Differential protection for reducing arc-flash hazards

[sonic02]

Hi there,

Some of the traditional methods for reducing arc-flash is to have to bus differential protection for quicker tripping and at times reducing incident energy on the bus.
Now, as far as I know, most of the electrical softwares (EasyPower, ETAP and SKM) allows you to model only overcurrent protection for a modern protective relay in a system ( Please correct me if i'm wrong here) and not differential.

Now, suppose you have a high incident energy (CAT #4) on a system bus that you modelled in an electrical software but due to the software capabilities, you are only able to reduce the clearing times for 50/51, amongst others, to reduce the incident energy on that particular bus. But you know that bus differential protection is being used but you are unable to apply it to the software to get a reduced incident energy. What is the best way to circumvent that? Do you typically not include differential protection for arc-flash calculations as that would give you the best case, and that is not something you want in an arc-flash hazard analysis ?


Thanks.

Sonic

 

[davidbeach]

Does the software in question allow the use of a user selected trip time? If so use that to produce the desired clearing time.

 

[jghrist]

SKM has a differential function. See

http://www.skm.com/faq_ptw12.html

 

[dpc]

In EasyPower you just enter a user-defined time for the buses inside the differential zone. Be sure to include the breaker clearing time along with the worst-case relay operating time (plus any other time required, like a lockout relay, and then add a little extra time as a margin of safety.

 

[sonic02]

Hi All...thank you for all your responses...

 

[burntcoil]

The incident energy levels are usually high on Low Voltage switchgears due to high bolted faults current and differential protections are not used on low voltage switchgears. If dedicated Arc Protection System is not installed, clients requiring Arc Protection tends to prefer Zone Selective Interlocking.

In terms of defining Incident Energy on softwares like SKM or ETAP, you can enter the clearing time and calculated the incident energy.

 

[rcw retired EE]

In earlier versions of Easy Power we put instantaneous overcurrent (50) relays with time delay to simulate the differential. It worked for radial feeds.

Burntcoil, - "The incident energy levels are usually high on Low Voltage switchgears due to high bolted faults current and differential protections are not used on low voltage switchgears."

I wish that was always the case.

I am attempting to commission a system that has several large 13.8 kV switchgear lineups and many double-ended 13.8kV- 480V substations, 1 MVA to 3.5 MVA with full transformer differential protection. All switchgear has GE Multilin B90 full bus differential, ABB REA optical arc flash protection relaying for each cubicle, Zone-Interlocking, Maintenance Switches for every breaker & bus to switch all overcurrents to minimum settings and Main-Tie-Main transfer schemes. Since the B90 has breaker failure functions as do the ABB RET & REF feeder relays, the motor protection SEL710's and GE Multilin 750 M-T-M relays, the client's consulting engineers decided to implement all those functions at all voltages.

But the bus differentials can’t pickup ground faults because all systems are high resistance grounded at 400A for 13.8 kV and 5A at 480V.

All 480V switchgear and MCC buckets have 50GS sensors that can sense the 5 A fault levels, alarm if it is the first fault and trip if it is the second ground fault.

With all this protection, the engineer’s ETAP arc flash protection program came up with very low hazard levels due to the “instantaneous” arc detection and the 3 cycle breakers, even with tie breakers closed during transfer. I am skeptical.

The specs also call for the system to ride through a 4 second voltage “dip” without tripping process motors. The dip was defined as any voltage less than 70% including zero. All motor starter coils are operated off UPS power to achieve this.

I wonder if I will ever get it energized and fully tested.

Will the client ever be able to keep it on line with all this protection?