In answering this question I tossed out that a candle lantern can provide as much as 10 degrees (F) of temperature difference while winter camping. I've heard this tossed about before, but is it true? I assume there are differences between snow-caves (well insulated) and tents (less so)... Anyone have any cold hard numbers?
I've gone winter camping several times, usually staying in a tent, and I prefer to avoid candles in a tent so have no data about that.
However, on one long-weekend trip I stayed in an igloo built from blocks of snow, north of Grand Rapids MN. After the four of us on the trip skied far enough into Suomi Hills (see map) to be well away from roads and trails, we tramped down an area of snow on a lake, then after a few hours cut out snow blocks and built an igloo, about 10 feet (3 metre) across outside and a bit less than six feet (1.8 m) high inside. After it was done, I left a small candle-lantern burning inside while we fixed dinner outside.
By itself the candle warmed up the air in the igloo to 40°F (4°C), which I would count as about 8°F (5°C) of warming, supposing the inside surface of the igloo to be about 32°F (0°C). The small entrance to the igloo was covered much of the time. The air in the igloo warmed up to 50°F (10°C) with all of us in it, while outside air temp varied from 24°F (-4°C) down to -12°F (-24°C).
Getting 10°F (6°C) temperature rise from a candle in even the smallest of tents is clearly nonsense. Do the math.
Figure a candle puts out about 80 W. Of course there is large variation from candle to candle, but this is in the reasonable range for a typical modern paraffin candle. Let's say 100 W to be generous.
Next we need to come up with the surface area over which this supposed 10°F (6°C) difference will dissipate across. About the smallest you could call a "tent" would need to be long enough for a person to lie down in with some extra room sideways and at the head and toes. Let's say the footprint is 8x3 feet (2.4×0.9 m²). That's "small" by most standards. Let's also say the bottom is insulated. That means the 100 W is dissipated over at least 24 square feet (2.2 m²) just due to the footprint alone. Obviously the height of the tent will add some to that. Again, let's be generous and say the surface area of concern is only 25 square feet (2.3 m²). That's very small.
Dissipating 100 W over 25 square feet (2.3 m²) means 4 watts per square foot (44 W/m²), or 13.7 BTU/h per square foot. At a insulation "R value" of 1 ft²·°F·h/BTU (0.176 m²·K/W), 13.7 BTU/h per square foot (44 W/m²) would cause a 13.7°F (7.6°C) rise. That means the tent fabric would need to have a R value of 0.73 ft²·°F·h/BTU (0.128 m²·K/W; to sustain the 10°F/6°C rise at that same power level. Not gonna happen. To put this in perspective, a 1/2 inch (1.3 cm) of plywood has a R value of 0.63 ft²·°F·h/BTU (0.111 m²·K/W), and 1/2 (1.3 cm) drywall of 0.45 ft²·°F·h/BTU (0.079 m²·K/W). Do you really think a few mils (~50 µm) of nylon are going to insulate better than a 1/2 inch (1.3 cm) of plywood?
And this is just looking at the conductive heat losses thru the tent wall fabric. Of course there will be some ventilation, so a considerable fraction of the heat power will be lost by convection. And, these were all quite conservative numbers, especially considering we'd be talking about a 4-season tent when this would matter, and those tend to be physically larger. Even taking the conservative 3x8 foot (0.9×2.4 m²) footprint and adding a side wall just 3 feet (0.9 m) tall all around adds 66 square feet (6.2 m²). 90 square feet (8.4 m²) of surface area would still be a small tent. Consider that is equivalent to a 9.5 x 9.5 foot (2.9 metre) sheet of fabric.
The point of the ultra-conservative numbers was to show that it's not even close with that, so the 10°F (6°C) rise from a candle in any real winter tent is totally absurd.
A small candle burns about 1/8 ounce (3.5 g) per hour. paraffin has 19,900 btu/pound (46 MJ/kg). So a small candle releases about 19900/(8*16)=155 btu/hour (45 W).
A 5 foot (1.5 metre) diameter hemispherical igloo (including the floor) has a surface area of about 235 square feet (21.8 m²). The change in temperature was 8 degrees F (5°C). A proper igloo is about a foot (0.3 m) thick. A foot of dry snow has an R value of 12 ft²·°F·h/BTU (2.1 m²·K/W). The standard heat loss calculation is: SF * dT / R = btu/hr. In this case: 235 * 8 / 12 = 156.67 btu/hour (46 W)
So a small candle can indeed maintain the temperature of this igloo at 40F (4°C) when it's 32F (0°C) outside.
Humans at rest release about 300 btu/hour (88 W) ... much of that by breathing. This could keep the igloo at 40F (4°C) when it is significantly colder outside.
A winter tent the same size insulated with Thinsulate G600 (about 28 pounds, 12.7 kg), has an R value of 5.29 ft²·°F·h/BTU (0.93 m²·K/W). To heat it by 8 °F (5°C) you would need 235*8/5.29=355 btu/hr (100 W). That's a little less than 3 candles.
A single layer nylon tent (R=.027 ft²·°F·h/BTU, 0.004752 m²·K/W) is the same size would need 69,630 btu/hr (20.4 kW). That's 450 candles.