Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

GENERATOR/UPS Power Disagreement

  1. Sep 25, 2007 #1
    I must start with some background. We have a client who has a 480/277 Service into a 1000A Breaker. Neutral/Ground bond is located here via a busbar. The ground is a ground ring of unknown size and distance. There is also a supplemental ground wire (#6) attached to a ground rod. The service is then routed through an Cutler Hammer ATS. The service is then routed to a distribution panel. The panel has 2 600A 480/277 breakers. It also has various other "Utility" breakers that feed other distribution panels that go to HVAC units and lights/receptacles through another transformer. The 2 600A breakers go as follows. 1 goes to UPS as Mains 1 and the other goes as Mains 2. I am not sure how the Piller UPS manages these two Mains but there is also a MOTS bypass. This is simply a manual transfer switch to bypass the UPS. Now for the problem. The UPS just runs the power through it until it sees a problem. Now, as mentioned before the entire bldg is through an ATS. This ATS is connected to an 800kW Diesel generator with a Marathon DVR2000 Voltage Regulator. When this generator turns on either by power fail or exercise the UPS automatically triggers itself by some interconnect relay to close an internal switch that switches all power to the rectifiers. Now even though the ATS is connected to the utility company the UPS is rectifying power. (maybe in anticipation of transfer, the UPS tech cannot convey why this is) Now when the ATS transfers to ER Power (GenSet) the power is rectified from the Generator. This is raising havoc on our generator. We cannot stabilize the voltage. The current appears to stay the same...around 360 and the Freq is stable at 59.7. However the voltage is floating between 420 and 490. This causes the UPS to switch periodically to batteries. The "powers that be" recommended that we put filtering on the Voltage Regulator to filter and stabilize the voltage. We installed MOVs to clamp the 3 phase sensing wires on the VR and a 3% Line Reactor on the sensing wires. This, as I thought, had no effect. The VR seems to be doing what it is supposed to. We have about 175kW of coming out of the UPS and the GenSet is supplying about 260kW. So this means that about 85kW is being used for the Utility. This means that 67% of the load is PURE Reactive through the UPS rectifiers. We have paralleled a Resistive load bank with the UPS at 250kW and were able to stabilize the voltage. This tells me that we need to add some resistive load to this unit. HOW? Oh yeah...the UPS tech says that his machine must maintain a SEPERATE neutral/ground bond. That makes (2) Two complete neutral ground bonds on the 480/277 buss.

    Test Results from Previous power monitoring
    Voltage Harmonics THD=2.17% H1=276 H3=5.01 H5=0 H7=0
    Current Harmonics THD=7.8% H1=86.95 H3=0 H5=3.86 H7=2.86

    Also at the time when the Static Switch on the UPS triggers the rectifiers this is the power
    Before static switch
    Apparent Power 19500 VA
    Real Power 19130 W
    Reactive Power (Ind) 0 VAR
    Reactive Power (Cap) 3734 VAR

    1 Min Later Gen Running withRectifiers
    Apparent Power 23960 VA
    Real Power 19700 W
    Reactive Power (Ind) 13640 VAR
    Reactive Power (Cap) 0 VAR

    How can we stabilize the voltage? IF more info is needed just ask.
  2. jcsd
  3. Sep 25, 2007 #2


    User Avatar
    Science Advisor
    Gold Member

    Well maybe you could just bypass the UPS with another ATS when the generator is running. i.e. after the generator get up to speed, switch the UPS out of the circuit until your service returns.

    I'm guessing here, but it sounds like the UPS is not seeing a clean sine wave. So maybe some filtering of the UPSs sense voltage can help. Or is that what you have already tried to do with the VR sense wires?
  4. Jan 4, 2008 #3
    (See edit below)

    I think that the problem you are seeing is that the UPS is a nonlinear thyristor load: it draws power in spikes. The reason it is a problem is that most GenSet voltage regulators are powered from the same generator (a shunt-excited generator).

    Adding a dummy load gives the AVR something to work into, which absorbs some of the overshoot and allows it to stabilize. But that's a waste of energy!

    The solution is an AVR that receives its power from a separate source, usually a permanent magnet generator on the end of the generator shaft.

    You probably also want an AVR with three-phase sensing rather than the single-phase sensing default that most gensets are delivered with. Three phase sensing is best for unbalanced loads, but it also helps with non-linear loads because three-phase sensing AVRs are usually RMS rather than average sensing.

    You will have to purchase a PMG bolt-on kit for your generator (assuming that is available, some generators have to be built with them) and the PMG AVR that goes with it. Some PMG AVRs (Stamford Newage) have a three-phase 120Hz PMG supply, others like Marathon use a single phase 300Hz PMG.

    Re-reading your post I realize I skipped over where you mentioned that you have a DVR2000, which is a three-phase sensing PMG AVR...

    Have you tried adjusting the stability or other settings on the AVR? Lack of good grounding of the AVR could cause instability due to spurious RFI/EMI caused the by the UPS' switcher input. I don't think the filtering on the AVR sensing is a good idea, it is RMS sensing and that could make things worse (but MOVs are good). I would also verify the power supply to the AVR -- and if all else fails, try replacing the AVR, it could be bad. Try a DVR2000E, it has more options and a serial port interface.
    Last edited by a moderator: May 3, 2017
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook