Can NGR reduces arc flash on the secondary side of delta-wye grounded transformer?

No, Arc flash hazard is calculate from a 3 phase short circuit.  In the IEEE 1584 equation, there is a factor involving if the system is grounded or ungrounded(NRG included), and the energy of the incident will be greater if the system is ungrounded

Just to be more specific, the system will be considered ungrounded if it's ungrounded or HRG (high resistance grounded).  The arc flash result will be greater in theses cases per IEEE 1584.LRG grounded is considered as a grounded system.

I found this on a website.

IEEE Std 141-1993 (Red Book) section 7.2.4 states "A safety hazard exists for solidly grounded systems from the severe flash, arc burning, and blast hazard from any phase-to-ground fault." For this reason, IEEE recommends resistance grounding.

IEEE Std 142-1991 (Green Book) section 1.4.3 states "The reasons for limiting the current by resistance grounding may be one or more of the following:

1) To reduce burning and melting effects in faulted electric equipment, such as switchgear, transformers, cables, and rotating machines.
2) To reduce mechanical stresses in circuits and apparatus carrying fault currents.
3) To reduce electric-shock hazards to personnel caused by stray ground-fault currents in the ground return path.
4) To reduce the arc blast or flash hazard to personnel who may have accidentally caused or who happen to be in close proximity to the ground fault.
5) To reduce the momentary line-voltage dip occasioned by the clearing of a ground fault.
6) To secure control of transient over-voltages while at the same time avoiding the shutdown of a faulty circuit on the occurrence of the first ground fault (high resistance grounding).

IEEE Std 141-1993 (Red Book) section 7.2.2 states "There is no arc flash hazard, as there is with solidly grounded systems, since the fault current is limited to approximately 5A."

As you can see, it is best to not only ground the neutral, but ground through High-Resistance (typically 5A) for all systems < 600V and most systems > 600V to 5kV. For systems > 5kV, Low-Resistance Grounding (typically 200A or 400A) is used.

I would say that use of HRG will greatly reduce the risk of arcing faults on a 480 V system, since the large majority of faults begin life as a line-to-ground fault.

But what the others are saying is that it does NOT reduce the hazard that can exist for a three-phase arcing fault, so the required PPE is still based on the worst-case three-phase fault.  This is pretty clearly addressed in IEEE 1584.  

There is still an arc-flash hazard since the HRG has no effect on phase-phase or three-phase faults.  

What dpc said.  All of the things that Humble2000 point out go to reducing the likelihood of an arc flash occurring, but if one does occur, the presence of the HRG does nothing to diminish the hazard.

Sure, I realize that. The situation I have right know is due to low arc fault current the clearing time for upstream fuse is long and therefore it exceeds cat 4 requirements. If I do the calculation based on three phase faults then the clearing time would be fine and the incident energy would be reduced since the fuse would clear the fault instantaniously.
So, in this situation NGR would limit the arc fault current since it is same as L-G fault .right?

Once established the arc will involve all three phases and the arcing current will be seen by your phase protection.  The arc current will not be limited by the NGR.

Humble, what fault current do you take?  Equation are only made for a 3 phase bolted short circuit.  If in this case the clearing time is very long, you can take 2 sec. max as per IEEE 1584.  

I could reduce it to 1.2 second using faster fuses , but thats it. however since the arcing current is only around 1200Amps on the 13.8 side the cal is around 42.
transformer is 4MVA,600 secondary.

Humble2000:

As dpc and others have repeatedly said, NGR does not come in play for L-L faults or 3 ph faults. Since the NGR does not have any effect on L-L fault currents, the PPE requirement do not change with grounded or ungrounded or impedance grounded systems.

Yes, in case of juat L-G fault the fault current would be less and flash/burn hazard would be lower, but basis of IEEE recommendations is that a L-G fault can quickly turn in to a L-L or 3 phase fault.

Also the impedance grounding is inteneded to limit the damage to the equipment not to the people necessarily. It may limit burn hazard to a person as long as he/she is not part of the fault circuit. But, if a human comes in direct contact with a live part and a grounded surface, the system with a NGR is no less dangerous than a solidly grounded system, as it will produce a current not enough to trip a protective device but more than enough to kill a person.

If you are trying to follow IEEE 1584 calculation procedure, only three-phase faults are considered.  The equations are not even valid for single-phase faults.   

With a high-res grounding system, it's not really possible to have much of a line-to-ground arcing fault.  Not enough current.

If this is for the 13.8kV side you can use an arc fault sensing relay, reduce you clearing time to 20 ms (So they claim).

There are two problems, first the primary protection is fuse , so I can not trip it in case of light sensor detection. the second is that the secondary breaker does not have shunt trip and I dont want to alter the wiring to trip it using only one trip circuit.

I suggested using fault fiter S&C product and adjust the curve as close as possible to tx inrush current. However I am vconcerened about the hot load pick up. This subject is discussed on the other post.

Zogzog, 20mS! Wow, that is fast.

Didn't we discuss a little while ago, that the fastest we've seen was the around 50mS?

We discussed arc flash maint switches for LVCB's, this is a light sensing arc flash relay, ABB claims 20ms trip times but as we discussed earlier, the relay is only as fast as the breaker it uses.

Humble2000,

What type of trip unit does the seconday breaker have?

Is the primary a fused disconnect switch? What feeds the switch?

When I read "reduce you clearing time to 20 ms." I knew that is misleading. Clearing time and repsonse time of a relay are two different things.

No intention to blame you, but just to avoid any confusion.  Especially to someone who doesn't have much experience in this area. They'll think the clearing time is 20mS.  No it is not.

I understand and agree with you, it can be misleading, this ABB literature say "Extremely short overall operating time <2.5ms" That can be misleading. The breaker will still take 3-8 cycles depending on voltage, vintage, and type.

However, it is a good solution for those low arcing currents that take longer to trip a overcurrent relay, because they operate based on the light from the flash, not the fault current.

Oh yes, I see what you mean about the operating on light versus current at those fault levels. Very good point :)

I forgot to post the link

http://library.abb.com/GLOBAL/SCOT/scot229.NSF/VerityDisplay/AED31E7860092CE3C2257069001AD58E/$File/REA10_751404_LRENe.pdf

The issue is not the main breaker itself. the point is available incident energy at the secondary side of transformer which is at the main breaker bus bar . This prevents any type of work carried out on the breaker , such as opening or closing. Breaker can not be remotly operated., so this option is out of question.
The only option is finding a way to reduce the arc flash. primary has a fuse disconnect switch combination., however available short circuit rating is more than switch rating to open under fault.
despite the other practical situation, I need more fault current here to operate the fuse faster.

Why cant the breaker be remote operated?

Cant you use a remote operator?
http://www.remoterackingsolutions.com/remote-switch-operators.htm

And a remote racking system?
http://www.remoterackingsolutions.com/

Ok Zogzog. This was great. suppose they can open it. How about closing it.
there is no electrical charging mechanizm on this breaker. so it needs to be manually charged.
Thanks for the links.

Oh Humble!! You need replace your old breaker to new one
with close and trip coil option. Isnt joke, Im serious.
Its simple and cheaper.
Regards.
Slava

You manually charge it then hook up the remote operator and close it remotely, just like other plants that use this system do.

Best I can tell from info in your posts you have an (indoor?) substation with a transformer of unknown size and type, delta-wye connected, I assume MV/LV. It sounds like you have a fused disconnect switch(?) on the primary fed from an unknown source. The secondary of this transformer feeds a main breaker of unknown type with unknow protective devices to some design of switchgear lineup.

Your problem is a problem at every industrial plant I have been in and there are several solutions, the more details about your actual problem the more we can help recommend a solution. We understand the coord issue you are having, it is very common.

Help us help you.

Replacing the low side breaker does not help, as Humble has already pointed out. Faster high side protection is needed, such as a breaker, circuit switcher, or transrupter. I wonder if you could use the flash detector to initiate the closing of a high side grounding switch and shave a few ms. You did say you needed more fault current.

Or work it dead. See my posting in thread238-210239: Arc Flash Reduction Methods On Secondary of Sub-unit for a suggestion for getting around the catch 22 involved in de-energized work in this situation.

How do you get the system dead? You have to open the breaker, possibly rack it out to remove key for primary switch, and verify dead with a meter. Arc flash rules apply for all of these so "Work it dead" dosent remove arc flash concerns.

Zogzog, you didn't read his post in the referenced thread.

Yes I did, he is saying you have a reduced arc flash hazard on the primary and is right but you need to open the secondary main forst (Might even be key interlocked with the primary switch) and need to follow arc flash rules to do so.

While stevenal's method is safer for the voltage check and grounding part, you still have a breaker operation on the seconday side.

Common practice in the field is also to ground and test both the primary and seconday side of the transformer prior to work, grounds on the primary side of a transformer may not protect a worker on the seconday side from a backfeed on the secondary.

Lots of speculation here. A keyed interlock that increases the arc flash hazard is no longer acceptable and should be removed or redesigned.

NFPA allows a reduced amount of PPE for breaker operation with doors closed or covers on. If still unacceptable, downstream loads can be removed one by one. Or move upstream to find a suitable device. Or use the remote switch operator you suggested. Once the primary is opened, tested, and grounded, assuming there are no other sources, the manual recharging can occur without PPE.

Talk about speculation!! NFPA allows for reduced protection if you are using the tables and you can only use the tables if you are within the tables fault current and clearing times listed which we obviously are not here with such a high Ei.

I am not speculating anything, you need to open the main breaker before you open the primary switch. Thats why I suggested the remote operator in the first place!!!

After I suggested the remote operator you came in with the work it dead idea (Which as I pointed out is a fine idea but you still have the problem on the LV side for opening the breaker)then you suggested removing an interlock as a safety suggestion??

Ok. Guys. I just found out that according to safety codes, if the upstream device has an load intrupting rating ,key intelock is not required.so there is no interlock present.

However, the breaker needs to charge mechanically, so it means working on a breaker withing the boundary even if I use remote trip device.with high calory incident energy.

Is charging the breaker allowed without using PPE?

Now there is speculation that the high side switch is not load break capable.  If it is there is no reason it can't be opened first.

Right, if it is load break capable than you dont need an interlock, but if there is an interlock then advising to remove those interlocks is a bad idea IMHO.

As far as chraging the breaker goes, that is a grey area, the tables refer to breaker operations, which implies per IEE def of operations opening and closing, no mention of charging. The 2008 70E will use the term "Interact with switchgear: which implies charging is interacting. I have asked the 70E committe (Some members) in an open forum about charging operations and got some answer like "If it presents a hazards the PPE should be worn". What dies that mean? I dont know.

I am on a special risk assessment group that will be looking at these types of items, I will report our findings. But I believe (Again, my opionon) that charging would not produce a signifigant arc flash hazard, I believe racking from disconnect to removed and vice versa does not either. It is the making and breaking of a circuit that poses the hazard.

Now there is a possibility that when charging the breaker it could close on you (Charging latch failure or misadjustment), it is a long shot but possible.

What working distance are you assuming for the switchgear (LV side)Ei calcs?

Zogzog,
My handbook only says to perform an analysis, not specifying how the analysis is done. The tables imply the presence or not of covers or doors and the nature of the task may come into play. Not as clear cut as it could be, I agree.

Seems Humble does not need to open the breaker first.

My comments regarding the speculated interlock were not meant for anyone with a wrench, but for engineers familiar with evolving codes and safety issues qualified to make the call.

If work practices mandate testing and grounding all sides of a transformer prior to clearance, the catch 22 will remain.

He needs to open the main if any work is to be done on the system, why would you open the load break switch if you were not going to do any work??

I have seen plenty of EE's overide or remove key interlocks in substations.