Most people asking this question want a yes or a no. The honest answer is that running a rooftop air conditioner on battery power is straightforward for two hours, expensive for four, and a serious engineering project for eight. The gap between those three answers is roughly $6,000 in hardware and about 300 lbs on your axles.
What follows is the arithmetic, done properly, with the numbers that vendors leave out. If you take one thing from this page, take this: the battery bank is not the hard part. Putting the energy back in is the hard part, and almost nobody models it before they buy.
The three numbers that decide everything
Every honest calculation for battery-powered air conditioning rests on three figures, and two of them are routinely misreported.
Running watts. What the compressor and fan draw once the unit is up and stable. For a standard 13,500 BTU rooftop unit this is 1,300–1,600W at 120V. Call it 1,450W as a working figure.
Duty cycle. What fraction of each hour the compressor is actually running. This is the number that gets omitted, and it changes the answer by a factor of two.
Conversion loss. Your battery is DC. Your air conditioner is AC. The inverter in between is not free — a good unit under real load returns about 87% of what you feed it, and it draws idle current the entire time it’s switched on.
Multiply those together and you get watt-hours per hour of clock time. Everything else is bookkeeping.
What a 13,500 BTU rooftop unit actually draws
The running figure is the easy one, and it’s reasonably consistent across brands: 1,300 to 1,600W, or roughly 11–13A at 120V. A 15,000 BTU unit sits at 1,500–1,800W.
The startup surge is where the specification sheets get interesting. A conventional single-speed RV compressor demands a brief but enormous jolt to overcome inertia — commonly 2,600 to 3,500W for a 13,500 BTU unit, lasting somewhere between 150 and 500 milliseconds. That surge is why a 2,000W generator that comfortably handles the running load will still fail to start the unit, and it is the single most common reason an inverter that looks adequate on paper trips out. If you’re specifying an inverter for this job, the surge rating matters more than the continuous rating — we’ve laid out how to size for motor loads separately, because the mistake is expensive and extremely common.
Surge does not, however, meaningfully affect your energy budget. It’s over in half a second. Duty cycle is what drains the bank.
Duty cycle is the number everyone omits
An air conditioner does not run continuously. It runs until the thermostat is satisfied, stops, and restarts when the interior warms back up. How often that happens depends on outside temperature, insulation, sun exposure, humidity and how many people are breathing inside the vehicle.
Realistic figures for a moderately insulated rig:
| Conditions | Duty cycle | Effective draw (13.5k unit) |
|---|---|---|
| 78°F night, shaded, low humidity | 40–50% | 580–725W |
| 85°F night, typical summer | 60–70% | 870–1,015W |
| 95°F afternoon, direct sun | 85–100% | 1,230–1,450W |
| 105°F desert afternoon | 100% (never satisfies) | 1,450W |
Notice what happens at the bottom of that table. Above roughly 100°F, a 13,500 BTU unit in a poorly insulated trailer stops cycling altogether — it runs flat out and still loses ground. Your energy consumption is then identical to the nameplate figure, and no amount of battery makes that sustainable.
Notice also that the mild-night figure is less than half the hot-afternoon figure. Anyone quoting you a single number for “how much power does RV AC use” is either assuming conditions they haven’t told you about or hasn’t thought about it.
The overnight arithmetic
Take a realistic hot-but-not-extreme night: 85°F, 65% duty cycle, eight hours of sleep.
Compressor + fan draw 1,450W
Duty cycle × 0.65
AC-side consumption = 943W average
Inverter efficiency ÷ 0.87
Inverter idle draw + 25W
Battery-side consumption = 1,109W average
Eight hours × 8
TOTAL = 8.9 kWhJust under 9 kWh from the battery bank, for one night’s sleep.
To put that in perspective: 9 kWh is what a well-equipped off-grid rig running lights, a compressor fridge, a laptop, a water pump and satellite internet consumes in roughly four to five days. One night of air conditioning costs more energy than most of a week of everything else combined.
Converted to bank capacity, at 12.8V nominal for lithium:
8.9 kWh ÷ 12.8V = 695 Ah
That is seven 100Ah LiFePO4 batteries, and it assumes you’re willing to run the bank to genuinely empty, which you shouldn’t. Add headroom and you’re at 800Ah — around 400 lbs of battery, before you’ve added the inverter, the cabling or the charge sources.
Sizing table: what each target actually costs
| Target | Battery-side energy | Usable bank needed | 12V equivalent | Array to recover next day | Verdict |
|---|---|---|---|---|---|
| 2 hrs (fall asleep) | 2.2 kWh | 2.6 kWh | 200 Ah | 700W | Achievable on a normal build |
| 4 hrs | 4.4 kWh | 5.2 kWh | 400 Ah | 1,200W | Realistic ceiling for most vans |
| 8 hrs (full night) | 8.9 kWh | 10.5 kWh | 800 Ah | 2,400W | Trailer, bus or fifth-wheel only |
| 24 hrs (desert full-time) | 30 kWh | 35 kWh | 2,700 Ah | 8,000W | Not a battery problem. Buy a generator. |
Array figures assume five peak sun hours and a 75% system derate covering panel temperature, controller losses, wiring, shading and imperfect angle. In June in Arizona you might beat this. In September in Michigan you will not come close.
If your rig can’t physically carry the array in the right-hand column, you are not running air conditioning on solar. You are running it on a battery you brought pre-charged from somewhere else, which is a completely different proposition and a perfectly legitimate one — but be clear with yourself about which you’re doing.
The recharge problem is bigger than the battery problem
Here is the constraint almost nobody models before they spend the money.
An 800Ah bank will get you through the night. Getting it back requires putting 8.9 kWh into it the following day — and you’re likely running the air conditioner during part of that day too, which means the array is fighting the load before it starts refilling anything.
A 1,000W rooftop array on a good June day in the Southwest might harvest 3.75 kWh. If the unit runs for four hours of that afternoon at 85% duty, it consumes about 5.7 kWh. You are 2 kWh further behind at sunset than you were at dawn, having run air conditioning for four hours and slept without it.
This is the daily deficit that turns a $12,000 electrical system into a disappointment. The bank is a buffer, not a source. If the array can’t refill the buffer, you’re on a countdown from the day you leave shore power.
Alternator charging changes the picture, but less than people hope. A 50A DC-DC charger delivers roughly 690W while the engine runs — about 0.7 kWh per hour of driving. Covering one night of air conditioning takes roughly thirteen hours behind the wheel. If you’re moving every day, this works. If you’re parked for a week, it doesn’t.
System voltage: 12V is the wrong platform for this job
At 1,450W of AC load, the inverter draws about 1,670W from the battery. At 12V that’s 130A, continuously, for hours.
130A continuous means 2/0 or 4/0 cable, a 250A+ Class T fuse, a busbar system rated accordingly, and lug terminations that need to be made properly or they will get hot. Voltage drop across the inverter run becomes a real design problem rather than a rounding error. Every 100Ah of capacity you add brings another set of parallel connections that all have to share current evenly.
The same load at 24V is 65A. At 48V it’s 33A — comfortably inside 6 AWG.
If you are building a system from scratch with air conditioning as a design requirement, build it at 24V or 48V. The cost difference in components is small, the cost difference in copper is substantial, and the failure modes at 130A are considerably less forgiving. Working out where your total consumption actually lands before you pick a system voltage is exactly what our off-grid load calculator is for.
Soft starters: what they do and what they don’t
A soft starter ramps the compressor motor up over roughly a third of a second instead of slamming it to full speed. The startup surge drops by 60–70%, which typically brings a 13,500 BTU unit’s peak demand down to the 1,000–1,200W region.
What that buys you:
- A 2,000–2,500W inverter can start a unit that previously needed 3,500W of headroom
- Two rooftop units can run on a single 30A shore connection
- A small inverter generator becomes viable
What it does not buy you:
- Any reduction in running consumption. None. The compressor draws what it draws.
- Any improvement in your overnight energy budget
- Any additional cooling capacity
Soft starters are sold in the same aisle as efficiency upgrades and are frequently discussed as though they were one. They aren’t. A soft starter is a compatibility device that widens the range of power sources that can run your existing unit. That’s genuinely useful and usually worth the $300. It will not extend your runtime by a single minute.
Cheaper things to do first
Before you spend $6,000 on batteries, spend $600 on the reasons you need them.
Shade. Parking under a tree or deploying an awning on the sun-facing side can cut duty cycle by 20–30%. It’s free and it’s the single highest-return intervention available.
Roof colour and reflectivity. A dark roof in direct sun can run 60°F above ambient, and every degree of that is heat conducting into the cabin.
Ventilation before refrigeration. A decent roof fan moving air across a bed at night handles a surprising share of what people reach for the compressor to solve — up to about 80°F ambient with low humidity. Above that it stops being an alternative, but below it, it’s a tenth of the power.
Sealing and insulation. A rig that loses heat slowly needs the compressor less often. This is the least glamorous and most consequential factor in the entire calculation, and it deserves its own treatment.
Reconsider the unit itself. If you’re building rather than retrofitting, a variable-speed DC unit modulates output instead of cycling on and off, which changes the energy profile meaningfully. Whether the marketed figures survive contact with a 95°F afternoon is a separate question, and we’ve tested those claims against the manufacturers’ own specification sheets.
The case for a generator that nobody wants to hear
There is an unfashionable answer to this problem, and it’s worth stating plainly because the arithmetic supports it.
A 2,200W inverter generator costs around $700, weighs about 47 lbs, and will run a 13,500 BTU unit indefinitely with a soft starter fitted. The battery-and-solar equivalent — 800Ah of lithium plus 2,400W of array plus the charge controllers, cabling and mounting — costs somewhere north of $8,000 and weighs several hundred pounds.
The generator wins on cost by better than ten to one, and it doesn’t care whether it’s cloudy.
What it loses on is noise, fuel logistics, and where you’re permitted to run it. Quiet hours at most campgrounds and many public lands rule out overnight operation entirely, which is exactly when you want the air conditioner. That single restriction is what pushes people toward batteries, and it’s a legitimate reason — not a technical one, a regulatory one.
The pragmatic build most experienced users land on is a moderate bank for silent overnight operation, and a small generator for hot afternoons when noise restrictions don’t bite. Pretending you’ll never need the generator is what makes the battery bank enormous.
What we’d actually build
Budget: comfortable sleep onset, mild climates. 200–300Ah at 12V, 2,000W inverter with soft starter fitted, 400–600W array. Runs the unit for two to three hours while you fall asleep, then you switch it off. This covers about 80% of what people actually want and costs about 20% of what the full solution costs.
Mid: serious summer capability. 400–600Ah at 24V, 3,000W inverter, 1,000–1,400W array, DC-DC charger sized at 50A minimum. Four to six hours of cooling, with a realistic chance of recovering the next day if you’re not also running the unit through the afternoon.
Full: air conditioning as a given. 800Ah+ at 48V, 2,400W+ of array, and honest acceptance that a generator is part of the system for consecutive 100°F days. At this point you’re specifying a small off-grid power plant, and the cost belongs in your total cost of ownership calculation rather than being treated as an accessory purchase.
The one configuration we’d argue against is the middle-sized 12V build — 400Ah at 12V, 3,000W inverter, and a hope that it’ll do a full night. It won’t, the cabling is unpleasant, and you’ll spend the money again when you upgrade the voltage. If you’re already committed to 12V and working out how much bank you actually need for shoulder-season use rather than peak summer, that’s a different and more forgiving calculation.
Frequently asked questions
Can I run my RV air conditioner directly from solar panels? No. Panel output varies second by second with cloud, angle and temperature, and a compressor needs stable voltage and a large instantaneous surge. Solar charges a battery; the battery runs the air conditioner. Any product marketed as “solar powered air conditioning” has a battery in it somewhere.
How many batteries do I need to run AC all night? For a 13,500 BTU unit over eight hours in typical summer conditions, roughly 9 kWh of usable capacity — about eight 100Ah 12V lithium batteries, or the equivalent at higher voltage. In mild conditions with good insulation you might manage on six.
Will a 2,000W inverter run a 13,500 BTU air conditioner? Only with a soft starter fitted. Without one, the startup surge will exceed the inverter’s capability and it will fault out. With one, a quality 2,000W pure sine unit will usually manage, though 3,000W gives useful headroom for anything else running simultaneously.
Is a portable power station a realistic option? For an hour or two, yes. A 2 kWh unit delivers roughly 1.5 hours of cooling in typical conditions and about 50 minutes in real heat. Stacking enough of them to cover a night costs substantially more per kWh than an installed system.
Does running the air conditioner from battery damage the batteries? Not if the bank is sized for the discharge rate. A 1,670W draw from a 200Ah 12V bank is a 0.65C discharge — within LiFePO4 specification but toward the upper end, and it will generate heat. From an 800Ah bank the same load is a gentle 0.16C. Undersized banks age faster under this kind of load.
Do 12V DC air conditioners solve this? They improve it. Removing the inverter recovers about 13% of the losses, and variable-speed compressors reduce cycling waste. They do not change the underlying physics — moving that much heat still costs that much energy.