How to Build a 10000 Watt Solar Generator for a Greenhouse: A 7-Step Emergency Checklist

In my role coordinating off-grid power builds for greenhouse growers, I get the same call every fall: "I need the solar running before the first frost, and no, I can't push the build." The February 2023 freeze changed how I think about greenhouse power. A client in Zone 7 lost 4,000 seedlings during a 46-hour outage—heat mats went dead at 10pm and the benches were below freezing by 6am. He ordered a system that spring, but the lesson stuck: in a greenhouse, solar isn't a sustainability talking point. It's crop insurance.

This checklist is what I hand a grower who needs a 10,000-watt solar generator built under deadline. I've used it on 40+ emergency off-grid projects since 2019, from a 3kW walk-in grower to a 25kW multi-hoop operation in Minnesota. It's written for the operator who's been told a 10kW build is "too small to engineer." It isn't. Seven steps, in this order. When I'm triaging a rush build, this is the order I work in.

Here's the kit you're aiming for: a 10kW PV array, two Morningstar Tristar MPPT 96A controllers, a 48V battery bank (40–50kWh usable), a 10kW pure sine wave inverter with a 20kW surge rating, and the correct disconnects and copper for a 200A+ battery circuit.

Step 1: Size the load with a winter-day number

You don't design a greenhouse solar system in July. You design it for the third week of December, when the lights run 16 hours, the heater blower wakes up at 2am, and the sun is a rumor.

For a typical 30×80 greenhouse—2,000W of LED grow lights, 400W of circulation fans, a 250W water pump, and a 3kW heater blower running about 6 hours on cold nights—the daily load lands around 45–55kWh. Use that number, not the manufacturer's "average" wattage.

Run a 7-day energy log before you buy anything. A $35 plug-in power meter is far cheaper than the array you'll oversize because you guessed 20kWh/day and were wrong by 2.5×. We sized one client's system at 8kW—or rather, 7.2kW after we re-checked the heater's duty cycle. That difference paid for half his battery bank.

Step 2: Pick the charge controller before the panels

Most people buy panels first. That's backwards. The charge controller is what makes or breaks a 10kW greenhouse build, especially in cold weather.

At this scale, don't buy PWM. A PWM controller clips your panel voltage down to battery voltage, which in cold months means throwing away 20–30% of the only power you're getting. You need MPPT. That's not a preference, it's math.

I generally spec Morningstar Tristar MPPT controllers on these builds. The published specs: 150V max PV input, 96% peak efficiency, and either 60A or 96A output. A 10kW array feeding a 48V bank pushes roughly 200A of charge current at solar noon, so you're looking at two Tristar MPPT 96A units in parallel (4,600W of charge capacity each). That's the cleanest path.

The Morningstar website has a string-sizing calculator and downloadable manuals for every controller. I use it on every build—even rush jobs, especially rush jobs. A 30-second check beats a service call in January.

Panel brand matters less than the voltage window. I recently paired a Tristar MPPT with a set of Aureus solar panels; it worked cleanly because we verified the nameplate Voc before wiring. Pair it with any panel whose cold-temperature Voc fits the 150V window—on the Aureus build, that check took ten minutes.

Wait— one cold-weather trap: PV voltage rises as temperature drops. Three "60Voc" panels in series looks fine on paper until −15°C pushes the string over the controller's limit. I learned that one the hard way on a client install, so now it's the first thing I check. (Should mention: in 2023, a client saved $300 on a no-name controller and then lost 25% of his winter harvest. If you ask me, the charge controller is the last place to save money on a 10kW build.)

Step 3: Build around a 48V battery bank

At 10kW, system voltage isn't a preference—it's a safety and cost decision. At 12V, 10,000W of load would be 833A. That's enough current to turn a 2/0 cable into a resistance heater. At 48V, it's 208A. Still serious, but manageable.

All the charge controllers and inverters in this build come in 48V, and the Tristar MPPT family auto-detects 12/24/48V. On the battery chemistry question, I'd take lithium if you can survive the upfront check: 80–90% usable depth of discharge versus 50% on flooded lead-acid, and roughly three times the cycles. At this scale, the usable-kWh-per-dollar math tends to favor lithium.

Even after I specified a 48V bank on a recent build, I kept second-guessing. What if the client already had 24V equipment and didn't tell me? I didn't relax until the spec was confirmed. That anxiety is normal—but don't let it delay you. Confirm the battery voltage and move on.

Step 4: Choose an inverter with surge headroom

Greenhouse loads spike. A 500W water pump can draw 2,500W for 150 milliseconds at startup. A heater blower's fan kicks on while the lights are already at 100%. If your inverter is sized at exactly its continuous rating, you'll trip it the first cold night.

Spec a pure sine wave 48V inverter rated 10kW continuous and at least a 20kW surge for 5–10 seconds. Also check the idle draw. An inverter that consumes 80W doing nothing burns nearly 2kWh a day before it powers a single grow light. "10kW" on the label tells you nothing about that number.

Step 5: Protect, bond, and ground before you energize

This is the step that separates a greenhouse that hums through a blizzard from one that burns. Every Tristar MPPT controller gets a DC breaker between it and the battery bank, and a disconnect between the array and the controller. The inverter gets an AC disconnect. The main panel gets bonded to a grounding rod.

Per NEC 690.8, conductors must be rated at 125% of continuous current. In practice, for a 208A battery-side circuit, that means 4/0 copper or 2× 2/0 in parallel, with a battery fuse or breaker at 250A. Don't downsize "because it's a short run." I've watched a 250A fuse open twice in a single build because someone thought 18 inches of 2 AWG doesn't count.

One overconfidence mistake on my part: I knew I should thermal-image a battery bank before we left a job, but we were rushing to catch a shipping deadline. That was the one time a lug was loose. We caught it during a final touch check—the battery terminal was noticeably warm—and fixed it before it became a fire. Now every commissioning day ends with a touch-test of every lug. Disconnect, tighten, reconnect, and check for heat after 30 minutes of charging. If any terminal is more than a few degrees hotter than the rest, stop and find out why.

Step 6: Set up remote monitoring before you need it

You will not be at the greenhouse when the problem happens. You'll be at dinner, or asleep, or 200 miles away. That's what Morningstar Portal is for.

Every Tristar MPPT has a communications port. Connected to the Portal—accessible through the Morningstar website—it shows array voltage, charge current, battery state of charge, and alarm history on your phone.

Last winter, a client called at 9:40pm worried that his greenhouse batteries were low. I pulled up his Portal feed while on the phone: one panel string was dead, and the battery had been slowly draining for three days. I told him which breaker to check. He reset it, and the system recovered overnight. Without Portal, that diagnosis would have cost him a week and a generator rental.

A 10kW greenhouse isn't big enough to justify weekly site visits. Remote monitoring is what makes a build this size practical to run. A small operation isn't a mini version of a big one—it's a system that has to run itself, and monitoring is how you let it.

Step 7: Run a 72-hour test before crop day

Commission the system two weeks before you actually need it, not two days. Run the real greenhouse loads through the inverter for three days straight, including overnight. Wake up at 2am on the coldest night and check the battery state of charge. If it's lower than when you started, you have a load-vs-generation problem, and you need to know that before the plants are in.

Keep a commissioning log. Date, battery voltage, array wattage, any alarms. It sounds old-school, but the log becomes your baseline when something feels off in December.

Common mistakes that will cost you a crop

A few things I see repeatedly when I'm called in to fix a failed greenhouse build:

• Wiring the charge controller output too small. The controller is happy to deliver its full 96A; the wire isn't. Result: voltage drop, heat, and a battery that never fully charges. Worth stating plainly: that's also a fire risk.
• Ignoring December sun hours. NREL's PVWatts v8 model (accessed January 2025) shows northern-state locations getting 5+ peak sun hours in July and closer to 1.5–2 in December. A 10kW array that makes 45kWh/day in summer may make 15–20 in winter. If your December load is 50kWh/day, a backup generator is part of the design. That's not failure; it's math.
• Going cheap on the controller. When I was starting out, the vendors who treated my $500 orders with respect are the ones I still use for $20,000 orders. Small budgets deserve real engineering—not a discount MPPT that wastes the only harvest you have.

That's the whole list. If you're building a 10000 watt solar generator for a greenhouse and the timeline is tight, follow the steps in order. The panels are the easy part. The design decisions—controller, voltage, wire, monitoring—are where you earn your crop.