at the local bank we just installed a 3phase generator (delta system)w/ auto transfer switch to keep the data processing part up and running during an outage. when the service tech fired up the generator, it runs great voltage holds steady. put it under load the needle starts bouncing all over the place, the flourescent lights start flickering, ups systems start getting unstable, and it fries boards in the large 3 phase sorter. you take and adjust the regulator so the needle holds steady everything is fixed except the ups systems still are unstable. the tech guy replaced the regulator twice to no avail. he feels its the ups systems causing harmonics. this generator just feeds the data processing part of the bank (PC's, Flourescent lights, small portable A/C, 2 servers, the sorter and a few other small odd things in the room. what can the tech do to correct this problem? is it harmonics?
Loads across the phases: A- 28 B- 2 (only 3 phase equipment was sorter) C- 26
the tech started throwing out ideas the banks "could do" to help reduce the "harmonics". but it shouldn't be upto the bank, they just spent $20k+ on the generator they should make it right. we gave the company all the loads (A,V,W,VA, # of phases per load, etc...) and then the company sized the generator according. thanks for the time!
Lots of possibilities. What is the generator rating as compared to the UPS rating? What is the generator rating vs the measured load rating? Can the loads be reorganized to provide more even load distribution among the three phases? Is there a dynamic filter available for the input the the UPS (call the manufacturer of the UPS)?
#50388 - 04/01/0501:00 PMRe: generator and harmonics?
If the output of this generator is delta where are you getting a neutral for the small single phase loads (PC’s, fluorescent lights, etc)? Has this generator been tested with a load bank before testing it on the building?
[This message has been edited by caselec (edited 04-01-2005).]
#50389 - 04/01/0503:35 PMRe: generator and harmonics?
I see a lot of this as I work a lot with generators supplying rectifiers and UPS systems. Affraid most of your problems lay in the design/planning or lack of it. I doubt there will be an inexpensive fix
Ron hit on some of the deatails, especially the generator rating versus the measured load rating. When a UPS is involved you need to be at a 2:1 ratio.
Fist step to do is balance the load evenly accross all phases.
If that does not work then the next step is to have a UPS designed to work with a generator, that has built-in or add on filtering, and a sensor input to tell the UPS when going on generator to "Walk-In" the rectifiers.
#50391 - 04/03/0502:35 PMRe: generator and harmonics?
Is the service a delta? Is this the reason delta was choosen for the generator? Does this UPS accomodate a high-leg, and phase correctly? I assume a high-leg delta is the reason for such an un-balanced load.
Anyway the load needs to get balanced there. Maybe a delta/wye transformer, and then you can start moving the loads around and re-identify them for better balance. With service, gennie, line side of tranformer, UPS and loads the other.
Or ditch that 3-phase sorter, and start over in single phase.
Mark Heller "Well - I oughta....." -Jackie Gleason
#50392 - 04/03/0506:41 PMRe: generator and harmonics?
I would have to say the UPS is creating havoc, in the form of Harmonic Distortion.
When tracing Circuits via my Amprobe-Pascar Circuit Tracer, the audible sound plus received signal would be extremely distorted on UPS Circuits.
The Circuit Tracer mentioned above, functions by Modulating the 60 Hz AC Circuit's Fundamental with a ± 2kHz signal (tone). The Transmitter is connected at the location to be traced - and uses the 60 Hz Fundamental as the Carrier Wave. The Receiver is normally used at a Breaker/Fuse, but may be used elsewhere. It picks up the tone via Inductive Coupling on the test circuit.
Normally, the received tone and scaled levels are crisp and clear. However, when testing UPS circuits, the tone + scaled levels are very distorted. Also, they become "distributed" around common circuits (like remaining two circuits of multiwire circuits, or same phases in panelboard, which are multiples of three in location numbers).
If the tester reacts this way, imagine what is being pressed against the Genny!
Just a hunch!
p.s. Ironically, the work being done in the above scenario (along with nearly 95% of projects between 1990 and 2002), was in the Banking Industry.
Scott " 35 " Thompson Just Say NO To Green Eggs And Ham!
#50393 - 04/03/0507:36 PMRe: generator and harmonics?
How on earth do you 'balance' a system where the three phase load is (apparently) only 2A per ph, against 24-26A on the single phase legs? The A.C. thermostat and the Staff at the Bank will keep switching all the single phase stuff on and off, ( the ungrateful idiots!!). Also, and correct me if I'm wrong, IF the generator is wired Delta, and IF there's thus no 'neutral' tapping, are'nt all the single ph.'neutrals' effectivly hot (live) by going phase to phase when on stand-by?
Wood work but can't!
#50394 - 04/03/0508:41 PMRe: generator and harmonics?
"the ungrateful idiots" probhably leave the stuff on all the time, and balancing it would be as easy as moving some two wire circuits from A and C to B, if possible. (If this were a wye system, it would be very easy)
Now more questions: What happens to the neutral in the transfer switch? Where does the neutral come from on the gennie? (Corner/center tapped delta?)
"put it under load the needle starts bouncing all over the place, the flourescent lights start flickering, ups systems start getting unstable, and it fries boards in the large 3 phase sorter." - Sounds like neutral drops out, and phase reconnect to me.
Mark Heller "Well - I oughta....." -Jackie Gleason
#50395 - 04/05/0506:44 PMRe: generator and harmonics?
As others have pointed out the various problems, the root cause sounds like a very bad design. Any generator used with a UPS has to be large enough to run the UPS at least 2to1 add all loads not supplied by the UPS at least at1.5 to1. And single phase loads have to have a neutral from the generator. This sounds like the archetict/engineer who designed this system did not do thier work properly. Have seen very simular sceneros in the past and it is always a real PITA. Have also seen several tries to use filters, and other means to work around the basic problem. Thede tries had varing degrees of sucess and personally have yet to see one work the first time. The only fix I know of is to rip it out and try again. With a properly designed system. Who eats the cost IMHO should be the designer. No cheap fix here. Good luck with it