Why Does Your Car's AC Make The Engine Work Harder?

A car's air conditioning compressor is bolted to the engine and spun by the serpentine belt off the crankshaft, so it takes mechanical power straight from the engine rather than electricity from the battery. Peer-reviewed measurements put that draw at 3.89 kW (5.2 hp) at peak and 2.10 kW (2.8 hp) once the cabin cools, against just 150 watts for the cabin blower fan. Oak Ridge National Laboratory measured the resulting fuel penalty at 9% to 27%, depending on vehicle and speed.

It is the middle of July. You have been parked in an unshaded lot since morning. Opening the door feels like opening an oven.

You drop into the seat, start the engine, and press the AC button. Something happens in the next half second. A soft click comes from under the hood. The idle sags a little. Pull away now and the car feels slower than it did last night.

Almost every driver has noticed this. Far fewer can say where the power goes. And the most popular explanation, the one that sounds perfectly reasonable, is wrong in a genuinely interesting way.

What Does Your Car's AC Actually Do?

Start with something you can feel. Wet the back of your hand and blow on it. It goes cold.

Nothing made cold there. The water took heat from your skin in order to evaporate, then carried that heat away. Cooling is not the creation of cold. It is the moving of heat from one place to another.

That is the one idea this whole article rests on. Hold on to it.

Your car does the same trick deliberately, using a fluid called refrigerant. The refrigerant expands and evaporates inside a small radiator behind your dashboard. It turns very cold. Cabin air blows past and gives up its heat.

But that heat has to go somewhere. To dump it outside on a 35 °C (95 °F) day, the system must first squeeze the refrigerant. Squeezing a gas makes it hot. Hot enough, in fact, to shed heat into hot outside air.

That squeeze is the whole problem. It is also what the word compressor means. Your home air conditioner and your refrigerator run this same cycle, on the latent heat of a fluid changing phase.

Here is the part that matters. Heat never flows from cold to hot on its own. Pushing it uphill costs energy, every time. Engineers track that cost with one ratio:

COP = cooling delivered ÷ work put in

COP is the coefficient of performance. NREL models a typical car system with a peak cooling job of about 7 kW and a COP of 2.33. Divide, and the machine needs roughly 3 kW of work fed to it continuously.

Three kilowatts has to come from somewhere. In your car, it is not the battery.

What Happens The Moment You Press The AC Button?

That click is not the compressor starting up. The compressor pulley has been spinning since you turned the key. It was simply freewheeling, doing nothing.

The button engages an electromagnetic clutch. A coil energizes. Its magnetic field yanks a steel plate against the spinning pulley. The two lock together. Only now is the compressor actually being turned.

This is why the load arrives so abruptly. It is not a dimmer switch. The compressor goes from free to fully coupled in a fraction of a second, and your engine feels all of it at once.

There is a second reason the dip is so sharp. GM's own engineering filings note that the torque needed to start the compressor turning is higher than the torque needed to keep it turning. The worst moment is the first moment. Modern engine controllers see it coming and quietly hold back a little torque in reserve, ready to release the instant the clutch bites.

The clutch is measurable, not metaphorical. Oak Ridge National Laboratory tracked it directly during its air conditioning tests. On the Ford Explorer, engineers monitored clutch engagement through the OBD-II port. On the Toyota Corolla, they measured the current flowing to the magnetic clutch.

Why Is The AC Compressor Bolted To Your Engine, Not Your Battery?

Open the hood. A long rubber belt snakes around a set of pulleys on the front of the engine. That is the serpentine belt. The crankshaft drives it, the same shaft the pistons push on.

The AC compressor hangs off that belt. So does the alternator, and the power steering pump, and on many engines the water pump too. When the crankshaft turns, they all turn.

So here is the answer, plainly: the engine is physically turning the air conditioning pump. There is no electricity in that chain at all. The only electrical part is the small clutch coil, and it draws almost nothing.

If the arrangement sounds familiar, it should. A supercharger works the same way. Bolt a pump to the engine. Spin it with the crankshaft. Accept that it eats some of the engine's own output. Only the purpose differs.

What about the fan? The blower pushing air at your face is electrical, and it does run off the alternator. That is exactly why the wrong answer is so tempting. But check its size. Two NREL and SAE papers, five years apart, both put the cabin blower at 150 watts. The compressor runs in the low thousands.

The fan is real. It is just the wrong suspect, by a factor of fourteen to twenty-six.

How Much Horsepower Do You Lose With The AC On?

This is what everyone actually wants to know, so here is a measured number instead of a workshop rumor.

NREL researchers built a peer-reviewed SAE study around a measured compressor power profile. Their figure: a peak of 3.89 kW, decaying over roughly ten minutes to a steady 2.10 kW.

Convert that to the unit the question usually uses:

1. 3.89 kW ÷ 0.746 kW per hp = 5.2 hp at peak
2. 2.10 kW ÷ 0.746 kW per hp = 2.8 hp once the cabin is cool

So it is about five horsepower at switch-on, settling to under three. On a 200 hp car, that is a small slice. On a small engine in traffic, it is not small at all. Which is exactly why the idle sags.

NREL put it more starkly in 2000. Farrington and Rugh wrote that a vehicle AC compressor's power draw "can be greater than the engine power required to move a mid-sized vehicle at a constant speed of 56 km/h (35 mph)".

Sit with that one. On a gentle 35 mph cruise, cooling your face can cost the engine more than moving the car does.

The same paper notes that a mere 400 W of extra load costs about 1 mpg. Your compressor pulls five to ten times that.

Does that blunt your acceleration, then? Less than you would expect. ORNL's data caught the compressor being turned off briefly "due to the engine state (hard acceleration)". Many cars simply drop the AC out for a moment when you floor it, and hand the power straight back.

Does Revving The Engine Make The AC Cool Faster?

Mechanically, yes. The reason follows straight from the belt.

Compressor speed is tied to engine speed. More revs means more squeezes per minute, and so more cooling capacity. ORNL saw this from the other direction. On one city-driving run, stop-and-go traffic "reduced the engine speed, slowed the compressor speed and reduced air flow through the condenser". The cabin took noticeably longer to cool.

Notice that two things fell at once there. The compressor slowed, and less air moved over the condenser at the front of the car. Both cut cooling.

So revving in a parked car is close to pointless. You speed the pump up. But you are not moving, so the condenser gets no fresh airflow to dump heat into. Driving does both jobs at once. That is the real reason a car cools down faster once you are rolling.

Is It Bad To Idle Your Car With The AC On?

Not damaging. But wasteful, and by more than most people guess.

The Department of Energy says idling burns a quarter to a half gallon of fuel per hour, depending on engine size and AC use. Its cost table splits the two cases. Roughly 2 to 3 cents a minute with the AC off. Three to 5 cents a minute with it on.

ORNL measured the same effect on real cars. Full-load AC added about 0.2 gallons per hour on the Explorer at idle, a 55% increase. On the Corolla it added 0.13 gallons per hour, a 60% increase. Idling with the AC on burns roughly half again as much fuel as idling without it.

Over a thirty minute wait, that is real fuel bought for zero distance. DOE's advice is blunt: "Don't idle with the AC running before driving. Turn the AC on after you begin to drive."

One more detail is worth knowing. In a follow-up study, ORNL parked cars in midday sun until cabins hit 53 to 59 °C (128 to 139 °F). Then they drove with the AC at maximum. The compressor ran at 100% duty for 7 to 25 minutes. Most short trips never leave the expensive part of the curve.

How Much Fuel Does Car AC Use Nationwide?

Multiply one car's penalty across a country and it stops being trivial.

NREL's 2018 analysis in the SAE International Journal of Passenger Cars found that vehicle AC burns 7.6 billion gallons of fuel a year in the United States. That is 6.1% of all light-duty vehicle fuel use. It works out to 30.0 gallons per vehicle per year, or 23.5 grams of CO₂ per mile.

Six percent of a nation's driving fuel, spent on staying comfortable.

Individual trips can be far worse than that average. DOE says that in very hot conditions, AC "can reduce a conventional vehicle's fuel economy by more than 25%, particularly on short trips." NREL's earlier modelling found mid-sized vehicles losing over 20%. Very high fuel economy vehicles can lose around 50%.

That last figure surprises people. The cooling load is set by the weather and the cabin, not by the car. So the more efficient your engine, the bigger a bite the AC takes out of it.

AC Or Windows Down: Which Actually Uses Less Fuel?

Here the internet's favorite rule of thumb runs into measured data.

You have heard the rule. Windows down wins below about 45 mph, and AC wins above it, because open windows wreck your car's aerodynamic drag. It is repeated everywhere. It is also not what the most careful public measurement of the question found.

In 2013, Huff, West and Thomas at Oak Ridge National Laboratory tested a 2009 Ford Explorer and a 2009 Toyota Corolla. They ran both on the road and on a chassis dynamometer at steady cruising speeds. Their result: at every steady speed from 40 to 70 mph, both vehicles burned more fuel with the AC at maximum than with the windows down.

The Explorer held that pattern past 70 mph and never crossed over. The Corolla only broke even at 75 mph. Windows-down only lost at 80 mph.

That is not 45 mph. It is not close.

So how does this square with DOE, which still advises windows at low speed and AC on the highway? Honestly, without pretending the gap away. DOE is writing one simple heuristic for every driver in every condition. ORNL tested maximum cooling, the harshest AC setting there is, on two 2009 cars at steady speeds. Real driving runs lower duty cycles. And a boxy SUV takes a very different drag penalty from open windows than a small sedan does.

One more caveat deserves stating plainly. This is two cars from a single model year, and both ran the hardware of their era. NREL's modelling of that same 2007 to 2010 period describes a belt-driven fixed-displacement compressor, clutched on and off to regulate cooling.

Newer cars often do it differently. In a petition to the EPA, General Motors explained that a fixed compressor's output "is only controlled by cycling the compressor clutch on and off, which creates waste". A variable-displacement compressor can instead "moderate cooling capacity and work less", avoiding the cycling altogether. The EPA's credit schedule prices that gap at 1.7 g/mi for an externally controlled variable-displacement system against 1.1 g/mi for a fixed one.

So the crossover speed on a car built today is a genuinely open question, and not one this study can answer.

What survives all of that is still useful. The AC penalty is bigger, and the open-window penalty smaller, than the folklore claims.

Why Do Electric Cars Change The Answer Completely?

Now take the engine away.

An electric car has no crankshaft, no serpentine belt, and no clutch to grab. Its AC runs on an electric compressor, a self-contained motor and pump fed from the traction battery.

The thermodynamics do not change at all. The refrigerant still has to be squeezed, and that still costs a few kilowatts. Only the source of those kilowatts has moved.

Which produces a rather lovely turn. In an EV, the AC genuinely is running off the battery. The answer that is wrong for a gasoline car is exactly right for an electric one.

And because an EV carries a fixed store of energy, the cost shows up as lost range. The Department of Energy measured this on a dynamometer in 2024. Against a 22 °C (72 °F) baseline, battery-electric range fell by an average of 14% at 35 °C (95 °F).

Now here is the genuinely surprising part. DOE found that hot-weather drop to be "an identical proportional decrease to that of an ICEV in the same conditions". In summer, an electric car and a gasoline car surrender about the same share of their range to staying cool.

Winter is where they part company, and badly. At −7 °C (20 °F), EV range fell 41%, against 10% for a gasoline car. The reason is the mirror image of everything above. A gasoline engine throws away most of its fuel as heat, so cabin warmth is free waste heat. An EV is efficient enough that it has no waste heat to spare, and has to pay for warmth out of the same battery that moves the car.

There is a consolation, and it comes from the same physics. The compressor is no longer chained to engine speed, so it can run at its most efficient rate at any road speed. It can also cool the cabin while the car sits plugged in. That is a real advantage the battery-powered layout buys back.

So, Why Does Your Car's AC Make The Engine Work Harder?

Because you asked it to run a heat pump. And there is no cheap way to move heat.

The chain is short and entirely mechanical. Press the button. A magnetic clutch snaps a steel plate onto a spinning pulley. The serpentine belt starts turning a pump. That pump squeezes refrigerant hard enough to dump your cabin's heat into 35 °C air. The crankshaft supplies every watt of it: about 5 hp at peak, under 3 hp once things settle. Your engine notices, so you notice.

The battery was never involved. Neither was the alternator, beyond a trickle for the clutch coil and the 150-watt fan. Blaming electricity is understandable, because the fan is the part you can hear and feel. But the fan only moves air that something else worked very hard to cool. Every one of those refrigerant molecules got cold because a belt off your engine squeezed it.

This is the part I find quietly satisfying. There is no trick in the system anywhere. No loophole, no free comfort. Those 30 gallons a year are the honest price of pushing heat uphill. It is the same price your fridge pays, and your house pays, and every cooled space on Earth pays. Only the payer changes. In your car, it is the crankshaft. In your kitchen, it is the grid. In an electric car, it is the battery, and a little of your range.

So the next time the idle dips as you press that button, you will know exactly what happened. A belt just got a new job.