250.30(A)(4), 250.30(A)(5), 250.104(D)(2) - SDS Grounding Design

[gbstand]

I designed a grounding/bonding diagram for a service-entrance rated 480V panel feeding a 480V/208Y/120V transformer, which in turn fed a 208V panel. Please see partial screenshot of the diagram in photobucket link below which shows the configuration. Please note the “building structural steel” shown on the diagram is not the same physical column, but representative of structural steel, which was exposed and accessible in the building.

https://i44.photobucket.com/albums/f31/gstandif/Grounding%20Diagram_zps5jnqzefg.png

250.30(A)(4) Grounding Electrode; 250.30(A)(5), Grounding Electrode Conductor, Single Separately Derived System, & 250.104(D)(2), Structural Metal

A question was posed by the electrical project manager: Why is a separate GEC connection needed at TX-1 (highlighted in red on diagram)? His point was that the EGC from MDP feeding TX-1 ultimately provides a connection back to the building's GES and the GEC connection at the TX-1 to structural steel appeared redundant.

The response I provided was that while I agreed that his point was electrically valid, it appeared that a reading of 250.30(A)(4), 250.30(A)(5), and 250.104(D)(2) required the connection to the exposed structural steel in this case. I indicated that I suspected the underlying reason, as seen in this case, was because if the EGC from MDP-1 to TX-1 were compromised or any connections further upstream of MDP-1 going back to building GES were compromised, then the earth connection to TX-1 would be lost and TX-1 secondary voltage would float. The basic point being that by requiring a local, direct connection to the GES at the transformer (or other SDS) the NEC decreases the risk of harm to people and property.

Is this interpretation/understanding accurate?


250.30(A)(4), Exception 1 & 250.30(A)(5), Exception No.2 & 250.104(D)(1) and 250.104(D)(2)

While re-reading the exceptions to 250.30(A)(4) & 250.30(A)(5) and also 250.104(D)(1) & 250.104(D)(2), it appears there is, however, a code basis for not specifying the a local, direct connection to the GES at a transformer (or other SDS) in cases where (a) structural steel is not exposed in the area served by the SDS or (b) grounded metal water piping system is not present in the area served by the SDS. 250.30(A)(4) Exception 1 and 250.30(A)(5) Exception No. 2 seem to indicate that the EGC from the panel feeding a transformer can count as connection to the "building or structure grounding electrode system" required by 250.30(A)(4) if the conditions specified in the exception are met, for example, the panel supplying the transformer is “equipment listed and identified as suitable for use as service equipment”, etc.

From a practical design and installation standpoint, it seems the purpose of the exceptions is to allow the EGC pulled in with the primary conductors from the source panel to serve as the connection to a building or structure GES in cases where local, direct access to the existing GES is not available at the transformer (or other SDS) location. For example, if TX-1 were to be installed inside an existing electrical room with no exposed structural metal or water piping, then specifying a local, direct connection to the existing GES would not be practical; and, the exceptions could be used to credit the EGC pulled in with the primary conductors as the connection to the building or structure grounding electrode system.

Is this interpretation/understanding accurate?

Thanks in advance for responses.

 

[infinity]

One issue is that the supply side primary EGC is typically smaller than the required GEC which is based on the secondary condcutors. IMO the GEC for a transformer within a building doesn't do much of anything.

 

[gbstand]

Thanks for the response.

Good point on considering transformer supply-side EGC vs. GEC sizing when using 250.34(5) Exception No. 2.

My understanding is that Exception No. 2. permits us to not specify or install a local, direct GEC connection to a transformer inside a building, which is generally required by 250.64, if certain conditions are satisfied. From an electrical standpoint, my understanding is that under this exception the transformer supply-side EGC essentially functions as the GEC connection for the transformer secondary system and must be "of sufficient size."


Code Analysis - EGC and GEC Sizing

For sizing GEC for a single SDS, 250.34(5) indicates that the GEC "shall be sized in accordance with 250.66" and 250.34(5) Exception No. 2 indicates to ensure "the [supply-side] grounding electrode conductor is of sufficient size for the separately derived system." T.250.66 provides GEC sizes based on "size of largest ungrounded [phase] conductors." For a transformer the largest phase conductors would be secondary conductors and the transformer GEC size would (reasonably) be derived from T.250.66 by using the secondary conductor size.

For sizing the minimum supply-side EGC for a transformer T.250.122 would be used. The EGC size would be based on the rating or setting of the OCPD feeding the transformer supply-side phase conductors.

For example, if TX-1 in the diagram above were a 480/208Y/120V, 75 kVA transfomer, the EGC would be be sized at 8 AWG (based on 100A breaker) and the GEC sized at a 2 AWG (based on 4/0 secondary phase conductors).

The transformer supply-side EGC conductor is smaller than the GEC conductor.

From a code analysis perspective, note, however, that 250.34(5) Exception No. 2 does not read that the EGC must be sized in accordance with T.250.66 but rather "the [supply-side] grounding electrode conductor is of sufficient size for the separately derived system." While T.250.66 would be reasonable to apply to determine "sufficiency", it appears its sole use is not prescribed under Exception No. 2.

Does anyone have a differing code interpretation?


Electrical Analysis - EGC "Sufficiency"

Without the local, direct GEC connection at the transformer an evaluation would be needed to determine whether the transformer supply-side EGC is sufficiently sized to comply in part with 250.34(5) Exception No. 2. In my understanding, this involves considering a ground fault involving phase conductor(s) downstream of the panel fed by the transformer and structural steel or equipment directly bonded to structural steel. In this scenario, the predominate ground-fault path would be through structural steel and the transformer supply-side EGC to return to transformer Xo. The term predominate path is used because ground fault current would in reality divide itself across all grounded components in inverse proportion to the path impedance back to the transformer Xo terminal. (There are other fault scenarios with this system but in my understanding they would not involve the EGC as the predominate path. For example, if a phase conductor insulation were "burned" during a cable pull and subsequently touched metallic conduit it would predominately return to Xo via the panel ground bar, supply-side bonding jumper, and system bonding jumper path. The EGC would carry some current, but would not be the predominate path in my judgment.)

There is likely a technical basis that could be developed to justify the sufficiency of a smaller sized supply-side EGC. Consulting the reference identified in 240.4, Protection of Conductors, Informational Note, ICEA's Short Circuit Characteristics of Insulated Cables, the ICEA maximum 1 cycle withstand value for 8 AWG and 2 AWG insulated copper conductors is approximately 6.7 and 27.2 kA-RMS, respectively. These are the currents from the ICEA insulation damage level to raise the conductor from 75 degrees C to 150 degrees C.

Although the 8 AWG EGC 1 cycle withstand is significantly less than the 2 AWG GEC, there seems to be case for sufficiency. If it were conservatively assumed (a) the transformer supply-side EGC carried all the fault current (i.e., no current division), (b) fault current was bolted sourced by the 75kVA transformer, and (c) no circuit impedance to limit the fault current, the ground-fault current would be less than 5 kA-RMS. If it were further assumed the 100A breaker operated within 1 cycle, which is reasonable assumption for a UL 489 listed breaker, there appears to be a technical basis for sufficiency. A final determination could be reached by considering and modelling cable damage curves and TCC plots of the specific breakers in a particular application.

Ultimately a evaluation of supply-side EGC sufficiency would be accepted or rejected by a particular AHJ who challenged the design or installation under 250.34(5) Exception No. 2. Does anyone see any errors or omissions in this methodology?


260.64, Grounding Electrode - What functions does it provide?

Changing gears back to the local, direct connection of transformer located in a building to the building's GES in accordance with 260.64.

Whatever the merits (which are debatable), it appears the functions that the local, direct GEC connection of a transformer to the building's GES provide to some degree are:

1. "Earthing" of the transformer secondary system by connecting Xo to the building or structure GES
2. Predominate ground-fault return path to transformer Xo in certain fault scenarios

From an electrical perspective, does anyone see other functions the 260.64 connection provides for a transformer in a building?

From a code perspective, is there a code basis for not specifying or installing the a local, direct connection to the building's GES at a transformer in a building where (a) structural steel is exposed and accessible and/or (b) grounded metal water piping system is present and available?