3 phase motor current imbalance test. 6

[dynets]

Does anyone know how I can easily create a phase current imbalance on a 3-phase AC motor? I am testing a product that claims to protect for motor current imbalances and I'd like to see the point (how much of an imbalance) at which the protection works. I would like to be able to create current imbalances from 5 to 60% while running an actual 3-phase 460V motor.

I was thinking about using a 3 phase variac and isolate one of the three coils. This will allow it to be independently adjusted from the other two to increase or decrease one phase voltage and create a current imbalance. But I'm not quite sure the best way to connect the coils. Also I thought I'd have a problem with inrush when I start the motor. PowerStats don't have much of an overload capacity unless they are oversized. I don't want to have to go buy a large 3-phase powerstat unless I have too (times are lean).

I was hoping there was a motor designer out there that might know of another way, perhaps with large power resistors or line reactors. Or maybe somehow manipulating or miss wiring the nine motor leads.

Any suggestions would be appreciated.

 

[electricpete]

A really cheap way of creating a variable impedance is using two metal plates submerged in a bucket of very salty water. Apply a terminal to each plate and mount them at an adjustable spacing using plastic (insulating) studs. Closer gives less impedance, further gives less impedance. I think what is created is an RC impedance predominantly resistive. The bucket of water will heat and slowly turn to steam. Be prepared to refill it and take due precaution for personnel/equipment safety.

You could try the same thing with pure water but you'd need very big plates to get any sizeable current through it even with few hundred volts across it. Also then you'd be more concerned about resonance.

Try putting that in series with one of the phases.

 

[nbucska]

To E.Pete:

Don't use aluminum electrode -- use iron in borax or NaOH
solution. <nbucska@pcperipherals.com>

 

[busbar]

There are a couple of ways to apply a powerstat that do not require having full VA-capacity of the motor.

Is the 480V service 4-wire wye?

What is namplate current and hp of motor?

 

[electricpete]

Nick Bucska - I'm interested... Why borax or NaOH solution? Isn't conductivity of the solution the relevant factor?
I don't think electroplating is an issue with ac.
Are you concerned about scaling if boiling occurs on the electrodes? Is there a long-term corrosion issue you're trying to address?

 

[nbucska]

Hi E.Pete:

Of course you don't want to corrode the electrodes too fast
and you don't want e.g. chlorine ( from NaCl ) or strong
acids to develop. Generally you don't want any gaseous
byproduct. Carbonates e.g. Na2CO3 would loose the CO2
and be converted to NaOH with increasing conductivity.

Borax was used in electrolitic caps and in the aluminum cell (see my posting about cheap rectifier)

<nbucska@pcperipherals.com>

 

[jbartos]

Suggestion: Three single phase rheostasts rated at 600V could be used. If you use other approaches, e.g. mentioned above, there is a need to observe safety since 480V is dangerous.

 

[dynets]

To EPete: I don't think I'mI daring enough to try the liquid approach... Thanks anyway

To: Busbar
Thanks for the response. I believe the power system is a 480V wye. I've measured each phase to ground and found 277V. I have many different test motors available: 5 to 30 hp. I originally wanted to use a 15 hp 18.5A at 460V but my plans may have changed… see below.

I just found a 3 phase 230V powerstat in storage. It has a rating of 30amps. I also have several isolation transformers (i.e. 230/460V). So I'm now thinking of using a transformer to step down the 460 to 230V and apply it to the input of the powerstat and use a smaller motor wired for 230 volts (5 hp 13.6 FLA). I would wire each output phase of the transformer (230V) to each powerstat coil and tie the N to the common of all three powerstat coils. The wiper from each coil would be tied to the 3 motor leads. I think this will give me a variable 0-230V to the motor leads. If I start out with all three phases of the powerstat at max (230V) I should be able to lower the one of the phase voltages by adjusting one of the powerstat coils. What do you think? I know a small voltage change will make large phase current change. The inrush when the motor is started is still the question. I could use a small motor sol that 6X motor current would be less than the 30 amp rating of the powerstat. But that smaller current level is below the phase imbalance sense level for the device I am testing. The min current is about 8 amps.

 

[peterb]

If your primary aim is to test the unbalanced current device, then you really don't need a motor. I suggest the following approach -
- Define the level of unbalance that the device needs to operate at
- Calculate the actual phase currents corresponding to the required unbalance
- Connect your powerstat to the 230 V 3-phase supply
- Connect the output of the powerstat to your device and short circuit the output leads together to the powerstat common
- This gives you a variable current supply, where you can set each phase current individually by the powerstat wiper
- A refinement to allow closer control would be to add reactors in series with the powerstat output; resistors would work but would be bulky and hot; a set of 230-120V transformers, short circuited on the secondary, would also work - there are a lot of possibilities

 

[electricpete]

dynets - can you give us a hint how your device works?
current sensing? flux sensing?

 

[busbar]

peterb, that is a great idea, but can I suggest taking it one more step? If only current is needed, why not use low-voltage-secondary transformers, like with 12-volt windings? Then the source voltampere requirements could be very low.

With 3-phase powerstats, there are two general configurations. (Look at Superior’s web info.) With two decks, they are connected in an open-delta arrangement with coils rated for line-to-line voltage. With three decks, they are connected in a wye arrangement with coils rated for line-to-neutral voltage. You are looking to create unbalanced conditions, and it’s possible to do that with one deck.

 

[dynets]

To: PeterB. Isn't that like shorting all three outputs of the powerstat together? The device under test is a pass through device, like a contactor. Won't I blow line fuses or something worse?

To: Electricpete, the device uses current transformers to sense/measure current.

To: Busbar, I have thought about the low voltage /current approach using step down transformers and a resistive load to get the current up. In fact I did do that originally and it works, but I came across a problem: the transformers I used were 240 to 5 volts, since the voltage is so low, the internal contacts in the device would not always conduct, (they are made for higher voltage). Plus I really wanted to use an actual motor to simulate a more real world application. But you did give me a good suggestion by using only one coil. Since I really only want to change the voltage in one phase why use three?? There are several single larger powerstats in storage that might do the job. I will check it out. Thanks.

 

[peterb]

I'm with you all the way on this, busbar. The lower the voltage the better - I was trying to work within the parameters of the equipment that dynets seems to have in hand. Note that individually adjustable Powerstats per phase are needed so that you can get unbalanced conditions.

Of course, the test device needs to be connected on the high current/low voltage side of the step-down transformer

 

[busbar]

Overload cycling capability of a powerstat [based on some old literature] is:

Toffmin = [Tonmax (Iol/Icont)]^2 – 1

where Toffmin = minimum off time
Tonmax = maximum on time
Iol = overload current
Icont = rated continuous current

http://www.superiorelectric.com/

 

[peterb]

dynets -
Yes, you are shorting out the supply (through an impedance, if you use a step down transformer), but you are controlling the current flow with the powerstat. If the device uses current transformers to measure current, then it shouldn't matter what the voltage is - or are we missing something here?

 

[SooryaShrestha]

Hi everybody. May I add some experience and some querry?
dynets -
May I know what type of unbalance are you expecting? Is it true unbalance or is it the unbalance as defined by IEEE or NEMA? IEEE or NEMA defines unbalance only in terms of magnitude. True unbalance requires the phase angle unbalance also.
electripete and nbucska -
I have used the load you mentioned in testing small hydro plants of size less than 2 Mw. We call it dummy loads. As mentioned by electripete, it has three big steel plates connected to the three phases and hoisted by a chain pulley. We continously add water to replenish the evaporated water and also common salt. We have not used the borax or NaOH, somehow because of cost and also I think salt is safer than NaOH. In this dummy load, it is not the separating distance that controls the impedance of the load, but the contact area of the steel and water. Changing the separating distance of the plates is practically more difficult compared to hoisting up or down the tank.
But getting exactly equal loads on the three phases is very difficult, as the resistance appreciably changes with temperature and the concentration of salt at any particular location of the tank. So some unbalance load is created.
As far as the unbalance of supply is concerned, I think, dynets as you have indicated, if you use only a variable inductor of size commensurate with your requirement, only on one phase and a variable capacitor on the other phase, maximum unbalance can be created.
Thanks everybody, it is very interesting post.

 

[dynets]

TO: SooryaShrestha
I never thought about imbalance in those terms. I guess I was looking at it from a plant electrician's point of view. If the device indicates it protects for motor phase imbalance in the range of 40-60% and a RMS clamp on is all you have it would be have to be magnitude. I.e. if I measured phase currents of 10,12 and 15 amps, the greatest difference is 5 amps (15-10)/15 = 33% unbalance, I would expect it not to fault.

 

[electricpete]

SooryaShrestha - It's a good point. I was referring to the same situation when I quoted Nailen as saying that imbalance current is used by NEMA as an approximation of negative sequence current, but I didn't emphasize it the way you did (which was good).

I would offer a minor suggestion on the terminology.
What you call &quot;true unbalance&quot; current, I would call &quot;negative sequence current&quot;. I've never heard the term &quot;true unbalance&quot; used in IEEE literature regarding motors but &quot;negative sequence&quot; is used.

Here's the way I look at it:

&quot;Unbalance&quot; is used to describe a parameter based on magnitudes alone as you say. As you say NEMA MG-1 14.36.1 defines voltage unbalance as the (maximum deviation from the average voltage)/(average voltage).

&quot;Negative Sequence&quot; - Can only be computed by vector methods using phase angles.

Here are some relevant excerpts from IEE c37.96-2000 (Induction Motor Protection) section 5.7.2.3
&quot;While unbalanced phase voltages or currents are readily identified, it is the negative sequence component that actually jeopardizes the motor. Hence, si ple unbalance measurements may not provide the degree of motor protection required.&quot;...

&quot;Several classes of relay are available to provide unbalance protection:

Phase balance relays (46) compare the relative magnitudes of the phase currents.....

Negative sequenc current relays (46) respond to the negative sequence component of the phase currents. The instantaneous version of this relay provides excellent sensitivity. Because of this, it will operate undesirably on the motor contribu6ion to uinbalanced faults on the supply system and therefore must trip through a timer or be directionally supervised).&quot;

The standard has some discussion about why negative sequence is the most damaging portion of the current. They point out that the reverse rotating stator field assoc with neg sequence stator current produces reverse rotating current (approx 2X Line freq)in the rotor which produces high heating on the rotor skin due to skin effect.

One other thing I've heard (although not mentioned inthe standard) is that negative sequence stator currents and resulting reverse-rotor currents actually give rise to a steady torque in the reverse direction. So this component of the current (negative sequence) not only does no work... it does negative work which means the positive sequence current must increase proportionally in order to compensate.

 

[jmcs]

dynets,
Renting a 3 phase protective relay test set would allow you to test the unbalance relay all day without damaging a motor.

 

[busbar]

e-pete, I appreciate your explanation of motor protection. This has been an interesting thread. In your September 4 posting here you mention that “Negative sequence current relays (46) respond to the negative sequence component of the phase currents. The instantaneous version … must trip through a timer or be directionally supervised.&quot;

Question … Do you have any feel for what should used be for supervision of the 46? Maybe a 32 device to inhibit 46 operation if watt flow is away from the stator? If no, in your opinion what would be the proper device to supervised the 46? If yes, this leads to: Is stator heating more or less of a problem with changing power-flow direction?

I appreciate your comments on this.