Carburetor Icing

by Dug Smith

As we’re getting into winter, it’s worth looking into carb icing. While lower temperatures themselves don’t help, the increased humidity is the real problem.

The Venturi effect states that if you push air through a constriction, its speed will increase, but its pressure will decrease – this is how our carbs work, after all. Unfortunately, this pressure drop also creates a drop in temperature, and that can be 30F or more. Gasoline evaporating lowers the temperature of the incoming charge even further.

As we know, when the temperature of the air cools lower than the dew point, water condenses. Even when it’s a humid 60-70F outside, the temperature drop will often cause ice to form. This ice can build up as a restriction in the carb throat, causing more of a constriction and decreasing the pressure even further, and that will cause the engine to run rich.

Ice can also form on the butterfly valve that’s used to restrict the incoming air flow and controls the amount of fuel the engine receives – left long enough, this will stop your throttle from working at all. Of course, those of us with 2 stroke engines are usually using slide carburetors, and won’t have this particular problem.

The solution is to warm the air entering the engine with carburetor heat, where heat from the exhaust manifold, coolant or an electrical circuit, is used to prevent icing.

As the air filter is often bypassed when using carb heat, it’s usually not a good idea to use it on the ground (however carb heat can help out when the air filter is clogged).

Carb heat will cause an engine to produce less power, as the incoming air is less dense. If applying it increases the power output, there’s a good chance that ice buildup had started. Water ingestion will cause an engine to run rough, but that doesn’t mean carb heat should be turned off.

Full throttle applications restrict the air flow less than partial throttle, so carb heat is generally less necessary, but long descents can be problematic because there is a large restriction in the air flow, and less exhaust heat to melt any ice that’s being formed. Clouds are formed by moisture in the air, so carb heat is more frequently used in and around them.

AOPA has this handy chart that shows when you can expect to see problems with carb icing. As I’m typing this on an early November day, it’s 60F outside, with a dew point of 40F, so I could expect to see icing at cruise power and serious icing in a glide.