Motor Flicker

[kashew]

I've been calculating motor starting flicker using the basic formula: LRA x (R cos0 + X sin0)
where cos0 = starting pf and R & X values are in ohms

I've run across a different method of calculation that I'm not familiar with and was wondering if anyone could provide some insight. Is this method more accurate?

The two methods do NOT give the same result.

The formula is:
%Flicker= 1-(sqroot(Zcable*2)/sqroot(Ztransf*2 + Zcable*2))

where *2 represents the value squared
replace Z with R+jX values for cable and transformer

An infinite system bus is assumed.

 

[alehman]

The first formula should give a reasonable approximation if R and X include the total impedance. I don't see how the 2nd formula takes into account the motor starting current.

 

[kashew]

Yeah, I know what you mean. That was my first thought. I've been doing some more research and found an EBASCO manual that has a similar method. They include the motor impedance in the formula. The motor impedance is calculated with the following formula %Z= 10,000/KVAstart on a 100 KVA base
The rest of the impedance for cable, transformer and system are calculated on the same base. EBASCO's method get the same results as I've described above. The answers are several percentage points less flicker than using the
LRA x (R cos0 + X sin0) method. I just can't quite picture it in my mind.

 

[ccjersey]

Hi everyone,

I want to understand the term "zero sequence current".

Can anyone elaborate or point me to a website that will explain.

Thanks,

Jim

 

[kashew]

In three phase systems there are three mathematical vectors that are used to analyze the system. The 3 positive sequence vectors rotate as A-B-C 120 degrees apart, the 3 negative vectors rotate as A-C-B 120 degrees apart and the 3 zero sequence vectors are 0 degrees apart...kinda like have three phases attached to a single phase source. This mathematical system is called symmetrical components and is used to calculated unbalanced systems as if they were balanced. Used predominantly in short circuit calculations.

 

[weh3]

On the original post, regarding the alternate equation, I would replace Zcable with (Zutility + Zxf + Zcable), and in place of Zxf use Zmotor. The flicker is simply a voltage divider between source and motor under starting conditions.

William

 

[kashew]

Hmmm ... you're right. I hadn't thought about it that way. But, why would the results be different using the two different methods? The voltage divider method gives a smaller flicker result.

 

[rbulsara]

Firstly I take it that you mean voltage dip at motor terminals by the term voltage flicker. I am not sure which point you are considering as infinite bus (which will not dip).

Ideally you want to include the source impedance Xs in the calcs, or if you have a service transformer use infinite but ahead of it.

You use the same method as if you are calculating SCC, with

the follwing in series (use per unit method and convert all to correct bases). You can use Z=R+jX if you know R, but all practical purpose, only R to be considered will be that of the cable.

Xs+Xt+(Rc+jXc) +Xm. Place a short at end of Xm and calculate starting I = 1/(Xs+Xt+(Rc+jXc) +Xm) where Vpu=1

Total voltage drop will be across each element can now be found by voltage devider rule.

If you use LRA as known current you need not include Xm and just find VD=IZ across known component.

If you ignore immediate trasnformer imepdacen upstream you will have very small VD.

Hope this helps.

 

[weh3]

kashew,

It all depends on which impedances you use: locked rotor, transient, subtransient, etc. These will give you different results at different times. For the overall picture, I would use the LRA from the NEC.

William