Sizing transformer

[HienN]

Hi all,

Today, I bring a new question that hopefully it is worth for a while thinking.
We got a new project of a Cryogenic plant. This is also my very first big project. The electrical equipments specification is as was following:
Air compressor (02 sets): 6.6kV_2056kW_Reactor starting
Recycle air compressor: 6.6kV_3500kW_Reactor starting
Oxygen compressor: 3sets; 6.6kV_800kW_DOL starting
Three options: Transformer with 6.6kV_15_17_20MVA (Z%=7)

During this very first phase, information is quite unclear. These motors will be started one after the other and could be explained as followed: AC 01; AC 02; and RAC (and then 03 Oxygen compressors).

The question come up is how to determine the initial transformer capacity to avoid the voltage drop to allowable value?

As the presence of many of experienced people here, theoretically and practically, I look forward to receive your suggestions to get forward.

Regards,

HienN

 

[rbulsara]

While there are rule of thumbs and experience, for this size of project a motor starting/transient analysis will be in order, among other studies that may be required, using a good software.

A lot depends on what is allowable voltage drop, what is the starting current and how strong or weak is the utility source (its impedance), etc. In theory it is a simple voltage drop calculation. You will need to know system impedance and those of motors, transformers and cables, etc.

Starting current on the other hand depends on the load characteristics. So you need to gather all relevant data.

Also transformer MVA rating can be a little deceptive as it can be increased very easily by adding fan cooling and forced cooling etc, while the voltage drop issue remains unchanged and is related to xfmr impedance, usually published at its base or natural cooling rating.

Rafiq Bulsara

http://www.srengineersct.com

 

[HienN]

Rafiq Bulsara,

Thanks for your first opening,

For software, I have been using ETAP400. But it seems so difficult to understand while information is sitll not enough.

For the voltage drop, as reference to the last project, the voltage drop during starting is within 5 to 10% at the 22/6.6kV transformer secondary side. Therefore, for this, we will also do the same.

How to determine the utility source is weak or strong?

For this initial calculation, I will ignore the system impedance, just base on the motor capacity and their starting method (I already got information for these starters: Let’s take an example: For Reactor: Starting current is 80% of DOL. For Autotransformer: Starting current is 64% of DOL. It is just for an presumption against transformer.

Regards,

 

[rbulsara]

HienN:

I would not ignore utility system impedance, in fact the utility source would be the culprit in many unacceptable cases. Without that the exercise is meaningless.

All you have to know is available short circuit current from the utility source. I am sure by this time there is a contact with the utility source. If not it needs to be made.

System with greater available short circuit current (lower impedance) would be considered stronger as its voltage drop will be less for a given current.

Also you need to know realistically how many motors will be started simultaneously. Hopefully not more than one.



Rafiq Bulsara

http://www.srengineersct.com

 

[waross]

Old school rule of thumb;
Sum of the full load motor currents plus all other loads plus 25% of the largest motor or motors which will start simultaneously if greater than largest motor. This will not replace a good motor starting/transient analysis but it will give you a quick sanity check. In some places it is the code minimum.
Your analysis should be close unless you have a very weak system.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[HienN]

Rbulsara,

Normally, the short circuit current available at the outlet of 110/22kV transformer (belong to Power company) in our location is limited at 25kA (already contacted). Can we utilize this information to apply to short circuit capacity of the inlet of our 22/6.6kV transformer. As I said previously, it is just an estimation not a detailed design. So the impedance could be ignored.

Waross: Base on principle, sequential start shall be applied for this plant. I forgot to mention, the starting currents are at 80% of TAP for both cases. So, the biggest current is when RAC starts. It draws about 1737A during starting (eff.~0.9, power factor~0.9). This case, we should take account of the starting current not only the running one.

Except the power line distance, is there any other relation contributes to the system impedance? If one power utility with many consumers connected and the supply voltage is less than the nominal one, is that also a weak system?

Regards,

HienN

 

[waross]

a lot of customers on a circuit may give you fluctuating voltage.
Think of the system reaction to a block load. The stronger the system, the less voltage dip due to block loading.
The weaker the system the greater the voltage fluctuations due to varying loads, as well as greater voltage drop for a given block load.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[rbulsara]

25kA at 22 kV is a mighty system, 952MVA of fault MVA, in fact sounds a bit high. But if true, it is a very stiff system and should help your cause a great deal. On the other hand the SCC rating of your equipment needs to be high too.

The fault MVA would give you the source impedance. Zpu=base mva/fault MVA. And yes you can plug in to line side of your xfmr.

I would say, the fault MVA is correct indication of stiffness of the system. Actual voltage may be just a function of taps and loads. However if the feeder conductors are too small and/or too long it would have a significant impact on the voltage drop regardless of the system behind it.

Rafiq Bulsara

http://www.srengineersct.com

 

[Marmite]

The fault level of 25kA given by the utility is the maximum fault level. This is the figure the utility would normally be asked to provide for a customer so he could spec his switchgear. You need to go back and ask them the minimum fault level under first circuit outage conditions.
If you ignore the source impedance the figures will be meaningless. You could be fed via several miles of spindly overhead line and the fault level could be nearer 5KA. Until you ask the question you will never know.
Regards
Marmite

 

[HienN]

Marmite:I did some calculation for transformer 110/22kV and found that the short-circuit current at the 22kV size would have been around 23.6kA. Then, consulted some other Vendor like Schneider (local), and knew that it was not that high (just around 10~16kA). Not satisfied yet, I switched back to ABB transformer manufacturer(also local), and they also gave the same comments. But refused to give the method of calculation.

Moreover, I requested them the short circuit current calculated at the busbar of 22kV side of their transformer. I think they knew what we meant. But caution is never redundant.

I see and will never ignore the system impedance when doing the detailed design. But, I am not sure that we can win this job or not? As I mean this is just a very first phase. Just get price and propose to customer.

Normally, the plant is located inside Industrial zone where 01 Subsation supplying power to the whole area. Normally, switchgear manufacturer designes their equipments at the rating from 23 to 25kA.

Regards

 

[rbulsara]

Equipment ratings are always much greater than actual available SCC. I agree that equipment SCC rating at 22Kv could be 25KA but that would not be the actual SCC, which is of real interest in the VD calcs.

Think it this way what size substation tranfromer would the utility co. use to fee this facility? 20MVA? 40 MVA?

I would use more restraining size during the initial phase.




Rafiq Bulsara

http://www.srengineersct.com

 

[HienN]

Rafiq Bulsara,

Utility company transformer is 63MVA capacity with 8% PU.

Overhead line distance is about 1.5km.

 

[rbulsara]

Great! That is the info you need to use in the model. It is substantial so it should be a positive. Make sure you use the model of the 1.5km feeder as close to reality as possible. Do not ignore it.

Rafiq Bulsara

http://www.srengineersct.com

 

[HienN]

But I was wrong, the impedance is not that small. It should be around 12 to 14% for 63MVA transformer. With this impedance, the short circuit current will reduce dramatically.

 

[rbulsara]

Now you are on the right track. Devils are in the details. Assuming something is always dangerous. It pays to invest time and engergy in gathering facts.

Good Luck.

Rafiq Bulsara

http://www.srengineersct.com

 

[rbulsara]

Also verify that 63 MVA is the base rating and not forced cooling rating.

Rafiq Bulsara

http://www.srengineersct.com

 

[davidbeach]

Also verify that 63 MVA is the base rating and not forced cooling rating.

Quite true in the ANSI world, quite wrong in the IEC world. IEC transformer impedances are on the top rating of the transformer rather than the self cooled rating.

 

[rbulsara]

Thanks davidbeach. I wasn't aware of that. More reason to verify the base rating related to the published %Z.

Rafiq Bulsara

http://www.srengineersct.com

 

[HienN]

Thank you,

If I dont ask, it may lead me to many misunderstading in actual application.

By the way, what is the difference between the base and forced cooling capacity? If this 63MVA is forced cooling capacity then I would ask for the base?

Moreover, at the same time with this, for 110/22kV, 63MVA; 7% impedance (equivalent one) transformer, my project manager applied only 10MVA to calculate the short circuit current. He said to me that it was the base capacity and all of the calculation would base on this 10MVA capacity. How about your suggestion?

I understand like this: We need to collect at least 02 facts: 1st is short circuit current (or impedance), and power line information. If I already had impedance voltage, there will be no need to ask for short circuit curent. Correct?

HienN

 

[rbulsara]

The question to ask is What is the "base" MVA for the published %Z for the transformer. Here the "Base" means the "basis" of the calculating %Z, which generally is the natural cooling rating but as davidbeach said could be different.

For example a transformer rated 40kVA at natural cooling and 63 kVA with forced cooling, the %Z would be different for each rating but the absolute Z in ohms remains the same. So you really need to know at what is the base MVA that relates to the %Z.

If you mean %Z as impedance voltage and you have the rating of the transformer, the answer to your last question (last line) is yes.

Rafiq Bulsara

http://www.srengineersct.com