Infeed + Motor Fault Contribution 2

[squasher]

I would be most grateful for the forums comments on the following:

a) Has anyone approached an MV switchgear manufacturer for an opinion on whether an MV busbar will withstand approx the first 6 cycle and ½ cycle peak when a short circuit study indicates that the infeed short circuit current with induction motor fault contribution exceeds the Fault rating of the switchgear by a small margin re: 8 % ?

Example: calc Ik” =28kA for approx 120ms (6cycles,UK 50Hz) and Ik =26kA for 3secs

b) Has anyone had a manufacturer provide a non-certified document on old MV switchgear to cater for this situation : a very short time rating re: 28kA for 0.2 sec (10 cycles) ?

c) Apart from the usual fault reduction methods of :bus section open on 2xinfeeds,fault limiting reactor,I limiter,uprating of busbar support structure etc are there any other alternatives other than total replacement of the MV switchgear?

I’ve assumed for the above that the MV Circuit Breakers can be replaced with a higher fault rating

IEC60909 Definitions for reference:
Ik” = initial symmetrical short-circuit current [For above=Calculated infeed s/c r.m.s value + symm r.m.s current from rotating plant]
Ib = symmetrical short-circuit breaking current
Ip = peak short-circuit current (Making Current 1st ½ cycle peak)
Ik = steady-state short-circuit current (rated short time current [3decs])

 

[waross]

If it is only 8% over, have you included the impedance of your cables? You may be able to drop a few percent of SC current with longer cables.
Also, when you're that close to the rating, it may be prudent to investigate the X:R ratio at which the switchgear Available Short Circuit Current rating is tested and determined.
Busbar bracing must withstand the stresses of the asymmetrical short circuit current. Switchgear is often rated at the symmetrical short circuit current at a particular PF or X:R ratio.

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

 

[squasher]

waross,
Thanks for the comments, much appreciated.

We have included the impedance of the cables,typically 11kV,400sqm sc, 1.5kM long between substations.Although as this is so close re: 8% over, as you indicated, its worth an accurate measurement on length and perhaps increase the lengths to gain a little more impedance.

The upstream older switchgear, which may have been built more substantially could well have some leeway,so point taken about checking the X:R ratio at which the switchgear was tested.

Hypothetically if the 1st cycle 1/2peak and subsequent 6 cycles of asymm fault current exceed the fault rating of these bars by this size of margin would the electromagnetic forces have fully stressed the bars to cause damage?

Whats your view on fault level tolerances ?

 

[waross]

The point of these ratings is to design an installation which will withstand possible fault currents without being destroyed. At 8% over, the magnetic forces will be almost 17% greater.
The greatest magnetic forces are developed by the peak asymmetrical current.
When the designers are building switchgear for a particular class of Available Short Circuit Current or Symmetrical Current they assume a conservative X:R ratio and design the equipment to withstand the corresponding Asymmetrical current. The Available Short Circuit Current is easily determined by taking the reciprocal of the PU impedance of the source transformer.
In your case you must determine the X:R ratio for which the switchgear was designed and from this determine the Peak Asymmetrical Current that the switchgear will withstand.
Then using the X:R ratios of your transformers and motors you may find the actual Peak Asymmetrical Current that may be encountered in service.
It is for you to determine if this approach is accepted by your local codes and practices.
Let's see if any others care to comment.

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

 

[squasher]

Hi waross,

Thanks again for taking the time to reply and nicely explained.

In general,two examples I have considered to obtain a realistic order of magnitude for motor fault contribution are:

a) Batch Plant: Where the motor fault contribution is calculated using the guidelines of the UK G74 recommendation for a new plant installation,but connected to existing older switchgear.
Where there are no accurate plant load profiles to go by but only "probable" plant loadings,then a typical load diversity of 0.65 was used re: Not all motors running

b)Continuous Process Plant:Again using the UK G74 guidelines, but in this scenario the accuracy of motor fault contribution was based on modelling actual steady state motor running loads from meter readings taken from the plant distribution system.

I very much take the point that where by calculative methods that can be substantiated and stand up to scrutiny that where asymm peak kA well exceeds the known Peak rating of the bars and if all other s/c current limiting methods have been considered,then this leaves no other alternative but to replace the switchgear.

My interest is where there are marginally borderline cases to see if:
a) the basis of the calculations reflect the operation of the plant and so a meaningful motor fault contribution
b) Tolerances in the switchgear,above which there is an absolute cut off

Would anybody else care to comment?

PS: I think this is an excellent forum

 

[jghrist]

If you exceed the switchgear rating by 8% when considering only 65% of the motors running, this is not tolerable. Is it possible that more motors are running? If so, this needs to be considered in evaluating the equipment.

The rating of the equipment is the absolute cutoff. If you don't consider this the limit, then what is the limit and why isn't that the equipment rating?

Are you willing to accept liability for damages if a fault occurs and the equipment fails after you have calculated that the fault current can exceed the equipment rating?

 

[squasher]

Hi jghrist,

Thanks for your input

Its probably unlikely in this case that 65% of the motors will be running this would be considered as conservative, more likely to be 50%.If it was possible to say what the peak load was and base the analysis that a fault developed upstream simultaneously this I think would be the worst case scenario.Probabilty- low but certainly possible. Point taken.

My approach to the equipment rating and absolute cut off would be to approach the switchgear manufacturer, put the scenario of the calc SC 8% over (for 6cycles) and request them to confirm whether this would be within the tolerance of the switchgear or not.

For both cases highlighted, this is existing switchgear is owned by operating companies.My view on accepting liability is that I would advise in these cases that dependant on the final outcome of the investigation & final calculations whether the fault rating of the switchgear would be exceeded and the possible consequences. I would suggest they have a second opinion before making any recommendations.

As in both cases this would involve not only substantial capital switchgear replacement costs but consequential costs due to loss in revenue from lost production due to downtime, then all avenues need to be explored.

Purely from a safety point of view it would be easier to make a recommendation for replacement of the switchgear on safety grounds.

 

[dpc]

request them to confirm whether this would be within the tolerance of the switchgear or not

I think we already know what the answer to this question will be. The maximum short circuit rating on the equipment nameplate is an absolute maximum, subject to the testing standards used to rate it.

If you are considering less than 100% motor load, I would insist on something in writing from the operations side that lists the maximum complement of motors that could ever possibly be running at the same time.

I would suggest looking for other operational solutions that would limit the maximum fault current. This is not always possible, but should be explored.

I have been similar positions in the past - recommending replacement of underrated switchgear and breakers - in some cases brand new equipment. It is never a happy situation, but you have no choice but to bring it to the attention of the responsible parties.

 

[jghrist]

I wouldn't consider peak load in determining maximum fault current. You need to consider the full rating of connected motors. Lightly loaded motors will still contribute full short circuit current.

 

[squasher]

Thanks to both for the last comments;

My reference to the peak load was relating to the estimated % maximum of connected motors at any one time rather than the base load + the largest motor(s) starting,but I take your point.

For the batch plant,part of the process is considered "new" re: unknown for load pattern then the estimate is unlikely to be that accurate,hence the conservative fig of 65%,this would be quoted in the studies.

In both cases the operating companies site electrical engineers are aware of the possible outcome.

I'll give an update and post again if there is any new light to shed

Thanks again

 

[waross]

You may be able to legitimately get your figure down the 8% by considering the impedance of the motor cables. I would do a calculation on one or two selected motor runs of average length. That should make it apparent whether it is worth while calculating all the motor cables. It would be a shame to spend thousands of dollars when a few hours of calculation may show it to be not needed.

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

 

[squasher]

Hi Waross,

A good point.

For the Batch plant all the large loads (<100kW) were modelled with cables.The smaller drives were "lumped" and the ratio of locked rotor current/FLC reduced from typically x8 to x5 - inline with the UK G74 recommendation
for the estimation of m.f.c.including connecting cables (simplifed).
I take your point,certainly worth the time- Thank you!

The situation in the process plant model where there are a number of relatively large (400kW+)regen drives-active front end,at the moment these have not been individually modelled ,again the mfc contribution could be reduced to x3 using the UK G74 guidance.

So there is further work that would certainly be of benefit.
This was a relatively high level initial study for a approx 20MW plant

 

[JBinCA]

I don't work in IEC standards, but I will say that the drives do not contribute to a fault in anything like the same way that the motor directly would.

The drives may even be able to limit 1/2 currents. I know this is true of modern solar inverters.

Definitely incorporate Waross's discussion about the X/R ratio that the rating is based upon. Correcting for your actual X/R will make things bettor or worse. Your program may be doing this for you.

You asked

Hypothetically if the 1st cycle 1/2peak and subsequent 6 cycles of asymm fault current exceed the fault rating of these bars by this size of margin would the electromagnetic forces have fully stressed the bars to cause damage?

From a physics perspective, this is a very valid question. From a standards perspective (in the US anyway) it doesn't matter at all. 65kA rated SWGR is 65kA rated SWGR because it didn't blow up when tested at that value - not because of a calculation.

Regards,

JB

 

[squasher]

Thanks for your input,

I presume the 65kA rated SWGR is the Symmetrical SC Current rating typically 1sec- designed to withstand that value for that period (without damage) and not the peak ?

My interest is in the 1st cycle peak and subsequent 6cycles (US 6x0.016 approx=0.1sec)with motor fault contribution after this period when steady state SC current has then reduced to below the 8% margin and is within the SC rating of the switchgear.During which time the protection may have been set to clear the fault re: 0.5sec

I know this does'nt change anything,regarding exceeding the peak rated value of the swgear.However I have in the past had a 6kV cable manufacturer confirm a higher SC current rating for a shorter time rating(0.2sec)compared with its maximum 3ph sym scc rating at 1sec.In this case with fuse protection.


So hence my interest in seeing whether this approach would apply to busbars in switchgear.

Only the actual switchgear manufacture for these cases can confirm.

 

[wcarter]

dpc and JBinCA is correct. I work for a breaker/swgr manufacturer. Anything over a rating is unacceptable. Then you would also factor in if you do an Arc-flash study. They couldn't. Now what might be able to be done is depending on what is limiting the rating, IE supports, many times we have been contracted to perform a study on supports to raise a gear's kA rating by bracing.

 

[waross]

There are both mechanical limits and thermal limits.
The maximum mechanical stress is developed by the peak asymmetrical current in the first half cycle. Ratings for a number of cycles are a thermal limit.
If you are limited by the peak current in the first cycle the current must either be limited or the breaker must be constructed stronger.

Now what might be able to be done is depending on what is limiting the rating, IE supports, many times we have been contracted to perform a study on supports to raise a gear's kA rating by bracing.

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

 

[ScottyUK]

From a physics perspective, this is a very valid question. From a standards perspective (in the US anyway) it doesn't matter at all. 65kA rated SWGR is 65kA rated SWGR because it didn't blow up when tested at that value - not because of a calculation.

Valid indeed. However the 65kA rated gear will likely have been tested at an assymetric peak somewhere north of 100kA, and the 65kA will be the RMS withstand current. The full test certificate should state the peak current during the test. For a 65kA board tested to IEC 60694 the peak would be 162.5kA (2.5x 65kA). If this old board was built to an earlier British Standard then you might want to obtain a copy of that standard.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[squasher]

Thanks wcarter,waross & Scotty Uk for your replies,

As mentioned previously for the case of the (Batch Plant) older Vacuum Circuit Breakers they can be replaced with a higher fault rating e,g, 26-> 31.5kA, the problem is with the switchgear busbar assembley supports etc.The original switchgear supplier has been bought over,fortunately the new supplier(large international) has since indicated they can strengthen the bracing etc to achieve the higher fault rating.

I am assuming once the additional bracing has been installed that the switchgear supplier will issue a "type" test certicate to indicate the switchgear is then suitable for the higher fault rating.

The bottom line looks to be (as all comments have indicated)
Quote dpc 21-Jan " I think we already know what the answer to this question will be. The maximum short circuit rating on the equipment nameplate is an absolute maximum, subject to the testing standards used to rate it"

e.g. no tolerances in testing