In March 2024, at 4:15 on a Thursday, I got a call I knew would turn into a long night. A commercial greenhouse in Ohio had just watched their lighting contractor disappear. The plant room was empty, the benches were clean, and 3,000 tomato transplants were scheduled to arrive Saturday afternoon. The grower asked a simple question: could we install 64 horticultural LED fixtures in 36 hours? I said yes. I said yes before my spreadsheet reminded me that I hate spreadsheets.
I say yes because that’s my job. I coordinate emergency lighting installs for commercial growers. Over the past five years, I’ve handled more than 200 rush orders. That number doesn’t make me faster than physics. It means I know the difference between a tight timeline and an impossible one. This one was somewhere in between.
The risk was real. Miss the installation window, and the transplants would sit in the dark, stretch, and lose time that a commercial crop can’t afford. The grower mentioned a $12,000 penalty if the project delayed the next planting cycle. That wasn’t even the worst part. The worst part was that the room could look finished and still be wrong.
The fixture list didn’t match the room
The original contract called for 60 Gavita HGC906410 RS 1900e grow lights. That fixture is a workhorse for commercial canopy lighting, and we had most of them in stock. The remaining four slots were listed as Gavita CT 1930e LED units. We had those too. Good, I thought. Then I laid the fixture plan over the actual ceiling measurements.
When someone asks about the Gavita CT 1930e LED coverage area, they usually want one number. I get it; one number is easier. But coverage area depends on target PPFD, hanging height, and the overlap between fixtures. So I printed the PPFD maps from Gavita’s website and drew the layout in CAD. The CT 1930e’s footprint is not identical to the RS 1900e’s footprint, and I wasn’t going to install two different footprints on the same spacing just because the invoice said four.
Every spreadsheet calculation pointed to the same answer: put the CT fixtures in the same grid as the rest, and the room average would be acceptable. My gut said something felt off. I’ve walked enough grow rooms to know that averages can hide dark edges. The low spots don’t show up on a chart when you’re looking at the center of the room. They show up on the first row of plants along the wall.
My gut won
We created a separate layout for the CT units. Different spacing, different hanger positions, different aiming height. That cost two extra hours in the schedule and a few unhappy electricians who wanted to get on with it. It also meant the room would be more even when the lights finally turned on. I would rather spend two hours checking than walk back in two weeks with a PAR meter and explain why one bench was 18% darker than the rest.
The building itself made us slow down even more. The greenhouse was an old converted theater, and the lobby still had one of those Titanic chandelier pieces from a banquet hall, complete with dozens of candelabra chandelier bulbs. It was beautiful, and it had nothing to do with growing plants. The facility manager asked, half joking, whether we could replace a few chandelier bulbs while we were there. I said we would if we had time. I already knew we didn’t.
Halfway through the install, a maintenance guy appeared with a roll of flexible LED strip. “Can I cut an LED strip?” he asked. It was four inches too long for a shelf under the germination bench. The short answer is yes, but only on the marked cut lines and with a compatible driver. You can’t slice anywhere on a sealed strip and expect it to stay safe. That question reminded me of the actual grow light layout. You can’t cut a coverage area to fit a room. You have to choose fixtures based on where the photons end up.
The expensive mistake we almost made
We finished hanging brackets at 11:30 on Friday night. The electrician finished terminating circuits at 2:10 Saturday morning. Before the delivery truck arrived, two of us walked every row with a PAR meter. In the row closest to the theater wall, the CT fixtures were within 4% of target. If I had ignored my gut and used the first layout, that same row would have been about 17% low.
We caught it before the plants arrived. That’s the whole lesson. I built a twelve-point checklist after a mistake in 2023, and that checklist is what saved us here. It forced me to pull the PPFD coverage maps and compare them to the actual ceiling trusses. Fifteen minutes of checking beats five days of rework when the benches are full.
One more thing: the grower asked if we could market the new room as a yield booster. I told him not to promise that. Per FTC guidelines, claims like that need evidence, and we had PPFD maps, not yield trial data. The lights should perform. The lights did perform. But a good lighting plan doesn’t give anyone the right to invent a number.
What I’d repeat
- Check the physical room before trusting the fixture count.
- Use the manufacturer’s PPFD maps to set spacing, not the wattage.
- Don’t mix fixture models without re-running the coverage plan.
- If you cut an LED strip, cut only on marked lines and seal the cut edge.
- Don’t let chandelier bulbs or decorative fixtures become the lighting plan for a crop.
The transplants arrived at 1:10 Saturday afternoon. All 64 lights were on. I finally sat down for the first time in twenty-six hours. That Titanic chandelier in the lobby is still beautiful. It just isn’t a grow light.