UPS : Replacing inverter static switch with contactor

[Universe_in_Flux]

For one project, we have a requirement of 10kVA parallel redundant UPS with common bypass. Till now we have only purchased UPS which have static switches in bypass and inverter path. Recently one vendor has offered a UPS which has a static switch on bypass path, however they have provided contactor on inverter output side. our philosophy is to operate critical loads on UPS only, and switch to bypass only when there is any problem with UPS.
Technical specifications of offered model meets requirements of our specifications, however I am hesitant to go ahead with this topology of UPS since it is new to us. I am specifically worried about impact on availability of UPS, since failure of coil or actuating mechanism may lead to loss of load due to contactor failure on shorted position.
Do you have experience of using UPS which do not have inverter side static switch? If yes, Have you faced any reliability or operational issues with this? Could you please suggest a good reason to accept or reject this topology?

 

[Kiribanda]

There are several reasons for haing a static by-pass switch on a UPS.
Therefore, if the static by-pass switch is replaced by a contactor then it is
a concern.
Could you please upload a copy of the single line diagram of the UPS offered
to check at which location the contactor had been installed?

 

[Universe_in_Flux]

Hello Kiribanda,
Please find attached SLD. Since this is preliminary, vendor has not shown contactors. I have shown red arrows at locations where there should be static switches, but vendor has replaced them with contactors.

 

[waross]

Over many years I have found coil burnouts and general contactor failures to be rare.
Most coil burnouts are the result of chronic low voltage or. worse, a voltage below the pull-in voltage.
There are probably millions of installed contactors with multiples of 10,000 operating hours.
Your contactors may only be used for a few hours a year during maintenance.
I wouldn't worry about contactor failure.
But, do check the specs in relation to the load, present and possibly future.
Undersized contactors do fail more frequently.

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Kiribanda]

From the SLD, it seems that the STATIC switch of the Inverter line is going to be replaced by
a contactor. Most probably to cut down the cost of the UPS unit.
Normally, the Inverter output & the By-Pass line are kept in sync by UPS software,
and therefore, the sensing and transfer times of the static switch in either direction
(Inverter to By-pass line & vice versa) shall not exceed 4 milliseconds (ms).
Can it be achieved by putting a contactor on the Inverter line? I doubt, because contactors
take a longer time to energize/ de-energize.
Therefore, in my opinion, the manufacturer has to do a factory acceptance test to confirm this
4ms transfer time, if a contactor is introduced.
Also, the contactor should be,
1) At least AC4 duty rated
2) Rated for UPS overloads. (Typically the static switch of the Inverter line is rated for 125%-Continuous, 200%-One minute & 1000% 50ms)

 

[Universe_in_Flux]

Bill & Kiribanda,
Thanks for your valuable inputs
@ Bill : Since contactor is installed on inverter output & UPS will be continuously on; contactor will also be continuously ON. I agree that contactors are reliable, but reliability of static switches is comparatively better for UPS applications.

@Kiribanda : Vendor has agreed with max 4ms transfer delay in sync condition. I need fast transfer only when transfer to bypass is required during any fault in UPS. In that case, This UPS has static switch in bypass which will ensure fast transfer within 4ms.
Contactor will affect the speed of retransfer i.e. from bypass to mains. During retransfer, UPS will check sync condition and close contactor first. In this case, I am okay with higher retransfer delay as momentarily both sources are on, and there in no break in supply to critical loads.

 

[waross]

2) Rated for UPS overloads. (Typically the static switch of the Inverter line is rated for 125%-Continuous, 200%-One minute & 1000% 50ms)

In DOL motor starting duty, contactors experience 600% or more overloads every time the motor starts.
Many years ago I did some research into contactor pull-in times. Nothing was close to 4 ms. The slowest took about 100 ms. I don't remember the fastest but a lot more than 4 ms.
BUT, if this is energized at all times, you may not need 4 ms.


Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[Kiribanda]

I understand your acceptance for a delayed re-transfer. I agree too.
But during transfer to By-pass, in most of the UPS units, the 4ms total transfer time is
achieved by having two static switches (Inv line static switch to open and then By-pass
static switch to close).
Therefore, I wonder how 4ms total transfer time is still achieved with a contactor + static switch?

 

[Universe_in_Flux]

@Bill : I Agree, Bill. Contactor closing/opening time are not close to those of static switch. That's why we're alerted when we knew that vendor has replaced static switch with contactors

@Kiribanda : During Output overload / Inverter fault, my concern would be to protect healthy loads + UPS. During these abnormal conditions, bypass static switch will close within 4ms(tests show that this is close to 0ms when normal and bypass supplies are in sync). Since bypass has greater fault capacity due to absence of electronics (Except static switch), it will feed the required let through energy required to trip downstream MCB and protect rest of the healthy downstream circuits and UPS Inverter. Transfer time is the time within which alternate supply is available to critical loads.
I understand that you are considering contactor opening time too in 'Total transfer time', However again in this case I am not concerned with this as I expect the output voltage waveform within curve 1 of IEC 62040-3 dynamic performance classification 1. My load continuity is maintained & UPS is protected.