Ferroresonance

[vtpower]

I have begun my career as an Electrical Engineer in the utility feed and have been thrown right into it. There has been a large push to correct power factor at our tie points to near unity. My question is, is there info or literature that puts ferroresonance in understandable terms? I have read papers on it, but still am uncertain on how exactly it is caused and what occurs. I want to be sure this is not an issue as I begin installing distribution capacitor banks. Thank You.

 

[electricpete]

My two cents (sorry, no links to papers).


You already know about LC resonance. Ferroresonance occurs when a power transformer or more likely potential transformer is pushed into saturation. Due to non-linearity, the the "effective" magnetizing imducatance while in saturation can vary over a wide range. This variation in effective magnetizing impedance creates more possibility for resonance with existing capacitance (which doesn't change with voltage).

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[jraef]

You may want to see if you can find a book written on the subject by Joseph Sola, the founder of Sola Electric and the purported "inventor" of the ferroresonant transformer still in use by Sola and everyone else today (his patents has expired).

Here is an article I found using his name in a search engine. There were a lot more hits, so you may find more. Good luck.

http://www.ustpower.com/Power_Quality_Notes_No5_December_2004.htm

"Venditori de oleum-vipera non vigere excordis populi"

 

[jghrist]

Ferroresonance is not normally a concern with applying capacitors on a distribution feeder. It can be if there were ungrounded transformer windings that are single-phase switched by a switch on the source side of the capacitors and a transformer winding can be in series with the capacitor. This would be unusual.

Harmonic resonance can be a concern. This is where the capacitor is in resonance with the system reactance at a harmonic, or multiple, of the fundamental frequency. If there are nonlinear loads that create a harmonic current at a resonant frequency, high harmonic voltage distortion can occur.

 

[Alex68]

This is the most exhaustive report I know:

http://www.schneider-electric.com/cahier_technique/en/pdf/ect190.pdf

Dear jraef
I studied Latin for several years but I can't understand it. Please could you explain it.

 

[cuky2000]

Dear Castonguay,

The easiest way to understand the ferroresonant principles is as described by Electricipete in his post above. However, to determine and detect ferroresonance conditions is not as easy as sound since this is a very complex random phenomenon generated by a wide variety of resonance interactions involving capacitors and saturable iron-core inductor that affect not only the frequency dependency between voltage and current but also on a number of other factors such as the system voltage magnitude, initial magnetic flux condition of transformer iron core, the total loss in the ferroresonant circuit and the point on wave of initial switching. (See sample of wave form from Dstar)
.

Some practical advise to considered as mitigation strategy again overvoltage caused by ferroresonance are: provide group operates three-pole switch, use resistance damping resistor for voltage transformers, wyes grounded transformer in the primary & secondary is less prone to ferroresonance, reduce the feeder length, specify arrester with sufficient energy capability to dissipate overvoltage.
NOTE: Many engineers had devoted their career trying understanding this phenomenon with limited success in term since the development of commercial electricity in the 1900’s. Up today the most accurate way to access this phenomenon is to model the system using sophisticates software such as EMTP or others as shown below.

http://www.pqsoft.com/ferroview/ferroview.pdf

http://www.dstar.org/P_R_Summary_Ferr_2.htm

A more modest approach is to use hand calculation, graphic solution or using modeling tool in Excel provided courtesy of Basler and other from Bordeaux University in the enclose links:

http://www.basler.com/downloads/Ferro_R1.xls

http://ferroresonance.free.fr/

 

[mc5w]

One of the things that Cleveland Public Power actually does right is to use 4 primary fuses for delta primary delta secondary transformers. This helps control funny overvoltage that occur when operating line fuses 1 at a time.

Trying to run unity power factor at your substations and tie points could be a big mistake. If you have too much capacitance your generator field UNDERCURRENT relays could trip or your generators could slip a pole. This could cause all kinds of problems. A typical generator cannot go past about 95% or 90% capacitive power factor and still be able to support the full torque from the prime mover.

I was told by somebody that a hydroelectric plant out west had a generator slip a pole which broke a 12 inch shaft and sent the shaft through 12 feet of concrete. In other words, slipping a pole tends to create a major FUBAR or SNAFU.

A better philosophy would be to run the substations at 90% to 95% inductive power factor during peak periods and more like 85% to 90% during off peak periods. Some of your reactive compensation goes into driving inductive loads and some of it tunes the transmission lines so as to lower their impedance. There is an excellent article in Engineering and Science in the Bell System on how reacitve compensation for transmission lines works. If a substation is at the long end of a transmission line then I might want to run more like 90% capacitive during peak periods but there would need to be a fast response controller to prevent overvoltage when a load is lost.

If you have a fast response SCADA system for capacitor control you could try to get 3/4 of system reactive power from switched capacitors, but you might be better off to get more like 60% or even just 50% from the capacitors.

 

[itsmoked]

mc5w do you have a link to that Bell System article?

 

[vtpower]

Thanks everyone for the valuable input. I have to say that this forum is a blessing for someone like myself just getting out of college and into the power world. Thanks again and I look forward to lots of learning!

 

[mc5w]

For the book Engineering and Science in the Bell System you would need to go to a well stocked technical library. Carnegie-Mellon Unniversity had both volumes when I was there in 1908 to 1982. It has all kinds of interesting things in it.

One of the rather interesting things was that digital T carrier was applied to open wire line out west. These lines were built to last about 100 years and many run through places accessible only by mule or helicopter. Under that circumstance it is a lot less expensive to get digital quality voice and channel cramming by grafting something onto the existing circuits that to try to string a fiber optic cable.

Actually, a lot of the lod knowledge is independent of whether you are using vacuum tubes or transistors. Cancelling noise and echoes and so forth is still governed by the laws of physics. Transistors just make problems easier to lick.

 

[mc5w]

I forgot to add 2 other rules of reactive compensation:

1. You have to apply at least Pi reactive compensators per wavelength of line at the highest frequency that you want to pass.

2. When you add reactive compensation the "speed of light" on a transmission line can slow down to as little as 20,000 or even 10,000 miles per second depending on how much reactive compensation there is. This is what is known as a propogation factor.

For longer subscribe loops that standard reactive compensation was 88 millihenries once every nautical mile. This gave a bandwidth of about 4,000 Hertz. In this case the reactive compensation brought the pair impedance up from 200 Ohms to around 600 Ohms so that the insulation would do its fair share of the work.

In undersea cables the wire quads for 2-way analog carrier such as Spiral4 were often wound around a steel wire to avoid having a loading coil once every 1/8 mile. The overhead Spiral4 quads that the military ran all over during world War 2 sometimes used a rubber insulated quad with copper clad steel wires enbedded in a rubber tube. This was essentially a 4-wire of telephone service drop cable. These carrier sets were applied by telephone companies after the war to quadruple the capacity of intercity telephone lines.

The pair gain devices that you see on some subscriber lines as an alternative to party lines work on the sane principle as Spiral4 and other analog carrier sets.

 

[GTstartup]

mc5w: when I was there in 1908 to 1982

Wow, and I thought it took me forever to graduate.


Sorry couldn't resist it

 

[itsmoked]

I always thought mc5w was brighter than that..

 

[TestBeforeTouch]

One of the easiest to understand references on ferroresonance is at:

http://www.cadickcorp.com/download/TB004a_Ferroresonance.pdf

 

[mc5w]

Er, I meant 1980 rather than 1908. My typing is not as accurate as it was in 1977 when I was taking a typing class in high school. I got up to about 60 words per minute with 100% accuracy and figured that all I needed for programming computers.

Turns out that I should have minored in secretarial work back in high school. There are quite a few companies that have refused to hire me because I got financial aid instead of working my way through college. Somehow, the people who went to college Before-Financial-Aid resent those who got financial aid. Not that much different from how New Jersey says that I have to have a $300,000 letter of credit from a bank in order to get an electrician's license.

 

[wrong100]

VTPOWER

MANY COMMENTS no answers.
Some gave you good ref.
CVT stands for constant volage.

In general it is a concept of resonance
transformers. Basic form has three
windings of:
WINDING:
1) input (the line input)
2) resonance winding (a winding
that the LC resonance fc
(winding + external c )is the same
as the input line fc. and can saturate.
3) Output (the ourput winding)
This winding accepts energy from
the input winding, but with mods from the
resonant winding.
The kicker is if the reso winding sats than
no energy is transfered
In concept if the line FC is correct it should
transfer power from winding 1 to 3 (in to out).

(THIS IS ABOUT THE MOST COMPLEX
DESIGN FOR A SINGLE MAG DESIGN )

This transfer of power is dependent
on the specific specs of the resonant
winding and its control/mag sat specs.
VIN,VLOAD, ILOAD, TEMP, ESR
+all cap and winding and mag variables
control pf ect.

Real problem for you is PF can be .1 to 1.
Tip real range is about .3 to .6.

Personal;
Simple (parts) control method
Can be reliable.
Min parts.
Reg % needs to be studied.
Problems (1970)
Hand tweek each unit
high circulating current
Caps explode noise

wrong 100

 

[cuky2000]

Wrong100,

I believe that this discussion is about ferroresonance phenomenon in power systems. You are describing the ferroresonance transformer used for voltage regulation.

Perhaps Vtpower could help us clarify the original question.

 

[vtpower]

cuky and wrong,
I was referring too ferroresonance as it occurs in power distribution systems, 12kV to be specific. Wrongs response is quite interesting as I never knew these kinds of designs even existed. Thanks for all the responses

 

[mc5w]

A lot of 12,000 volt systems here is the U.S., such as the one that Cleveland Public Power has, are 6,900Y12,000 volts solidly grounded at the supply substation that steps down from say 138,000 volts. The distribution transformers are still connected phase to phase. On Cleveland Public Power's 12,000 volt system quite a few circuits do have a full size neutral but it is only used as a equipment grounding conductor, not to run any loads.

On these systems ferroresonance does not happen as much as on a 12,000 volt system where the only "grounding" is through the capacitance to ground.

The reason why the Russians invented the zig-zag set was to convert 3-wire ungrounded transmission and distribution to solidly grounded wye without having to replace existing supply transformers.

Both solidly grounded wye systems and resistance grounded 3-wire systems have a lot of advantages over ungrounded systems. On resistance grounded systems the grounding resistor is connected to the neutral of a wye source or a zig-zag set or on the secondary of a grounding transformer.

 

[jghrist]

From my experience with several utilities as a consulting engineer, I would say that Cleveland Public Service is unusual if most distribution transformers are connected phase-to-phase. Most US utilities in my experience use phase-to-ground connected single-phase transformers and grounded wye primary three-phase transformers.