Question about reactive power flow 1

[Pafi]

Dear All,
I have a problem related with power factor correction in an aluminum producing plant. It is supplied from 110kV network, 2 different supplies (equal supplies, 2 autotrafos 220kV/ 110kV). The rectifier transformers are supplied directly from 110kV, the scheme is with 48 pulses for each 110kV supply. The current regulation is from taps on transformer primary (automatic, on-load) and saturable reactors. Also from the same 110kV supplies there are 2 service transformers for each supply, 110/ 6 kV. The harmonic filters have been installed in the secondary (6 kV) of each service transformer (4 equal harmonic filters, mainly to block harmonics for producing resonances, 6 Mvar each). We have measured the currents, power, etc. on each harmonic filter, everything is OK.
But the measured power factor on 110 kV did not reach expected value. It was like we have introduced not 24 but 18 Mvars into circuit. The only thing that was different was the voltage that reached 113 kV from 110 kV before. Could this be the only reason why the expected PF has not been reached?
Any suggestion welcome (excuse poor english!)
Pafi

 

[Skogsgurra]

Do you have the same low power factor also when your saturable reactors are fully on? I.e. not developing any voltage drop.

See no problem with your English, by the way.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[wareagle]

Were the capacitors rated for 110 kv or 115 kv?

 

[sberbece]

I guess it must be something with the measurement instruments you use, or the point of measure. Since there is voltage rice (ie. 110 to 113 kV), that means the current circulation is less, therefore part of the reactive power required is supplied by the capacitor bank, consequently and improvement of the PF.

Regards,
Stefan

 

[sibeen]

Just to make sure I have this correct.

You have installed harmonic filters on the secondary of your 110/6 kV transformers (x 4).

What form of harmonic filters are they?

The measured power factor, is this true P.F or displacement P.F?

I'd suspect that the harmonic filters, though rated at 6 MVA, do not provide 6 MVA of capacitive loading.

 

[Pafi]

1. The regulating scheme is automatic, we have no control or knowledge of what the control system is doing (bigger steps with taps and smaller steps - between taps - with saturable reactors)
2. The harmonic filters have been installed on secondary - 6 Kv - of service transformers (these are not the rectifiers transformers, but supplied from the same 110kV supplies)
3. Yes, there is improvement of PF but not to the expected level . It is like 25% of the introduced reactive power is vanishing somewhere.
4. The accord frequency of harmonic filters is 4.6 (x 50Hz as fundamental frequency) equals 230Hz (for 50Hz fundamental). The PF is true PF and we have measured the currents on every harmonic filter, they are arround 595A in every phase with system voltage 6kV (losses are no more than 15 kW according to capacitors and reactors manufacturers). The harmonic content of current is low.To give you an example, the IRMS is 585A, I1 (fundamental) is 583.8A.
On the other hand, the THDI on the 110kV supply is also low, under 3%.
I shall try to measure directly the reactive power on filters, not only currents and voltages.
Pafi
Pafi

 

[slavag]

Hi.
Maybe reactive losses on the trafos.
Difference is 6MVar on the 2 trafos, 3MVar per trafo.Hmm, possible too.
What is a size of trafo, what is a %impedance?
Higher voltage---->less current with same MVA, less current---->less copper reactive losses.
Same Bill (Waross) explained in other thread.
Regards.
Slava

 

[Pafi]

I thought of that but seems to be too big reactive power for such transformers. There are 4 trafos, 2 of 25 MVA, uk=10.5% and 2 of 31.5 MVA, uk =11%. But I put all this in a power flow software programe and the difference in reactive power (with harmonic filters and without) is arround 0.3 - 0.4 Mvar for each trafo (total of less than 2 Mvars.
Pafi

 

[Bronzeado]

Pafi,

It seems to have current circulation between the transformers caused by unequal taps and/or saturable reactors condition in each transformer. This will cause a big reactive losses on the trafos (as Slava has guessed).

I suggest to have a look in the regulating scheme to see what the control system is doing.

Best regards,

Herivelto Bronzeado

 

[slavag]

Hello.
1. Pafi, please check Herivelto's point. Are trafos operate in parallel?
2. I'm not sure, if calculation 0.2 0.3Mvar per trafo is correct, from my point of view is too small.
Could you please attach LVcp losses or LWcop losses from trafo mnf. sheet.
Try use formulas
Qlosses(copper)=Qlosses(mnf) x ( I actual/I rated)^2.
Qlosses copper= sqr((VA trafo x %impedance/100)^2- (LWcop losses)^2).
simple calculation w/o LWcoplosses ( is not big value).
1. 25x10.5/100=2.6Mvar, OK, assume 2MVar
2. 31.5x11.5/100=3.6MVar, assume 3Mvar.
You can see is not 0.2-0.3Mvar per trafo, is ten times in full load.
Regards.
Slava

 

[Pafi]

Yes, all trafos are operated in parallel (4 rectifier transformers and 2 service transformers on each 110kV supply - 2 such 110 kV supplies).
The service transformers are running with small load, approx. 1/3 from their rated power (I don't know why, but that is it). Anyway, I think that finnaly you are right and I shall try to effectively measure the reactive power of each service transformer on it's own 110kV supply by measuring with and without harmonic filter connected in it's secondary side. This will at least give a precise value for the reactive power supplied by each filter in the 110kV supply.
All this will have to be done next week.
Best regards an thank you to you all.