With incandescent bulbs, in general, there is a trade-off between light output at a given wattage and how long the bulb lasts until it burns out. But I have also noticed that some light bulbs were designed to last much longer than others, without any sacrifice in efficiency in comparison. So there must be other factors besides just the filament thickness. I am wondering whether the quality of the filament has anything to do with it. There is also a rumor that in recent years GE and Phillips were allowing traces of air into the bulb to intentionally cause premature failure, so consumers would have to buy more bulbs. One accusation is that they did this as part of their intense effort to convince politicians/consumers that the old bulbs should be phased out in favor of the new "longer lasting" lighting technologies. They seem to have stopped manufacturing their more expensive "long life" light bulbs around the same time they began lobbying for the phase out.
Another possibility is that some of the "long life" bulbs simply had a higher lifespan rating because the filament has supports. But would this not only affect the lifespan if the light bulb was being exposed to vibration? How could filament supports increase the life span if the hot filament was not being agitated in some way? And how much can filament supports really extend lifetime, if this is the case? Something else I have been wondering is whether some of the bulbs were actually being filled with krypton gas, instead of the usual argon-nitrogen mix. Perhaps the manufacturers simply did not bother to mention this fact, as it was too technical for consumers to understand, or perhaps the technical details of the bulb was considered something of a trade secret the company did not want to be advertising to its competitors. Krypton gas typically offers the most benefit for bulbs rated under 40 watts. For a 100 watt bulb, the addition of krypton can improve the efficiency of the bulb by 4% and extend the filament life by around 50%. Krypton is also more expensive than argon, not by much, but it could significantly increase the cost of such an already inexpensive bulb (0.25-0.50 USD). Obviously a biger light bulb requires more gas than a smaller flashlight bulb. A standard sized A19 industrial krypton light bulb could be bought for 2.75 USD.
I have a package of 100 watt Sylvania light bulbs rated 1500 lumens with a 1500 hour rated lifespan.
I also have another package of 100 watt light bulbs (from another company) also rated 1500 lumens, but with a 2000 hour rated life span.
It seems one is no longer able to by these bulbs anymore. All the 100 watt light bulbs being sold the year before the phase out (when all the 100 watt bulbs dissappeared from store shelves) only has a rated lifespan of 750 hours! And I can state from experience that they also seemed to burn out very fast in my home. Good thing I have a stash of the old light bulbs tucked away.
I am also wondering whether the new halogen bulbs are really much of an improvement over the old light bulbs. As I mentioned, there is a trade-off between efficiency and lifespan. A typical "100 watt replacement" halogen manufactured by Sylvania is 72 watts, rated at 1480 lumens, and a 0.9 year (983 hours) lifespan. The "double life" version by the same company (also 72 watts and still claiming to be a 100 watt replacement) is somewhat less bright, rated at 1280 lumens and 1.8 years (1965 hours). Sylvania does not make any higher wattage long life halogen bulbs. In fact 72 watts is the highest wattage they manufacture a standard A19 sized bulb now. Bulbrite, however manufactures a 95 watt halogen with a 2000 hour rated life span. The company explicitly advertises that the capsule is filled with xenon (which is even better than krypton), so one would expect this bulb to have the highest efficiency for its given life span. It is rated 95 watts, 1600 lumens, 2000 hours. (the bulb is not in the shape of a standard pear-shaped A19, but rather in the shape of the old BT15 halogen double enveloped bulbs). This is a premium halogen bulb being sold for over 3.50 each.
But the efficiency of this 95 watt xenon halogen bulb is 16.8 lumens per watt, while the old 100 watt bulbs I have are 15 lumens per watt, both at the same life span. In other words this premium extra-efficient halogen bulb is only 12% more efficient than the old good quality incandescent bulbs. Is it worth it? In other words, if we simply took a good quality normal incandescent bulb that consumed 72 watts of power, and gave it a thinner filament to incandesce at a higher temperature, it should be able to give off 1302 lumens of light and last for nearly 1000 hours. Again, I am wondering whether these 100 watt bulbs I have were filled with krypton, or possibly whether the propaganda from the big 3 lighting companies are overexaggerating the inefficiency of the old incandescent bulbs.
Just for anyone who is wondering, I bought a pack of krypton traffic bulbs a while back. They are rated at 116 watts, 8000 hour lifespan, and 1260 lumens. So incandescent bulbs certainly can be made to last much longer. One would think with such a long-lasting bulb the light would be very yellowish because of the lower temperature, but surprisingly the color temperature is rated 2850K, which is a little higher even than some halogen bulbs.
UPDATE: I have tried these bulbs, and it seems their actual color temperature is lower, about the same as a 130v rough service bulb being used in a 120v outlet, just a little too yellowish for my preferences. I guess that is to be expected. You can't expect an 8000 hour lifetime from an incandescent bulb without some tradeoffs, apparently.
If you have a difficult to reach lamp fixture on a high ceiling in your living room, and really want the excellent quality of light from incandescent bulbs, these traffic bulbs may be an option to consider, if you are willing to pay the higher electric bills over many years. Of course, this might not matter so much if you live in a cold climate, since you will just be paying for extra heat. (a little advice if you want to buy one, be sure you get one with a rated voltage that exactly matches your outlet voltage, most of the traffic bulbs are rated at a slightly higher voltage, and this will significantly decrease both power consumption and light output, and make the light more yellowish)
I believe the color temperature of the incandescent bulbs most people have come to regard as "normal" is 2750-2850 K.
However, some Americans may now be using 130v rated bulbs without realising it, in which case the color temperature would indeed be a more yellowish 2700 K.Electronic Formulas and Calculations
The total power consumption of an incandescent filament is inversely proportional to the resistance. W= 1/r
A higher wattage bulb has a thicker filament. Keeping the resistance constant, a shorter length and thicker filament means higher temperature.
Some of you might be wondering why a regular wire does not get incredibly hot, since it is designed to have minimal resistance. The answer is that the current it carries is only a fraction of what it could potentially carry. If the wire was attached to large enough supply of electric current, it would indeed act as a heat filament. Ordinarily, wires are only connected to a complete ciruit through a current limiting load, such as the filament in a light bulb. Furthermore, as the tungsten filament is heated, within a fraction of a second its resistance increases to around 15 times the resistance value when cold. (another smaller factor is that tungsten has twice the resistance of aluminum used in the wire) So, theoretically, any given length of wire should be releasing 1/30th the ammount of heat that a similar length of the incandescent filament is releasing. We do have to remember that the tungsten filament in a typical incandescent lightbulb is a double coil, and the actual length is quite long, a whole 2 meters in a 60 watt bulb! So, if we used 60 meters of wire to connect our light bulb to the power source, both the light bulb and wires would be giving off the same ammount of heat! It is no wonder then higher voltages are used to distribute electric power in power lines, as power losses are proportional to the the current, and that is keeping the wattage constant, since the increasing voltage does not increase losses. (it is the square of the current if we are just calculating heat generated taking into account volts and amperage).
Here is one calculation that says a 100 foot AWG extension cord itself consumes about 20 watts of power in heat losses if it is powering a 600 watt load:http://www.answerbag.com/q_view/12772
There is also voltage drop to be considered. Making a conducting wire thicker will reduce its resistance and reduce how much voltage it consumes relative to the primary load.
For example, if we have a 10v power supply, wires with a total resistance of 2 ohms, and a load with a resistance of 3 ohms, the total current able to flow through will be 10/5 = 2 amps, and then the voltage drop in the wires will be 2x2 = 4v, so there will only be a potential of 6 volts left across the primary load. Doubling cross section area (thickness) of the wire would cut the resistance in half to only 1 ohm, and then the voltage drop in the wires would be only 2 volts. This would mean less heat wasted in the wires.
So in conclusion, the wattage of an incandescent bulb is determined by the resistance value of the filament. A filament can be made thicker, but it also has to be made longer to maintain the given resistance value. A longer filament means the heat is spread out over a larger area and means a decrease in filament temperature. This will lead to longer lifespan because the tungsten will evaporate off the hot filament at a slower rate, and it will take more time for the evaporation to lead to eventual rupture, because the filament is thicker.