If a generator is loaded with appliances that have a power factor lower than .8 is it safe to use power factor correction capacitors to improve this number?
1\$\begingroup\$ Power factor by definition is <= 1. It can't be as high as 8. \$\endgroup\$– TransistorSep 29, 2017 at 23:58
\$\begingroup\$ The pf of 0.8 does not define if the polarity of reactive loads. Is it safe to assume it is inductive loads that need capacitive correction?, probably. Are the capacitors reliable? who knows. \$\endgroup\$– Tony Stewart EE75Sep 30, 2017 at 0:40
\$\begingroup\$ I think it makes a different on abrupt changes in voltage if there is a transmission line delay to the said positive reactance and where you put the negative reactance , may matter on the transient reponse. It is like equalizing a lumped C to balanced and remote L so one might try to only partially compensate or locate it closer to the load before the switch and not after. \$\endgroup\$– Tony Stewart EE75Sep 30, 2017 at 0:47
\$\begingroup\$ If the motor is infrequently switched, jogged or reversed, then the best location is near the load before the over current trip breakers.. Otherwise further away before the starter circuit is better. \$\endgroup\$– Tony Stewart EE75Sep 30, 2017 at 0:56
\$\begingroup\$ I do that about year ago. I leaved the capacitor connected to the generator and one day when the inductive loads (motors) disconnected the voltage raise to more than 300V and the LED lamps blow up. (I don't know why the voltage jump maybe because resonance ?)You have to using automatic PFC capacitor switching \$\endgroup\$– M.A.KNov 17, 2019 at 12:13
Assuming the load is inductive, then shunt capacitors will improve the load power factor. However, with a stand-alone generator it's rarely necessary, and there are several considerations.
Do Not Do This on the output of a square wave or 'modified sine' inverter. Shunt capacitors will overload the inverter on the fast edges.
Power factor correction doesn't save you any power (fuel consumption) in the generator unless the transmission lines to the load have a significant loss compared to the load. It's not worth adding shunt C for the fun of it.
If you do decide to do it, then under-correct rather than over-correct, don't try to hit nominal spot-on.
Mount the shunt capacitors with their loads so that the shunt C changes when the load changes. It can be dangerous to add shunt caps to a generator and then run it off load, due to voltage rise with its residual inductance.