-
The Nine Checks
-
1. Confirm available fault current before you argue about anything else
-
2. Match the enclosure type to the real environment, not the catalog language
-
3. Check the bus, not the breaker
-
4. Open the contactor control box and read the BOM
-
5. Compare the as-built drawing to the submittal, line by line
-
6. Look at the gasket and the hardware on a weatherproof distribution board
-
7. Audit the cable entry before anything ships
-
8. Labels, torque tables, and the paperwork nobody asks for
-
9. Witness the test, or specify it in the contract
-
1. Confirm available fault current before you argue about anything else
-
Mistakes I See Repeatedly (Including Mine)
-
A Note on Small Orders
I'm the quality and brand compliance lead at a power solutions company. Every enclosure, panelboard, and control assembly that leaves our floor passes my bench first — roughly 200 units a year, plus incoming inspections on gear we buy from outside low voltage switchgear manufacturers. In 2024 I rejected about 18% of first deliveries. The biggest single reason wasn't sloppy workmanship. It was enclosure ratings that didn't match what the spec sheet claimed.
This is the checklist I run. Nine checks, in order. It works on a 400A power distribution board panel and it works on a $300 outdoor weatherproof connection box. It's written for facility managers, electrical contractors, and industrial buyers who have to sign off on gear they didn't design themselves.
One caveat before we start: this is a specification and acceptance checklist, not a substitute for the engineer of record's drawings or your AHJ's ruling. Where code applies, code wins.
The Nine Checks
1. Confirm available fault current before you argue about anything else
Everything downstream depends on this number. Get it from the utility or from the transformer data they publish, then compare it to the short circuit current rating (SCCR) of the assembly — not just the interrupting rating printed on the breaker. Those are two different things, and vendors blur them constantly.
A panelboard covered by UL 67, a switchboard under UL 891, and a custom industrial control panel under UL 508A can all land on the same job with three different SCCR values. If available fault current is 22 kA and the assembly is braced for 10 kA, that isn't a close call. It's a failure waiting for the right Wednesday. Even when you're ordering something as ordinary as an ac breaker box for a light commercial job, the fault current number still decides the rating.
Ask for the series rating documentation in writing. If they can't produce it, that's your answer.
2. Match the enclosure type to the real environment, not the catalog language
This is the step people skip, and it's the one that costs the most.
I once got this line in an email from a supplier:
"All our enclosures are weatherproof."
"Weatherproof" is not a rating. It's a word. Under NEMA 250, Type 3R means protection against rain, sleet, and snow with a drain — it is not dust-tight and it is not corrosion-resistant. Type 4 is watertight and dust-tight. Type 4X adds corrosion resistance. Type 12 is an indoor rating; put a Type 12 box on an outdoor pad in a wet climate and you've bought yourself a future service call.
IEC 60529 IP codes run on a separate scale. IP65 is dust-tight and jet-resistant; IP66 handles more forceful jets. Neither one covers corrosion, ice formation, or UV degradation. So a coastal installation calling for a weatherproof distribution board usually needs Type 4X or 316 stainless — not a painted 3R.
Reverse validation on this one: everyone told me to read the enclosure spec line by line. I only believed it after I didn't, once, and ate a $900 redo on a batch of boxes that got spec'd for an outdoor pad. The third time an enclosure mismatch showed up, I finally built a rating check into our approval form. Should have done it after the first.
3. Check the bus, not the breaker
Here's the thing: breakers get all the attention, but bus bars cause most of the arguments.
Ask three questions. What's the bus material — copper or aluminum? Is it tin-plated? And is the ampacity rating based on a 40°C ambient or a 50°C one? A rating that quietly assumes 40°C, installed in a mechanical room that hits 50°C in August, is a de-rated panel you didn't pay for.
UL 891 and IEC 61439-1 both set temperature-rise limits for assemblies, and the manufacturer should be able to hand you the test report without a week of email tag. If the answer is "our engineering team says it's fine," that's not a test report.
4. Open the contactor control box and read the BOM
For any contactor control box, I want a bill of materials with actual manufacturer part numbers — contactor series, coil voltage, auxiliary contacts, overload relay class. Not "or equivalent."
"Or equivalent" is a yellow flag rather than a deal-breaker, but it means someone has the option to substitute. If they do substitute, I want it in writing with the submittal revised before the build starts — not discovered at the factory test.
Check traceability too. Can the vendor tell you which production batch a contactor came from? If a relay fails six months in, that question has a dollar value attached to it.
5. Compare the as-built drawing to the submittal, line by line
This takes forty minutes and it's the highest-return forty minutes in the whole process.
When I compared Submittal Rev A against the as-built for two vendors side by side — same scope, same price, same delivery window — I finally understood why the cheaper one had been quietly costing us money. Fourteen small deviations. None of them individually fatal. Each one about twenty minutes of field rework. The gap between the two quotes was way smaller than the gap between the two outcomes.
Also worth knowing: getting the first submittal back took about three weeks — or rather, closer to four once you count the revision cycle. Build that into your schedule, not into your temper.
6. Look at the gasket and the hardware on a weatherproof distribution board
Nobody photographs the gasket. The gasket is where the enclosure rating actually lives.
Things to check: closed-cell vs open-cell foam, continuous vs seamed, whether the gasket is field-replaceable, whether hinges and latches are stainless, and whether the door uses quarter-turn latches or screws you'll lose on the first service visit. On Type 3R, confirm the drain and breather are present and clear. Powder coat thickness should be measured, not assumed — thin coats fail at the cut edges first.
The test is simple. Ask for a replacement gasket part number. If the vendor can't produce one, the enclosure was never designed to be serviced.
7. Audit the cable entry before anything ships
Cable entry is the most common field problem I hear about, and it's fully preventable at the drawing stage.
For an outdoor weatherproof connection box: are the gland plates removable, or are you drilling into a painted enclosure in the rain? Are the knockouts sized for the actual cable outside diameter, not the conduit trade size? Is bottom entry mandatory, and does the box come with a proper hub for any top entry? On a rated enclosure, top entry without a listed hub voids the rating — and the inspector will notice.
Specify sealing washers and listed hubs in the order. They cost almost nothing and they're the difference between a passing inspection and a second mobilization.
8. Labels, torque tables, and the paperwork nobody asks for
Three documents I want before I sign.
The arc flash label, per NFPA 70E, with incident energy and the study date on it. A nameplate showing the assembly SCCR. And a terminal torque table inside the door.
That torque table is the one people roll their eyes at. It's also the reason behind roughly half the loose-lug failures I've investigated. If the electrician in the field doesn't have the torque value, they'll use what feels right, and "feels right" is not a specification.
Ask for the factory test record too — dielectric withstand, continuity, and functional test for anything with controls. A certificate of compliance is a piece of paper. A test record is evidence.
9. Witness the test, or specify it in the contract
If the order is big enough to justify a factory acceptance test, go. If it isn't, put the FAT requirement in writing anyway so there's a record of what was agreed.
Honestly, I'm not sure why some vendors hit a FAT date inside a week while others slip three weeks on a comparable build. My best guess is internal buffer and how many other jobs are on the floor that month. Since I can't predict it, I build two weeks of float into any custom assembly schedule and treat an on-time FAT as a bonus rather than a baseline.
Mistakes I See Repeatedly (Including Mine)
Assuming the breaker's AIC is the assembly's SCCR. The assembly rating is limited by its weakest component plus the bus bracing. A 65 kA breaker in a 10 kA panel gets you a 10 kA panel.
Ordering spare breakers without checking the mounting type. Plug-on and bolt-on are not interchangeable, and this is a mistake you make exactly once, because the freight on the return is memorable.
Missing the third-party listing question. A UL 508A shop label and a field evaluation are different things, and some jurisdictions won't accept one where they'd accept the other. Ask before you order, not after the inspector arrives.
Buying on lead time alone. The shortest lead time and the lowest price are rarely the same vendor, and neither one is automatically correct. Three weeks faster on a product that fails the enclosure check is three weeks faster to a problem.
Skipping the ambient temperature rating. It's buried in the fine print and it matters in every mechanical room in the country.
A Note on Small Orders
When I was starting out, the vendors who took my $200 order seriously — quoted it the same week, answered questions without making me feel like an inconvenience — are the same vendors I now place $20,000 orders with. That's not sentiment. That's arithmetic.
Look, if you need one 6-circuit outdoor weatherproof connection box, that box is not a small project to the person installing it. I've had suppliers quote 30 days on a single panelboard and others quote 10 days on an identical spec. The difference was never capacity. It was whether they wanted to look at it.
That said, I'm not going to pretend a single unit should cost what a 50-unit run costs — setup, tooling, and testing time don't disappear because the quantity is small. What should be identical is the straight answer and the willingness to quote at all. A fair price and an honest lead time cost a supplier nothing.
One limitation worth naming: everything above assumes a domestic supply chain with standard products and predictable freight. Our situation is domestic batch manufacturing, and that shapes how much float I build in. If you're sourcing internationally with customs clearance and ocean freight in the mix, there are variables here I'm not going to pretend to understand better than someone who lives with them daily.
Bottom line: nine checks, forty minutes of drawing review, and one uncomfortable question about enclosure ratings. That's cheaper than any redo I've ever approved.
Standards referenced: NEMA 250 (Enclosures for Electrical Equipment, 1000 Volts Maximum), UL 67 (Panelboards), UL 891 (Switchboards), UL 508A (Industrial Control Panels), IEC 61439-1 (Low-voltage switchgear and controlgear assemblies), IEC 60529 (Degrees of protection provided by enclosures), NFPA 70E. Verify the current edition and your local code requirements with the AHJ before specifying. Pricing references are general and for context only — actual prices vary by vendor, specification, and time of order.