36W vs 48W Tri-Proof Lights, 100W Floodlights and Emergency High Bays: A Buyer's FAQ

I'm not a lighting designer. I'm the person who gets the work order when a fixture dies at 4:40 on a Friday.

I've been the office administrator handling facilities purchasing for a 340-person company since 2020 — roughly $180,000 a year across eleven vendors, and lighting is about a fifth of that. Warehouse tri-proofs, parking lot floodlights, high bays, office troffers, and a running case of LED bulbs for the residential units above our main building. I report to operations, and finance audits me quarterly.

What follows are the questions I actually get asked. By my ops manager. By the maintenance tech. By the controller who wants to know why we didn't take the cheaper quote.

We'll get into: 36W vs. 48W tri-proofs. LED bulb payback in a mixed residential and commercial building. IP ratings. Emergency high bays. 100W floodlights. And the spec that isn't on the sheet at all.

1. 36W or 48W tri-proof — which one do I actually need?

Neither number tells you what you're getting. Wattage is input power. What lights the floor is delivered lumens, and how those lumens land is the optic.

For a typical aisle in a 900 m² warehouse at 4–4.5 m mounting height, a 36W tri-proof in the 3,600–4,300 lumen range usually covers it. Wider aisle, taller racking, or mounting above 5 m, and 48W is the honest answer. A 48W fixture at 4 m is light you've paid for twice — once at purchase, once on every utility bill after that.

The trap is comparing watts to the fixture you're replacing. We swapped old twin-lamp fluorescent tri-proofs for 36W LED units on the theory that the spec sheets matched. They didn't. The old fixtures had a lensed diffuser doing the spreading. The new ones were clear polycarbonate. Same claimed light on paper, completely different floor in practice.

If you're comparing two fixtures, ask for the IES LM-79 photometric report, not the marketing summary. That's the industry test method for measuring an LED luminaire's real output, distribution, and efficacy — and it's the only number worth arguing about.

2. Do LED bulbs actually pay back in a mixed residential and commercial building?

Yes, but the payback math most people run is the wrong math.

The simple version: a 9W LED replacing a 60W incandescent saves about 51W per socket. On our commercial rate, that's roughly $9 a year per socket at 6 hours a day. Run that across 180 sockets and it's a real number.

Here's where it went sideways for us. The numbers said go with the cheaper lamp — 40% less, same claimed lumens, same claimed color temperature. My gut said stick with the brand we already had in the warehouse. I went with my gut, and I'm still not certain that was the right call, because the reason turned out to be something neither of us had on a spreadsheet.

We replaced 180 lamps in the downlights of the residential units above our main office. Over 14 months, 22 failed. Not the LEDs — the drivers. Enclosed downlight cans trap heat, and the driver is the part that dies. The fixture was never the problem. The lamp's thermal design was.

If you're buying LED bulbs for residential and commercial lighting in the same building, the useful filter is this: check whether the lamp is rated for enclosed fixtures, and check what the warranty actually covers. A 5-year warranty in a hot can is a different promise than a 5-year warranty in open air. ENERGY STAR lists lamps, but the listing alone won't tell you whether one survives your ceiling.

3. IP65 vs. "wet location" — do I need both?

They're two different languages, and people treat them as the same word.

IP65 is an ingress protection rating under IEC 60529 — dust-tight, and protected against water projected from a nozzle. "Wet location" is installation language from the electrical code and UL, describing the environment where the fixture goes. A fixture can carry one and not the other, and the code language is the one your inspector cares about.

IP ratings, briefly: IP65 = dust tight + water jets. IP66 = dust tight + powerful water jets (high-pressure washdown). IP67 = temporary immersion. The second digit is progressive, but a higher number doesn't automatically mean a better fixture for your application.

I still kick myself over our wash-down bay. We specified IP65 tri-proofs — correct for the rating. What failed wasn't the fixture. It was the cable gland, which was plain IP44. Six months of hosing later, we had water inside three fittings. The fixture was IP65. The installation wasn't.

Two things to add to any spec: the gland and entry-point rating, and an IK impact rating under IEC 62262 if the space sees forklifts, pallet trucks, or anything that swings. Honestly, we've lost more fixtures to impact than to water.

4. If our high bays have emergency backup, can we skip separate emergency lighting?

I thought so. That was the expensive assumption.

An emergency high bay — a high bay with an integrated battery pack — is a genuinely useful thing. It keeps light on when the circuit drops. What it does not do is satisfy an egress lighting requirement.

Emergency lighting under NFPA 101 and NEC Article 700 is about illuminating the path people actually walk out on, for a minimum duration — 90 minutes is the standard figure — plus marking the exits. A high bay at 9 m above the racking lights the top of the racking and the aisle around it, not a continuous strip along the egress route. Ours runs along the wall. We ended up buying wall-mounted emergency units anyway.

The surprise wasn't the cost of the emergency high bays. It was that they didn't get us out of the other fixtures we'd been told to remove.

Two practical notes. Emergency equipment in the U.S. is generally listed to UL 924 — ask for the listing, and ask whether the unit self-tests. And budget for the batteries. They're a wear item, typically 5–7 years, and almost nobody puts them in the capital plan. Verify your specific requirements with your local authority having jurisdiction.

5. How do I compare two 100W floodlights that look identical on paper?

Stop comparing watts. Watts are the input. Lumens per watt is where the money is.

A 100W floodlight at 120 lm/W puts out roughly 12,000 lumens. A 100W floodlight at 90 lm/W puts out 9,000. Same number on the box, 25% difference in light — and the cheaper one is very often the dimmer one.

Then check the distribution. A 100W unit with a wide beam floods a wall evenly and does almost nothing useful on a pole at 8 m. A narrow-beam unit on a pole throws light past the area you're trying to cover. Ask for the beam angle and the light distribution diagram, not just the peak candela figure.

Three more filters that have saved us money:

• Lumen maintenance. "L70 at 50,000 hours" is a projection under IES TM-21, calculated from LM-80 test data. If a vendor can't produce the test reference, the number is decoration.
• Surge protection. Pole-mounted outdoor fixtures take strikes. 10 kV surge protection is a reasonable spec in our region. Ask what's actually inside the housing.
• Driver replaceability. Ask whether the driver is a standard part you can still buy in five years. We've replaced entire fixtures because the driver was a sealed, discontinued module.

If you're chasing a utility rebate, check the DLC Qualified Products List — and check the listing date, plus whether your exact wattage and color temperature variant is on it. A listing for the family isn't a listing for your part number.

6. What's the most important thing that isn't on the spec sheet?

The warranty process. Not the length — the process.

Everyone advertises five years. Almost nobody tells you what happens in year three when three fittings fail at 8 m over a racking aisle.

Ask these before you sign anything:

• Who pays the labor to swap a failed fixture, and how long a window do you have to file?
• Do you stock the driver for this model, or do I wait on a shipment?
• Will you ship the replacement from the same production batch so the color matches?

That last one matters more than people expect. LED color shifts slightly batch to batch, and a replacement fixture that's 200K off its neighbors looks wrong in a way that's hard to explain to a customer walking through the space.

7. What do I check before I actually place the order?

This is the part I'm paid for.

In 2021 I found a great price from a new vendor — about $3,100 under our regular supplier on a fixture order. I placed it. They sent a handwritten receipt. Finance rejected the expense, and I ate it out of the department budget. Now invoicing capability is the first thing I verify. Before price.

My pre-order checklist, in order:

1. Can they invoice properly? Net 30, PO-referenced, line-item part numbers that match the quote. No exceptions.
2. What's the committed ship date? In writing. "Typically 6–8 weeks" is not a ship date.
3. Can I get spares? Order 3–5% extra from the same batch, especially for anything visible to customers.
4. Who is my warranty contact? A name and a phone number. A support portal is not a contact.
5. Does the quote include the parts that make the rating real? Glands, brackets, mounting hardware, surge modules.

On our last warehouse refresh we standardized on Zumtobel high bays for the main aisles after running three options on one aisle for a month. It was not the cheapest quote. It was the one where the driver, the photometric file, and a warranty contact all existed before I had to ask for them.

That's the pattern, honestly. When I say the cheapest quote has cost us more, I do not mean occasionally. I mean roughly half the time, across 200-plus fixtures over four years — and usually not on the fixture itself. On everything around it.

Prices referenced are as of early 2025; verify current rates. Code and regulatory details vary by jurisdiction — confirm with your local authority having jurisdiction before specifying.