Transfer switch delay...

Re: Transfer switch delay...

fca1 wrote: »

The secondary power source will always be on before the change (GRID for now), so will it worth to put some cut load before change source so that my engines will stop and restart from zero ?

You'll probably want to put a UPS on the critical loads like computers, sensitive electronics and lighting that you want to stay on. Fast transfer switching insures almost zero chance that your backup power source is in phase with the primary, meaning huge surge amps to your running motors. Three phase motors usually can handle this in stride. But single phase capacitor start motors with a separate start winding can suffer severe damage if the motor's rotor slows to the point of engaging the centrifugal starting capacitor switch while the rotor and field is in one phase state and a different phase state power source "hits" it.

In the absence of a phase synchronized backup power source, you will need to use timed delay switching. With timed delay systems you will have to time the delay so refrigeration compressors can restart without being rotor locked due to excessive trapped pressure on the compressor head. While locked rotor amp draw on refrigeration systems might not bother a virtually unlimited backup source like the grid, it will affect standby generators because they have limited surge capacity.

Typically with standby generator systems the UPS will keep critical circuits alive while the generator is started and warmed up, and then the load transfer takes place. There are two types for standby generators so it might be worth going over that, as I suspect you are designing your system to be eventually used with a generator instead of the grid:

- facilities that operate human life support systems, like hospitals, require standby generators in dual redundant mode. Those type start two generators (typically over 1.0 MW each) that frequency and voltage synchronize with one another. Both generators synch with the UPS and both are required by law to be started, online, sync'd and accept full campus load from the UPS within 10 seconds of a power outage. These types of paralleling switch gears and gensets are very, very, very expensive. You don't even want to know what they cost. Suffice it to say we're talking in excess of $5 million just for the UPS that keeps surgery, ICU, etc powered up while the standby's are coming online.

Residential and other normal backup systems used timed delay switching. Again, a UPS can keep critical systems like computers and elevators powered up during the switch. The transfer gear waits for usually at least 10 seconds to evaluate whether or not the primary power is coming back. Then it starts the generator and warms it for usually 2 minutes or so (depending on ambient temp). This delay automatically reduces the surge load for things like refrigeration compressors because it lets them "bleed down" so their rotors aren't locked. Then the transfer gear makes the switch and the backup power source accepts the load.

I don't know the particulars of your loads and maximum surge requirement. But in the absence of a sync'd backup source I'd vote for a timed delay switch gear to prevent damage to your single phase motors and overload of your backup source if you are running refrigeration systems.
--
Chris

Re: Transfer switch delay...

BB. wrote: »

For refrigerators, it is actually part of the design requirement that if they get a short power hit--The motor stalls on restart, kicks out the thermal protection device, cools (and pressure balances) and restarts in 5 minutes or so (after motor/motor starter) cool down.

It depends on locked rotor amps for refrigeration systems. If this transfer switch is being designed to be used with a generator eventually, even the smallest window AC units will kill a 4 kVA generator dead in its tracks with the compressor motor stalled. With grid power it will never be a problem. But with generator transfers the timing for restart of refrigeration compressors is pretty critical to not overloading your generator and kicking it offline.

If you have several refrigeration systems going (like 'fridge, freezer, AC/heat pump) and they all happen to be running at the same time when the transfer is made, the results can be disastrous for standby generators. With a very fast transfer (<16 ms) the compressor will more than likely stay running and only take an out-of-phase "hit". But I've seen Outback FX-series inverters (that have a pretty fast transfer time when switching to gen power) running a window AC unit that wipes the generator right out when the transfer is made. Then the Outback takes over and tries to restart it, the Outback kicks out due to overload, and everything goes dead. And that's with a fast <16 ms transfer.
--
Chris

Re: Transfer switch delay...

BB. wrote: »

I don't know how fast the rotor field decays in an induction motor

-Bill

You can't really know it either, because it depends on the particular motor and the application.
Worst-case scenario: the motor has enough inertial mass to retain the magnetic field and operate momentarily as a generator, forcing power back against the inverter. Not a problem with grid power which has so much electrical inertia almost nothing can affect it, but an inverter might object to suddenly finding an out-of-phase AC input on its output if only for a moment.

Used to have trouble like this with 240 VAC motors being run through step-up transformers that weren't phase synchronized. Short power outage would cause back-feed through the transformer. Most of the time it didn't cause trouble, but some more sensitive things attached to the same wiring ...

On the whole, the 'blip' of out-of-phase power from switching from one source to another won't be noticed since it will be gone in one RPM. Compressors would be most likely to 'feel' it, but even they tend to have enough inertia to keep going. Biggest problem would be if a motor tried to start just at the switch-over. Highly unlikely.

Re: Transfer switch delay...

Or in simple words: the bigger the motor the bigger the potential for problems.
(Not coincidentally larger motors are the ones that tend to have both start and run caps as they benefit the most.)

Re: Transfer switch delay...

inetdog wrote: »

Since a 16 ms transfer is only one 60Hz cycle, chances are good that the compressor is still spinning, but that the out-of-phase current surge is what is killing the generator. Not sure how you could avoid that, other than by stopping and waiting.

I know it's the out-of-phase relationship of the source and load when the transfer is made in an Outback that causes this. With Xantrex inverters you never see the problem because the inverter syncs with the generator before connecting it.

I just wanted to point it out because the transfer switch being designed here by the OP will probably work perfectly fine when transferring loads from the inverter to grid with a fast switching. But I had understood that the grid backup was only for testing, and temporary. I guess I maybe assumed that the transfer switch would eventually be used for a generator. And generators, because of their limitations, pose some different challenges for transfer switch design.

Quite interesting project. Not many folks build their own transfer switches
--
Chris

Re: Transfer switch delay...

Indeed, working with the big stuff amplifies the problems. The big Cummins Quantum K's can deliver 2.5x their prime rating (slightly over 6,000 hp) for up to a half hour. In the old days we used to just throw the fuel at 'em and turn the sky black to get it. But then EPA came along and said you diesel guys are a lot easier to regulate than the coal powerplant guys, and we're looking for something to regulate. So only coal power plants can smoke. No more smoke from diesels because we said so. So in engineering it was like, "Hey boys, we's bein' regulated - cut the top end fuel rate back a little bit on them G-drive engines!"

In the old days of Big Smoke we could do 'er with one genset. But after the EPA No Smoke Directive we had to start using dual sets and cut the top end back on 'em.

Still, in smaller residential applications the things I've seen that can cause problems are things like well pumps, bench grinders, ceiling fans and refrigeration systems. No, the fan motor in your microwave, or your washing machine motor probably won't care.
--
Chris