New Family of Relay Curves

[apowerengr]

Question, with rule 410 in the 2007 NESC providing a table for PPE cal/cm^2 ratings based on clearing times and fault current availability has anyone heard if microprocessor based relay manufacturers or IEEE standards will develop a new family of relay curves based on the table? In other words, assuming you have 5 cycle breakers, you should be able to develop a curve for a 4 cal/cm^2 system with the correct maximum clearing times at 5, 10, 15 and 20 kA. Same for the 8 and 12 cal/cm^2 levels as noted in the table. If your breakers are faster - all the better. Or, maybe the table is based, in part, on a particular existing relay characteristic?

 

[arcflash99]

Haven't heard of new curves coming out.

I would be cautious using fast clearing times for breakers unless they have been timed and are well maintained.

 

[dpc]

I don't see why you would need a new family of curves. New relays should have the adjustability already to accomplish this. And with many, you can create custom curves.

You will still need to coordinate with downstream recloser and fuse curves.

 

[stevenal]

If you have a microprocessor relay, program it so your hot line hold (non-reclose) toggle enables instantaneous protection. You may be able to stay below the 2 cal/cm^2 level so that FR clothing is not required. Coordination is sacrificed for safety during live line work.

I see at least two papers dealing with arc flash at next month's Western Protective Relay Conference.

 

[oldfieldguy]

Some of my clients have used the ability to program multiple setpoints groups in order to set up for faster trip times in work situations: one setpoint group with settings for normal coordination/protection requirements and a selectable second setpoint group withe much faster trip times for use when work may take place on the equipment downstream.

Thye did this on GE Multilin SR750's. Any of GE/Multilin's microprocessor relays have the capability.

 

[davidbeach]

Pretty much any numeric relay has the capability of multiple settings groups.

 

[stevenal]

The transition from one group to another can cause a brief loss of all protection. I prefer to assign an input as a torque control. Transition is shorter, and the unaffected elements remain in play during the transition.

 

[davidbeach]

Steve, use the right relays and there is no loss of protection between settings groups. Use other relays and there can be a loss of protection for a few seconds. Torque control works well if you're going to lose protection, but if not the settings groups can be much cleaner.

 

[apowerengr]

I appreciate the insights and am familiar with the topics described. I don't necessarily agree with stevenal regarding being able to program the relay such that you can guarantee less than 2 cal/cm^2 level, however I do believe settings can help manage arc-flash energy.

My question is perhaps better stated this way:

Does anyone forsee relay manufacturers or the IEEE developing a set of relay curves based on table 410-1 of C2-2007?

Does anyone know from the code standpoint if meeting the required clearing time at the maximum available fault current level meets the intent of the code? An example: I can program a relay with a 9kA instantaneous trip to meet the 8 cal system rating for a 25kV system with a 10kA maximum fault current. This does not mean that the relay clears in the required time at 5kA to again meet the 8 cal system protection requirements. Since we can't guarantee bolted faults how do you set relays as a means to provide "engineering controls" for arc-flash energy management? IEEE 1584 methodology recommends arc-flash energy calculations at multiple fault current levels to determine the "worst case". C2-2007 rule 410 does not seem to address this?

 

[jghrist]

Building a custom curve based on cal/cm² would enable meeting the arc-flash energy requirements with maximum space for coordination with downstream protection. The application would also result in minimum space for coordination with upstream devices.

In most situations, I suspect that conventional inverse characteristics combined with instantaneous or definite-time elements could do just as well. The upstream and downstream devices probably will have inverse characteristics that would make coordination easier with conventional inverse characteristics.

Using a hot line hold toggle as suggested by stevenal would seem to be a lot easier if coordination during hot line work can be sacrificed. This could be accomplished either by torque control or alternate settings.

 

[davidbeach]

All you need is a relay that lets you define the curve. When you can enter the coefficients for the trip curve, as is possible with at least one brand of relays, you can make an I²T type of curve that you can then set for what ever level you want. You will need to set the curve to allow breaker time before exceeding your cal/cm² and you will find that the relay has a minimum operate time. Once you are above the breaker/relay minimum operate times, you should be able to have what ever tripping time you want for any current. That doesn't mean that you will coordinate with any other trip characteristics in the system.

 

[stevenal]

"I don't necessarily agree with stevenal regarding being able to program the relay such that you can guarantee less than 2 cal/cm^2 level.."

I believe I said "may", it's up to you to determine if and where it's possible on your system. The table doesn't include a 2 cal/cm^2 column, so you would need another method to find the amount of fault current for your instantaneous clearing time.

I've been programming our feeder microprocessor relays as I suggested above since we've used them. The linemen appreciate the fast tripping regardless of the NESC version in effect.

Bolted faults seem to be the rule for these voltage levels per 1584, just look everywhere on the system.