Light Bulb Lifetime and Wattage
Country: United States
Date: May 2007
Does a light bulb's life depend on its wattage?
Light bulbs fail primarily because, over time, the tungsten filament
actually evaporates slowly at the very high temperatures it operates
at. When there is an area that becomes a little thinner than the rest
of the filament, that area has a little more resistance, and as a
result gets hotter. The increased heat at that point causes a more
rapid rate of evaporation of the filament there, causing more heat
and more evaporation, and so on. Eventually that area gets so thin
and hot, it burns out.
A 40 Watt light's filament (for example) operates at much the same
temperature as the filament in a 100 Watt light. The difference between
them is that the 100 Watt light has a thicker filament and uses more
current. But their filament temperatures are much the same, so their
rate of evaporation of their filaments is similar.
You might think that the slightly thicker filament of a higher wattage
light would allow a longer time before it evaporated enough to burn
out. However, because the 100 Watt light uses more current, the minimum
thickness of the filament before it will overheat and fail, is thicker
than for a 40 Watt light. So the result is the lifetime is much the
same. Of course this comparison only applies for lamps that are
similar to each other except for wattage. You cannot compare, for
example, a Christmas tree light with a 100 Watt lamp because there
are many other things that are different about them.
Wattage is but one variable in determining the "lifetime" of a light
bulb -- there are numerous others. Some examples: geometry of the
filament (assuming an incandescent light and not a fluorescent bulb,
an LED, or a laser), the operating temperature, the filament material,
the "on - off" cycling of the lamp, and other variables too. As a rule,
higher wattage lights "tend" to have a shorter lifetime, but it is
difficult to make a hard and fast rule.
No, the lifetime does not depend directly on the wattage (power). Rather,
the lifetime depends most directly on the temperature of the filament, all
other things being equal.
The hotter the filament, the shorter the life of an average incandescent
A bulb that is powered so that the filament is glowing a brilliant
blue-white will soon burn out, regardless of the power. And a bulb that is
powered so that filament is glowing a dim yellow, will last for years,
regardless of the power.
Commercial bulbs are rated for a certain voltage. That is, the manufacturer
specifies the voltage necessary to make the bulb operate properly - to give
a reasonable bright light without burning out too fast. It might be 120V I
a house, or maybe 12V in a car.
The light output and lifetime of lamp filaments is known pretty accurately
as a function of voltage. Briefly: the amount of light that a bulb produces
goes as the 3.5 power of the voltage ratio. The lifetime goes as the
inverse twelfth power of the voltage ratio. This is for a normal light
bulb. Also, the current goes as the 0.55 power of the voltage ratio.
Voltage ratio is the applied voltage divided by the rated voltage.
So, for example, if a 220V bulb is operated at 110V, the voltage ratio is
0.5. The light output will be 0.5 raised to the 3.5 power, which is 0.088.
It will be very dim! It will make less than one-tenth as much light as if
it were run on 220V. It will also be orange in color.
The bulb will last a LONG time at this low output. The lifetime will be
increased by 0.5 raised to the inverse twelfth power, or 2 raised to the
twelfth power, which is 4096 times as long! It will last for thousands of
Light bulbs for photographic work are sometimes overdriven to make a lot of
light, but they do not last as long.
For example, suppose a 120V bulb is driven at 140V. The voltage ratio is
140/120 = 1.167. The light output is 1.7 times as much. The actual power
consumed is 1.27 times as much, so the bulb is more efficient. However, the
lifetime of the bulb is cut down by a factor of 6.35.
For fun, I ran a 12V light bulb on 22V. It was very bright. It lasted 2
Click here to return to the Engineering Archives
Update: June 2012