The Human Heat Balance

Your body is a slow furnace. Even at rest it turns food into roughly 100 watts of heat - about the same as an old incandescent bulb - and that heat has to go somewhere, continuously, or your core temperature drifts. Human physiology defends a core of about 37 C fiercely, so the body is always shedding heat to its surroundings. Comfort is simply the state where it can do that effortlessly: no shivering, no sweating, no conscious sense of being too hot or too cold.

That is why comfort is best understood as a heat balance, not a temperature. On one side is the heat you generate (your metabolism); on the other is the heat you lose to the room. When the two match at an easy, unforced rate, you are comfortable. When the room makes you lose heat too fast, you feel cold and your body clamps down blood flow and shivers. When the room will not let you lose heat fast enough, heat builds, you sweat, and you feel hot. A comfort simulation is, at heart, a calculation of whether an average person's heat balance can settle comfortably in a given room - and it needs six inputs to do it.

Four of the six comfort factors belong to the room. The other two belong to you, and they change the answer completely.

Metabolic rate (met) is how much heat you are producing, set by activity. The unit is the met: 1 met = 58.2 W/m2 of body surface, roughly a seated, relaxed adult (about 100 W over a typical 1.8 m2 body). Sleeping is about 0.7 met; quiet office work about 1.2 met; walking briskly around 2.5 met; heavy manual work 3-4 met or more. A gym and a library at the same air temperature are not the same thermal problem, because the people in them are generating very different amounts of heat.

Clothing insulation (clo) is how much the room can pull that heat away. The unit is the clo: 1 clo = 0.155 m2K/W, roughly a business suit. A light summer outfit is about 0.5 clo; typical indoor dress around 0.6-1.0 clo; heavy winter layers 1.5 clo or more. Clothing is a controllable, personal insulator - which is exactly why a person can restore their own comfort by removing a jacket, and why a comfort model must be told what people are wearing. Assume the wrong clo and the whole prediction shifts by a degree or two.

The body sheds its heat by four physical paths, and a room can block or open each one independently. This is why air temperature alone is a poor guide.

Radiation is heat leaving your skin as infrared toward every surface around you. If those surfaces are cold - a single-glazed window in winter, an uninsulated wall - you radiate heat to them and feel cold even in warm air. If they are hot - a sun-baked concrete roof, a west wall in the afternoon - they radiate back and you feel hot even in cool air. This is captured by the mean radiant temperature (MRT), and it is often the factor beginners forget.

Convection is heat carried away by air moving over your skin. It depends on air temperature and, crucially, air speed: a breeze of 0.5-1.0 m/s strips heat far faster than still air, which is why a fan cools you without changing the air temperature at all.

Evaporation is heat lost as sweat turns to vapour - the body's emergency cooling, and the one that humidity governs. In humid air, sweat cannot evaporate, so this path chokes and you feel sticky and hot even at moderate temperatures. In dry air it works freely.

Conduction is heat lost by direct contact - a cold floor underfoot, a metal chair - usually small, but not always. A fifth, quieter path is respiration: you breathe in cool air and exhale warm, humid air, carrying away a steady trickle of heat and moisture. Together these paths, plus the metabolic heat you make, are the whole balance sheet.

There is a subtlety worth naming early, because it trips up beginners. Comfort is not only about the whole-body balance being right; it can also be spoiled by local discomfort even when the average is fine. A cold draught on the neck, a warm floor, a strong temperature difference between head and ankles (vertical stratification), or one very cold surface radiating on one side of you (radiant asymmetry) can all make a person uncomfortable in a room whose average is neutral. This is why comfort standards set limits on draught, floor temperature and asymmetry separately, and why 'the average is fine' is never the end of the comfort question.

Put it together and there are exactly six factors that decide thermal comfort - the reason a single thermostat reading can never settle the question:

1. Air temperature (dry-bulb) - the temperature of the air itself. 2. Mean radiant temperature (MRT) - the average temperature of the surfaces around you, driving radiant exchange. 3. Relative humidity - how much water the air holds, governing whether sweat can evaporate. 4. Air speed - how fast air moves over your skin, governing convective and evaporative loss. 5. Metabolic rate (met) - the heat you generate, set by activity. 6. Clothing insulation (clo) - the insulation you wear.

The first four are environmental (the room); the last two are personal (you). Change any one and comfort changes. This is the deep reason simulation matters: you cannot judge a room from a single number, because the six factors trade off. Cold surfaces (low MRT) can be offset by warmer air; high humidity can be offset by air movement; heavy activity (high met) demands cooler or breezier conditions. A comfort model is the bookkeeping that keeps all six honest at once - and every model in the next three lessons takes these same six as its inputs.

In a simulation, the six factors come from different places. Air temperature, MRT, humidity and air speed are outputs of the building thermal model (EnergyPlus, or a Ladybug/Honeybee energy model) at each hour - the physics of the envelope, glass, sun and ventilation produce them. Met and clo are assumptions you supply, based on how the space is used: an office at 1.1 met and 0.6-1.0 clo, a bedroom at 0.7 met.

A useful shorthand the tools compute is operative temperature - roughly the average of air temperature and MRT, weighted for air speed. It is a single number that folds radiation into air temperature, which is why comfort standards specify operative temperature rather than air temperature. When a room has cold or hot surfaces, air temperature and operative temperature diverge - and it is operative temperature the body actually feels.

Because met and clo are assumptions rather than measurements, honest practice tests a small range rather than a single value: an office might be run at 1.0-1.2 met and 0.5-1.0 clo across the seasons, and the comfort result reported as a band, not a false-precision point. This matters because a change of about 0.5 clo shifts the comfortable operative temperature by roughly 3 C - a bigger effect than most envelope tweaks - so the clothing you assume can dominate the answer. The discipline of the whole module is to keep all six factors visible and to be explicit about which are physics outputs and which are human assumptions. The whole point of the next lesson is a model (Fanger's PMV) that takes all six factors and predicts, on a scale, how an average occupant will vote - turning this heat-balance picture into a number you can design against.