Generators vs Computers

[augie47]

I was asked a "broad" question and without any specifics it may not be possible to answer this question, but I thought I would see if there were any opinions.
A correctional facility has a fair amount computerized equipment and they are experiencing
problems when they switch to their generator supply. They have been advised the problems are due to spikes and have received suggestions to install: (a) additional surge protection, (b) UPS systems, and (c) isolation transformers.
Anyone have any input as to which route would provide the best solution.

 

[dereckbc]

Use a Dual Conversion UPS like telecom and data center does. The UPS batteries are designed to carry the load for about 15 to 20 minutes to give a generator time to come up on line and take over.

 

[mgookin]

I vote for UPS as the solution; that's what it's for - to provide stable power supply to the load during times of adverse line side supply.

 

[kwired]

Problem with a generator is voltage will see a more drastic change whenever load changes compared to utility power. The utility generator has much larger "flywheel" and the change of load at one customer (some large industrial loads maybe an exception) doesn't have much impact on the voltage output of their generator, but smaller onsite generators this same load is a much higher percentage of the generators capacity it does create a significant change in output voltage, even if only for a short duration. This happens when adding or removing load from the generator, and is possible for governing devices to overcorrect making even more fluctuation when there is a sudden change in load.

A good UPS is probably the best way to go, it will draw from the UPS battery when needed for voltage sags, and should filter out over voltages as well leaving a fairly clean output to the supplied equipment.

 

[junkhound]

Yep, it certainly is a 'broad question'

Well, UPS would take care of problems if sized correctly, even without knowing the real cause of the 'problems' or even what type of problems - note the 'sized correctly' implies that the root cause of the 'problems' is known. One would need the specifics to correctly size an UPS.

A statement such as "have been advised the problems are due to spikes" is so generic that it implies the 'advisor' is incompetent.
e.g what type of 'spikes', voltage or current? Magnitude? Cause and effect? etc.

Did anyone even bother to put a scope or spectrum analyzer on the power with all the 'computers'?
For instance, what is the harmonic content - high harmonics due to non PFC input computers are well known cause of marginally sized generator instability.

Surge protection - 99% sure would not be applicable to this problem even with the limited details.
Isolation transformer - a delta wye isolation transformer would short out and burn off the triplett harmonics in the delta primary, but would need to be sized well above just the generator rating due to heating.

 

[ATSman]

CT ATS

CT ATS

All good stuff being discussed here. If I might throw into the mix: closed transition (CT) ATS. Most ATS's are open transition type: break before make. Sure most have the inphase monitor feature (synch check) that only allows transfer when the two sine waves are in a certain threshold window (usually less than 10 electrical degrees) but there is still a voltage disturbance to the critical load. With CT it is a make-before-break transfer of load where the gen is actually paralleled with the utility. In this case the transfer is "seamless" and most server front-ends are not affected during transfer, especially ones that are not backed up by a UPS. It needs to be mentioned here that this design does not help during a normal power failure: you still need UPS backup during transfer to gen. It only function on re-transfer to normal power or during a routine test of the ATS (hot-bus-hot-bus transfer.) when both sources are available.
I am surprised that the original design did not include UPS backup when this "computerized equipment" was installed. Or maybe it was and it's just that the batteries have not been maintained and are failing.

 

[GoldDigger]

If the loads are supplied and wired as 120-0-120 or three phase wye, you may want to specify an OT ATS which breaks the phase lines before the neutral and connects the alternate source neutral before the corresponding phase conductors.
This prevents an unbalanced load from seeing an open neutral condition momentarily during transfer.

Tapatalk!