Before delving into the world of efficient use of Peltier devices, it’s important to understand what we mean by the word “efficient”, not only by definition, but also conceptually.

For the purists, a percentage represents a portion of quantity, and so the numbers must always be bounded between 0 to 100%. For example, when measuring the efficiency of an electric kettle to boil water for a cup of tea, we know that the heating efficiency can never be less than 0 or more that 100%, because the energy in the cup is a portion of the energy put into the kettle.

However, another popular use is as a ratio of quantities. For example we might express the increase in sales from one year to the next as a percentage: it’s not saying one year’s sales are literally a portion of the previous year’s sales. Being a ratio, there is no boundary, it can be anything. In this use, it’s important to know exactly what the ratio is to avoid confusion, for example “103%” could mean just a little bit more or basically double: without a definition, it’s impossible to be sure.

In the case of heat pumps, when efficiency expressed as a percentage it usually means the cooling side power (Pc) as a ratio of the driving power or “waste heat” (Pw): efficiency = Pc/Pw x 100%.

Since this is a ratio, there is no law that limits this to 0-100%. As a silly example, I can carry the above cup of hot water from the kitchen to my office, and in doing so I am “pumping” about 50kJ of heat from one room to another. I might use about 2kJ of body energy to do this. This means my “efficiency” as a heat pump is pretty darn good at 2500%. I can achieve this because the energy in the cup is not taken from the energy I’m using walking from one room to another.

As compressor heat pumps improved, claimed efficiencies of 300% or more were not unusual, at which point the purists started complaining about the abuse of percentages, leading to the adoption of “COP” or coefficient of performance, which simply removes the percentage part and just refers to the ratio. There’s an additional twist for heat pumps in that the so called waste heat often combines with the pumping heat, which can be useful depending on the situation. For Peltiers, this is usually the case, for example, a Peltier that can sink 18W on the cold side with 6W input (COP = 3), will output 24W on the hot side (COP = 4). Thus, when discussing COP, or efficiency of heat pumps, it’s important to know if it is hot or cold side based. It’s also interesting to note that for Peltiers, the COP heating = COP cooling + 1.

Another point to note is that for COP and efficiency, either power or energy can be used, as long as it’s consistent. In some situations power makes more sense, while in others, energy is reasonable.

When people talk about Peltiers being only 5% efficient, they typically mean there is only 5% cooling power relative to the input power. For example, with 10W input, there would be 0.5W of cooling (and 10.5W on the hot side). But, this figure is so low as to be unrealistic. It is more likely that they are referring to the cooling of an object external to Peltier, like a can of Coke. In this case, the numbers can be very low, mainly because of the poor thermal coupling.

Let’s say the hot side of a TEC1-12708 Peltier is effectively cooled to room temperature by a nice water pump, while the cold side just has a thermocouple to measure the temperature. With a current of 3A/20W, the thermocouple quickly goes below -10°C in about 20s. It looks cool, a nice party trick. But if you calculate the energy efficiency, it turns out the thermocouple only needs about 0.2J to cool to -10°C, while the input energy was 400J (20s @ 20W), giving an efficiency of just 0.05%. Terrible!

But is this because of the Peltier? No, it’s just the set up. It’s like measuring the open circuit voltage of a battery, which is not a useful condition to talk about efficiency. Many cooling applications are like this, there are inherently high thermal resistances involved and it’s wrong to blame any system inefficiency on the heat pump itself. A can of Coke initially at room temperature will still take hours to cool in a compressor based fridge, and the efficiency will be terrible, in the order of 5%, despite the compressor system itself having a COP of 3 or more.

The big mistake of experts make when assessing Peltiers is comparing the COP of a compressor system under optimal loaded condition (low resistance thermal coupling, low temperature difference) against a Pelter with a crappy thermal coupling to a can of Coke and a large temperature difference.

Peltiers can easily achieve COP of 10 or more at low power, and cooling COP of 2 or more at useful power levels with useful with temperature differences. And in beverage cooling, COP is often not that important, since cooling generally needs to counter the leakage heat which is typically in the order of 10W or less. Under these conditions, whether the COP is 1, 2 or 3 isn’t a big issue. But a Peltier can be far more efficient on size, weight, controllability, cost and environmental impact compared to a compressor, and the world of solar and battery, can directly run off dc without needing big inverters, or using big dumb high power factor motors.

There’s a lot of potential here!