Was sitting in on a conference call 2

[oldfieldguy]

Was sitting in on a conference call today and one of the participants said that the recent installation of a high resistance grounding system resulted in an increase in arc-flash hazard. I've installed and worked around dozens of HRG systems in 480-volt systems and have done a few analyses of those installations. I've never seen a case get worse with the application of HRG.

Any ideas or comments?

old field guy

 

[waross]

A phase to phase arc should not be affected by the ground resistance.
A single phase fault may have a lower current times a longer clearing time.

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

 

[Parchie]

Agree. An HRG bordering on the "ungrounded" area would have the same behavior; lower ground fault current resulting to ground overcurrent relays tripping a bit delayed.

But there are always ways to prevent those from happening. IMHO, it is always best to use DT trip relays when ground fault is sensed, a far better proposition that having IDMT!

 

[dpc]

The mostly widely used calculation method for arc-flash hazards is IEEE 1584. In the original 2002 version, results for ungrounded systems are 10% to 15% higher than for grounded systems. These results always seemed odd because all of the calculations in IEEE 1584 are based on three-phase faults. But this difference was built into the method. Fast forward to the new version (IEEE 1584-2018) released late in 2018. Based on vastly greater test data and a careful analysis of the old test data, it was determined that there really is no significant difference between grounded and ungrounded systems, so this distinction no longer exists in the new version of the standard.

So... if 1584-2002 method is used, results for high-resistance grounded systems (treated as ungrounded per the standard) are indeed higher than grounded systems. But if the new IEEE 1584-2018 is used, there will be no difference.

Also - most faults (70% to 80%) at least start as ground faults. They often escalate to three-phase faults in solidly-grounded systems. In a HRG system, there is no possibility of a ground fault escalating to anything else. There isn't enough energy. So HRG does not change the calculated incident energy, since it is based on 3-phase faults. However it greatly reduces the risk of arcing faults occurring in the first place.

Cheers,

Dave

 

[Mbrooke]

Is it possible he is assuming a grounded phase that is never taken care of when the second ground fault hits?


I'm curious how the IEEE arrived on the original conclusion of HRG having a higher incident energy.

 

[dpc]

I'm curious how the IEEE arrived on the original conclusion of HRG having a higher incident energy

Testing was done on ungrounded and grounded wye systems. The resulting difference in energy was MEASURED. The equations used in IEEE 1584 are 100% empirical, derived from the test data. Even though they didn't have an explanation, the data showed a difference, so it was incorporated into the methodology.

When the standard was updated, additional testing was done. The original 2002 version was based on about 300 tests. The new 2018 update had over 1800 tests. In the newer testing, they did not see the difference between grounded and ungrounded. Their investigation into this discrepancy pointed to some issues with the measurement techniques and test setups in the original testing. In any event, the distinction between grounded and ungrounded no longer exists in IEEE 1584-2018, so it is really a moot point if using the current standard.

Quoting from Jim Phillips, who is on the IEEE 1584 committee, at his Brainfiller.com website:

"Grounded/Ungrounded
The 2002 edition made a distinction in the results based on whether a power system is grounded or ungrounded/impedance grounded. During the arc initiation stage when the fuse wire is melting, the arc can be very erratic i.e. unbalanced. The erratic arcing created a difference in the arcing current and incident energy during this very brief period that could be influenced by how the system is grounded. The new model is based on the more stable arc so there is no difference in the calculations based on grounding."

HRG systems were assumed to be more like ungrounded than grounded so the ungrounded factor is used for the 2002 version calculations.

Cheers,

Dave

 

[Mbrooke]

Interesting!