×
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Log In

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips Forums!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!
  • Students Click Here

*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.

Students Click Here

Jobs

Quiz: Non-simult. contact closing- effect on max instantan. current?

Quiz: Non-simult. contact closing- effect on max instantan. current?

Quiz: Non-simult. contact closing- effect on max instantan. current?

(OP)
Let's say that upon starting a motor, the contacts close at different times separated by up to ¼ cycle.   

QUESTION 1 – How much do you think the peak instantaneous current can increase compared for non-simultaneous contact closure compared to simultaneous closing?
QUESTION 2 – What timing characteristics create the worst case?  

To facilitate uniform terminology/conventions for #2, I assume phase A is the last to close.  There is no system response when the first contact closes (no current flow until 2nd contact closes), so the relevant time interval to characterize non-simultaneous closure is the time delay between closure of 2nd and 3rd contact.  I'll assume that phase A closes last and that the time between closure of B/C and A is called "TswitchA".

Additionally for a given time delay between closure of 2nd and 3rd contact, we can shift the phase of the applied voltage.  For A/B/C phase sequence, I will capture this in parameter "ThetaB" which represents delay of B phase voltage compared to sin wave.  Therefore the voltages will have the form:
Vbn = V * sin(w*t – ThetaB)
Vcn = V * sin(w*t – ThetaB – 120Degrees)
Van = V * sin(w*t – ThetaB – 240Degrees) = V * sin(w*t – thetaB + 120Degrees)
These are voltages to power supply neutral (not motor neutral).   B and C phase contacts are assumed to close at t=0 and A phase contact assumed to close at TswitchA.  

So question 2 boils down to worst case combination of "ThetaB" and "TswitchA".

I have already analysed this question by simulation (Krause's transient model) as well as circuit analysis  = modeling the motor as a 3-phase-wye R-L circuit.  The results are qualitatively similar for both models (same combination of thetaB and TswitchA to create the worst case) and quantitatively fairly close if we select power factor of the R-L circuit to match the motor starting power factor.

I'll share my results/conclusions on Friday (unless you want more time... let me know).   Until then, I'll be interested to hear any thoughts you guys have.
 

=====================================
(2B)+(2B)'  ?

RE: Quiz: Non-simult. contact closing- effect on max instantan. current?

(OP)
We have in the past corrected some test data on non-simultaneous closing of motor starter contacts which I have attached.

Here is a general background of the circumstances under which the data was collected:  Back in 2007/2008 we had some intermittent instantaneous trips of a 60hp 460vac motor.  It was about 3 trips spread over a year and probably 50+ motor starts (we did a lot of starts for troubleshooting).  Testing the breaker, contactor, motor, we never found any smoking guns.  Eventually we increased the instantaneous setpoint and went about our business, no further trips. In the course of that evolution we did hook up a recorder to the contactor with line side deenergized (monitoring the main contacts as dry contacts), and monitor timing of the contact closing with results attached.  There were 2 contactors involved labeled "old" and "new".  I know we had trips with the "old" contactor and none with the "new" (although I think we increased the setpoint shortly after installing the new).  Also I am thinking we had replaced the contactor once before following the very first trip, so that the trips occurred with two different contactors, but only one of them has data here (the one labeled "old".

Reviewing the data in the "data tab", we had not only non-simultaneous closing, but also a lot of bounce on both the old and new contactors.  For purposes of determining TswitchA  (time between closing of 2nd and 3rd contacts), I have disregarded any contact bounce.  My reasons for disregarding the bounce are:
#1 One might expect that the bouncing contact will arc but not interrupt the current (?)
#2  Review of motor starting current waveforms (a different test with power applied) shows clear evidence of non-simultaneous contact closure, but no evidence of contact bounce.  Perhaps confirmation of #1
*3 – It increases model complexity to add a bounce and would be impossible (for me) to model the arcing.

Results shown in the summary show the old contactor had Tswitch (playing the role of TswitchA) as 0.1 – 2.3 msec for the "new" contactor and 0.2 – 3.9 msec for the "old" contactor.  This shows substantial variability and also the old contactor had half as many results which were higher than the highest of the new contactor (2.3 msec).   So I guess there is some possibility that IF non-simultaneous closure increases our peak instantaneous current (the quiz question) THEN this might have been contributor to our trips.  
 

=====================================
(2B)+(2B)'  ?

RE: Quiz: Non-simult. contact closing- effect on max instantan. current?

(OP)
It's Friday so here is my take on these questions.

Is the peak of motor starting current higher for non-simultaneous closing of contacts
Yes.  

These results are summarized on slide 2 of attached.

For starting power factor of 0.2, the worst case peak for non-simultaneous closing is 14% higher than for simultaneous closing (compare slides 7 and 8)

For starting power factor of 0.0  (an idealized case), the worst case peak for non-simultaneous closing is 19% higher than for simultaneous closing (compare slides 3 and 4).

Note 1 - all results discussed are normalized such that the peak of the sinusoidal component of starting current is 1.  Stated another way, all results are normalized by dividing the peak instantaneous current by (sqrt2 * LRCrms).

Review how the worst-case peak current is produced for the simultaneous closing case
We look at the simpler case of power factor = 0 which represents closing of an inductive circuit.
Let's say v(t) = sin(w*t-theta)  reprsents the phase to neutral voltage
v = L* di/dt
i = (1/L) * Integral {v(t)} dt
Integral is the area under a curve... in this case the area under a sinusoid.   To get the highest area under the curve we want to integrate for the longest time interval without changing sign (when voltage passes through 0, the peak of the current occurs, and current begins decreasing).

The longest time we can integrate a sinusoid without hitting zero is 180 degrees.  To get that longest 180 degree integration period, we must begin integrating at the voltage zero and integrate all the way to the next voltage zero where the peak occurs.
This worst case of theta = 0 degrees is shown on slide  3.  It gives a full  dc offset of 1.0 on  the starting currnet,  and the peak is 2.0.   (twice the peak of LRC,  2*sqrt2 times the LRCrms number).

For thetas farther away from 0 we reach a lower peak that the worst case. On a given phase, the lowest-peak possible occurs when theta = 90 in which case we integrate for 90 degrees and there is no dc offset.   (we are closing the circuit at the phase angle where current would naturally be zero anyway since current lags voltage by 90 in an inductive circuit).

If we increase the power factor, the result is that the offset is not true dc, but exponentially decaying dc (time constant L/R).  The offset has decayed some by the time the first peak is reached, so the peak is not as high (lower worst case peaks as we increase power factor).   You can see comparing slides 3/5/7/9 for p.f.=0/0.1/0.2/0.3 that the offset is decaying away quicker giving lower peak.

How is the worst-case peak current is produced for the NON-simultaneous closing case?
Again we start with the the simpler case of power factor = 0
thetaB and tSwitchA were defined above and also shown in slide 1.
The highest starting current as shown on slide 4 can be created using
thetaB = 30 degrees
tSwitchA = 90 degrees

Looking at the curve we can see that phase B current is increasing from 0 until 210 degrees.  (That's longer than the 180 degree max possible with simultaneous closing).

But how can current possibly increase continuously for 210 degrees if the voltage reverses sign every 180 degrees?..... The answer is that there are two different voltages to consider: the relevant voltage prior to time TswitchA is the phase-to-phase voltage ....and the relevant voltage after TswitchA is the phase-to-neutral voltage.

The phase-to-phase voltage between B and C which is driving the phase B current prior to switching is the difference between the red (phase  B) and blue (phase C) voltages on the plot and it is indeed greater than zero starting at time 0 and continuing at least through the moment of time tSwitchB    After time tSwitchB, the B  phase phase-to-neutral voltage becomes more relevant and it remains positive all the way through t=210 degrees.... which we can see is in fact the peak of the phase B curren twaveform.

Now I  think you can see why thetaB =30 was chosen.... it is the point where the plotted phase B and C line-to-neutral voltages cross, so it was a zero-crossing of the phase-to-phase voltage.  Selecting zero of the phase-to-phase voltage creates a worst case in analogous manner that selecting zero of phase-to-neutral voltage created the worst case when we looked at simultaneous closing.

Now why is TswB = 90 degrees the worst case?  That is not quite as obvious.  The explanation that I have lies in slides 13 through 15_.  These slides represent a superposition approach to solving the non-simultaneous closure of the R/L circuit.  It greatly simplifies the problem, because instead of solving three interrelated R/L circuits, we only have to solve two R/L circuits separately.  Each circuit has the complexity of a single-phase circuit and we simply add the currents from the two circuits together to get the total current.   I developed this approach simply for purposes of solving the problem to create the graphs.  But it had the unexpected (lucky) benefit of answering the 90 degree question which bothered me for awhile.  Let me know if the slides require further explanation.



Are these results (higher current for non-simultaneous closure) obvious
I first heard this discussed on eng-tips by someone... I think it was Keith (itmsoked) or Bill (waross).  I could never visualize exactly what effect non-simultaneous closure  would have, or even how to arrange the switching sequence and voltage phase to create the worst case.    I just used brute force trial and error simulation initially to find the conclusions.  Once you look at the curves, I think that it is now more intuitive to see what's going on.

The 30 degree and particularly the 90 degree numbers were initially very mysterious to me because based on my trial/error they are very exact numbers which continue to predict the worst case even as power factor changes.  I don't think there is any other obvious intuitive explanation for the 90 degree worst-case tSwitchA,  unless you stumble into the superposition approach by dumb luck as I did.

Is it important?... What is the significance?
Clearly it creates a higher possible "worst case" scenario if  up to degree delay occurs between between 2nd and 3rd contact closing.  (90 degrees seems plausible from what we saw which was up to 3.9 msec.)

We saw a clear difference in tSwitchA between or "old" and "new" contactors, so when troubleshooting those mysterious trips, then the differences between behavior of the contactors has to be considered as one possible "cause".   

The subject of instantaneous settings for low voltage motors seems to carry some controversy at least at our plant, and some others I have seen. With better understanding we can make better decisions.

In the unlikely event that you were inclined to perform a statistical analysis of expected peak current based on a set of peak current measurements,  you could reach an incorrect conclusion by neglecting non-simultaneous closing.  Specifically if we considered there are two random variables:  thetaB and tSwitchA, then that would create higher  variability of peak instantaneous current compared to considering tSwitchA as a constant 0 (simultaneous closure) and only consider only the effects of thetaB alone. (By the way I'm not saying such a statistical analysis is wise or that there are no other sources of variability.)
 

=====================================
(2B)+(2B)'  ?

RE: Quiz: Non-simult. contact closing- effect on max instantan. current?

(OP)
1 - Sorry for some typo's in my post directly above.  Everywhere I wrote TswitchB should be TswitchA (I never defined anything called TswitchB). There are no typo's in my attachment directly above.

2 - There is a much more straightforward explanation for the choice of TswitchA = 90 degrees to create worst case (given that we have already chosen ThetaB=30).  The driving voltage prior to switching is 0.5*(Vb-Vc) which is added to the plot attached as a green line.  The driving voltage after switching is Vb.  90 degrees is where these two voltage curves cross. So the worst case switching time ensures the higher of these two voltages is applied.

=====================================
(2B)+(2B)'  ?

Red Flag This Post

Please let us know here why this post is inappropriate. Reasons such as off-topic, duplicates, flames, illegal, vulgar, or students posting their homework.

Red Flag Submitted

Thank you for helping keep Eng-Tips Forums free from inappropriate posts.
The Eng-Tips staff will check this out and take appropriate action.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!


Resources