On a Tuesday in March, our most patient client asked a question that sounds simple: how to replace fluorescent tube with led in a 40,000-square-foot office building. I typed that exact phrase into a search bar and assumed I'd have a purchase order written by lunch.
I didn't. The project ended up with a mix of LED T8 tubes, a Legrand sensor switch in one very awkward conference room, and a conversation about zigbee wifi that made me rethink how I buy lighting controls.
For context, I manage lighting materials procurement for a mid-sized electrical contractor. I've been doing budgets and vendor reviews for seven years, mostly commercial lighting retrofits and tenant improvements. My job is not to be the electrical engineer. My job is to know when the cheapest quote is really the most expensive one.
The Ballast Problem Nobody Mentions
The building had about 325 fluorescent fixtures. The facility manager said, “Just get us LED tubes.” I believed him for about an hour. Then I read the fine print on retrofitting: there's T8 and T12, instant-fit Type A, direct-wire Type B, and external-driver Type C. The phrase “LED tube” hides a lot.
I called our usual supply house. The rep asked: “Are the fixtures using original ballasts?” I said, “Probably.” He said, “Get the fixture model and a few photos.” This slowed everything down. That annoyed me until I realized why it mattered: old electronic ballasts sometimes aren't compatible with plug-and-play LED tubes. If a ballast is near end of life, you can end up paying an electrician to diagnose a fixture that doesn't light. And that labor cost eats whatever you saved on the lamps.
People think an LED retrofit is about replacing a lamp. It's actually about the system around the lamp: the socket, ballast, switch, and how the light is controlled. The lamp is the visible part. The rest of the system is where money gets hidden.
The Quote That Almost Fooled Me
I compared four vendors. The lowest quote on tubes came from an online seller: $3.49 per lamp, 325 lamps, free shipping, for a total that looked great. I was about to approve it. Then I asked myself the question I ask on every order: what happens when it goes wrong?
The online seller had a phone number but no local rep. The lamp manufacturer's warranty said the lamp had to be used with specific ballasts. We had a ballast population from 2009 and no inventory of ballast models. That's not a risk; that's a bet.
The local lighting distributor quoted $5.80 per lamp for a ballast-bypass LED tube and explained the extra work: remove the ballast, rewire the fixture correctly, and test the sockets. That scared a few people on our side. But the distributor also included a sample installation after the electrician tested three fixtures. The added cost was around $700. That looked like an unnecessary add-on until I thought about the cost of a failed installation in a finished office.
I knew I should have done a test install before signing anything. I almost skipped it because “it's only a tube swap. What are the odds?” Well, the odds caught up with me during the trial: one fixture in the sample set would not turn on. We assumed the LED lamp was faulty. It wasn't. The ballast was too old to start it reliably. The vendor replaced the tube, and the fixture still flickered. Then the electrician put an old fluorescent tube back into the fixture, and it flickered too. In other words, the tube was fine—the ballast was the problem. I'd been blaming the wrong component.
A few people think the answer is “just remove all ballasts and use Type B.” That's a legitimate approach, but not every building has rewiring in the budget. For this one, the pragmatic route was to replace the tubes that made sense, remove the old ballasts on the floors receiving new sensors, and leave the existing wiring where we weren't touching anything. It wasn't the lowest unit cost. It was the lowest responsible cost.
The Legrand Sensor Switch and the Zigbee Question
Here's where the scope grew. The office had several conference rooms where the lights were left on all night. The facility manager asked for occupancy sensors. I assumed that meant ceiling-mounted sensors and someone running low-voltage wire through the finished ceiling. That was the old way I knew—and honestly, it's expensive. Traditional wired controls still make sense in exposed shells, but in a finished office, opening walls and ceilings is a different budget conversation.
Our distributor suggested testing a Legrand sensor switch in the worst room. It fit into a standard wall box and controlled the LED load directly. No low-voltage panel, no extra wiring through the ceiling. The wall switch behavior felt normal to the people using the room, which matters more than anyone in procurement wants to admit. The Legrand switch replaced the existing wall switch, gave the room automatic shutoff when unoccupied, and did not add a separate controller that the tenant would have to manage.
The phrase zigbee wifi came up during the controls meeting. At first I thought we needed a Wi-Fi-based system because the building already had Wi-Fi. The engineer said no. He asked whether we cared about the protocol. I honestly did not, until he explained that Zigbee and Wi-Fi aren't rivals; they're different tools.
What we needed was a simple way to group sensors and switches so the lights could follow occupancy, without adding a dedicated network. The wireless controls we evaluated use the protocolo Zigbee—on the Spanish datasheet, it literally says protocolo Zigbee—and that's a low-power mesh protocol, not Wi-Fi. The devices talk to each other directly. The Wi-Fi options we looked at would have relied more heavily on the building network and the office IT team. Zigbee did not need that. Zigbee and Wi-Fi can coexist in the same building, but they solve different problems.
Did I do a perfect job explaining that to the client? No. I said, “Zigbee is like a walkie-talkie network for the lights, and Wi-Fi is the office internet.” The client looked at me like I was translating from another language. But the purchasing decision was clear: buy the controls that fit the building and the budget, not the controls that win a buzzword argument.
The Light Almond Interruption
One side quest taught me the same lesson. Around the same time, the client asked for new receptacles in the break room. I typed the lower-case search term legrand light almond receptcle—yes, I misspelled it—and got a wall of options. The pictures looked similar, but I could not tell if the new finish would match the existing wallplates.
Color matching in a retrofit is one of those details that doesn't show up in the energy analysis. But if the new receptacle is slightly different from the old wallplate, the client sees it every day. I made the sales rep give me the catalog number for each finish. That forced me to slow down and think about the total appearance, not just the product name.
That interruption was small, but it was the same principle as the lighting project: the easy answer is rarely the complete answer.
So What Did This Cost?
I don't like publishing energy savings percentages without a baseline and a metering period. What I can say is that the building's measured lighting load dropped significantly after the retrofit, but more importantly, my original “low cost” plan would have left aging ballasts inside fixtures that were supposed to last for years. That's the real danger of thinking the lamp is the entire project.
The final order included:
- LED T8 tubes specified after sample testing
- Legrand sensor switches in rooms with wasted lighting
- Zigbee controls for the areas where wireless grouping saved the most labor
- An unreasonable amount of time spent confirming finish codes
I would not call the project perfect. We did have one delay because the phrase “sensor switch” meant something different to me than it did to the electrical foreman. I meant an occupancy sensor controlling the light. He thought I meant a switch that could be controlled by an external sensor. We caught the mismatch before the walls closed, but it cost us a day. Lesson: say what the device does, not just what it's called.
What would I do differently? I would run a three-fixture pilot on any LED tube replacement, no matter how urgent. I would also call the network integrator before buying controls, not after. And I would be more careful with the assumptions I make when someone says “just swap the tubes.”
The industry has changed in the seven years I've been doing this. What made sense for lighting controls in 2018 is not necessarily what makes sense in 2025. LED replacement doesn't mean what it meant in 2015. But some fundamentals are the same: understand what you're putting into the existing system, measure the total cost, and leave room to catch mistakes early.
If you search how to replace fluorescent tube with led, you'll get a lot of quick answers. Some are fine. The better answer is more honest: it's not just a tube. It's a fixture system, a control decision, and a procurement lesson all wrapped in one.