Load Flow for in per unit

[mutimuti]

I am an educator and find that many examples for load flow given in textbooks are in per unit. The software which I use uses physical data (ohms, km MW KV etc)and there is no option in my software for this. I tried to use 1kV, 1Km (impedance is given per km) and 1 MVA so that I would not have to change anything but this only worked for certain cases, for other cases there was no accuracy?. Any advice?

 

[itsmoked]

I'm still not clear on your question... Perhaps an actual example? What software?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[jghrist]

I would suggest converting the textbook examples to physical units, then using these with your software. Then convert back to per unit values for comparison with the textbook examples. If no base values are given in the textbook, then use some typical values like a base voltage of 115 kV and a base power of 100 MVA. Use standard per unit conversion formulas. Be sure to use the proper voltage bases on each side where the examples include transformers.

 

[mutimuti]

Thanks for the quick responses. I am using NEPLAN 5.2. The impedances are given per km. Rated Voltages are in kV. Power in MVA, MW. So I figured out that if I make my base voltage 1kV and my base MVA, 1MVA then the actual ohms and the per unit ohms would be numerically the same. I also figured out that if I make the length of each and any line 1km, then the actual ohms and per unit ohms would be the same. I then tried this reasoning on a fully worked 3 bus network with a PV, PQ and Slack bus in the per unit format and got identical solutions. However, the technique seems to have failed on another example in another book (a 4 bus example with susceptances. To go back to jghrist's suggestion, I say why not just use 1 kV and 1 MVA then there will not be any conversion as I am trying to do and I have succeeded on one occasion?

Is my logic above technically sound?

 

[rbulsara]

If you are an educator, please invest in (buy) IEEE "Buff" book "Recommended practices for power system Protection and Coordination" to learn the PU system and coversions to actual values. Refer to

http://www.ieee.org

 

[mutimuti]

Rbulsara, thank you for mentioning the IEEE book, I will try and find it. It looks like a useful book. I know the PU system very well and I also teach it.
Z (pu) =Z(ohm)*(Vbase (kv))/(Sbase (MVA)^2). Clearly, if you make Vbase =1kV and Sbase=1MVA, then Z(pu)=Z(Ohm). Actually may be I exaggerated the extent of the descrepancies in the results, the Real Power flows balance exactly, the reactive powers by about up to 1 to 5%. The voltage solution is correct. I am beginning to think now that in the example which I thought had descrepancies, my (NEPLAN) results are correct and I will continue with my method for more examples and compare. I will continue like this until I am 100% sure since it will help greatly if I do not have to convert literally hundreds of input data by hand and in my opinion for no apparent reason.

mm

 

[jghrist]

The problem may be that the textbook examples are based on much different (and more realistic) base voltages and powers. This may result in rounding errors.

If you are combining impedances of transformers or machines with line impedances, you will not get the textbook answers by assuming all lines to be 1 km long. Transformer and machine impedances will not be scaled by the same factor.

 

[ijl]

You must be careful with the transmission lines and other impedances, if there are transformers in the network. The pu-impedance of a transmission line on the primary side is different from the pu-impedance of an identical transmission line on the secondary side.

 

[motorspert]

multimuti
I am not sure if i understood you correctly - but I dont think units will ever tie up, eg a 1MVA 1kV system has a per unit current of 0.766A (1000/1000*1.732).
This gives a pu resistance of 1000/0.577 = 1732ohm. To get a 1km line to have such a high resistance would need a small section - I think you would have to many fiddles that the results will not be reliable. Thats why i like the pu system, much simpler to understand.

 

[davidbeach]

I'm not sure it would ever be possible to take per unit system system data and arbitrarily change the base values and be able to come up with meaningful physical units. The original systems probably have a base of 100MVA, maybe 10MVA and the a base voltage equal to the system voltage for one voltage level. The base voltage beyond each transformer from the first base voltage will be related to it by the voltage ratio of the transformers, which means it may not be the same as the system voltage on its side of the transformer.

You really need to find the original base information and use that. I believe that most competent system analysis software would allow impedances to be input in PU if the appropriate base values can also be entered. But you can't just make up the base values.

 

[mutimuti]

Thanks all of you who are in this debate. The text book examples are all per unit representations of the systems involved and in my opinion when this is done lines cannot be distinguished from transformers. The caution from jghrist is noted. I have not had to convert physical units to per unit to do the analysis because this would be unnecessary in any case, since most Power System Analysis software operates directly from physical data. This effort is to help with my teaching a process which needs validating everything.

mmt