Advantages and Disadvantages of Transformers connections 6

[mnnc]

Hi, I would appreciate it if anyone who can explain what are the advantages and disadvantages of transformer connected delta on the secondary side? connected Wye on the secondary side?

 

[jtkirb]

Depends on what the end load is and what the primary connection is.

Transformers facing a generation plant are often delta for 3rd harmonic and fault reasons.

Transformers facing a industrial plant are often grounded wye to use the neutral on single phase loads.

Transformers that are delta for an industrial plant are always a pain due to the ground detection schemes required. I'd stay away from those if possible. It really doesn't provide much advantage.

 

[RalphChristie]

Check

http://www.beckwithelectric.com/powerlines/powerlines-37.html

for some pros and cons for the different connections. (Interconnection Transformer Winding Arrangement Implications on IPP Protection)

[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[waross]

The one connection to avoid is a Star Delta. It is plagued with problems. Phase loss on the primary causes backfeeds, circulating curents, overloads, burnouts, and/or blown fuses.
In some situations, residential customers on the same circuit will experience refrigerator and freezer burnouts.
Star-Open Delta is a good connection when appropriate, but avoid Star Closed-Delta (Three Transformers)at all cost.
If any one wants, I can explain the various scenarios complete with calculations for the various failure modes of a Star Delta connection.Star/

 

[ScottyUK]

avoid Star Closed-Delta (Three Transformers)at all cost

A somewhat brazen statement: do you mean as final distribution transformers, or in all instances? Virtually every utility-level generator step-up transformer is dYN configuration, whether it be a single unit or a three-phase bank. Surely the power plant desingers haven't been wrong all these years...?


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

Your body might be a temple. Mine is an amusement park...

 

[waross]

Hi ScottyyUK
I primarily meant distribution. But.
The small utility that I advise, (2 x 1.5 MW diesel sets)
uses a Delta Wye step-up transformer bank. With this small system, it is impossible to keep the phase currents balanced. As a result the phase voltages are unequal.
If we were to use a Wye Delta bank for step-up, the voltage unbalances would cause large circulating currents in the Delta. Also, we would not then have a neutral available for our distribution system. Double the number of fused cutouts and lightning arrestors would be required.
If I was compelled to design a Delta Wye installation for a power plant, I would be aware that I must be able to assure equal voltages to avoid circulating currents. Something like three single phase voltage regulators or three single phase tap changers.

 

[mnnc]

hi waross

Can you explain the various scenarios with calculations for the various failure modes of a Star Delta connection.Star

Thanks.

 

[waross]

Hi cmarnix;
Wye/Delta Transformer problems.

This will be in several parts, which may be in the following order;
1> The % impedance voltage of a transformer and how it affects the currents in paralleled transformers when the voltages are not equal.
2> Circulating currents in Wye/Delta Transformer Banks
3> Different problems arising from Wye/Delta connections, and causes of unbalanced voltages.
4> Coments on Delta/Wye connections.
5> Systems that work, and why they work.
6> Some history.
Transformer connections; Open/Delta, Broken/Delta.

1> The % impedance voltage of a transformer and how it affects the currents in paralleled transformers when the voltages are not equal.

Definitions for this discussion.

Primary; The transformer winding that receives power in.
Secondary, The transformer winding that delivers power out.
Note; If a distribution transformer is used in reverse as a step-up transformer, then I consider the low voltage winding as the primary and the high voltage winding as the secondary.
Note; Wye/Delta = Wye connected primary, Delta connected secondary.
Delta/Wye = Delta connected primary, Wye connected secondary.

Example
We will start with an infinite transformer. That is, a transformer large enough and efficient enough that we can neglect it’s losses and voltage drops.
We will parallel this with a transformer of the same voltage ratings.
We will change the open circuit terminal voltages of one or both transformers with the voltage adjusting taps and see what happens.
Ratings; One transformer is infinite, so the kva of the other is unimportant for now. For now we will use % Impedence voltage and % of full load current.
Taps; We will specify taps at +2.5%, +5%, -2.5%, and -5%
% impedance voltage. The last few transformers I checked had a % impedance voltage of 2.7% . I have seen ratings as high as 7%.
For this example we will spec the % Impedance voltage at 2.5%. We may later do some checks at other % impedance voltages.

As you know, % Impedance voltage is an expression or measurement of the voltage that must be applied to the primary of a transformer to cause 100% rated current to flow in the secondary. Also, the % impedance voltage is the voltage drop of a transformer from no load to full load.. It is also used in some short circuit calculations to determine the available fault current. (Symetrical) Full load current / % impedance voltage = available symmetrical fault current. We have a code requirement that service equipment be rated for the available fault current. This was the method accepted by the inspection branch.
Example. 50 KVA, 240 Volt = 8333 amps. Available fault current. Service equipment would have to be rated in excess of this value.
Another handy feature of % impedance voltage is that unequally sized transformers with equal % impedance voltages may be paralleled and will share the load in proportion to their KVA ratings. Someday ask me about the calculations for paralleling transformers with unequal % impedance voltages, but first let’s finish Star/Delta.
Another interesting point is that if a transformer is re-rated, as for instance by adding cooling fans, the % impedance voltage changes. The explanation, If the KVA rating is changed then the full load current rating is changed. If the rated current is increased, then it will take a greater percentage of rated line voltage to cause the increased, rated current to flow in a short circuit.

Now we will connect the transformers in parallel. The voltages are equal, so there will be no circulating current. Now we will set the tap changer to -2.5%.
With a difference of 2.5% voltage, full load current will circulate in the transformers. No we will change the tap to +2.5% Still full load current circulating, but in the opposite direction. Now let’s go to 5% tap setting. Now we have 200% of full load current circulating between the transformers. The primary fuses of distribution transformers are typically 250% to 300% or higher to allow for inrush currents. The fuses cannot be expected to clear a distribution transformer at 200% load.
Please note, this is an example and I am using type K fuses, not the more expensive dual element or time delay fuses. Also, the government power utility in this country uses only type K fuses. (Plain and simple, the cheapest available).

If we use a transformer with 5% impedance voltage, the current becomes 50% at the 2.5% tap and 100% at the 5% tap.
A transformer with 7% or 7.5% impedance voltage will be even more forgiving.
However, the lower the % impedance voltage, the less voltage drop under load and the higher efficiency. Also the point of this is to avoid Star/Delta connections, not how to live with them,.

Now lets put two equal transformers in parallel. We are back to 2.5% impedance voltage. Now the circulating current must flow through both transformers so the % impedance voltages of both transformers must be added. It’s easier for me to visualize the currents first. If we have 100% current circulating, then the voltage drop or rise of the transformers must equal the impedance voltages. So one transformer has a drop of 2.5% and the other has a rise of 2.5%. Total percent impedance voltage 2.5% + 2.5% = 5%
Taps +2.5%, -2.5% for a difference of 5%. (If you think that +2.55 and –2.5% = 0 then just use one 5% tap and leave the other transformer on the 100% tap)
A voltage difference of 5% now causes full load current to flow.
Now it is not usual to have two transformers on the same phase with unequal voltages. Without the taps, we would have to do this example with some sort of variable voltage feeding one of the transformers. In the real world this may have an application as a warning to check the voltages and taps when paralleling transformers, but how often are distribution transformers paralleled any way.
However, and notwithstanding, when we move from single phase to three phase, several things change. The most important is that we now have (On a Star system) three independent but related sources of power that are nominally of equal voltage, but in fact are sometimes not equal.
In the next post, we’ll look at circulating currents in a Star/Delta Bank.
When we get to the systems that work I will comment on
craft (Electrical) 10 replies 8 Jan 06 (6 Jan 06) corner grounded Delta system

 

[waross]

Hi cmarnix;
Circulating currents.
Lets look at an open delta transformer bank with two equal transformers. We can consider the two transformers (“A” phase and “B” phase) to be equivalent to one transformer on “C” phase.
Let’s call that our phantom transformer. Now, lets put a real transformer in parallel with the phantom transformer on “C” phase. Please note, the current supplied by the phantom transformer is actually supplied by “A” phase Transformer and “B” phase transformer. With no load, no voltage will flow, but, suppose the voltage of one primary phase changes. Now we have a current that must circulate through the other two transformers, “A” phase and “B” phase.
If there is 100% full load current circulating, through three unloaded transformers, 1/3 of the voltage drop or rise will be across each transformer. 3 times 2.5% = 7.5%
Some of the gurus who know more math than me (and there are a lot of them on this forum) may not agree with my simple calculation.
But I will take a chance and state that a 7.5% voltage drop on one phase of a Star/Delta bank will cause a circulating current of 100% of full load current in three equal transformers with impedance voltages of 2.5%

Suppose that the transformers are loaded to 100%.
Now any unequal primary voltages will cause an increased current in the transformer(s) with the higher voltage and a reduction of current in the transformer(s) with the lower voltage.

Earlier this year I checked a Star/Delta transformer bank that had been in operation for less than a month.
This was a new installation on an old job. Two transformers hit by lightning. Replaced with three transformers and the connections changed from Star/Star to Star/Delta.
After they were through, the Automatic Transfer Switch wouldn’t work. That’s when they called me.
The engineer had reversed the phases on the new transformer connections and the Automatic Transfer Switch had phase reversal protection. The owners people said:
“No! That’s not possible. The engineer checked the connections with a phase rotation meter.”
I used one of the reverse phase protection relays out of the Automatic Transfer Switch, and proved conclusively that the incoming phases were indeed reversed.
System; Large Utility.
Transformer Specs:
34500 Volts Primary.
50 KVA
240/480 Volt Secondaries.
One transformer was warm. One transformer was Hot. One transformer was very hot.

The load was reasonably balanced.
Don’t worry about the actual voltage. The significance is the Percentage difference, and the fact that the voltages were changing as I moved from one part of the site to another.
If the primary voltage had been stable, reading #1 and reading #2 should have been equal.
I took reading #3. I then de-energised and changed the taps on a dry type transformer. When I re-energized, I took readings #4. In that time the voltage differences had changed. The load on the dry type transformer was negligible.
In different parts of the system I recorded phase voltages of

1> 435 Volts, 438 Volts, and 445 Volts 2.24 % Difference.
2> 437 Volts, 440 Volts, and 440 Volts 0.68% Difference.
3> 187 Volts, 190 Volts, and 195 Volts 4.10% Difference.
4> 203 Volts, 205 Volts, and 211 Volts 3.79% Difference..
The current differences were not enough to account for the voltage changes.
I had to assume the primary voltages were changing.
This assumption is supported by the different transformer tank temperatures.
The job site was 10 or 15 miles from the point of supply of the 34500 Volt circuit.
The34500 Volt loads onb the primary circuit are mostly single phase residential and agricultural.
The line to the job site taps off the main line and runs about a mile. There are a couple of single phase loads on the line to the project. One phase continues past the job site to a single phase load.
There are a few transitions on the main line but possibly not enough. I don’t know how many transitions there are and I don’t know how many there should be. The line continues another 10 miles or so to a large industrial load.
Remember, I was there for only a short time, and I probably did not see the maximum voltage differences.

ONE TRANSFORMER FAILED WHILE I WAS STILL ON SITE..

Two or three local experts swore that Star Delta was the only connection that could be used, and I was dismissed by the owner.

The failed transformer was replaced and the Star/Delta connection was retainned.

ABOUT A MONTH LATER A SECOND TRANSFORMER FAILED..

They are now running open Delta. When summer comes and all the air conditioning kicks in, there may another burnout from overload. The original transformer bank was sized fairly close to the load.

Another anecdote; A month or so later, I was visiting the little Utility that I advise. A visit involves a plane ride and hotels. They call me when they need me. I was walking with the manager and spotted a new three phase installation connected Star/Delta. I reminded him that we had agreed some years ago to ban Star/Delta connections from the system. He explained that the installation was owned and planned by a government agency, and we were stuck with it. I predicted transformer failure. The manager told me that they had already had a transformer failure. The first one lasted about a month. Shortly after the transformer was changed, the utility lost one of the transformers feeding that area. There are no three phase transformers in the system. They jumped two overhead lines together to maintain residential service until they could arrange repairs. The bank in question ran open Delta with no problems. When I saw the bank, there were two 50 KVA transformers, and a 37.5 KVA transformer..
I believe that originally, before the burn-out, there were three 50 KVA transformers.
The load on the bank was about 20 or 25 KVA single phase on “A” phase, and one three phase battery charger that draws a whopping 17 KVA.
The Utility is served by a 1500 KVA diesel generator.
The lessons here;
1 > Unequal supply voltages can cause overloading and burnout of Star/Delta transformers.
2> To estimate the circulating current, in three equal transformers , that is voltages equal, KVA equal, impedance voltages equal, and on the same taps,
One third of the Worst Voltage Difference divided by the Impedance Voltage times 100.

Voltage Difference = 9 Volts
Impedance Voltage = 2.5%
9/3 = 3
3 / 2.5 = 1.2
1.2 x 100 = 120% of full load current.


If your still with me, the next post will be;
Implications of one phase missing.
The implications of two phases missing.
Back feeds.
The human factor.
In my experience, voltage unbalance has not been the primary cause of Star/Delta Burnouts.
If I've inadvertanly used voltage, when i should have said current, please help me out of my confussion.

Dyslectics of the World UNTIE!!

 

[wfowfo]

waross,
I'm fascinated my your last two posts.
Confused, but fascinated!

I think I'm following your reasoning, but I'd like to ask one question. In the star/delta configuration (of which we have many) how much (if any) of what you just explained is dependent on whether or not the common point of the star is grounded?

My understanding, which is limited at best, was that this configuration was fine as long as you floated the common point of the star primary instead of hard grounding it.

 

[waross]

Hi cmarnix;
Hi wfowfo


Star/Delta Transformer Failures.

While we have shown that unbalanced primary voltages can cause circulating currents of sufficient magnitude to cause burn-outs in a Star/Delta transformer, unequal primary voltages are not the only cause of burn-out.
In some areas, the primary cause of burn-out is phase failure.
Consider what happens when a Star/Delta bank loses a primary phase.
For example let’s imagine a primary circuit that extends for a mile. The load is about 1200 KVA or 400 KVA per phase.
Somewhere along the line is a Star/Delta Transformer bank with three 50 KVA transformers.
A fuse blows on “A” phase.
Now that no voltage is being supplied to “A” phase, It’s secondary will be fed from “B” phase and “C” phase which are now an Open Delta circuit. “A” phase transformer will now operate as a step-up transformer and produce rated primary voltage at the primary terminals. This will feed back into the “A” phase line. Subtracting the 50 KVA transformer in our Transformer bank, that leaves 350 KVA connected. Our 50 KVA transformer will bravely try to supply this 350 KVA load. I think that’s a 700% over-load.
You may expect the "A" phase primary fuse to blow. It may not, for one of two reasons. The human factor, and what’s happening to the other two transformers.
“B” phase and “C” phase transformers are still energized, and are still supplying their loads. By virtue of the open delta connection, (Two transformers energized). The “A” phase transformer has now become a load supplied by an Open Delta transformer bank.) “B” phase and “C” phase transformers are sharing the load normally supplied by the “A” phase transformer.
The load on the energized transformers is now;
a> The normal load.
b> A share of the “A” phase load.
c> The same circulating current that is flowing in the “A” phase transformer to supply the back feed.
That gives us the sum of;
100% of full load current,
Plus current for the load normally carried by the “A” phase transformer. (I think that it is 86% of full load current, but it’s been a hard week. (If any one wants to jump in here with the exact figure, great. Other wise I’ll fill in the proper number over the week end after I get a little rest.)
Plus The 700% current to supply the back feed.
I make that almost 900% of full load current.
“B” phase or “C” phase fuse may blow first, but we still have the “Human factor”.

The current won’t actually be that high. The heavy current will cause a voltage drop at the secondary terminals. Let’s say 800% times 2.5% impedance voltage, that’s about a 20% voltage drop. We can expect a 20% lower voltage at the terminals of the “A” phase transformer. Let’s say it’s actual load is 600%. That’s a voltage drop on the primary of 600% times 2.5% impedance voltage, or a further 15% voltage drop. That gives us a back feed voltage of about 68%. The lower voltage will probably cause some load shedding that will reduce the actual backfeed voltage and current.

Now suppose that “B” phase fuse blows at our Star/delta bank. One of the fuses was going to blow, we just didn’t know which one.
Now “A” phase is out at the supply end. “B” phase is out at the transformer bank.
Only “C” phase is now energized. “A” phase and “B” phase transformers are in series across the “C” phase transformer. “A” and “B” phase transformers now become a voltage divider. “A” phase is is still trying to back feed the circuit, but it is now looking at it’s normal load in addition to the back feed load. That gives us about 400 KVA load on “A” phase transformer. “B” phase transformer has about 50 KVA. The simple voltage division will about (400/450 = .89%) of normal voltage at the terminals of “B” phase transformer, and (50/450 = 11%) of normal voltage at the terminals of “A” phase.


This is now single phase.
“A” phase in our panel will have about 11% of normal voltage.
“B” phase in our panel will have about 89% of normal voltage.
“C” phase in our panel will have about 100% of normal voltage.
This sequence of events gives rise to the real possibility of property damage both in our plant and for the loads fed by the back feed on “A” phase.

We still have to consider the human factor.
Later

 

[waross]

Hi wfowfo
Grounding the Star point is not a factor.
A neutral connection to the Star point is an issue. The distribution systems with which I am familiar are all 4 wire systems. Three over head phase wires and a lower mounted neutral wire. The neutral has to be connected to the star point and extend back to star point of the supply transformers.
Floating from ground, not an issue. A fully floating star point, that raises issues about harmonic currents with no way home, and voltage division with unequal loads. I think that if you lost the load on the secondary of one phase of a Star/Delta bank, with a floating star point that you may straighten your phase angles out like a Bow String.
I have never seen a Wye/Delta bank with a fully floating Star point.


I wonder if your

 

[waross]

Hi Cmarmix & wfowfo & others;
A couple of clarifications;
In a previous post, I referred to a 34500 Volt primary.
This is phase to phase voltage. The actual transformer voltage is 19920 Volts.
The 11% Voltage on “A” phase is also the voltage now being back-fed into the system..
I use the term “Amps-per-Volt” in place of the more cumbersome term, “The Reciprocal of the Impedance at a Given Voltage Level.”

The load on “A” phase will change as the voltage Drops.
If the load on this circuit is residential and small commercial. The load will be mainly Incandescent Lighting, Resistance Heating, Small Motors, and Florescent Lighting.
Incandescent Lighting
The amps per volt of Incandescent Lamps may increase as much as 10 to 1 as the voltage drops.
That is, the cold resistance of lamp filament may be 1/10 the hot resistance.
The amps per volt of heating loads may be fairly linear. The change in resistance with heat will be less with a water tank heating element than it will be with a red hot element on a cook stove.
I would expect the current draw of any single phase motors to increase with dropping voltage until the motor stalls. From the stall point I would expect the amps-per-volt to be constant as the voltage falls further.
The point is, estimate the back-feed current. To determine the connected load, and the current draw of the connected load at reduced voltage is impractical.
Either make a “Well Educated Guess as to the Current” or “Estimate the Current”.
My understanding is that an “Estimate” has a few more significant figures than a “Guess.”.

The Human Factor.
In this area, the fuse most likely to blow first is indeterminate. The Fuse LEAST LIKELY TO BLOW is the last fuse changed. The line crews know that Star/Delta transformer banks tend to blow fuses, and tend to over fuse. Whatever size fuse has failed, replace it with a bigger value.
Anecdote,
A three phase Star/Delta bank on 13800/7967 Volts.
100 KVA Transformers.
Calculated Full load current; 12.55
Expected fuses; 25 Amp to 35 Amp
Actual fuses; 100 Amp

Equipment failure. A complaint that I often heard on the Island with the small utility was that,
‘The power company is always burning out refrigerators and freezers.”
Note, I know that a system this small is not a valid model for a larger system. However, there may be many circuits of this capacity that are part of a larger system.
At the time of the Star/Delta problems, the main transformer bank was three 333KVA transformers. 480/7967 volts. Delta Star.
There was one large industrial customer with a Star/Delta transformer bank. The bank was three 100 KVA transformers. The significance of the 300 KVA Transformer bank is not it’s size in relation to the supply, but it’s relation to the magnitude of the other loads. At that time, the load WITHOUT the industriasl plant would not often exceed 600KW. The point at which a second generator had to go on-line.
In case of a power outage, all the three phase motors in the industrial plant tripped off. The motors all had to be restarted manually. Also, the industrial plant would phone the power station before starting their motors to verify that capacity was available. They often had to wait until another set was started and paralleled. The plant never had problems.
The only available switch in the station was on the primary of the transformers and would put the whole load on the system at once. It was used to go off line, but the generators could not withstand the block loading of 100% load at one step. The lines were switched on manually, one phase at a time with a “Hot Stick”. The 100 Amp fuse cut-outs were used for switching.
It was not uncommon for two plant operators to be outside looking up into the rain. One man with a Hot Stick and the other with a hand light.
They would close “A” phase.
This would energize one third of the residential customers. It would also energize “A” phase of the industrial transformer bank.
“B” phase transformer and “C” phase transformer would then be in series across a phase.
They would back feed about one half voltage into “B” phase and “C” phase of the system.
If the power had been off for any length of time, all the freezers and refrigerators on “B” phase and “C” phase would try to start. With only half voltage, the compressors would accelerate slowly, and many would stall from the increasing back pressure before getting up to speed. Sure a refrigerator has a thermal cutout to save it in the event of a stalled rotor. I would suspect that many thermal cutouts in the developed world are only called on to operate one or two dozen times in the life of the refrigerator.
The plant once ran almost a year without an outage, but that was only once, a long time ago. A more realistic figure would be a couple of outages a month.
The fact is there were enough failures that it was a matter of public concern, and the burnouts were invariably associated with power outages.
I remember when my hotel room was on the third phase to be energized. The light bulb would glow dimly for awhile. Then it would get bright. Then as the last phase came on, it would get just a little brighter.
Remember also, we are in the third world and many of the refrigerators here are built in the third world to the finest third world standards. The thermal overload devices may not be as dependable as hoped for.
Interesting but what is the application to a “Normal” system.
Well, suppose a tree branch or a palm frond comes down on the lines and blows out two fuses. If we have a large Star/Delta transformer bank on the part of the system that has lost two fuses, we can expect about half voltage or less to back feed on the two lines with blown fuses.
Regardless of the load distribution, the voltage will not exceed rated voltage. For example, suppose one transformer burns out and short circuits completely. Now we will have one transformer in series with a shorted transformer that we can consider to have zero ohms impedance.
The voltage to the transformer,and the resulting back feed voltage will equal the supply voltage minus voltage drops. If the shorted transformer has some appreciable impedance, this will serve to further lower the back feed voltage.

Still to come, why some systems work, and some of the factors to consider.

 

[itsmoked]

waross; fascinating...

Too bad they can't cut their Island system into a couple of blocks. So they could dump all three phases on at once.

Do please continue.

 

[waross]

itsmoked, Thanks for the feed back..
I agree with you 100%
Actually we were working towards that, but the system is small, and the budget is smaller. The load has increased, and the old 600's and 350's have been replaced by a 1500 KVA Generator.
The 1000 KVA transformer bank has had a second 1000 KVA transformer bank added. We now have two blocks and can physically "dump" either one of them on the line.
In the real world, the operating procedure is now to "Dump" the smallest block on the line, and then go out in the rain with the "Hot Stick" to connect the remaining load phase by phase. Our original intention was to do exactly as you suggested, but the load grew too big. Geography precludes adding more blocks. It took a couple of years of prodding, to get permission, but the Star/Delta bank was reconnected Star/Star years ago. It's been years since anyone has stopped me on the street to complain about a refrigerator burnout.
More later

 

[waross]

Hi folks; I’m going to try a different format to summarize the effects we have been discussing.
I hope you enjoy it and find it informative.

Consider a distribution circuit protected by fuses.
I will call these the system fuses.
On this circuit there is a Star/Delta transformer bank.
I will call the primary fuses the transformer fuses.
I will call the load on the transformer bank the panel load.
These comments are valid whether the load is three wire or four wire Delta. Grounded or ungrounded.
What happens when the various fuses blow.

“A” PHASE TRANSFORMER FUSE BLOWS .
RESULTS;
The effect may well be transparent to the users. The transformer bank will become an Open Delta bank and continue to supply three phase power to the panel.
The capacity of the transformer bank drops to 58% of the original value.
If the load is over 58% of rated full load, then transformers “B” and “C” will over heat and the life expectancy of the transformer will be shortened.
If the load on the transformer bank is less than 58%, the condition will probably go unnoticed. The only indication may be that one of the fused cutouts is hanging down.
Safety consideration. The cutout or fuse holder will have full primary voltage on the line side and almost full line voltage on the load side of the primary fuse holder, or cutout.
The voltages on both side of the fuse holder will be in phase and there will only be a small voltage difference across the fuse holder.

“A” PHASE AND “B” PHASE TRANSFORMER FUSES BOTH BLOW.
RESULTS
“C” phase in the panel will have full voltage.
“A” phase panel load will be in series with “B” phase panel load. The voltages on the “A” phase bus and the “B” phase bus in the panel will be in proportion to their respective load impedances and the sum of the “A” phase panel voltage and the “B” phase panel voltage will be equal to the “C” phase panel voltage. The magnetizing current of the "A" phase and "B" phase transformers will be part of the panel loads on the respective phases.
The “A” phase transformer and the “B” phase transformer will both back feed a voltage to the load side of the fuse holders. The voltages will be in the same proportion as their respective secondary voltages. The sum of the back feed voltages will be nearly equal to the normal primary voltage.
All the voltages, both panel voltages and the back feed voltages on the transformer primaries will be in phase with “C” phase.

“A” PHASE SYSTEM FUSE BLOWS.
RESULTS
The “A” phase transformer will back feed “A” phase of the circuit back to the blown fuse.
FIRST CASE
If the Star/Delta transformer bank is large enough in relation to the balance of the load on the circuit, it may successfully supply the circuit with no ill effects. Single phase users on “B” and “C: phases may notice a slight drop in their voltages, manifested by a very slight dimming of their incandescent lamps.
The dimming on “A” phase will be a little more, but still very slight.
After the initial small voltage drop, the only evidence of trouble may be the “A” phase system fuse holder hanging down.
SECOND CASE.
The magnitude of the connected load is such that the Star/Delta transformer bank is over loaded, but not enough to blow the fuses. The voltage drops will be more noticeable. The voltage drop on “A” phase may be about twice that on “B” and “C” phases.
The lamps on “A” phase may or may not dim enough to be noticeably dim after the initial drop. This will not be a voltage dip, but the steady state voltage will drop a few volts and then be stable at the slightly lower level.
The “A” phase system fuse holder will be hanging down.
Eventually one of the transformers in the Star/Delta bank may fail. Simple logic would suggest that because “B” and “C” phase are the heaviest loaded, one of them will be the first to fail. However, More experienced logic will realize that it is quite probable that the transformers have suffered life shortening overloads in the past, and in the real world, the transformer that fails may be the one with the most load, or the one with the most severe prior damage.
The voltage may be low enough to cause problems for refrigerators trying to start. Those refrigerators already running will have fewer problems.
THIRD CASE
The Star/Delta transformer bank is small in relation to the connected load on “A” phase.
There will be a very noticeable voltage drop on “A” phase. One of the fuses in the transformer bank will blow quite quickly. Because of the relatively fast clearing, there may not be any serious overheating of the transformers.

FURTHER DEVELOPMENTS OF THE THIRD CASE.
FOURTH CASE.
The “A” phase system fuse has blown.
Then “A” phase transformer fuse blows..
RESULTS
System implications. “A” phase is de-energized. “B” phase and “C” phase are energized and normal.
Plant implications. The same as “A” transformer fuse blows. The phase failure will be almost transparent to users in the plant.
The incandescent lamps will increase in brilliance slightly when the “A” phase transformer fuse blows and the back feed ceases.

FIFTH CASE
THE “A” PHASE SYSTEM FUSE AND THE “B” PHASE TRANSFORMER FUSE HAVE BLOWN..
Similar to “A” PHASE AND “B” PHASE TRANSFORMER FUSES BOTH BLOW
BUT “A” phase circuit load will now be in parallel with the “A” phase panel load.
Note; The “A” phase transformer fuse hasn’t blown in this instance, the “A” phase system fuse has blown.
The “A” phase panel voltage and the “A” phase system voltage due to back feed will both be close to zero volts. The “B” phase panel voltage will be close to the “C” phase voltage. “A” phase panel voltage and “B” phase panel voltage will be in phase with “C” phase panel voltage and the sum of “A” phase and “B” phase voltage will equal “C” phase.

“A” PHASE AND “B” PHASE SYSTEM FUSES BLOW.
There will be a back feed of voltage on both “A” phase and “B” phase.
The respective voltages will be proportional to the impedances of the loads on “A” phase and “B” phase. The sum of the voltages will equal line voltage minus the voltage drops due to the transformer impedances. The panel loads on “A” phase and “B” phase are part of the load on the system, but the sum of their voltages will be equal “C” phase panel voltage.
This is the failure mode that is most responsible for damage to refrigerator motors.
The transformer currents at half voltage are quite a bit less that at full voltage, but there may be exceptions. Particularly if single phase motors are a large part of the load.
Still relatively large transformers may last indefinitely.
Small transformers will blow the fuses.
There will be a size of transformer that can be expected to overheat and eventually fail.
Can I pose a question?
We have a four wire delta system. The single phase loads on 120/240 volts are connected to the "A" phase transformer The Fuse blows on "A" phase. The Transformer bank is now Open Delta and the transformer feeding the single phase loads has no primary supply voltage.
What is the effect on the plant users?
Comments and suggestions welcome.

To come;
What our local utility does about Star/Delta problems.
Why some systems work.
Open Delta and a bit of history..



Note; Why do I use the word Blown rather than failed to describe an open fuse?
Well you may have heard the story of the Triangle player. For 6 months he didn’t miss a single rehearsal. He was, of course, there for the big concert. He waited patiently through the 2 ½ hour program. Then, in the final crescendo of the final number, he played the one note that he had prepared himself for.
Likewise, a fuse is designed for one purpose, installed for one purpose, protected, nurtured, and finally, the event he has waited his entire life for; A current exceeding his design parameters. He has been prepared. He has one chance. Can he do it? Yes, the fuse has interrupted the excessive current. The event that he has waited his whole life for. He has fulfilled his life’s purpose. And now he’s gone.
I don’t call that a failure, I CALL THAT A RESOUNDING SUCCESS!!
That and “The fuse blew” is so much easier to type than “There was a successful execution of the intended purpose of a fuse.”

 

[Skogsgurra]

I am so sorry that I didn't have a look at this thread before. It seemed to be another one of those DYn=good/YYn=bad threads. But this is so much more!

I now have a feeling that it is all too late. The Good and Bad Practice seems to have developed in some mysterious ways that I cannot follow. But it makes some mind-boggling reading. Anyone get anything useful out of it?

Or is it solely about how to load a small islanding generator? Is that what the OP asked about?

Gunnar Englund

http://www.gke.org

 

[waross]

Thanks for the feed back, skogsgurra.
I plan to make some comments on my view of history of The Star/Delta connection. Should help with your comment about the development of Good and Bad practice.
This is not just about a Small Island system. I have a little control over the Island system. The same problems are present on the large utility system on the mainland, where I am just an interested and sometimes affected observer.

 

[cuky2000]

Below is a summary of most common transformer connection used in the power industry

 

[waross]

A little history as I know it.
I understand that in the early part of the 20th century, 2300 volt Delta was a popular distribution voltage in north America. In the economic, industrial and consumer expansion during the late 40's and early 50's electric system expansion increased greatly. Utilities started converting their distribution systems from 2300 Delta to 4160 Star. I understand that up to that time low voltage Delta was almost universal in industry.

Back then, residential loads were very light compared to today. Many areas heated water and cooked the dinner with wood or gas. A typical residential load would be the light bulbs and a radio. The kitchen full of high consumption appliances was yet to come.
A large industrial plant was usually the largest load or the only load on a circuit. In the event of a voltage unbalance, the Delta bank would be a large enough component of the load that it could act to balance the voltages without overloading. Remember, the old 4160 volt line probable had a relatively high voltage drop on the line. (Compared with the percentage voltage drop on a 34500 volt line today.)
The older transformers had higher impedance voltages, and the single phase loads were a much smaller component of the overall load.
The higher line losses and the lighter single phase loads tended to reduce the magnitude of the unbalanced voltages on a circuit.
The higher line currents at the relatively low 4160 volts and the higher impedance voltages of the older transformers would tend to lower the circulating currents that result from unequal primary voltages.

As the old 4160 Volt circuits were updated over the years to 12,000 Volts 13,200 Volts, 13,800 volts, 25,000 volts, or 34,500 Volts and the residential demands increased dramatically, the possibility of potentially damaging voltage unbalances increased.
But the old systems still work.
Some of the factors that will help to avoid problems are;
1> Three phase protection on feeders, laterals and transformer banks rather than single phase fuses. Where I live, it's almost all fuses.
2> Reclosers will help to avoid many phase loss conditions.
3> Time Delay or Dual Element Fuses will help to avoid overload damage in the event of unequal voltages on the primary.
You will have "Outages rather than transformer failure, and if you have chronic unbalanced primary voltages you're out of luck.
Dual element fuse technology has not reached this part of the world yet. The only primary fuse links availlable in the country are the plane old type "K"s that have to be sized at 250% to 300% of full load.
4> Protective relaying to detect excessive currents in the Star/Delta bank and trip the transformer bank off line.
5> Transformer impedance. The lower the impedance voltage of a transformer, the greater will be the current with a given primary voltage unbalance. The older transformers typically had higher impedance voltages than the newer transformers.
6> Throw the fuse away. WHAT?? Yes I did say it. Throw a fuse away.
The line crews have a new procedure for blown fuses on Star/Delta transformer banks. Instead of over sizing the fuse as they used to, they now just take away one of the fuse holders and leave the bank on open Delta.
The burnout rate is down and back feed damage is a thing of the past.
I can stand in one spot downtown and count four Star/Delta transformer banks with one Fuse holder missing.
7> Change your connections to Star/Star.
If you have been having overloading problems with a Star/Delta condition be very carefull removing the transformer top to change the connections. The insulation may be damaged and brittle from overheating. Even a gentle tug on the primary leads when you push the tank cover back may be enough to fracture the brittle insulation and lead to very early transformer failure.

The power in this area was upgraded from 4160 Volts to 34500 Volts 6 or 7 years ago. When I first came here in the early 90's a common topic was "Voltage Surges and the damage they did". Almost every commercial establishment that I went into had some form of phase failure relay on the three phase equipment.
I don't hear about "Voltage Surges" anymore. I think that the voltage surges were more likely low voltage back feeds rather than actual "Surges" The surges and burnouts were always associated with power outages.
I think that with the change from 4160 Volts to 34500 Volts, we have a new set of impedances and currents and have moved from back feeds to transformer burnouts.

I don't see any takers on the question about the transformer bank with a fuse missing on the transformer that feeds the single phase loads.
The answer is absolutely nothing. The secondary of the transformer that is missing a fuse is connected across the other two transformers that are in Open Delta. Because the transformer has full voltage applied to it, it acts as an auto-transformer. The other two transformers supply the 240 volt loads, and the transformer that normally feeds the single phase 120/240 volt loads acts as an autotransformer to develop the neutral. The users don't know the difference. It would be difficult to detect the difference even with a voltmeter.