-
Who this checklist is for
-
The 8-step lighting installation checklist
-
Step 1: Verify delivered hardware against the spec sheet before you open anything
-
Step 2: Measure PPFD at canopy height—not on a table
-
Step 3: Check spectrum consistency across units
-
Step 4: Validate controller integration—and test the failover
-
Step 5: Verify the mounting structure before you hang anything
-
Step 6: Account for secondary lighting—including sensor downlights
-
Step 7: Walk the electrical path—including switch work
-
Step 8: Calculate total cost of ownership before you sign the invoice
-
Step 1: Verify delivered hardware against the spec sheet before you open anything
-
Common mistakes and what they cost
Who this checklist is for
If you're the person who has to sign off on a commercial grow light installation—whether you're a facility manager, a quality lead, or the owner-operator—you need a verification process that catches problems before they get expensive. I've reviewed lighting installations for indoor cultivation facilities for the better part of a decade. In Q1 2024 alone, we rejected roughly 18% of first-pass installations because of spec mismatches, mounting issues, or electrical documentation gaps that would've been a nightmare to fix after the fact.
Here's the thing: most of these problems aren't dramatic. They're boring. A 12% wattage discrepancy. A driver that fails at 90°F ambient instead of the rated 104°F. A sensor downlight wired to the wrong circuit. Each one looks small until you're three months in and half your crop is stretching because the PPFD map was optimistic. This checklist is the eight-step process I use before I'll approve a lighting installation. It's written for a mid-size commercial room—say 1,000 to 3,000 square feet of canopy—but the logic scales up or down.
The 8-step lighting installation checklist
Step 1: Verify delivered hardware against the spec sheet before you open anything
First pass: cross-reference the packing list against your purchase order and the manufacturer's spec sheet. I'm looking at model numbers, wattage, spectrum designation, beam angle, and driver brand. Takes about 20 minutes and catches the errors that are easiest to fix—before anything gets mounted on a ceiling.
Last year we ordered two pallets of Spider Farmer Glow30 LED grow light bars for a propagation room. The PO specified the Spider Farmer 100 watt version. The boxes said 100W on the outside label. But when I pulled the driver nameplate on three random units, they were pulling 87 watts at the wall. Twelve out of forty units in that batch came from an older production run with a different driver. The vendor's explanation was "within acceptable tolerance." Our spec said ±5%. 87 watts is 13% off. We sent the whole batch back.
Take it from someone who's been burned: the label on the box is not the spec. The driver nameplate is.
Step 2: Measure PPFD at canopy height—not on a table
Datasheets give you a PPFD map based on ideal conditions: a single fixture in a dark room, measured at a fixed distance, usually with a specific reflector and no interference from other fixtures. Real installations are not ideal.
It's tempting to think that if you buy a fixture rated for 1,500 µmol/m²/s at 12 inches, you'll get 1,500 at your canopy. But the moment you put eight fixtures in a row, the overlap changes everything. The center gets hotter, the edges get weaker, and the datasheet number is only useful as a starting point. The "just trust the spec sheet" advice ignores fixture spacing, mounting height, and the reflectance of your walls and floor.
I use a quantum sensor (we run an Apogee MQ-610, calibrated every two years) and measure at nine points across a 10-foot grid at actual canopy height. Target for a flowering room is typically 800–1,000 µmol/m²/s, but I'm less interested in hitting the target than I am in the uniformity ratio. If the edge of your canopy is getting 60% of the light the center is getting, you've got a problem—and it's usually a spacing or mounting-height issue, not a fixture issue.
Step 3: Check spectrum consistency across units
Spectrum is harder to verify than intensity because you need a spectroradiometer, and most of us don't have one sitting in the office. But you can catch obvious problems with a simpler approach: turn on every fixture in the room at once and look for visible color differences between units. Take a photo with your phone—auto white balance off, manual exposure locked. If you see a cluster of fixtures that looks warmer or cooler than the rest, you've probably got a batch-mixing problem.
We had a situation in 2023 where three out of twenty-four fixtures had a noticeably different red-to-blue ratio. Turned out two different LED bins had been mixed in the same shipment. Spider Farmer's GGS controller actually logs spectrum settings per zone, which made the investigation easier—we could see the three outliers were on a different firmware version.
I have mixed feelings about this one. Part of me thinks visual inspection is too crude to be meaningful. Another part knows the human eye is surprisingly good at picking up CCT differences of 200K or more, and that's usually enough to catch the egregious cases.
Step 4: Validate controller integration—and test the failover
If you're running a smart controller—the Spider Farmer GGS system, for example—this is where a lot of installations fall apart. The controller might work perfectly in the demo, but fail when you try to schedule a sunset ramp across four zones with different fixture types. The gap between "connected" and "functional" is wider than most people expect.
The test I run: program a week-long lighting schedule that includes at least one sunrise/sunset ramp, one abrupt on/off transition, and one zone override. Then simulate a network failure by unplugging the Zigbee coordinator for five minutes. Does the controller revert to a safe default? Does it hold last state? Or does it do something unpredictable that leaves half the room dark?
Document the answers. If the answer is "unpredictable," that's not a controller you want managing a $40,000 crop.
Step 5: Verify the mounting structure before you hang anything
This is the step most installers want to skip. "We've hung lights a hundred times." I've heard that line. I've also seen a chandelier stand rated for 50 pounds carry three fixtures that weighed 22 pounds each because nobody did the math on wire rope tension at the anchor point.
Two things to check: the working load limit of every component (anchor, wire rope, carabiner, stand), and the actual weight of the fully assembled fixture including driver, heat sink, and any added reflectors. The manufacturer's shipping weight is not the mounting weight—shipping weight includes packaging.
A chandelier stand or overhead truss system is only as good as its anchorage. If you're anchoring into wood joists, use lag bolts with a pull-out rating at least 4x the total hanging weight. If you're anchoring into concrete, use wedge anchors and verify embedment depth. If you're anchoring into drywall—stop. Don't do that. I'm serious.
Step 6: Account for secondary lighting—including sensor downlights
Growing spaces need more than just grow lights. You need work lighting for maintenance and safety lighting for egress. This is where sensor downlights and other auxiliary fixtures come in, and this is where I see a ton of sloppy work.
A sensor downlight wired to the same circuit as your grow lights will turn on every time you walk into the room during the dark period—which is exactly when you don't want white light hitting your plants. I've seen this mistake more times than I care to count. The fix is simple: put work lighting on a separate switched circuit, or use a motion sensor with a manual override and a tight field of view that doesn't spill onto the canopy.
Code-wise, you need egress lighting independent of the grow light control system. Per NFPA 101 (Life Safety Code, 2024 edition), means of egress must be illuminated to at least 1 foot-candle at floor level during occupied hours. Verify with your local authority having jurisdiction—requirements vary by state, and commercial cultivation facilities sometimes land in unusual occupancy classifications.
Step 7: Walk the electrical path—including switch work
This is the step that makes people nervous, and it should. If you're changing a light switch in a grow facility, the correct procedure is to turn off the branch circuit breaker that feeds that switch, lock it out if possible, and verify dead with a non-contact voltage tester and a multimeter before touching anything.
But here's the reality that the "just turn off the main" advice ignores: in a commercial grow facility, you often can't turn off the main panel. Environmental controls, irrigation controllers, and security systems are on the same panel. What you can do—and what you should do—is identify the specific branch circuit, isolate it, and verify it's dead. That's the real version of "how to change a light switch with power on"—the power is on elsewhere in the building, but the circuit you're working on is de-energized and verified. Don't confuse those two situations.
Don't work live. I don't care how experienced you are. Arc flash incidents at 120/240V are rare compared to industrial voltages, but they happen, and the consequences aren't worth it. If you can't identify the branch circuit or can't verify dead, you call an electrician. That's not a failure of skill—it's a decision about total cost.
Step 8: Calculate total cost of ownership before you sign the invoice
Before I sign off on any lighting installation, I run a total cost of ownership calculation that goes beyond the invoice. The invoice is the easy part. The harder part is the stuff that shows up later:
- Replacement drivers (typically 5–8% failure rate within 3 years for mid-tier fixtures; higher for budget units)
- Labor for fixture swaps (2 hours per fixture at $65/hour, plus lift rental if needed)
- Yield loss from spectrum or intensity drift (hard to quantify, but real)
- Energy cost difference between fixtures with different efficacy (µmol/J)
- Controller replacement or upgrade (3–5 year lifecycle on smart controllers)
An example from our own facility: we compared two fixtures for a veg room. Fixture A was $189 per unit; Fixture B was $245. Fixture A had a 2-year warranty and a driver with a documented failure rate of 7% at year three. Fixture B had a 5-year warranty and a failure rate of 2%. Over a 5-year horizon, assuming 40 fixtures, Fixture A's TCO was roughly $11,200 higher after driver replacements and labor. The "cheaper" fixture cost 15% more.
I have mixed feelings about TCO calculations because they can be manipulated to justify any decision. But the framework itself is sound: if you're only comparing unit prices, you're not comparing the right numbers.
Common mistakes and what they cost
Using shipping weight as mounting weight. I've seen a 14-pound fixture become a 22-pound installed load once the driver, heat sink, and reflector were added. The anchor failed within six months. The repair cost more than the fixture.
Wiring sensor downlights to the grow light circuit. Work lights turning on during the dark period can trigger herming or revegetation. Fixing it means rewiring—at minimum a half-day of electrician time, plus the crop stress you can't undo.
Skipping the voltage verification. A non-contact tester is $25. A hospital visit is not. Test every conductor before you touch it, even after you've flipped the breaker.
Comparing unit prices without TCO. The $500 quote isn't cheaper than the $650 quote if it needs a driver replacement in year two and the other doesn't.
If you're about to sign off on a lighting installation, run through this list. It'll take you a few hours. It'll save you a lot more than that.
One last thing: document everything. Photos of nameplates, PPFD readings at each grid point, controller schedule screenshots, mounting hardware specs. If something fails six months from now, that documentation is the difference between a warranty claim and an argument you lose.