UPS reactive power supply ? 3

[tomad]

Per manufacturers' data, the output power for a UPS is limited by its apparent power output [kVA] and its real power output [kW].
EX. for a 100 kVA rated UPS, its real power output is listed as 80 kW. That would be equivalent to assume that the UPS can deliver a maximum of 60 kVar. Where does that reactive power come from ? Adjustable capacitors on the output ?

 

[sibeen]

tomad, you should read what I write:

"In the UPS these currents will flow in the inverter of the UPS. As mentioned in one of my posts, this is why the switching devices in the inverter have an anti-parallel diode. "

The reactive currents DO flow in the inverter, and this is what the anti-parallel diode is for - to provide a current path.

You don't need capacitors on the output of the inverter, except for a small amount required for filtering.

"Some UPS systems also place extra capacitors, above and beyond what is needed for filtering, on the output of the inverter."

Again the key word is SOME. Most manufacturers have gone away from this topology as the load profile has changed due to better power supplies.

 

[slavag]

Bill Great!!!!!!!!!

Guys, please don't forgot, lot of our formulas are only math, not some phisical issues. Reactive power, symmetrical components, harmonisc, etc.. are math.
We have only current, votages and angle between them. It's all.
Best Regards.
Slava

 

[rbulsara]

Star for Bill for his last post!

 

[tomad]

I agree that what finally matters are currents and voltages. But I think that in the operation of each device, whatever the brand, manufacturer, etc and the registered or non-registered trademarks used, in the end, everything can be traced back to basic laws of electrical engineering (applied physics).
So, as mentioned in my previous posts, the question arose from a practical situation: the load to be supplied by a UPS had a power factor which exceeded the indicated output power factor stated by the UPS manufacturer, a fact that, in turn, questioned the suitability of the UPS for the envisioned load(by a blank statement that something will be overloaded). Usually, a better power factor is desired everywhere, for obvious and clearly stated reasons. It seems that in this case, a lower power factor is preferred/acceptable (i.e. 50 kW at 100 kVA would be considered OK), a fact for which we tried to find an explanation.
If anybody has been involved in factory load testing of a UPS, I would appreciate if he/she could send me a set of measurements. I hope that those measurements will not be treated as trade secrets, since those tests could be reproduced by a user with proper equipment. The results, if properly gathered and interpreted, have to match the ones guaranteed by the manufacturer.

 

[alehman]

Even if the load is within the published kW and kVA ratings, if the power factor is not within rating it seems like you should consult the manufacturer. The UPS may not be able to handle the poor power factor for reasons other than capacity (I'm not an expert). It may be able to handle load with lagging power factor, but not leading. I don't know, but you should ask.

As for the math vs. reality discussion, I wouldn't be so quick to dismiss the math part as merely a tool. Most of the mathematical entities we use to are readily observable. Real and reactive power can be measured by physical meters. Some utilities install induction disc VAR meter for power factor billing purposes. Where reactive power comes from and goes to is a matter of semantics, just as with real power but less intuitive. Normally inductive loads are defined as a 'consumer' of reactive power and capacitive loads are 'producers'.

Harmonics can be heard by ear (ever listen to a transformer supplying a large rectifier load?) Negative sequence current causes motors to overheat because it is rotating in the opposite direction as the rotor, etc. Mathematical tools are not always just tools.

 

[sibeen]

So, as mentioned in my previous posts, the question arose from a practical situation: the load to be supplied by a UPS had a power factor which exceeded the indicated output power factor stated by the UPS manufacturer, a fact that, in turn, questioned the suitability of the UPS for the envisioned load(by a blank statement that something will be overloaded). Usually, a better power factor is desired everywhere, for obvious and clearly stated reasons. It seems that in this case, a lower power factor is preferred/acceptable (i.e. 50 kW at 100 kVA would be considered OK), a fact for which we tried to find an explanation.

tomad, a lot of this just comes down to what UPS manufacturers have been designing for the last 30 years or so. Most manufacturers haven't changed which is why you still have the ratings of 100kVA @ 0.8 P.f.

When the system is designed the power sections, rectifier, boost section (if used), batteries, inverter and all associated cabling is then calculated using the real power, in this case 80 kW, that the UPS has to supply. If the output power exceeds this 80 kW you then begin to overload these sections of the UPS.

It is purely a cost equation.

 

[tomad]

Sibeen, this was exactly the path we tried to reason. We proceeded by eliminating parts that had to be sized based on the kVA rating (input & output cables, rectifier, etc). We eliminated the battery, since that part is an emergency source that kicks in (normally) only when losing the power grid, so the main path of power does not run through it. This brought the next question: if real/active power is supplied via the inverter (assumed sized based on the kW rating) where comes the reactive power, etc, etc... that triggered the whole chain of posts.
We tried to contact manufacturers for clarification, but were usually directed to (local sales)representatives whose main task is to market a product, technical hotlines experts were unfortunately not more helpful, so here we are.

 

[sibeen]

tomad, have a look at the PDF I've included. It shows what an IGBT in the inverter looks like.

The red arrow shows the current path of current that is in phase with the voltage - the IGBT is turned on.

The black arrow shows the current path for the out of phase current (reactive) and shows the current path for when the IGBT is turned off.

 

[waross]

The power comes from the inverter. The current from the inverter, as well as supplying the load, flows in the inductive element of the load. Because of the nature of an inductance fed from AC, the portion of the current that flows in the inductance is slow to start. Now, an inductance is capable of storing energy for a short time. Go back to basics.The inductance tends to keep the current flowing. As a result, the current is trying to flow after the IGBT turns off. Sibeen has shown the circuit and particularly the diode that allows the current to continue flowing. When an AC current flows through an inductance, the current lags behind the voltage. This is called a phase shift. We need to be able to predict the results of this phase shift. It is quite easy to calculate using power factor and an imaginary entity called reactive power.
It is really time to get a third year electrical apprentice textbook and brush up on inductance and phase shifts.
That this issue has been fought so long may indicate a lack of necessary basic knowledge.
Don't fight the problem!
Hi alehman:
A couple of friendly comments.

Even if the load is within the published kW and kVA ratings, if the power factor is not within rating it seems like you should consult the manufacturer. The UPS may not be able to handle the poor power factor for reasons other than capacity (I'm not an expert). It may be able to handle load with lagging power factor, but not leading. I don't know, but you should ask.

That is a valid question. With current flowing when there is no voltage, and at times the current and the voltage flowing in the opposite direction, how does the lagging current get through the switching devices? The diodes that sibeen shows are passive and always available to provide a path for the reactive current. These will have a current rating, but as long as you stay within the maximum current determined by the maximum KVA rating you should be within the current rating of the diodes.

As for the math vs. reality discussion, I wouldn't be so quick to dismiss the math part as merely a tool. Most of the mathematical entities we use to are readily observable. Real and reactive power can be measured by physical meters. Some utilities install induction disc VAR meter for power factor billing purposes.

The reactive current and the voltage are 90 degrees out of phase. The product of two quantities acting at 90 degrees to each other is zero.
For many decades power companies measured the imaginary reactive power with off the shelf kilo-watt-hour meters. There was no such thing as a kilo-var-hour meter. They fooled the meter with a set of coils with low resistance and high inductance. These little phase shift sets were used to shift the phase angle of the voltage supply to the voltage elements of the kilo-watt-hour meters. They were kilo-watt-hour meters and thought that they were measuring kilo-watt-hours. They were fooled into indicating KVARHr consumption.
Until the advent of digitizing and computer algorithms, all VAR meters that I am familiar with used some technique to fool the meter by shifting the phase of the voltage 90 degrees.

BUT, when you have a phase shift, it may take 100 amps to do the work normally done by 80 amps.
The work was called watts.
The product of the current and the voltage was called Volt-Amps.
How to explain the discrepancy? When the phase angle was observed on a dual trace scope, it was noticed that the phase angle and the watt and VA vectors formed two sides of a right triangle. Pythagoras came through and they were able to reconcile the third side of the triangle. Because it was a product of the reactive current they called it reactive power. But power does work and reactive power does no work. It can't, it's imaginary.
BTW a star for siben.

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

 

[alehman]

Sibeen,
Thanks for the sketch. Do you think we can safely assume that all UPS's use the diode scheme you show? I'd still want to run it by the manufacturer if possible.

Bill,
I agree, except perhaps the last two words. We call reactive power "imaginary" based on the mathematical tool used to represent it. It is imaginary (in the Layman's sense) only in that it does no useful work, unless producing heat is useful for some reason.

Using your Pythagorean analogy, think about pushing a car up a hill. You are pushing parallel with the hill, along the hypotenuse of the triangle. That force is translated to a horizontal force component against the hill and a vertical force which takes some of the weight off the tires. These forces can be represented mathematically by a horizontal ("real") component and vertical ("imaginary") component. Both force components are very real, but only the horizontal does any useful work (moving the car forward).

Sorry for the off-topic diversion.

Alan

 

[sibeen]

alehman, all PWM style inverters, whether for a UPS or a motor drive, will have the diode in the inverter switch; whether the switch be an IGBT, mosfet etc.