I am looking for a good technical article discussing the use of transformers with different Iz (or voltage regulation)in a delta-delta, Y-delta, and GrdY-GrdY.
Looking only at a GrdY-GrdY application, I would like to see comments on the use of a pole mounted bank using 3 units of the same rating but different Iz.
Also, if have a 150kVA bank, (made of 3-50kVA) and a unit fails, can I replace it by a 75kVA (same voltage rating) I have in the yard?
The regulation may be more important than the impedance. But if %Z is different regulation will probably differ also.
With different regulation, when the transformers load up, the voltage drops will not be equal. Unbalanced voltages cause disproportionately unbalanced motor currents and motor heating.
If a service was designed based on the installed transformers, and you replace one or more transformers with units of less impedance, you may be exceeding the Available Short Circuit Current that the service equipment is rated for. Exceeding service ASCC ratings may also apply to installing larger transformers.
That can lead to panels exploding on faults rather than clearing on faults.
Bill
--------------------
"Why not the best?"
Jimmy Carter
Bill... thanks; Indeed voltage regulation is the key here. Can you comment on the following?
Still looking at a pole mounted GrdY-GrdY bank : if a 1% maximum unbalance can be tolerated, a local Utility has established a rule of thumb stating that the maximum Iz deviation for any unit from the average of the 3 units is 0.5
As example, if they have 3 units at 2.6%, 1.9% and 1.9% the average is 2.1% so the banking of these is OK as the 2.6% unit is within the 0.5 rule. On the other side, many other utilities indicate the Iz should be within 7.5% Do you know where this 7.5% is coming from? Is it just the ANSI tolerance on the garanteed Iz, so they select units based on Nameplate Value. Let's assume for ease that regulation and Iz are moving the same way.
I think that the rules may be similar.
2.6% + 0.5% = 3.1%
2.6% + (2.6% x 7.5%) = 2.8%
We are getting into "seat of the pants engineering here".
We have seen questions where the answer is;
"It depends."
Here we may have "(It depends)2"
It's nice to be working with a major, well financed utility or plant with all the correct equipment available.
That is not always the case.
Sometimes, we have to get by with what we have available.
Some comments;
1% unbalance, I assume voltage, is quite good for a lot of systems.
Transformers with different %imp. may be different because one is a high efficiency transformer. This kind of upsets the Z:R relation ship, but %regulation will always be less than %imp.
Motors and rectifiers;
Motors will take energy (both real and reactive) from the higher voltage phase(s) and pass it to the lower voltage phase(s) to try to match the voltages. The current unbalance is much worse than the voltage unbalance. The motors run hotter. Sometimes they fail.
Rectifiers will take current first from the higher voltage phases. At light loads, a rectifier front end may take almost all the current from one or two phases.
Furthermore, as the power factor of a load drops, the %imp becomes more significant and regulation becomes less significant.
What does this mean in the field?
A plant with large heavily loaded motors wants a good voltage balance. Usually there is a large economic penalty associated with the failure of a large motor. They cost of downtime to change a failed motor may be several times the cost of the motor. Even a trip on overheating with no motor damage may incur heavy downtime costs. One of my first responsible positions was in a plant where the stated downtime losses were about 800 times my hourly rate.
A warehouse with a mostly single phase lighting and a few lightly loaded conveyor motors may run happily on a relatively large unbalance.
Faced with a failed 50 KVA transformer in a bank feeding a balance sensitive load, I may search the system for a single phase load fed from a suitable transformer and switch that unit out for the spare. Then I have a better matched transformer to install in the three phase bank.
I hope that this helps you.
Bill
--------------------
"Why not the best?"
Jimmy Carter
I believe the 7.5% figure is the ANSI tolerance on impedance, so, if you received a shipment of like units, you could get a 7.5% variation.
The 0.5% rule of thumb I believe is tied to a particular connection like delta delta. If you connected three like transformers, with a three phase load, you get no unbalance. If, however, you have to replace one of those with a different impedance unit, under load, the secondary voltage on the different unit will be different than the other two and this difference will force circulating current in the delta. I think the 0.5% is suggested so that you don't have to severely derate the transformer bank.
As you mentioned, the regulation is most important, but this is determined from the impedances and the power factor of the load.
Furthermore, with your wye wye connection, you may have single phase and three phase load which complicates the situation and can lead to further voltage imbalance.
You might look in the IEEE papers for one by D. R. Smith. Dave did quite a bit of work with transformer connections and had a few excellent papers on the subject.
A few more precisions... the avg voltage was calculated as the avg of the three line to ground voltages; the 0.5 is the absolute Iz spread that would still allow 2 pu overloading; for a 1.3 pu overload they would use a spread of 0.77.
Magoo2, the 0.5 is a plain absolute value, and applied to a GrdY-GrdY bank.
I have a few of Dave Smith papers as I used to be at Westinghouse while he was there. I wish I had kept them all! Thanks for pointing at the IEEE explore!
On the 0.5 factor, I normally deal with the impedances in % so if my % impedances are 1.9, 1.9 and 2.6, then the 2.6 % is within the 2.1% average impedance plus or minus 0.5%. Agree?
The other point referred to the 2 pu overloading. I think you meant 2 pu loading. 2 pu overloading means 3 pu loading.
If you can get your hands on an ABB Distribution Transformer Guide, there's a lot of loading and unbalance info relating to the different transformer connections. I think Dave had his hand in much of this portion.
You are correct on all points; its just that dealing in % can be misleading. I have the Distribution Tx Guide and it goes back indeed to the Westinghouse days.
A comprehensive tutorial on parallel operation rules for transformers is given under clause 6 of IEC60076-8 Power Transformers-Application Guide with formulae and typical calculations.The extent of variation allowable in %Vz depends on the relative rating of units.
When equal rated units are considered, normally 10 % variation in Vz is allowed,considering that transformers are suitable for 10 % continuous over loading and IEC allows that much variation in rated impedance when 5Z is less tahn 10 %.
You can replace 50 kVA by 75 kvA in three phase bank, provided %Vz of 75 kVA is 50 % more than that of 50 kVA.But if you want to parallel operate a 3 phase 50 kVA with 75 kVA unit,then %vz of units shall be the same.
Then your question is how much variation in % Z can be tolerated between units in a three phase bank.Normally 1-5 % among the units is the rage demanded by utilities.
PRC, thanks for this note; the IEC document is quite detailed for parallel operation indeed. I have found some data on paralleling in IEEE C57.105-1978. I wish there was as well a detailed document for banking single phase in a three phase bank
Of course when replacing a failed 75 kVA in a 3 ph bank made of 3 - 50kVA units, I understand I have to put the 75 kVA Nameplate %Iz on the same kVA base as the others,ie divide the 75kVA nameplate %Iz value by 50/75.
PRC, thanks for this note; the IEC document is quite detailed for parallel operation indeed. I have found some data on paralleling in IEEE C57.105-1978. I wish there was as well a detailed document for banking single phase in a three phase bank
Of course when replacing a failed 75 kVA in a 3 ph bank made of 3 - 50kVA units, I understand I have to put the 75 kVA Nameplate %Iz on the same kVA base as the others.
On a historical note;
Although there may not be standardization from ne manufacturer or even product line to another, there is a standardization of sorts in that the efforts to build transformers economically sets a de-facto limit on transformer parameters.
I have a very old text book which develops the vectors for two transformers in parallel. Although the transformers have equal impedances, they have unequal X/R ratios.
They share the load in inverse proportion to their impedances, but the sum of the transformer currents is more than the load current.
Bill
--------------------
"Why not the best?"
Jimmy Carter
Absolutely correct... and the same goes for Voltage Regulation as it is a function of %Ir and %Ix
There are many good articles on paralleling, but do you (or anyone) have any article that develops single phase banking in a three phase bank, specifically for GrdY application? The idea is to develop a guide for operations... specially when a transformer of the bank fails and need replacing. A duplicate may not be in stock.
ffont-Regarding balancing of impedances between single phase units in a three phase unit,the major power company in India specifies for Yd Generator transformers as " Impedance of any single phase unit shall be within +,-5 %of impedance of other units"
