There is the following problem in a textbook on Thermodynamics:

A heat pump is an electrical device that heats a building by pumping heat in from the cold outside. In other words, it’s the same as a refrigerator, but its purpose is to warm the hot reservoir rather than to cool the cold reservoir (even though it does both).

I have a hard time understanding the principal difference between a heat pump and a refrigerator. In a refrigerator, heat is drawn from a cold substance into the refrigerator cabin and then excess heat is pumped out of the device.

Now, in the case of a heat pump, heat is also pumped from a "cold substance" (the cold outside) and it goes into the "cabin" (the building interior). But then how does the building get heated and not refrigerated? I'm really confused by this problem statement, it couldn't be less clear.

  • 1
    $\begingroup$ Heat pumps have been used for many years and many are "reversible" in that they can heat buildings in the winter and cool them in the summer. A large river is a very good thermal reservoir. designingbuildings.co.uk/wiki/CIBSE_Case_Study_Kingston_Heights $\endgroup$
    – Farcher
    Commented Jul 9, 2017 at 7:16
  • 6
    $\begingroup$ It's a refrigerator installed backwards... A refrigerator moves heat from inside the fridge to outside. A heat pump moves heat from outside the building to inside. (Note: Where I live heat pumps are made to work in both directions, so they are forwards refrigerators as well as backwards refrigerators depending on which button you push) $\endgroup$ Commented Jul 9, 2017 at 10:12
  • $\begingroup$ I can't readily accept the notion of pumping coldness. Rather I find it easier to understand that heat is pumped even from a cold place to a warm place. However, like the electron hole is still effectively an electron transmission in the opposite direction. So saying pumping cold is like observing the movement of an electron-hole. $\endgroup$ Commented Jul 10, 2017 at 4:33
  • $\begingroup$ @CynthiaAvishegnath I think what is being pumped is not cold but whatever heat there is in the cold. That is, whatever kinetic energy remains in the air, say, even if it's close to 0 K. $\endgroup$
    – sequence
    Commented Jul 10, 2017 at 6:23
  • $\begingroup$ @CynthiaAvishegnath As well as you should not. Cold is a sensation. In physics, there is only heat or the absence of heat. Any concept of 'pumping coldness' is only going to make this harder to understand. $\endgroup$
    – JimmyJames
    Commented Jul 10, 2017 at 14:35

4 Answers 4


The principal difference between a heat pump and a refrigerator is a matter of perspective. A heat pump is simply a refrigerator where the inside of the refrigerator is outside the house. In other words, a heat pump attempts to cool down the air outside the building, and in the process heats up the air inside the building.

Put differently, a refrigerator is simply a heat pump which takes heat from inside the refrigerator and transfers it outside the refrigerator.


Before answering the question let's talk about the word "efficiency".

In the meaning we care about here the word is used to described the effectiveness with which some consumed resource is converted into some desired effect. In the context of thermodynmanic engines, refrigerators and heat pumps the desired outcome and the available resource are one of

  • heat moved between the warm source and the mechanism
  • heat moved between the cold source and the mechanism
  • work (done by or on the mechanism)

all three of which are denominated in terms of energy, so we use a ratio $$ \text{efficiency} = \frac{\text{desired outcome (in energy)}}{\text{resource used(in energy)}} \;.$$

In an engine we feed energy in (heat from the hot reservoir to the engine, and get work out, so $$ \text{efficiency}_\text{engine} = \frac{\text{work done}}{\text{heat supplied}} \;.$$

In the case of a either refrigerator or a heat pump supplied work is the resource used up, but the desired outcome varies: for refrigeration you care about heat removed from the cold source, while in a heat pump you care about the heat added to the hot source. (For historical reasons we also use the phrase "coefficient of performance" instead of "efficiency", here.) \begin{align*} \text{CoP}_\text{refrigerator} &= \frac{\text{heat removed}}{\text{work supplied}} \\ \\ \text{CoP}_\text{heat pump} &= \frac{\text{heat supplied}}{\text{work supplied}} \;. \end{align*}

So, to finally put the answer in words, the only difference is what you are trying to accomplish.

  • $\begingroup$ Excellent explanation; and honestly it should be written like this in the original text. +1 $\endgroup$ Commented Jul 9, 2017 at 7:14

I think that part of your confusion stems from the fact that the statement you highlighted is rather poorly phrased. A couple of fact that might clarify things:

1.) a heat pump is a fancy science word for any refrigeration device or unit, be it a kitchen refrigerator, air conditioner, "swamp box" or anything else. I'm not sure why the author of the quote is drawing some kind of difference between a refrigerator and a heat pump. (Somebody correct me if I'm wrong on this point).

2.) Somewhere in your quote, the author seems confused about how "heat pumps" are normally configured. Every "heat pump" I've ever seen in the real world was used to cool off a room or the inside of a freezer, etc.

3.) Lastly, every device whose authentic purpose was to heat a room was called a heater. (My sarcasm is aimed at the author of the quote, not you SEQUENCE). Granted, a traditional food refrigerator may incidentally heat a room up by cooling off the inside of the "cabin", but this is incidental, and should be presented as such.

I agree that the author couldn't be less clear. The author seems to need to re-write this section of the book.

  • 1
    $\begingroup$ Concerning 'Every "heat pump" I've ever seen in the real world was used to cool off a room or the inside of a freezer, etc. ' and 'every device whose authentic purpose was to heat a room was called a heater' In the building trades a "heat pump" is a mechanism that can be used to either heat or cool a dwelling by exchanging thermal energy with the surroundings (usually through a coil buried int he ground). $\endgroup$ Commented Jul 9, 2017 at 5:28
  • $\begingroup$ And concerning "I'm not sure why the author of the quote is drawing some kind of difference between a refrigerator and a heat pump." Of course a refrigerator is a heat pump in a general meaning (in many cases—including the advanced heating and cooling system for buildings—it is exactly the same mechanism as a refrigerator), but in the technical language of thermodynamics there is an exact difference which relates to what outcome you care about (see my answer). This distinction is entirely standard in introductory thermodynamics texts. $\endgroup$ Commented Jul 9, 2017 at 5:28
  • 2
    $\begingroup$ As an aside, I typically ask student to determine how much more energy efficient it is to use a heat pump for warming a building than a resistive heater such as an electric fire. The answer can be quite startling and the exercise is worth your while. $\endgroup$ Commented Jul 9, 2017 at 5:31
  • $\begingroup$ Unfortunately, a lot of the problems in this book are poorly phrased, so that it takes a lot of time to actually understand what a problem is asking. But the book is the reference book for the course. $\endgroup$
    – sequence
    Commented Jul 9, 2017 at 6:50
  • $\begingroup$ @dmckee - my apologies. Point about heat pump lingo as used in the building trades is conceeded. Moreover, I had a feeling that there was a certain amount of blurring the lines between what we normally see in the HVAC world and the strict terminology of thermo books. $\endgroup$
    – the_photon
    Commented Jul 9, 2017 at 6:56

Although other people focused on the fact that when performing a single operation they are essentially the same, no-one has pointed out that heat pumps are often built to be reversible, i.e. they can be used both to cool down and to warm up an environment (they are easily switchable between those two configurations). Refrigerators are built to perform just that function.

See also the relevant Wikipedia article on heat pumps.

In particular this excerpt (emphasis mine):

Heat energy naturally transfers from warmer places to colder spaces. However, a heat pump can reverse this by absorbing heat from a cold space and releasing it to a warmer one. Heat is not conserved in this process and requires some amount of external energy such as electricity. In heating, ventilation and air conditioning (HVAC) systems, the term heat pump usually refers to vapor-compression refrigeration devices optimized for high efficiency in both directions of thermal energy transfer. These heat pumps can be reversible, and work in either direction to provide heating or cooling to the internal space.

Bottom line: from an engineering perspective, an heat pump can be easily reconfigured either to warm up or cool down the conditioned area, whereas a refrigerator can't.

  • $\begingroup$ Can you please explain what conservation of heat means and why additional work needs to be put in in the case of an A/C? I know that $\Delta U = Q + W=Q_h - Q_c + W$ (with $Q_h$ being the heat dumped to the hot reservoir and $Q_c$ the heat taken from the cold reservoir). If something works the other way around (an engine, say), then it's the work that is expelled as a result of the engine's operation. However, I still don't conceptually get what conservation of heat exactly means. $\endgroup$
    – sequence
    Commented Jul 9, 2017 at 23:23
  • $\begingroup$ There is no general rule requiring the conservation of heat, only the conservation of energy as expressed in the first law (you should be aware that there exist different sign conventions to be applied to $W$, so in some texts the mathematical expression of the first law may be different from the one that you have written). $\endgroup$ Commented Jul 10, 2017 at 5:36

Not the answer you're looking for? Browse other questions tagged or ask your own question.