pose a counter-question to trigger their thought process, leading them to their own conclusions and understanding.

 

[Taylorleelai]

No, not broken neutral. Let me try to further explain so that it is more clear.

Imagine a main feeder (3-phase), and at some arbitrary junction, a 3-phase tap is pulled off the main feeder. At this junction, the lateral tap is protected by a bank of fuses (3). Think of this tap in the same way you would a basic circuit that has a parallel path. Suppose a single LG fault occurs down line from this fuse bank and takes out, say, C phase. Also down line from this same fuse bank, a 3-phase XFMR (winding configuration is not of paramount concern) is servicing a commercial/industrial complex (doesn't really matter, just take note there's a 3-phase service that is now only energized from the grid on phases A and B). Also on this same lateral tap with the blown C phase fuse, a small residential development is serviced by the blown phase (C phase). What would you expect these residential consumers to see? Will this XFMR backfeed onto the grid and put load on the "open" phase? If not, why won't it? I'm asking more for concepts, not numbers.

This will most likely give it away, but what's different between a single phase XFMR and a 3 phase XFMR? No, not that it's 3 phases and not 1, think more along the lines of circuitry, connections and internal design. How is flux generated in a XFMR?

Does this make sense? If not, I'll continue until I make it clear for you. I've tried to upload pictures on here before but have been unsuccessful.

Customers on c phase if it’s a open delta will still be energized (which is dangerous especially on a storm call without knowing the bank). If it’s a wye-wye bank with a phase still hot then more then likely a customers generator is hooked up wrong or bad transfer switch causing to back feed with primary voltage going out. Correct?

 

[Cram For The PE]

You can read about real life events on the situation you are describing here:

https://www.nrc.gov/reactors/operating/ops-experience/open-phase-electric-systems.html
A famous example of this occurred on Jan. 30, 2012, Unit 2 at the Byron Station.

A single phase left open can go undetected depending on several factors. I was heavily involved in open phase detection including new technology to help detect it. Let me know if you have further questions after reading about the operating experience.