Power Factor Correction

The p.f. correction capacitor improves the p.f. of upstream line not the down stream line. Thus, to reduce the current hence the copper losses of Transformer + line losses capacitor should always be installed at load end.

Anything more u want know ?

That is correct. You may find the following link useful.

http://ecmweb.com/mag/electric_primer_power_factor/

Thanks for the explanation.

My worry was the 3.5km length of cable carrying the 3.3kv to load whether the cable length had an effect since the utility meter is by the 33kv/3.3kv 350kva transformer which reads the pf and we get penalised if pf is below 0.9.

Thanks

PF correction capacitors should be installed near the reactive load and switched with the load.

For the loads, you connect the required capacitors at the end of the cable. But, to compensate for the reactive power of the transformer itself, you connect a capactor of a rated kVAR of about 6% of the transformer kVA rating, right at the secondary terminals of the transformer. This will avoid your penalty problems.

Dear Mr. electricalexpert65, how these 6% is quantified ?

In this case it will be 6% of 350kva 33kv/3.3kv transformer.

Thanks,

Joseph

Usually the reactance of the transformer will be somewhere between 3-5% so the EE must be trying to overcompensate it by putting the 6% compensation. However nobody infact tries to compensate the inductance of the transformer. The cost may not be commensurate with the value. And this value is for compensating transformer and not load.

The cables are inherently capacitive, so no effect of load inductance of 3.5Km long cable.

As already suggested put the APFC near the load. large distributed loads - individually, large/small concentrated - near the load distribution point.

"Nobody, in fact, tries to compensate the inductance of the transformer" - Dear Mr.Guest! That -exactly - is the problem. I have faced problems in batch operating plants wherein, even after providing proper capacitive compensation for the connected loads, the customer ended up paying penalty for poor power factor, for, when the loads were OFF, the APFC Capacitors were also OFF, but the no-load reactive power of the transformer was reducing the net power factor. It was 0.99 lag for 12 hours in day time and 0.5 lag for 12 hours in night time (when the plant was not in operation) and the average daily power factor fell to 0.7, causing power factor penalty.

When compensation for the no-load reactive power of the transformer was implemented, the customer was able to maintain his power factor to 0.99 lag. It definitely justifies the investment. Precise calculations can be made for the pay back period and the average pay back period is from 6 months to 2 years, depending upon the transformer capacity, the load cycle of the transformer and the enrgy costs.

At our place (almost a batch production) the line configutration is

132KV (from Grid) → 11KV (at our factory substation) → Individual local substations(hubs) (11KV/415V and 11KV/ 6.6KV) → individual machines.

Our APFCs along with some fixed capacitor banks, relics from the old age, are connected in the end hub portion for compensating the load power factors (mostly machines and few Induction for melting Iron as well as other furnaces)

We work in 16 Hrs/ day and no we do not pay penalty for the low pf of the 8 Hrs. When the plant is not working then of course the KVAr is high but then the total KVA is low.

As per definition, the pf for which penalty is levied is based on

a) peak load pf - which is in control

b) Average power factor (as per tribunal report and clarification)

The average power factor is the ratio of KWH to the KVAH consumed during the billing month/period. Billing Power factor shall be the average PF recorded in ETV meter. In case the same is not available the ratio of KWh to KVAh consumed during the billing period and in case of non-availability of the above also, the PF obtained during the rating shall be taken.

If your customer has paid penalty for poor, he should have gone to tribunal and challenged rather than paying the penalty or spending in the capacitors. Definitely 0.99 lag(12 hrs on load) + 0.50 lag (12 Hrs on lo load) will not add up to 0.70 lag on average unless one takes time average and not the pf average.