If a balanced (symmetrical) 3 phase/4 wire power source is connected to a balanced 3 phase/4 wire load (impedance), I1 + I2 + I3 = 0 and In (neutral wire current = 0). In this case, the neutral wire can be removed and nothing changes.

If the power source is unbalanced, for example, In (neutral wire) would draw the imbalanced current.

Now, my question is - what happens in an unbalanced power source connected to same 3 phase load except with NO neutral wire connection (floating neutral). Seems to me I1 + I2 + I3 still wants to be zero, but with no neutral to carry imbalance, the imbalance will cause circulating line-to-line currents to flow such that the line voltages alter to satisfy the sum (I) = 0 formula. If this is correct, then the unbalanced power source will become even more unbalanced. Is this correct analysis?????

The reason I ask is because an end user is using a 3 phase autoxfmr with floating neutral (440 to 420V stepping). They have an unbalanced power
source. The drive mfg is recommending they either ground the floating neutral or install an isolation xfmr with grounded neutral. However, I'm not convinced grounding the neutral specifically eliminates the imbalance that originates at the power source.

Thanks.

 

Replying to TR Gianni's questions (Thu, Aug 1, 10:21 am):

A) Par 1. We agree.

B) Par 2. What is meant by unbalanced source... Unbalanced voltages?
Unbalanced kVA phase capacity of the power source transformer?

C) Par 3. Your assertion that the algebraic sum of the line currents is zero is correct. It means that the three line currents will have
different magnitudes and phase angles. However, moderate unbalance current should not materially influence source voltage.

D) Par 3. What do you mean by "circulating current? Usually, it refers to that current that, due to transformer action, actually circulates in
the transformer tank. Whether or not it is appreciable depends on transformer connection and construction... eg, wye-wye vs delta-wye, and
3 sets of primary and secondary windings in one tank vs 3-individual transformers.

E) Par 4. You must not ground the neutral of the auto-transformer, especially if it's relatively small compared to power source transformer.

Conclusions:

1) A severely unbalanced load connected to the auto-transformer could result in what is called "neutral inversion," resulting in its failure.

2) Installation of a transformer for the VFD installation will not adequately compensate for the unbalanced source voltage. If a system
neutral must be established for VFD grounding purposes, then, consider a zig-zag transformer.

3) If source voltage unbalance seems to be multiples of 2-1/2 percent, then tap settings are suspect.

If you need additional info, please provide pertinent parameters.

Regards,
Phil Corso, PE
(Boca Raton, FL)

 

Phil Corso wrote:
> <snip>
>
> B) Par 2. What is meant by unbalanced source... Unbalanced voltages? Unbalanced
> kVA phase capacity of the power source transformer?
>
> C) Par 3. Your assertion that the algebraic sum of the line currents is zero is
> correct. It means that the three line currents will have different magnitudes
> and phase angles. However, moderate unbalance current should not materially
> influence source voltage.
> <snip>

Moderate unbalance in supply voltages as Phil suggests is not generally a problem and is in fact normal. Power transformers don't have three phase symmetry because of the cost and impracticallity of constructing a symmetrical core. Even if power transformers had three phase symmetry loading patterns on the different phases
would still result in moderate voltate imbalance at supply points.

Vince

 

Attn: TR Gianni...

Additional food for thought:

In the spring of 2001, the List carried two topics that discussed unbalanced voltages associated with ungrounded or "floating" power
systems. The end user you cited could be experiencing the same problem.

The thread subjects were titled:

1) "Power Distribution Woes":http://www.control.com/983813443/index_html , and

2) "The Physics of Electrical Failures":http://www.control.com/984504035/index_html .

Regards,
Phil Corso, PE
(Boca Raton, FL)

 

Phil's reply helped me realize my original circuit analysis was incorrect. I know Kirchhoff's Current Law can't be violated, which prompted my need for an explanation to the dilemna. Below is my reply to Phil:

*********************************************************************
You helped me answer my own main question! I was in the process of posing my question another way when it occurred to me I was analyzing the circuit wrong - in a system with no neutral connection, you have to analyze the circuit with "line-to-line" superposition vs. line-to-neutral (for 4 wire system). When I used superposition for circuits L1-L2, L1-L3 and L2-L3, I see that Kirchhoff's Current Law (KCL) is not violated regardless of whether the voltages are symmetrical 120 degree sinusoids or NOT! What I now realize is that even though non symmetrical voltages, without a neutral, the total line to line circuit couples the L-L voltages such that
KCL still works.

Comparing different power systems:

1. Symmetrical/balanced 3 phase/4 wire source-load system: I1 + I2 + I3 = 0, and In = 0 (no neutral current)

2. Same as above except either source or load unbalanced: Sum Ix not zero and In draws imbalance.

3. Symmetrical or non-symmetrical 3 phase source with balanced or non-balanced 3 phase load with open neutral: Sum Ix must be zero (and is)
per KCL.

Therefore, regarding my customer's situation, grounding the 3 phase autoxfmr will NOT magically balance the unbalanced voltages per drive
"serviceperson". Adding the neutral wire will simply result in such wire carrying the imbalance.

Question, why in your point E below must you NOT ground the neutral of an auto-transformer?

For your information, the drive does not need a neutral connection. It takes 380/460V lines and PE connection only. I think the REAL problem at
the user site, based on collaboration with drive mfg and their service report, is that the mains is [at times] very distorted (metals working
plant) and has "soft" line characteristics. The ground is probably also poor quality and noisy. The net effect is that the drives occasionally
fault out on input power/noise related error codes. Other similar competitors drives on the same machine don't fault out, but, our drive
supplier's drive [by design] is more complex and hence more sensitive to power quality.

Drive supplier "engineering" is leaning towards an isolation transformer to replace autoxfmr. If correct kVA chosen and with electro-static shield, it may solve problem by reducing distortion and noise.

Thanks in advance,

 

In order to further broaden the knowledge-base of interested List members, attached is my reply to Tom's direct query:
- - - - -
Tom,

Re E) Although the system is "ungrounded" the phases are coupled to "ground' via system capacitance. Thus, if the neutral of the auto-transformer were effectively grounded, then, any net current due to unequal phase capacitance has to be carried by the AT. The subject of
floating systems was discussed in great detail last year.

A second, and more compelling objection, is the fact that the neutral would have to carry any ground-fault current that could arise. If the
system is large, then the current could easily exceed the AT's rating.

Personally, I believe that the source voltage you alluded to is being caused by system capacitance. A simple test will prove it... measure the phase to ground voltages.

Th use of an isolating transformer may exacerbate the problem because it can't compensate for source unbalance.

Regards,
Phil Corso, PE
(Boca Raton, FL)

 

Following are some additional thoughts, but from a "technically correct" viewpoint:

F) The service-person is technically correct. Grounding the AT's neutral could reduce triplen harmonics. Note I said could, not will, because
reduction is possible only if the AT is the source of harmonics.

G) An isolation transformer, as was specified, could ameliorate the problem, only if the primary-side supply "voltage triangle" is not severely unbalanced. If balanced but its neutral is severely shifted, then I recommend the primary winding insulation be upgraded to 1.8 times
the nominal design value for its voltage class.

One more very small comment... at least in light of life's real problems... and I will stepdown from the "technically correct" soapbox! The second most irritating word in electrical engineering jargon is IMbalance. The proper adjective is UNbalance when discussing such
phenomena, or is it phenomenon!

BTW, guess what the most irritating word is?

Regards,
Phil Corso, PE
(Boca Raton, FL)

 

I have a question similar to the one that started the discussion in this forum.

When ground power is used on an AIRCRAFT, 3-phase wye-wound ground power generator is connected to the 3-phase aircraft buses, with the neutral of the generator connected to the earth terminal of aircraft supply (earthed to aircraft structure).

If the ground power unit is connected to earth, and if the connection between the generator neutral and the aircraft earth terminal is open (floating ground on the aircraft),

1) What will be the impact of three-phase unbalanced load?, and

2) What will be the phase-to-earth voltage on the aircraft side?

 

Responding to NV Rajan's 03-Sep-09 (03:08) questions... the following Thread may have the answers you seek:

http://control.com/thread/1026242921 "Neutral Interruptus!"

If you want an example illustrating what happens when the neutral conductor is inadvertently open-circuited, then contact me On- or Off-List.

Regards, Phil Corso (Cepsicon [at] aol.com)