Maximum Transformer Impedance 1

[kartracer087]

For a power transformer that is built to IEC standards say 132kV to 11kV and 40 or 50MVA top rating what is the maximum percent impedance? Is it possible to build a transformer that has roughly 30-35% impedance at the top rating? I know in US I’ve heard figures as high as 20% are possible at the self cooled rating which is usually 3/5 of the force cooled rating. So hypothetically a 30/50 MVA unit could have as high as 33.3% at the ONAF rating of 50MVA. Is this consistent with IEC maximums?

 

[davidbeach]

Oh, but why?

When one this sentence into the German to translate wanted, would one the fact exploit, that the word order and the punctuation already with the German conventions agree.

-- Douglas Hofstadter, Jan 1982

 

[kartracer087]

Limits on switchgear mainly. 25kA limits on field switchgear mean higher impedances may be needed.

 

[prc]

Any impedance is possible. But then you must consider the higher terminal voltage drop(regulation) with loading. Many times, this will not be acceptable. In IEC 60076-5, minimum impedances are specified in Table 1. 25-40 MVA - 10 %; 63-100 MVA -12.5 % and above 100 MVA > 12.5 % etc.
There is an optimum percentage impedance for every MVA rating of the transformer. Any increase or decrease from this range will increase costs to uneconomic levels. At 50 MVA level, this level is 10-12.5 %

 

[7anoter4]

The minimum short-circuit voltage is stated in IEC 60076-5 and it depends on transformer rated power.
According to Table 1 for 40 to 63 MVA it is 11%.
Maximum short-circuit voltage depends on the difference between maximum permissible voltage and minimum permissible.
So, if rated voltage is 11 kV, according to IEC 60909-0, the maximum voltage factor [c factor Table 1] it is 1.1.
The minimum voltage is stated also in IEC 60909-0 and it is 1.
Then the Vmax=1.1*11=12.1 kV and Vmin=11 kV.
The minimum short-circuit impedance -seeing from secondary winding Zsc=11^2/50*11%=0.2662 ohm
Ir[transformer secondary rated current]=MVA/sqrt(3)/kV=50/sqrt(3)/11=2.6243 kA
Vmin=Vmax-sqrt(3)*Ir*Zsc=12.1-sqrt(3)*2.6243*0.2662=10.89 kV [c=0.99].
That means you have to go to less ukr% then 10%
However, they are national standards allowing only 0.95 of rated as minimum.
In this case ukr=15% it could be.

 

[waross]

what is the maximum percent impedance?

The OP is looking for the maximum, not the minimum possible impedance.

25kA limits on field switchgear mean higher impedances may be needed

Often line impedance will limit the ASCC to within the rating of field switchgear.
Available Short Circuit Current, Switchgear is rated in ASCC, not actual possible fault current.
If some close-in switchgear is still looking at too high an ASCC the ASCC may be reduced for that switchgear by the application of air core reactors.
This will leave your system with the benefit of the better regulation of lower impedance transformers.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[7anoter4]

In order to limit the short-circuit current you may increase the transformer impedance but also to limit the transformer power -in order to keep the minimum voltage at full load. For instance, if you reduce the I”k3 to a half you have to increase the ukr to 20% [instead of 10%].In this case the maximum load will be 25 MVA in order to keep the 11 kV.
If the transformer load is constant then you may supply a reduce minimum voltage and to reduce the rated voltage of the supplied system. You have to check if in an empty state the supply voltage is not too high. An on-load tap-changer will be welcome.

 

[kartracer087]

Well I was looking at a utility that has paralleled 3 50MVA transformers on the 11kV bus so of supplying downstream ring main gear close to the station that is rated 25kA the issue is you need to be at 63kA peak or less for the load break switches. By my calcs with 150M of 240 mm 3C cu cable you end up needing transformers at 33% impedance at 50MVA base. Considering a stiff source of 30kA at 132kV side.. That could be right at the upper limit of impedance. The utility in question is in Dubai (DEWA).

Or you can run lower impedance in the 27-29% range and end up needing regular breaker gear with 31.5kA ratings at the distribution substations rather than Ring Main Gear. I’m just curious what the upper limits would be on transformer impedance. Pretty sure they are using LTC transformers and they keep power factors high at the sub close to unity to deal with the higher voltage drop.

In my eyes though a better design would be to operate with split bus and use more standardized impedance values. Paralleling 3 transformers creates a very high short circuit current.

 

[FPelec]

Hi kartracer087,

I fully agree: split bus operation seems to me the most efficient solution. In the absence of other information, connecting 3 x 50 MVA transformers to the same 11 kV busbars, IMHO, is not a good solution. Even if it is possible to design 50 MVA 11 kV transformers with arbitrarily high impedance, voltage regulation will become more of an issue.
If necessary, I would rather suggest using series reactors instead of increasing the transformer impedance. This way, in the future, when changing the 11 kV switchgear or the operating scheme to a split bus configuration, you could simply remove the reactors.
Otherwise, you would still end up with transformers having very high impedance and thus difficulties in voltage regulation.



Si duri puer ingeni videtur,
preconem facias vel architectum.

 

[7anoter4]

First, I have to appologize since in the drop voltage through the transformer I used a wrong formula.
DV=sqrt(3)*Zsc*Ir
Actual formula it is:
DV=sqrt(3)*Ir*(Rtrf*cosfi+Xtrf*sinfi)
A 50 MVA transformer presents η=99.5% efficiency. Let's consider the all power losses as load losses.
Then load losses=(1-0.995)*50=0.25 MW
Rtrf=LossP/3/Ir^2=0.250/3/2.6243^2=0.0121 ohm
If ukr=18% then Zsc=11^2/50*18%=0.4356 ohm
Xtrf=sqrt(Zsc^2-Rtrf^2)= 0.4354 ohm
cosfi=0.85; sinfi= 0.52678
Ir=50/sqrt(3)/11= 2.6243 kA
DV=sqrt(3)*2.6243*(0.0121*0.85+0.4354*0.52578)= 1.0873kV
Vmin=12.1-1.0873=11.0127kV
However, I think the transformers are connected separately with the RMU.
They are 3 transformers: one for each bus of the RMU and the middle transformer is a reserve.
Usually Sw1R, Sw2R and SwRMU are open.
If one of bus bars of RMU is not energized then through SwRMU it can be energized. A single transformer may supply all the consumers, but usually both RMU bus bars are connected each one with one transformer only.
In my opinion, you can supply only 75% of the maximum load in a case of emergency with a single transformer[ or get OFAF instead of ONAN].
So,no need more than 25 kA per one transformer.

 

[prc]

A more straightforward way is to express r and x in %. r= load loss in kW/ rating in kVA x100 X= square root of ( impedance squared - R squared )
Since the load loss of a 50 MVA 132 kV unit- 150 KW; r= 0.3 %

at unity power factor= voltage drop= R % + 1/200 x square of( X %)
at zero lagging PF = x% - 1/200x square of (R %)

The secondary terminal voltage will drop by 0.3 %(approx) at unity PF load and 12.5 % at zero lagging PF load assuming X=Z =12.5 %

 

[kartracer087]

I actually think they are running their transformers in a split bus configuration at the substation with the bus ties normally open. If you look in this article, I think they only parallel the 11kV breakers in switching operations for the low voltage AC supplies which they call LVAC to prevent circulating currents in the 11kV bus per this article:

https://www.slideshare.net/slidesho...-scheme-in-dewa-transmission-system/265349397

In that case, it means the transformers can be at a more standardized level of impedance realistically and the system operation would limit the downstream currents to RMU's under normal operation to levels 25kA and below (and likely even less than that because normally fed out of one substation bus (one 50MVA) unit at a time. I know they do use some older style Oil RMU's on their system as well which have close in levels of 50kA peak.

Thanks,

 

[kartracer087]

And in thinking about this a bit more, with this configuration saying 25% impedance on 50MVA base with 11.5kV secondary, you would be able to run two transformers in parallel on a bus indefinitely and not exceed the peak ratings of the ring main switches and also not exceed the 25kA for 1 second with proper high speed relaying. This is figuring you have about 150 meters of cabling between the substation and the first RMU and about 30kA short circuit current available on the 132kV side.

I looked at this with a C factor of 1.05 as well and still is within limits of the equipment. Transformers with 25% impedance on the full load ONAF cooled base are available. A neighboring utility uses 40MVA 11.5kV units with 25% on 40MVA base.

I think long term loading with (2) units in parallel is possible but could create a degree of asymmetry on the bus loading since one of the busses would be expected to operate with no more than 1/3 of the (n-1) rated load considering the other bus loads being same balance. If the station is a sectionalized straight bus design and designed for a maximum of 100MVA peak load (n-1 condition), ideally each of the 3 bus sections should be carrying no more than 33.33MVA load to account for a loss of transformer. This is without figuring any "beyond nameplate" loading that we would typically do here in the USA in contingency scenarios. Having a higher impedance unit limits your ability to do this and be within the range of a typical +10% / -10% on load tap changer if there are greater than -5% voltage fluctuations on the transmission supply.

Another note - to parallel 3 units with this setup would require station circuit breakers with 40kA (100kA peak) capability since your bus fault levels would be roughly 31.5kA and X/R in the 30's. I do believe 40kA breakers are in use.

Paralleling 3 units would exceed the ring main unit close in ratings but that is acceptable since these switching operations with 3 paralleled would be for a very brief period and otherwise the system would operate with a maximum of 2 transformers paralleled on the bus long time. Many of the ring main units use current limiting fuses with 40kA symmetrical interrupting as well so should a transformer fail in the middle of a station switching operation, the fuse would still be able to interrupt all of the failed distribution transformer's fault current. It would also be possible to put a small relay delay on the instantaneous region of the ring main breakers, should they be equipped with breakers, so that they will not try and interrupt a very high fault current. Instead the station breaker would be responsible for clearing very high fault currents, those above 20kA. In essence, the instantaneous pickup on the ring main units is slower than the feeder breaker for faults detected that are greater than the interrupting rating of the breaker. There would be a faster 50 function (instantaneous override) that would arm on the substation feeder breaker relay when fault currents at the substation exceed the ring main unit capabilities so the higher rated substation breaker will clear the fault and not the ring main breaker. This could be done with bay control logic that senses when all bus ties and all transformer mains are closed and enables the quicker 50 function.

Ring main spec would ideally be use 25kA/1s units (62.5kA peak) load break transformer switches with fuses or breakers with 25kA, 63kA peak, and at least 36% dc component capability. You could run 2 station transformers paralleled on the 11.5kV bus and be within the ratings of this equipment very close to the substation. Running a single transformer per bus (no paralleling) would meet the ratings of basically all modern RMU's with a 20kA spec.

Just a lot of hypotheticals here but just me doing some back of the envelope capability engineering as design reference.

 

[kartracer087]

I actually followed up with some engineers over there and got some additional information, for record:

They use transformers with impedances of around 33.5% impedance on 50MVA base, maximum short circuit level at the station bus is about 25kA. They do parallel all (3) 35/50MVA units together under normal operation. 31.5kA breakers or 40kA breakers can be used. The peak is in fact being limited to around 62.5kA when a couple hundred meters from the substation which allows a few manufacturer's ring main units to be used as long as they have 25kA for 1s and 62.5kA peak ratings. Most of the other neighboring utilities are staying within that 25kA maximum level due to the limitations of the distribution equipment. A bit farther away from the substation when there is some cable other lower rated ring main units can be used.

Followed up with my ABB/Hitachi rep who also said transformer impedances can vary greatly and there aren't hard limits to the design. He mentioned he has some transformers with up to 40% impedance on the ONAF base. He did mention designing transformers with such high impedance adds cost and the losses will be higher. Paralleling the units helps with the losses somewhat due to the fact that the current through each transformer is reduced.

Another neighboring utility uses 40MVA units with 17-18% impedance and 2 parallel on the bus also achieving 25kA station fault levels.

They also use Earthing Transformers that are 500kVA so that limits the single line to ground fault current as well.

 

[SilverfoxUK]

You say field switchgear. The impedance of the cabling between the transformer and the switchgear although likely minimal will cause the fault level to drop very quickly and likely to end up lower than your 25kA. Increasing the cable route length slightly might easily get you to where you want to be.
Getting desperate but an inline reactor could also help.

Increased impedance will increase operating costs through losses.

 

[kartracer087]

Yes you are correct, the added cabling is figured into the equation figuring in about 100 meters or more which would be the closest transformer station and switchgear.

CL reactors are an option though they can be expensive. I had one priced recently 400A and 0.5 ohm with enclosure it was $110,000 USD. If used per feeder that is pretty big money. You could decide to only use them on the shortest feeders and save some cost there since you wouldn't be installing them on all feeders but I was shocked at how expensive they are.