I wanted to bleed my radiators by myself for the first time and I was wondering what is the reason that bleeding radiators saves energy?

Most of the guides on bleeding radiators mention that this is important for saving energy but none explain the exact reason why. I understand that by bleeding you increase the surface area through which the radiator can heat the room and probably also the surface area through which the hot water heats the radiator.

  • 8
    Don't you think that increasing the surface area available to heat transfer would make the transfer more efficient, therefore requiring less heating of the water and, therefore, spending less to heat the water? Where, exactly, are you lost in the process?
    – FreeMan
    Commented Jan 10, 2023 at 12:59
  • 16
    It's a fair question. A lower surface area results in lower heat transfer on a local basis, but why would the system overall be less efficient? @FreeMan, your response doesn't explain why less heating would be necessary. How does air in a radiator affect heat cycles or total heating time?
    – isherwood
    Commented Jan 10, 2023 at 13:58
  • 2
    I used to read a lot of misinformation in automotive forums regarding radiator function in performance cars. Heat cycles and energy efficiency are often oversimplified or simply described incorrectly. It's not always as intuitive as it may seem.
    – isherwood
    Commented Jan 10, 2023 at 14:02
  • 3
    Just a note: you'll likely need to bleed it again. The fresh water that comes into the system will have air in it. It takes a few times after if it's been drained and refilled. Also, I've learned to protect any carpet or rugs near the valve so they don't get stained.
    – JimmyJames
    Commented Jan 10, 2023 at 21:41
  • 1
    Wait, are we talking forced-water or steam?
    – JimmyJames
    Commented Jan 10, 2023 at 21:43

10 Answers 10


For maximum efficiency, many modern heating systems (condensing gas boilers, heat pumps) are designed to output heat at a low water flow temperature. The unit can generate heat more efficiently at (for example) 40C than at 70C. The heat emitters (radiators etc) are sized to achieve a certain deltaT (difference from input to output water temp) at that flow temperature - ie the amount of heat supplied to the room.

Air in radiators reduces the usable surface area of the radiators. That reduces the heat flux provided by the radiator. It means the water is returned hotter back to the heater, reducing the deltaT. In extremis the heating system may be unable to keep up with heat losses from the room.

If the surface area of your emitter is reduced, the unit will have to work harder to maintain the same temperature in the room. If it can't successfully heat the room, it'll either work longer (increasing distribution losses as mentioned) and it'll extract less heat from combustion (a higher return temp means less heat is exchanged with flue gases, meaning more is wasted in the exhaust). Or, if the unit has some kind of dynamic adjustment, it'll decide to heat the water hotter, which reduces its efficiency.

  • 8
    One aspect that you don't mention here is that losses in the system (e.g. heating voids in the walls or an unfinished basement) are greater when the temperature is higher. And the pump(s) are not working as hard to move the same amount of water when the system is full. Consider a siphon: it won't start if there's too much air at the top.
    – JimmyJames
    Commented Jan 11, 2023 at 15:40

If you get less heat because of air in the radiators, the losses in the system can add up to a greater %.

Depending on where the pipes run, some heat might be lost in useless places. This is worse if it runs longer to provide the same heat. Also worse if you return warmer water through one of those useless places.

You also need energy to pump that water. If you need to pump more water, you need more energy.

Not really an answer but I suspect it gets repeated so much because most people don't understand how this works and directly equate "colder radiator" with "lost energy".

  • 1
    That's a good point about "Depending on where the pipes run, some heat might be lost in useless places". In one of my earlier houses - a 2 story Victorian (or Four Square?), all the pipes ran along the inside of a wall (not the interior of the wall). So any heat lost by those pipes just went into heating a different room.
    – SteveSh
    Commented Jan 10, 2023 at 18:55
  • 2
    "If you get less heat because of air in the radiators" - Isn't that the core of the question: unless the radiator is completely empty, there will still be heat transfer, but at what point does that become an issue?
    – MikeB
    Commented Jan 11, 2023 at 8:11
  • 3
    If the room is too cold, people turn the heating up. Commented Jan 11, 2023 at 9:41
  • 1
    @SteveSummit thing is, the room being heated up is not a perfectly sealed container - energy is lost. If the energy source in the room is less efficient at heat transfer (ie the lower amount of water in the radiator coupled with the insulating properties of the air replacing the water), less energy is imparted to the room at any given moment - the room heats up slower or not at all, resulting in people trying to add more energy by increasing the water temperature. Making the heat transfer more efficient by increasing the volume of water means you dont need to impart more energy to the water.
    – Moo
    Commented Jan 11, 2023 at 23:45
  • 2
    @SteveSummit every water-filled-radiator-connected heating system I have ever used has had a temperature setting on the boiler that is independent of the thermostat in the room being heated (standard UK setup tbh) - thats the one which gets changed in my experience... Has the side effect of making the next person washing their hands yell swear words...
    – Moo
    Commented Jan 12, 2023 at 0:36

A radiator works best if the water inside it slows down and lingers to emit the heat into the room. Flow speed depends on the cross section of the pipe and the volume per time being pushed through, a radiator is equivalent to a much wider pipe section.

A massive airbubble in the radiator decreases the effective cross section of the radiator. meaning that less heat transfer happens in that radiator and that you are returning hotter water back to the boiler which means more heat loss from the return pipe.

Another factor is that you lose quite a bit of surface area that could be used to transfer the heat into the room.

  • 3
    This is wrong. At least the first point is wrong. It's the same fallacy I've heard in the automotive community regarding engine radiators. Heat transfer is proportional to temperature differential, so slowing things down reduces heat transfer because the average differential will be smaller.
    – isherwood
    Commented Jan 11, 2023 at 20:12
  • 1
    and it's proportional to the amount of time the heat differential is present. Things will cancel out there. Therefore keeping the water in the radiator slow will help heat transfer. The other option is to zigzag the pipe around. Commented Jan 12, 2023 at 3:51
  • 2
    No. Over time heat is lost, reducing the differential. With higher flow the differential remains higher, on average. If what you're saying is true these systems would have trickle pumps. The only relevant factors are input temperature and surface area.
    – isherwood
    Commented Jan 12, 2023 at 14:37
  • @isherwood You seem to be confused. In automotive (as well as here) there are 2 delta Ts that must be managed to produce the best efficiency, engine to water and water to air. If we maximized water to air efficiency as you suggest, then when water returned to the engine it would be very close to engine temp, and engine to water efficiency would be minimized.
    – stonemetal
    Commented Jan 13, 2023 at 12:22
  • In this case, furnace-to-water is not constant. It's an on-demand situation. There's nothing to maximize from the radiator. Also, in the case of the automobile, the engine expels massive amounts of heat to the air directly. The concept is cut-and-dried. Slowing flow in either case is detrimental.
    – isherwood
    Commented Jan 13, 2023 at 13:50

Net energy consumption won't be affected very much if some radiators have some air in them. But if you look at the efficiency of the system's ability to heat the house, and to heat each room appropriately and quickly, THAT efficiency is reduced significantly if parts of the system are essentially disabled by being full of air.

In a ideal system each radiator is sized so that it provides the right amount of heat to the room it is in, each room has a suitably sized radiator (accounting for the particulars of the room eg north/south facing, windows, etc) and the boiler is sized to provide the correct amount of heat flow to all the radiators. If there are zones, the boiler is capable of working efficiently with any zone or multiple zones.

There are not a lot of ideal systems. There are lots of things that can throw the ideal out of whack. Someone can open a window. A room with large windows could have the curtains left open in the sun. The occupant of a room could turn the radiator off, or there could be thermostatic valves on radiators that turn them off automatically.

If one radiator is partly or totally full of air, that's no different to the radiator being partly or totally turned off at the valve by a user or by a thermostatic valve. It does not make the entire system more efficient or less efficient. Except at heating the particular room it's in ... it's less efficient at that, because its capacity is reduced by up to 100%.

If ALL the radiators have a little air in them, the entire system has less capacity to emit heat, so it will heat the house more slowly. That isn't necessarily more or less efficient from an energy-consumption perspective but it's certainly less efficient in heating the house quickly. On a very cold day it may not be able to keep up.

If ALL of the radiators have so much air in them that the house just can't be heated, then the boiler will cycle rapidly and may have trouble finding a stable temperature. The problems with that are bigger than efficiency.


You've hit on some major reasons why you should bleed them so I won't repeat them. Another reason is to balance the heat in all the rooms. If the thermostat is in a room with an un bled radiator, it will take longer for that room to heat up and kick the thermostat off and the boiler will run longer and the other rooms will be hotter than the operator wanted.

  • 2
    Neither the question nor this answer address efficiency, assuming all radiators have a similar water level. You've described one possible edge case only.
    – isherwood
    Commented Jan 10, 2023 at 13:59
  • As a corollary - if the radiator in the room with the thermostat is too big, the thermostat will turn OFF too soon for other rooms to warm up, so this may actually improve things!
    – MikeB
    Commented Jan 11, 2023 at 8:13
  • @isherwood the goal here is a warm house, not actually heat transfer, so a system that uses more energy to make a warm house can be said to have less efficiency. Commented Jan 13, 2023 at 17:57

It messes up with the water circulation in general in the system. People have mentioned efficiency, but that depends then on the piping layout. If the house and piping is isolated from the outside all of the heat still ends up in the house. But the thermostat on the radiator wont work right if there is air at the top either and that could result in a hotter house overall instead of the temperature you meant the house to be at, the thermostat area of the radiator remaining cooler than the rest - and that would cost more money.


A radiator produces heat. That's the least efficient possible use of energy. To make a radiator more thermodynamically efficient, we would have to turn it into an engine that turns some of the energy into work, so that less of it is is wasted as heat.

A radiator system full of air bubbles will not move heat where you want it, though, or not very well. If the system doesn't circulate properly, the wrong spaces will be heated. If you're heating unoccupied spaces, while people are cold in occupied species, that's a waste of energy. That's what is inefficient: low benefit for the same cost/energy.

Returning to my first point, though, it can't be the case that you're getting less total heat out of the system for a same energy input though, just because there is air in the lines.


Given all the conflicting answers here and the fact that some seem to be partially correct but not completely, I've thought about this, come up with my own answer and then had someone I know with a graduate degree in thermodynamics and an esteemed multi-decade career as an engineer in the HVAC industry confirm my understanding.

The fundamental thing to realize here is that the point of a home heating system is to move heat generated at the heating source to the living areas of the home. In the case of a forced-water hydronic system, the water in the system is the 'carrier' of the heat which moves through the pipes of the system and delivers it to the radiators. Efficiency in this context is basically defined as what percentage of the heat generated at the source to the radiators.

At a very simple level, the presence of air in the pipes means there is less water. That is, the capacity of the system's ability to hold, and there more deliver heat is reduced. In addition (as my expert explained) the "air will go into solution with the water which reduces the heat transfer properties."

The secondary issue is that if there's enough air, it can create gaps in the system which makes moving the water more difficult. That is, instead of cycling water in a gravity-neutral way, the pump needs to work to move water from (typically) the lowest point in the home to the highest point. If this isn't intuitive, note that a siphon can move water with only the force of gravity but only if the tube is sufficiently (i.e.: nearly completely) full. While this extra work is undesirable from a wear and tear perspective, it's the reduction in flow that is going to really have an impact of the efficiency.

In a nutshell the air in the system means heat being produced at the source is not distributed as quickly and/or as evenly as it could be. Here's an analogy, consider a bus-system that travels around a city and we can define efficiency as the maximum number of people that it can move each day per bus. If we were to remove half of the seats in each bus and change nothing else, the efficiency of the system would clearly be reduced. The hydronic system is like a bus-system for heat and air reduces the amount of heat that can be delivered over a given time span. This means more heat loss at the source (such as an unfinished basement because it remains there longer) as well as in other non-living area such as in voids that the pipes pass through.


Two reasons:

  1. Water has a 4x higher heat capacity than air. If your boiler puts out 1000 units (joules, whatever), and your system is 50% air, you end up with 625 units in your system to be circulated around your house.

  2. Water is more efficient at transferring heat than air, by like 24x. If you have 50% air in your system, the radiator will only transfer about 52% of the energy out.

Obvs there are transfer efficiencies that aren't accounted for here, just using some numbers as examples of why you don't want air in your radiator system.

  • 5
    Well, if the air is trapped in the radiator (like at the top), then it does not circulate.
    – SteveSh
    Commented Jan 10, 2023 at 17:53
  • 1
    @SteveSh makes an excellent point. Typically air in a system is localized at the top of any given radiator - not traveling through the system. Every bleed valve I've seen is at the top of a radiator for that reason. So if 25% of a radiator's volume is air you're losing about 25% of heating effectiveness in that room.
    – HoneyDo
    Commented Jan 10, 2023 at 20:52
  • @SteveSh Precisely - which is why you have to bleed every rad, not just one.
    – MikeB
    Commented Jan 11, 2023 at 8:16
  • 3
    If 625 units go "into the system" then where do the other 375 go???
    – MikeB
    Commented Jan 11, 2023 at 8:18
  • @SteveSh sounds like you just corroborated both of my points...
    – Huesmann
    Commented Jan 11, 2023 at 21:52

Bleeding may safe money, if you happen to live in an appartment that shares the central heating with other appartments, because the sensors that are used to distribute the heating costs assume equal efficiency. If your radiator is less efficient because of air inside it, you're billed more heat than you actually use.

Bleeding does not save energy, though - as long as you heat your room to the same temperature, it doesn't matter whether you use one radiator or two or half of one or none (use a server rack instead, perhaps :-), the energy required stays the same.

  • 2
    I'm not sure I follow why it wouldnt change the energy consumption. For that to be true, I feel like you would need to have a few assumptions that dont hold here. If the heating system converted 100% of the energy input to heat and the rooms were completely thermally isolated from their surroundings maybe that would apply. But energy is also required to move the water, and not all heat sent to the radiator will necessarily go into the room and stay there.
    – JMac
    Commented Jan 11, 2023 at 13:44
  • If we follow the logic of this to its logical conclusion, there would be no need for water in your hydronic system at all.
    – JimmyJames
    Commented Jan 11, 2023 at 15:30
  • @JMac you're right, of course, this is an abreviated depiction. My point is - as long as the source of energy is within the house and temperature in the room stays the same, it doesn't really matter how the energy is transported. What matters is whether the radiator works at 50% efficiency or 100% efficiency, because at 50% efficiency the energy is effectively twice as expensive (if you have cost distributing sensors).
    – Haukinger
    Commented Jan 12, 2023 at 8:03
  • @JimmyJames at least, if you reach the same room temperature without any water at all... I doubt the burner will work at all without enough water... as long as you heat your room to the same temperature is the important limit on the way to the logical conclusion
    – Haukinger
    Commented Jan 13, 2023 at 9:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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