Same heat every year?

Brother jack let it be know that Bangkok's horrible heat and humidity was the game, and that he expected solutions from me. "You're the guy from MIT," he said.

I had a couple of days before leaving and instead of buying gift chocolate figurines of Mickey Mouse or Parodi cigars as gifts for all the people I was going to meet I went for the UCLA library and loaded up on data. I found a huge collection of Bangkok weather data from some American military spy agency and almost missed my flight transferring all the files to a CD.

Our ultimate concern here is the weather within a room, or between your shirt and your skin, not the world at large. Although we are not particularly interested in grand questions such as global warming or pressure ridges in China -- I know nothing of these matters -- one of the first things that I asked myself was whether the weather, or more specifically the heat, which is my enemy here, is always the same in Bangkok, year by miserable year.

This extract from my massive data file (which you can have, for free, if you have some good purpose)gives a quick answer to this question. It shows the daily mid afternoon temperature, which is almost always the hottest temperature of the day, for fifteen years, starting in 1982.

Given that you are going to start to sweat at 30 degrees even at rest, you can see that a brisk walk at three pm is not going to be enjoyable.

Especially if it happens to be 1983.

But will jogging be more pleasant at six in the morning?

Night heat in June

We saw the other day that in theory the roof on brother Jack's house retains a critical amount of heat resulting in a room temperature that is two or three degrees higher than the outside temperature -- enough to make the difference between a good and miserable sleep.

How does this check out in practice?

Here you can see my notes from a recent stay at the house. I measured the temp of the interior wall surface with a radiant thermometer until about 3 am, at which time the outside air temp was about 29 and the interior temp hanging up at 31 -- too warm for comfort (the orange colored zone). This was a pretty typical June night (June 8, 2007) ending with a dawn temperature of 27 to 29, (dawn temp also happens in almost all cases to be the daily minimum.). This matches the experience that Jack has been reporting, too hot in room in the middle of the night many nights in June. (It doesn't help that Dah insists on closing the windows at night for fear of intruders).

The interesting thing here is that the the discomfort comes from only a small amount of temp elevation. If we can drag the temp down just a couple of degrees, we will be as happy as the farmer sleeping under a cool nighttime tin roof. The challenge is this-- how can we get this modest amount of cooling without resorting to the overkill of air conditioning?

Is a hot tin roof cooler?

Interestingly enough, the answer might be yes if we are talking about the time that matters most to many folks, that being at night when the outside temperature is slowly falling from the low thirties (too hot!) to the upper twenties (reasonably cool for sleeping). A "tin" (galvanized iron) roof is fairly light, perhaps a kilogram per square meter and has a very low "R" value. So hot as it might be even at sundown, it quickly cools to the ambient air temperature, even two or three degrees below if the humidity is not too high and the effective sky temperature is low, say 5 degrees or so. So you may even get some free radiative air conditioning from such a roof at night.

Lets look at Brother Jack's roof:

This roof weighs about 10 kg/sm and has an "R" value of about 5, and each square meter retains perhaps 10,000 joules of heat per degree. If you want it to cool as fast as the outside air, say 1 degree per hour, you've got to make it discharge 10,000 joules per hour per sm, or 3 watts/sm. If the "R" value is 5, this will require a temperature gradient across the roof from inside to outside of about 3 degrees. In other words the interior temperature will be 3 degrees higher than the outside air temperature--which is what we saw earlier on our April 3 measurements (vertical difference between red and blue lines). Therefore even when the outside air dips to a nice 29 degrees, the inside of the room will still be a sweaty 32 -- too warm for sleeping. And if in your desperation to keep cool in the daytime you had installed some bulk insulation instead of the foil insulation, the "R" value might be double, and the required temperature to push 3 watts out through the roof even more, maybe 5 degrees, an even worse situation! Meanwhile the poor farmer next door will be happily snoozing with his tin roof at or below outside air temperature! His red line will closely track the blue line, maybe be even lower.

So the joke is on you, Mister high-tech fahlang--you tricked nature in the afternoon, but she got back at you when you were trying to sleep!