UPS Nuisantly Transfers to Bypass 5

[nightfox1925]

I am presently working on a problem with one of the UPS systems in the plant. There are two Single UPS systems (480-208/120) feeding two UPS buses (A and B) separately. The two separate bus circuits feeds a couple of 208VAC-24Vdc rectifiers which in turn individually feed as redundant power supplies to downstream PLCs, SIS, etc.

Of the 24Vdc rectifiers, the 24Vdc, 400A rectifier is the largest. The UPS systems always shifts to bypass everytime the UPS output MCCB is closed or this Rectifier is energized. The UPS alarm shows "Output Overload". We suspect that it was the downstream in-rush which is causing the UPS to nuisantly transfer. An idea by one of the contractor was to put a reactor on the input of the rectifier. A 5% impedance reactor was connected in series but still the UPS keep to transfering to bypass.

It is observed that the UPS output transformer secondary is wye and the downstream rectifier isolation transformer primary is delta...is it possible that the problem is also caused by phase mismatch between the UPS and downstream charger?

is there a possibility that the UPS settings may be too sensitive?

I am gathering all the data so I can post them here soon. But any help or guidance will be very much appreciated. Thank you.

 

[sibeen]

Actually, there may not be a problem at all.

If the rectifier does have a substansial transformer at its input then you wold expect a decent amount of inrush current when it was energised. If the size of the rectifier is a decent percentage of the size of the UPS then I would expect the UPS to transfer to bypass (reserve) when the rectifier is turned on. Of course, once the rectifier is up and running and the UPS has transferred back to inverter, then any power interruptons shouldn't cause any problems.

I have seen many UPS instalations where downstream transformers are used, for harmonic cancellation etc, and in the vast majority of cases the UPS will transfer to bypass when these transformers are turned on.

I suspect you may be chasing something that really doesn't need fixing.

 

[nightfox1925]

Hello Marke. I will be suggesting the UPS be tested further at site to determine what is causing it to sense an "output overload".

1. Energize the the 24Vdc rectifier with the converter isolated and only the input transformer is energized. If UPS do not transfer,
proceed to item 2 below.

2. Energize the 24Vdc rectifier with both input transformer and inverter connected but the 24Vdc loads and batteries isolated. If UPS
do not transfer, proceed to item 3 below.

3. Energize the 24Vdc rectifier with both input transformer, inverter and batteries connected but the 24Vdc loads isolated. If UPS
do not transfer, proceed to item 4 below.

4. Energize the 24Vdc rectifier with both input transformer, inverter, batteries and 24Vdc loads connected.

I will proposing these operational tests to determine which current in-rush is making the UPS transfer to bypass (as waross suggested) and then from there propose some mitigation.

By the way, I heard from Marke about putting a DC choke coil on rectifier's converter output as an option..(in case its a load in-rush as a culprit) to mitigate peak currents. Does anyone also have further explanation on this?

Thanks for all the support.

 

[sibeen]

nightfox, I take it from the choke suggestion that the rectifier is a standard six pulse style thyristor controlled rectifier. Placing a choke on the output of the rectifier is equivalent to placing a choke at the three phase input.

Draw it out. In a six pulse rectifier, two of the thyristors are turned on during each switching cycle. Now you can either put a choke at the input of the rectifier and therefore have a choke followed by a thyristor to the DC bus; or you can put the choke on the output and then have a thyristor followed by a choke.

Electrically equivalent

 

[nightfox1925]

To update everybody, we will still go to site tomorrow to get all the infro we need on these equipment and will gladly post the diagrams here.

I have said before that the UPS output transformer is wye and the downstream rectifier input transformer is delta.

Is possible to mitigate the downstream in-rush by placing a specially manufactured wye-delta transformer so designed to have a wye primary winding (in consideration to the downstream in-rush)which will draw a lesser in-rush below the overload capability of the transformer?

 

[nightfox1925]

Hello gentlemen,

The UPS in question is a 100kVA, 208VAC, 288.3A, 0.98 P.F. lagging, 103.8kVA, 3Ph, 3W input and 208VAC, 277.6A, 100kVA, 0.8P.F. lagging, 3Ph, 4W ouput. The UPS rectifier component is rated for 360Vdc to charge the UPS batteries and feed the UPS inverter. The Inverter has an output overload capability of 105%~125% for 10 mins. and 126%~150% for 10seconds with voltage regulation maintained. The converter and inverters are insulated gate bipolar transistor (IGBTs). The UPS is a Mitsubishi model 9700 series UPS.

The UPS feeds two Staticon rectifier units:

1- 208VAC, 60A input and 24Vdc, 500A output Staticon
rectifier unit, delta connected primary input X'former
with series saturable reactor to control the x'former
input.

1- 208VAC, 7A input and 24Vdc, 50A output Staticon
rectifier unit, delta connected primary input X'former
with series saturable reactor to control the x'former
input.

It was observed that everytime the 500A staticon charger is energized, the UPS transfers to bypass. We are looking at it as an in-rush problem.

The solutions being considered is to adjust the overload setting of the UPS. The present setting was verified to be at 100%, hence measuring the 500A rectifier in-rush and adding it to the other 208VAC continuous loads will determine the %overload to overrride (if the setting allows it). My question, will this endanger the electronic circuitry and protection of the UPS?

The other solution is to consider an isolation transformer, delta-delta between the UPS output and the rectifier input. The isolation transformer may be designed with an impedance and magnetic property that will absorb or damp the in-rush from the rectifier input through it secondary winding and still maintain an acceptable value of primary in-rush at its primary winding. The purpose is also to provide electrical isolation between the UPS output circuit and the rectifier input circuit. Will this probably work?

I appreciate any comments and guidance that would help us resolve the problem. Thank you.

 

[waross]

Settings. I would not like to exceed the 150% 10 second rating. Have you checked the actual magnitude of the inrush?
If the series reactor didn't work, don't let the same person calculate the transformer specs.
More on the transformer;
Are you able to check the action of the saturable reactor? Is it providing any current limiting at startup or is it allowing full conduction?
Adding a second transformer to avoid the inrush of the first transformer may be counter productive. The added transformer will probably have a similar input unless it is wound with a high resistance primary. That will cost you a lot of heat losses and give poor regulation.
I'll through this out for comments from the other regulars.
It may be cheaper and easier to follow another course of action.
How about using resistors to limit the surge and then shorting out the resistors. Either an electro-mechanical contactor, back to back SCRs, IGBTs or some other electronic device.
It may be time to rent the equipment to record the current profile during turn on. Remember the difference between peak and RMS.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[nightfox1925]

Hi Waross,

The saturable reactors connected in series to the staticon rectifier input transformer are part of the rectifier package. These reactor is a series synchronous switch (analog to 2 anti-parallel SCR's). The power control is derived from the concept of equal ampere-turns. These reactors are used as to regulate the DC output of the rectifier.

Do you have a typical model of measuring instrument (preferably portable) that can measure the in-rush within say less than 6 cycles?

We will conduct an in-rush measurement next week. If you have a contact of a third party test group within the edmonton or within alberta, we will be happy to get in touch with them.

Did you say resistors as an option? Is this gonna be connected temporarily in series with the line? If you won't mind sharing the concept and a contact person or link where I can start with...

Thank you.

 

[Jasonarc]

The 500Amp staticon charger has an input iso-tx that is about 20kVA right?
The inrush from an iso-tx is from 10 to 16 times nominal (so about 500-600A).
The UPS practically has a short circuit on the output during transformer energisation.
The Static switch SCRs will fire on a quarter cycle output undervotage detection.

I've had 500kVA 600V output UPS systems transfer to bypass upon energising a 75kVA distrib ution iso-tx.

If your UPS has a seperate bypass feed to it, try opening the bypass feed breaker. Now the UPS will not see the bypass availibilty and will leave the inverter to current limit the inrush rather than let utility take the hit.

You will probably see an output UV alarm but load will not be dropped.

The IGBT inverter is super fast and will react to the inrush, but if the bypass is available it will transfer the load to and from utility.

 

[waross]

Do you have a typical model of measuring instrument (preferably portable) that can measure the in-rush within say less than 6 cycles?

I'll let some others answer that. There are a lot of people here with more knowledge of test equipment than I.

We will conduct an in-rush measurement next week. If you have a contact of a third party test group within the edmonton or within alberta, we will be happy to get in touch with them.

Did you say resistors as an option? Is this gonna be connected temporarily in series with the line? If you won't mind sharing the concept and a contact person or link where I can start with...

The idea is to limit the inrush by connecting resistance in line. A short time later the resistance may be shorted out. The transformer will now be energized with two smaller inrushes instead of one large one.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[sibeen]

You can use the resistor in-line method to limit the inrush current, with a few caveats.

Normally with this type of circuit the in-line resistors are shorted out by a contactor or motorised circuit breaker shortly after energisation of the transformer.

With the rectifier you would need to be sure that the rectifier doesn't turn on and start drawing substansial current before the resistors are shorted out. Otherwise you'll quickly find that your resistors are now charred stumps which have released a considerable amount of smoke.

I will reiterate a previous statement.

I still think you are chasing and trying to fix a problem which 'really' doesn't cause any issues with the site reliability. Once the transformer has been energised the UPS should then operate correctly during all grid outages etc. There are countless sites around the world where it is accepted that the UPS will transfer to bypass as it first energises a load and then operate 'correctly'. Of course you still need to test that the UPS operates as expected, but that is just standard commisioning practice.

 

[davidbeach]

Come on guys, this is way too much effort to solve a non-problem. The UPS is doing exactly what it is intended to do. UPS systems have a self supplied fault current capability below 2 per unit and go to bypass for everything above that level. On the assumption that this inrush condition only occurs occasionally, why saddle the system with components, and associated losses, that are only needed during inrush and provide no benefit at any other time.

When the rectifier is started, just let the UPS go to bypass for the few cycles/seconds the inrush lasts. Months/years later when the rectifier is started the next time let the UPS go to bypass again.

Just because a "problem" can be solved doesn't mean that it needs to be solved.

 

[nightfox1925]

I understand davidbeach's and sibeen's logic on this issue. However, I also explained this to the client but it seems they are still insistent on trying to mitigate it. It may sound absurd on our part as electrical practitioners but this is not how the operations people look at it. As I agree that the bypass transfer dur to in-rush is part of the UPS designed operation, it still gives off a nuisant alarm that makes the operation folks at the control room jumping out on their seats and yeah they sure do not like the fact that this happens and are somehow concern on the fact that it may create a wrong impression aside from causing some panic to them everytime an alarm happens.

While I also agree that placing some mitigation also causes additional maintenance and "point of failure", I do intend to present them the disadvantages of making external mitigation and what I gather here are also essential to help or guide me the way (or a better way to explain to them).

The specs read as "The Inverter has an output overload capability of 105%~125% for 10 mins. and 126%~150% for 10seconds with voltage regulation maintained."

If I set the UPS overload setting to 150%, does this mean that the UPS algorithm will pickup automatically in 10seconds? Is the time to pickup already inherent to the UPS and all its protection components (fuse or breakers) already coordinated to trip within 10 seconds should a continuous overload of 150% persists (more than 10 seconds). I would like to have an understanding on this since the vendor supplied specs does not explain this? Anyone's idea is appreciated. Thank you.

 

[davidbeach]

What it means is that the UPS won't exceed 150%, and after 10 seconds it will back off to 125% and after 10 minutes it will current limit at 100%. The inverters just don't have much overload/fault capability.

If this is really in inrush condition, lasting less than a second, why not just delay response to the alarm? Make the control system wait until it has had the bypass alarm for more than 1 second (or what ever is necessary) before annunciating that alarm point.

 

[nightfox1925]

By the way Waross, is the sizing of the resistor (or reactor) for in-rush the same as if sizing a limiting reactor for short circuit limitation?

Taking the system P.U. impedances and then determining the P.U. R (or X) with given specified maximum allowable in-rush? What company does this type of design? Postglover supply this type of resistors?

Sibeen, in your statement for the in-line resistors, do you mean that the converter should be fitted with an input circuit breaker so that the converter be isolated when the rectifier input transformer is energized.

Waross, what is the disadvantage of setting the overload to say 150% to override the in-rush (if possible)? Trying to understand what the specs read out.

The UPS technical guy was indicating that placing an isolation transformer in between will provide mitigation and circuit isolation...however, it will be special design as I perceive it. I do hope you can expand this more.

Bottom line, if such suggested mitigation are having a lot of counter hickups that may out weigh the advantages, I will appreciate much for any further comments and guidance that would help my justify to live up with the situation and provide a written procedure for the rectifier start-up.

 

[waross]

For a proof of concept, I would buy some replacement element coils for a residential clothes drier and cut and parallel them to get the resistance I needed.
As for settings, make sure that the UPS settings are the maximum allowable.
If you go with the resistors, I would assume a value for the energization surge (200%?) and use enough resistance to set the voltage drop to 50% at that current value. If the UPS is limiting the current to 150%, then the effective load impedance is the parameter of interest.
That way you will have two surges, one on energization and the second when you switch out the resistors, but both will be 50% of the present surge.
I'm concerned about a special design transformer. If it works I expect that it will waste a lot of energy and reduce the useful output of the UPS.
BUT we still should know the cause of the inrush. If it is a reflected load from the DC bus, and is real current, you want resistors limiting it. If it is magnetizing inrush it may be highly reactive and respond to reactors.
As it did not respont to the installation of the reactors, it may be a real current inrush or the reactors may have been improperly sized. We should know the quadrant of the inrush current before spending any money on transformers reactors or resistors.
I would prefer to solve a problem of this type with an adjustment of the present parameters. Transformers, reactors, resistors contactors. This seems kludgy if the system can be adjusted to make the customer happy as it is.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[nightfox1925]

Hi david,

we could have loved to delay the alarm response. The problem is that the UPS vendor do not want to change their UPS software algorithm..according to them, the alarm response is fixed as per their manufacturing standard. For the overload seeting, if I set the UPS overload say, 150% and then a continuous overload happens at 150%, does this mean that the UPS will start to pick up and then after 10seconds, it will resort to bypass? I may have mis-interpreted the term current limit...how does it actually work. (I apologize but, the manual we received does not explain this term further).

Hi waross,

What do you mean adjustment of the present parameters..do you mean the "overload" adjustment or maybe alarm time delay? Somebody even mentioned decreasing the UPS sensor gain (will this work also)?

Going back to the series resistors. Did they call these "buffer" resistors before? Is the concept actually putting the resistors in series while energizing the transformer so as to create a voltage drop on the line, consequently reducing the voltage input at the transformer primary and avoid saturating the transformer primary to decrease the starting flux generated in the core based from formula:

Flux (max) = 2Vmax / (2*pi*f*Np)

Is my understanding valid?

If ever an isolation transformer is provided to isolate UPS output circuit from the rectifier input circuit, then the isolation transformer should have a designed primary resistance (+ reactance) so as to develop a lower in-rush. The set-back of this is that, an increase primary resistance will increase copper loss on the transformer and may also create poor voltage regulation on the transformer.

is my understanding on your previous post correct? If there is a downstream in-rush on the transformer secondary, will this secondary load in-rush add up (via transformation ratio) to the primary magnetization in-rush?

 

[nightfox1925]

In addition, we will conduct in-rush measurements tomorrow and simulate the following scenarios:

With the UPS output breaker on:

1. Energize the downtream 500A, 24Vdc rectifier with its
associated 24Vdc loads and batteries disconnected.

2. Energize the downtream 500A, 24Vdc rectifier with its
associated 24Vdc loads disconnected and batteries
connected.

3. Energize the downtream 500A, 24Vdc rectifier with its
associated 24Vdc loads and batteries connected.

Check UPS alarm and pickup level. Keep everybody posted about this. Thank you.

 

[waross]

Hi;
First, we should know the source of the inrush. Is it reactive or real? Different possible solutions.

Parameters, I meant any adjustments that may be made to the UPS. This would include anything that may be done with the rectifier unit to lessen the inrush.

Is the concept actually putting the resistors in series while energizing the transformer so as to create a voltage drop on the line, consequently reducing the voltage input at the transformer primary and avoid saturating the transformer primary to decrease the starting flux generated in the core based from formula:
Yes. I'm not sure about the formula and I am away from my library and can't check.

If ever an isolation transformer is provided to isolate UPS output circuit from the rectifier input circuit, then the isolation transformer should have a designed primary resistance (+ reactance) so as to develop a lower in-rush. The set-back of this is that, an increase primary resistance will increase copper loss on the transformer and may also create poor voltage regulation on the transformer.
Yes. You are combining a reactance and a resistor in one unit. The losses will be permanent and the voltage regulation will be poor.
The disadvantages of the transformer are:
The extra losses and the poor voltage regulation.
It cannot be bypassed like the resistors or the reactor.
The losses will subtract from the available capacity of the UPS.
The advantages of the transformer are:.
No switching.

The UPS technical guy was indicating that placing an isolation transformer in between will provide mitigation and circuit isolation...however, it will be special design as I perceive it. I do hope you can expand this more.
The UPS guy is really the one to answer this question.
BUT, if he will gaurantee that the transformer will work that is a plus.
A heads up here. It is easy to lose sight of the overall picture when you are wrestling with a problem like this. If the next step may be buying a larger UPS to handle the losses of the transformer, consider going all the way to a UPS large enough to handle the inrush.

I once had an issue with some small control transformers. The transformers would blow the factory installed fuses if the load was connected when the transformers were energized.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[nightfox1925]

To all, can anybody explain how the "current limiting" function of the UPS works...for my benefit of understanding.

What if we leave the system as it is, and then for some reason, the bypass input supply was out of service. can we still be able to energize the staticon charger through the converter + inverter UPS path? Will it withstand the in-rush and then normalize and sends off an alarm during the event of transient overloading? I just want to check out that the inverter will still assume the transient in-rush load.

Overload capability of 150% at 10 seconds, is the 10 seconds count inherent to the UPS even if the UPS is set at 100% overload?

 

[nightfox1925]

Waross, is sizing the series reactor to mitigate the transformer in-rush similar as if you are sizing a reactor for short circuit limitation?