4

I have a 25kW liquid cooled Generac for whole house, using transfer switch RXSW200A3 (200A, service rated). Is there a way to get a "dead time" before the transfer back to the utility source when it comes back hot? I'd like at least a tenth of a second dead time (longer fine too) so I can avoid the out-of-phase kick in the face to my motors/compressors and to the generator itself.

I have an SEL-735 PQ meter on my house and captured this couple days ago. We had a power outage and were running on the Generac. I happened to be outside when I heard the transfer switch "thunk" and my outside HVAC unit fan started making a sound like a stick had been stuck into the fan blade. I turned it off and upon restart it was fine. But, out-of-phase closing is a cumulative damage - it will take life off of motors etc.

enter image description here

I have gone through the manual on the transfer switch and I don't see a way to get a dead-time (I have RXSW200A3). My dealer says that the Cummings machines he has put in have a transfer switch that does have the dead-time ("break before make" when the utility comes back) and it is around 0.1 seconds. If I can't find a way of getting my present setup to have dead-time then I'll go the route of replacing with a Cummings (will have to convert my 4-wire Generac control wires to 2-wire but that is apparently doable).

Anyone know how to adjust RXSW200A3 transfer switch (or Generac controls) to give the much needed dead-time?

thanks, russ

UPDATE 7-12-2022: Based on @Jasen and @Gil input I now see that in is indeed a "break-before-make" switch action, but the out-of-phase transition is still present since the voltage on the induction motor load (compressor) cannot decay to zero immediately due to trapped flux in the rotor.

As soon as the Generac is disconnected (but prior to utility closing in) the frequency drops due to the induction motor slip. In this case the slip was around (60.12 - 51.3 = 8.82 Hz). You can see this (and I suspect that is what clued @Jasen to it) as the time between zero crossings gets stretched out. This is an immediate response to the induction motor being disconnected from the source.

enter image description here

The decaying voltage of the induction motor is a voltage source that is slowing down and reducing in magnitude quickly - but not fast enough to avoid the out-of-phase transition. Because this voltage is not in-phase with the utility the effect is that when the utility closes in these 2 voltages are additive (based on their phasors amplitude and relative phase position). So, the voltage across the motor may be higher than rated. A rule of thumb many designers historically used for a maximum momentary voltage applied to a motor is 125% of rated. This applies equally well to single-phase motors as three-phase.

Question: As I understand it, when my thermostat (Honeywell th8321wf1001) loses AC it causes a 24 V contact to dropout and disconnect the compressor from AC. When power comes back there is a delay before it re-energizes that contact to reapply AC to the compressor (e.g. 5 minutes). But, what I need to know is - how long does the loss of AC need to be to ensure that contact dropout occurs and my compressor is disconnected (letting pressures equalize)? Is it a tenth of a second? Is it half a second? I'm fine with a 5 second dead-time if need be to ensure compressor is out of the way before I close the utility back in - just need to figure out how to implement this and get the dead-time I want.

Also, below is the phase-to-phase voltage that @Gils points out will be what the compressor experiences. Remember that the voltage across the motor is this voltage +/- some portion of the residual voltage still on the motor due to trapped flux.

enter image description here

relayman357
  • 143
  • 4
  • 1
    I don't know enough to answer your question, but I do admire the question! Anyway, I know product recommendations are OT here, but a couple of years ago I did a ton of research on whole house standby generators for my son's new build. I have an Onan (25KW, 200AMP transfer switch) which has been bullet proof for the past 15 years (and used quite a lot). I was going to suggest that, but discovered Onan was purchased by Cummings a number of years ago. Online reviews said customer service was clueless in dealing with any sort of software issues...they are basically a commercial supplier. cont. – George Anderson Jul 11 '22 at 23:15
  • I went with a Kohler generator for my son's house, came with a 200 amp xfer switch. Nice pieces of equipment and great reviews. Good software too. Generac gensets don't have the best reputation. So if you are considering swapping, I'd seriously consider Kohler. (disclosure: I have no financial interest in Kohler, just like their genset), As usual, things change and Cummings might have gotten the customer service on the retail side in order, so yeah, do your homework. – George Anderson Jul 11 '22 at 23:20
  • 1
    Thanks @GeorgeAnderson, I'm happy with the Generac - it is the transfer switch I'll be replacing if can't find solution. Appreciate your input. – relayman357 Jul 11 '22 at 23:57
  • Is the graph am measurement you made? I didn't think transfer switches were that fast. – Jasen Jul 12 '22 at 01:08
  • Yes @Jasen, it was each half (120V) of my service (a typical split-phase 120/240V service in U.S). I just labeled the A & B for convenience. The SEL-735 was sampling at 128 samples/cycle. My switch is clearly do a make-before-break. You see how the blue line was headed upwards and then BAM, got jacked back down the other way? That is way out of phase. If the transfer had waited another 4 ms it would have been 180 degrees out-of-phase. – relayman357 Jul 12 '22 at 02:59
  • That's crazy, synchronization would be easy. Just measure when the AC voltage between utility and gen L1 descends to zero, then SNAP! – Harper - Reinstate Monica Jul 12 '22 at 04:06
  • I can now you can see from the graph that the generator switch opened about 20 ms before the line switch closed. – Jasen Jul 12 '22 at 06:51
  • 2
    Hi @Harper-ReinstateMonica, the control is not that smart apparently. It is actually pretty hard to do a timed-closed on two systems that are out of phase and frequency that much. Years ago Beckwith Electric designed a high-speed sync check relay (M-0425) that could do it - they hired an analog electronics expert from NASA as I recall. Was impressive. Normal sync check relays as we use in utility are looking at much slower moving targets. I don't want to try and tie in sync - i want a flat out dead voltage time until my HVAC contactors are open and out of the way before reclose utility. – relayman357 Jul 12 '22 at 15:16
  • 2
    If anyone wants to look at the shot themselves I put it here: https://relayman.org/temp/genshot/ Download both the CFG and DAT files into a folder and open with a COMTRADE viewer. You can get one for free from https://www.electrotek.com/pqdiffractor/ You just have to register. Tom Grebe of Electrotek wrote it and it is good. He wrote TOP years ago and it is good comtrade viewer too: http://www.pqsoft.com/top/ – relayman357 Jul 12 '22 at 15:27
  • 2
    Relayman, I have to wait 1.5 seconds with auto start gear or I can snap drive shafts on 3 phase equipment , I have 32 closed transition ATS’s at my main facility and 9 open transition in the smaller / older buildings, I would make sure you have off delay’s on your contractors so they cannot pull back in until 1.5 seconds or the motor may be running 180 out (the noise you heard) easy to do with single phase but if the voltage is allowed to decay they will normally start correctly after 1 second we add an extra .5 to reduce equipment damage. – Ed Beal Jul 12 '22 at 15:34

2 Answers2

2

I know of no transfer switch that is not break before make. That automatically gives you dead time. How much, depends on the switch. Mine and many others are mechanically interlocked switches and the dead time is not adjustable. Imagine what would would happen if your genset was connected to the utility feed. That is why they are break before make.

Also consider what happens to an HVAC and other compressors if they lose power for a few cycles, they stop then try to restart on high head. I do not understand how you can get a single phase motor out of phase. Your wave forms show two separate phases referenced to neutral. The compressor motor only sees phase to phase.

As far as the fan blade be sure it's set screw it tight to the motor shaft.

Gil
  • 5,880
  • 1
  • 5
  • 19
  • 1
    you can see from the graph that the generator switch opened about 20 ms before the line switch closed. – Jasen Jul 12 '22 at 06:51
  • 1
    Thanks @Gil, that is helpful (+1) - so I was wrong when i assumed a make-before-break had occurred. The end result is the same, the motors had voltage on them when the utility came back out-of-phase position with them. The induction motor has trapped flux in the rotor when you disconnect it from the source. This flux decays depending on mechanical loading, electrical loading, and is extended if capacitors are present. Due to slip it immediately is at a reduced frequency (this is evident in the shot now that I looked at it closer thanks to @Jasen). – relayman357 Jul 12 '22 at 15:17
  • @gil look up closed transition transfer switches. They are out there so the only power bump is on a loss of power. some plants have large MG sets (motor generators) with heavy flywheels once up to speed they can carry the load for the few seconds before the generators are up to speed and in sync. code requires hospitals to be online within 10 seconds. We use closed transition so we can test our times each month under load. My main plant is generating 6Mw in 8 seconds when the line voltage is stable and we allow the system to return to line power then the switches transfer back no bump 1 by 1 – Ed Beal Jul 12 '22 at 20:36
2

What you need is called a closed transition transfer switch (no power bump) going back and no power bump for monthly testing. This ATS verifies the line and generator are in sync prior to transferring, most don’t go to the extra expense and just use open transition (power bump).

Closed transition cost more because the ATS monitors both line and generator and waits for the generator to sync to the line, then transfers.

If you just want open transition as in some of the locations I support a 1.5 second delay is all that is needed on my gear with much larger gen sets.

Ed Beal
  • 101,306
  • 4
  • 72
  • 150
  • 1
    Thanks Ed. Can you point me to a specific switch manual for a "no power bump" transfer switch? Appreciate the help. Yes, 1.5 seconds would be plenty I would think - should be long enough to let the thermostat drop out the compressor AC circuit. – relayman357 Jul 12 '22 at 15:31
  • 2
    ASCO 7000 series switches is what most of mine are. I think my smallest one is 200a , that may just be the feeder my largest is 4000a but this is a large facility ~1.2 million square feet with 6Mw gen set going to 8 eventually. – Ed Beal Jul 12 '22 at 15:41
  • 1
    Much obliged Ed! – relayman357 Jul 12 '22 at 15:50