Mike Dolt Enterprises Logo Mike Dolt Enterprises

Topic: Size of bonding jumper to equipment rack

User: Matt Donovan, Location: United States

Asked on: Monday, September 30, 2024 at 4:20 pm EDT
Answered on: Thursday, March 6, 2025 at 3:11 am EST

Question: I have an equipment rack on a rooftop which holds an 800A panelboard and a few pieces of equipment fed from that panelboard. The rack is made of unistrut, and is bolted to the roof with OMG anchors, but not bolted to the building steel (so, not electrically connected to the building structure). Do I need a bonding jumper from the panelboard to the equipment rack? If so, what size does that jumper need to be?What everyone tells me is that I need the jumper if the rack is "likely to become energized". I can't imagine it becoming energized, since there's no exposed wiring on it - wiring is all in conduit. How do most AHJs interpret this?In terms of sizing, it seems "obvious" to size the jumper for the size of the feeder equipment ground. I have a #3/0 feeder EGC. Do I really need a #3/0 bonding jumper landing on the unistrut rack? Usually I see #6 for this stuff, but don't know what the justification is for that sizing.

Okay, this is a common situation and a point of confusion. Let's break down the requirements for bonding the equipment rack on the rooftop and address the sizing issue.

The Need for a Bonding Jumper: The "Likely to Become Energized" Conundrum

You've hit the core of the problem. The NEC uses the phrase "likely to become energized" which is subjective and open to interpretation. Here's a breakdown of how to think about it and how AHJs typically interpret it: The NEC's Intent

The intent of bonding is to create a low-impedance path for fault current to return to the source, allowing the overcurrent protective device (breaker or fuse) to trip quickly and safely in the event of a fault. If a metallic structure
could become part of that fault current path, it needs to be bonded.

Why "Likely to Become Energized" Matters

The phrase is designed to differentiate between situations where the structure is inherently part of the electrical system (like an enclosure) and situations where the risk of energization is very remote.

Common AHJ Interpretation

Most AHJs will interpret a metallic equipment rack supporting electrical equipment as likely to become energized, even if the wiring is in conduit. The reasons are:

Conduit Failure

Conduit can be damaged (weather, physical impact, etc.). While designed to protect the wiring, it's not foolproof. A fault within the conduit could energize it, and subsequently the rack.

Equipment Failure

Equipment inside the rack (the panelboard itself, or anything fed from it) can fail. A loose connection, insulation breakdown, or other fault could energize the enclosure, and again, the rack.


"Reasonable Foreseeability"

AHJs tend to use a "reasonable foreseeability" standard. Is it reasonably foreseeable that something could happen that would energize the rack? The answer is usually yes, given the presence of electrical equipment.

Your Argument (Conduit is a Barrier)

Your argument that everything is in conduit is valid,
but it's unlikely to sway most AHJs. They will point to the possibility of conduit failure and equipment failure as reasons to require bonding.

Therefore,
generally speaking, you will need a bonding jumper from the panelboard to the equipment rack.
Sizing the Bonding Jumper

This is where it gets interesting, and where you're right to question the typical practices.
NEC 250.92(A)(1)

This is the main governing rule.

(1) General

This section describes that the rack, in addition to other items, shall be bonded to an equipment grounding conductor. This isn't really a sizing section, it's more of a "what needs bonded" section.


NEC 250.102(C)

This is the sizing section. There are three options for sizing your conductor:

(1) Supply-Side Bonding Jumper

Size the bonding jumper per Table 250.102(C)(1). This table sends you to the
unfused overcurrent protective device (OCPD) for the system.

(2) Load-Side Bonding Jumper

Size the bonding jumper per Table 250.122, based on the size of the circuit OCPD, not the service OCPD.

(3) Grounding Electrode Conductor Connection

If you're bonding the rack to a grounding electrode conductor, you need to use 250.66 to size it.


So Which Do I Pick?
Since you are bonding the equipment rack to the panelboard, which is tied to the feeder, option (2) (Load-Side Bonding Jumper) is the correct method. You are effectively creating a short circuit to ground to trip the breakers. Table 250.122 is what you use to size your EGC, which you said is #3/0. Table 250.122

This table provides the
minimum size of equipment grounding conductors based on the rating of the overcurrent protective device (breaker/fuse) protecting the circuit. The table has columns for copper and aluminum.

Your Specific Case

You mention a #3/0 feeder EGC. Let's say the panelboard is protected by an 800A breaker. If your main breaker is 800A, then the minimum size for the bonding jumper is a #3/0 copper conductor.

You are correct, in that the calculated minimum is #3/0.
Why You Often See #6 AWG

This is where the confusion arises. Here are possible reasons why you commonly see smaller bonding jumpers: Misinterpretation/Misapplication

Sometimes, electricians incorrectly apply the rules, or simply don't realize the proper way to size the jumper.


Local Amendments

Some jurisdictions may have local amendments to the NEC that permit smaller bonding jumpers in specific circumstances (though this is rare for equipment racks). Check your local codes.

It
Could Be Okay: Load-Side Bonding Jumper

Remember, you can run a smaller bonding jumper based on the
individual circuit OCPD if you follow 250.102(C)(2). However, that seems impractical.

Recommendations
1. Consult Your AHJ

Always discuss this situation with your local AHJ before you install anything. Explain your reasoning, but be prepared to follow their interpretation of the code. They have the final say.2.

Document Everything

Keep detailed records of your calculations and justifications. This will be helpful if there are any questions during inspection.3.

Consider a "Belt and Suspenders" Approach

If the AHJ insists on a #3/0 bonding jumper (or even larger based on an 800A OCPD) and you're concerned about practical issues of terminating such a large conductor on the rack, you might consider:


Multiple Smaller Conductors

Use two or more smaller conductors (sized to equal the ampacity of the required #3/0) run in parallel. This can make termination easier.*

Bonding to Multiple Points

Bond the panelboard to multiple points on the rack. This can distribute fault current.

In summary: While it seems counter-intuitive that a #3/0 bond is required on a rack where all the wiring is in conduit, based on most interpretations of the NEC, this is likely what will be required. Always check with your AHJ before installing anything!
Flag for review

Previous Question Next Question