Control panel design: why your 65 kA panel is actually rated 5 kA

The panels I get called out to look at almost always started with the same brief: build it as cheap as possible. So the shop used whatever was in the bin. I have no problem with that, I do it myself. What bothers me is the nameplate. Blank where the short-circuit current rating goes, on a panel hanging off a service that can push serious fault current. Control panel design gets treated as a layout job, wire routing and duct fill and where the drives sit. The thing that stops an inspection is arithmetic nobody did.

A back panel like this is normal work. The blank rating plate on the door is what fails a control panel design review.

SCCR is the smallest number in the box

Short-circuit current rating is "the fault current at a nominal voltage to which an apparatus or system is able to be connected without sustaining damage exceeding the defined acceptance criteria." UL 508A's own wording, Chapter 2.44. NEC Article 409.110(4) puts that rating on the panel label, either matching a listed and labeled device or established by an approved method, and Supplement SB of UL 508A is the method.

Then the part that catches people. It is not your main breaker's rating, and it is not an interrupting rating. Whatever sits lowest in the power circuit, that is your panel. Rockwell writes that the total SCCR is that of the lowest-rated component, citing SB-4.4.2(c)(1), in publication SCCR-AT002B-EN-P. Eaton's control panel design guide (BR033001EN) says the smallest short-circuit value of all components in the main circuit determines the rating of the entire assembly. Siemens puts it the same way in reference manual A5E02118900A. Three manufacturers, one sentence.

The power circuit is every circuit minus the control circuits: disconnects, supplementary protectors, branch circuit protective devices, bus bars, fuse holders, load controllers, overload relays, receptacles, terminal and distribution blocks. Transformers, reactors, current transformers, dry-type capacitors, resistors, varistors and voltmeters are excepted.

One trap sits inside that. SB3.2.1 pulls the control circuit's primary protective device back into the power circuit count, while everything downstream of it stays out.

The default table is the part that hurts

Unmarked components get no benefit of the doubt. UL 508A Table SB4.1 assigns them a value, and those values are built to be conservative.

UL 508A Table SB4.1, reproduced as Table 6-15 in Siemens A5E02118900A.

Read the top row again, then remember SB3.2.1. An unmarked supplementary protector on a control transformer primary caps the whole panel at 200 amps of fault current. Two hundred. Not the control circuit. The panel.

65 kA of hardware, 5 kA on the nameplate

Rockwell's worked example in SCCR-AT002B-EN-P is worth walking through, because the shape of it keeps showing up in the field.

Feeder breaker, manufacturer rating 65 kA. Power distribution block, unmarked, 10 kA default. Three of the four motor branches carry tested manufacturer combination ratings of 65 kA each. The fourth is a breaker, a contactor and an overload relay, none of them marked, so all three sit at 5 kA.

Panel SCCR: 5 kA at 480V.

One branch out of four drops the rating by a factor of thirteen. Their next example is the one that gets the careful people. Every motor branch is a tested 65 kA combination, nothing wrong on the motor branches at all, but the feeder breaker with its distribution block is a combination rated 35 kA. Panel SCCR: 35 kA. The weak link moved upstream and every branch rating stopped mattering.

One unmarked branch, defaulted under UL 508A Table SB4.1, sets the rating for the entire panel.

Three ways up, and only one of them needs a purchase order

First is paperwork. A rating comes from the component's marking, from Table SB4.1, or from the manufacturer's tested combination data. Same part, different combination, different number, which is why Rockwell tells you to verify the rating for every component or combination you are using (SCCR-AT002B-EN-P). Siemens has the clean example: a 3RT2015 contactor marked 5 kA reaches 65 kA at 480V when it sits behind a 3RV2.1 or 3RV2.2 combination motor controller rated no higher than 16 A, per its UL Certificate of Compliance. The contactor did not change.

Second is a current-limiting device in the feeder. Eaton runs a single one-line diagram through three versions. As drawn it is 5 kA at 480Y/277V, held there by a branch breaker marked 5 kA. Add a Class CC fuse with peak let-through under 5 kA in the feeder, step that branch breaker up to a 10 kA part, and the panel becomes 10 kA. Change the branch device again for a 65 kA unit and step up the main: the same panel reads 65 kA at 480Y/277V. Identical motor contactors in all three drawings.

Two rules on that, both from the same guide. The current-limiting element goes in the feeder and cannot be the branch protective device of the branch you are strengthening. And every feeder protective device needs a breaking capacity at least equal to the panel SCCR you are claiming, because UL 508A does not recognize series ratings for back-up protection.

Third is replacing the weak device, which is what everybody reaches for first anyway.

Control panel design with surplus parts

Here is the version that costs me sales. A used contactor in good condition is electrically fine. What often does not survive the trip is the label, the instruction sheet, or a readable nameplate. In control panel design terms an unreadable marking is the same as no marking: default table, 5 kA, and no amount of "it's a good breaker" moves that number.

So buy the documentation with the part. Ask for a photo of the nameplate before it ships. Check the catalog number against the manufacturer's own combination data instead of the figure you remember from the last job. On older drives, keep in mind that UL 508C was replaced by UL 61800-5-1 in February 2020 (Rockwell CMPNTS-SR002D-EN-P), so surplus stock may carry only the legacy listing.

None of that argues against used parts. It argues against undocumented parts, which is a different problem, and the fix is usually a phone call.

The numbers that end the shop-floor arguments

A few control panel design debates have a table sitting behind them.

Control circuit conductors, UL 508A Table 38.1: 10 A wants 16 AWG, 7 A wants 18 AWG, 5 A wants 20 AWG. And 20 AWG or smaller is permitted only for control circuits feeding electronic programmable I/O and static control with no moving parts. Everything else in the control circuit is 18 AWG minimum. So a 10 A supply wired out through 20 AWG is two sizes undersized, and the supply's internal current limiting does not fix it. Protection sizes to the conductor, not to the device on the end of it. Table 66.5 caps it: 14 AWG in-panel control wiring tops out at an 80 A branch device, 60 A if the control is remote.

Power circuit conductors are copper, 14 AWG minimum, sized for at least 125% of full-load current on a single load. For a mix: 125% of all heater loads plus 125% of the largest motor plus the full-load amps of everything else that can run at the same time.

NEC Table 310.15(B), as printed in Eaton BR033001EN.

Motor branch protection, as a percentage of full-load amps: non-time-delay and Class CC fuses up to 300%, time-delay fuses other than Class CC up to 175%, inverse-time breakers up to 250%, self-protected combination motor controllers at 100%.

Spacing is measurable, so measure it. UL 508A Table 10.2, feeder circuits at 251 to 600V: 1 inch through air, 2 inches over surface. Table 10.1, branch circuits at 301 to 600V: 3/8 inch through air, 1/2 inch over surface. Section 10.8 applies the branch figures to fuse holders wherever they sit.

The old habit of leaving room for fat wire has a table too. NEC Table 430.10(B), for the terminals of enclosed motor controllers, specifies nothing at all for 14 through 10 AWG, then 1-1/2 inches for 8 and 6 AWG, 2 inches for 4 and 3, 2-1/2 inches for 2, 3 inches for 1 AWG. Which is exactly why nothing bites you until #8.

Below #8 the code stops leaving control panel design to your judgment and gives you a number.

Labels and the five-year test

Every power circuit conductor gets labeled at its termination with the letters or numbers from the wiring diagram. Requirement, not courtesy.

The nameplate is where the compliance actually lives. NFPA 79 wants the manufacturer's name or trademark, a serial or production number, rated voltage with phase count and frequency and full-load current for each supply, the ampere rating of the largest motor or load, the maximum rating of the short-circuit and ground-fault protective device, the panel SCCR, and the electrical documentation number. Permanently attached, still visible after installation and commissioning.

UL 508A lets you fasten it with screws, rivets or adhesive, and lets the text be die-stamped, silkscreened, etched into metal or plastic, or printed in indelible ink on an adhesive-backed label. That last option is the one I have watched fail, and not only on nameplates. The same adhesive label shows up on wire markers, and I have opened panels to find blank white rectangles where the wire numbers used to be, because the cabinet ran warm for a few summers. The standard permits it. Heat does not care what the standard permits.

Colors are where house practice and UL 508A part ways. Black is any ungrounded conductor, AC or DC. Red is ungrounded AC control. Blue is ungrounded DC control, white with a blue stripe is grounded DC control. Orange means the conductor stays live with the main disconnect open. Rockwell's own factory-built panels match that exactly: blue for 24V DC control, blue and white for common, red for 120V AC control, black for 120V AC power (publication 1711-TD001A-EN-P). Run red and black for 24V DC and the black half is defensible, the red half is not. It only bites when the panel has to be listed, which is why the habit survives where listing is rare.

Two more from that publication, both worth stealing: slotted duct sized for 60% fill maximum, and end anchors on every DIN-rail-mounted component. Terminal blocks that can slide will slide.

The standard permits this label. It does not promise you will still be able to read it in five years.

Frequently asked questions

How do you calculate SCCR in control panel design?

Four steps, per UL 508A Supplement SB. Establish each power circuit component's rating off its marking, out of Table SB4.1, or from tested combination data (SB4.2). Look for current-limiting components in the feeder that let you modify it (SB4.3). Take the lowest surviving value as the panel rating (SB4.4). Mark it on the nameplate (SB5.1).

Does every industrial control panel need an SCCR on its label?

NEC Article 409.110(4) requires an SCCR marked on the panel, and Article 409 now covers panels at and under 1000 volts rather than 600. Since the 2017 code you also field mark the date the calculation was performed, per Article 670.5(2).

Can you use surplus components in a UL 508A panel?

Yes, when the rating is documented. The standard cares about a component's established SCCR, not its age. An unmarked or illegible component drops to the Table SB4.1 default, which is 5 kA for most breakers, overload relays and small motor controllers.

Which control panel design mistake costs the most?

Assuming the main breaker's interrupting rating is the panel rating. The lowest-rated device in the power circuit sets the number, and it is usually something small and cheap that nobody looked at twice.

Closing

Most of what makes control panel design pass or fail on inspection day is documentation, not hardware. The parts themselves are usually fine. If you need a contactor, breaker, terminal block or control transformer to finish a build, we keep surplus and refurbished versions of most of them in stock, and we will send the nameplate photo up front so you can check the rating against your own calculation instead of taking anyone's word for it.

https://iqelectro.com/collections/all