Thanks for the update, jbfan. It’s not clear who is saying that the genset size is adequate. That is a critical bit of information that, if it hasn’t already, needs to be put in writing. The verbal finger pointing will continue until someone’s feet are held to the fire. It sounds like the genset is new, versus a decade-old+ UPS. It also sounds like the genset voltage regulator doesn’t know what to do with a nonlinear/nonsinusoidal load. Permanent-magnet excitation helps a little, but a 300kVA genset can have trouble with even a purely resistive load near its nameplate rating. I’m not up on the latest genset capabilities, but it may be difficult for some gensets to simply handle a resistive {1.0PF} 100-to-200kW-load step change. That can be a hard thing for a genset {and its AVR} to negotiate—much less if the load is nonsinusoidal.
For your testing session you might want to look at phase voltages and currents with something like a Fluke 41 or Tektronix 700-series, to get a clearer idea of how the two major components are interacting.
Ideally, for AC circuits, distorted voltage points to the source, and distorted current points to the load. In this case, though, it’s more complicated, for the source does not present a very low impedance to the connected load, and UPSs are {in the fine print} promised to work acceptably with a low-impedance source, like a nice utility-owned 5%Z padmount transformer. Aside from that simple low-impedance source, decent voltage regulation and the ability for downstream overcurrent devices to operate reliably may be seriously limited.
[This message has been edited by Bjarney (edited 03-14-2003).]
