Tuesday, 13 August 2013

Stand-by power and air-source heat pumps

I have been aware for some time that my 5kW air-source heat pump (ASHP) has been using about 3/4 of a kWh/day whilst doing 'nothing' over the summer (no circulation pumps running), but recently I have put energy monitors on other ASHPs and was shocked see that these known-brand units are using 2kWh/day when doing no heating at all - I bet the average family use less than that showering! (an 8kW shower for 15 mins = 2kWh).

Whilst in standby mode, is there any reason or excuse for it consuming so much? I think not.

There has, in recent years, been a general focus on standby power since this 'constantly dripping tap' can tot-up to a sizeable amount. Highlighting this problem has been very productive – the standby power of most new products has now plummeted.

Let us look at some figures to get a handle on the quantities. Here is what I recorded using pulses monitored from my kWh meter
ASHP on standby          31 watts (average),       0.75kWh/day,     272kWh/ year,      (£38 / year)
My A++ larder fridge   7.6 watts (average),     0.182kWh/day,        67kWh/year,      (£9.30/year)

Yes you have that right - my fridge, in use, used one quarter the power of the heat pump on standby!

Standby power is however quite hard to measure accurately using any instantaneous devices. This is in part because the current (amps) wave form from modern electronic devices bears no resemblance to a sine wave, and in part since the values are relatively small . For my own satisfaction, so I borrowed a calibration unit to verify the accuracy of the ‘real power’ OpenEnergyMonitor internet-connected devices that I am also using. http://openenergymonitor.org

Here is a plot of the 9kW ASHP that is consuming 2 kWh per day (Y axis is in kW).
It alternates between about 35-40 watts and 200 watts.
 Below is a plot that shows an 8 hr period..  Scale is 0 to 5kW input power.

The buff-coloured area on this 8hr plot relates to energy use, and shows a hot-water cylinder heating 'hump' (above 18:44) . This illustrates at-a-glance that the buff-colour area of the standby 'blips' (on the bottom axis) forms a significant chunk of the total.

I have previously decried ground source units with controllers that use 15 watts, so why are these particular air-source heat pumps using so much?

The oil inside the compressor can absorb some of the refrigerant. This not only thins the oil, but it can cause oil-foaming at certain times – both can lead to bearing failure due to inadequate lubrication. One answer is to keep the oil warm, and a small electric compressor heater (crankcase heater) can be used for this purpose.

Ground source units are situated indoors, so usually stay relatively warm. I don't know of any GSHP with a compressor heater. However, air-source are sited outside, and they can experience extreme 'swings' of temperature. The worse scenario would be a cold sub-zero night followed by a warm day. In this scenario, the lightweight heat-exchangers warm up, but the heavy compressor can remain the coldest part of the system. Some refrigerant can then migrate to the coldest part (in effect, it condenses in the compressor like water condenses on cold windows). If this cold refrigerant-laden compressor were to start, then there is risk of damage.

The picture shows the compressor with strap-around element heaters. The compressor was quite warm, but one wonders why the lagging around the shell was open at the top; possibly to aid compressor cooling for times when the compressor is working at it's upper temperature limits?

What is the solution?

There is a lot of scope here.

Most of these Japanese-type heat pumps are very impressive, they are clearly very well developed and very reliable. Computer-aided fan and heat-exchanger design has led to highly efficient quiet products, and most defrost mechanisms are very well optimised, but are all aspects up to scratch?

It seems already (from graph) that the heater is pulsed, and not simply left on, so why cannot a little more intelligence be added to the controller so that the switching can be controlled more responsibly.

Lagging the compressor should also be possible, and should help, but now one might need to introduce some form of cooling to deal with overheating whist running. There are many options here, and it would take time for a designer to work out the best reliable and cost-effective option that did not compromise over all energy use. I have certainly spent many hours in the past designing crankcase heaters out of systems. The phrase 'could do better' springs to mind, and this sentiment should really be directed to the manufacturer's design office.

A similar situation existed with washing machines many years ago. All for the sake of a thermostat, the hot water feed was blended with cold, just in case the hot feed happened to be at 70C , and could therefore potentially damage clothing. No manufacturer wanted to be blamed for harming any clothes so they all played-safe and kept it simple, to the detriment of energy-efficiency.

My feeling is that we are at the same point with some aspects of ASHPs. Whilst standby power is a small fraction of the total annual consumption, it is still a detail worthy of attention since I’m sure it could easily be improved significantly.

I can think of a simple solution. Make it mandatory to print the standby power on the product label. This would push the manufacturers to give it a little more attention, and sooner or later this unnecessary waste would disappear.


  1. Hi John,

    Our entire gross electricity consumption is ~4.5kWh/day for the family, so 2kWh/d 'standby' plus maybe another 1kWh/d for DHW tank losses would be a bit ugly.



  2. My Daikin LT 11Kw heat pump has a standby of .4/.5 Kwh, this is an yearly cost of €700 just wasted.
    This not a green pump

  3. Newer altherma 8kW has 15W standby usage.
    This is measured during summer mode and mid season when it only runs during night.
    Temps during daytime standby are around 3-5°c while measuring.

    The higher output models are less efficiƫnt in standby mode.
    They do not have the same standby mode.
    When there is no heat demand for a while, part of the exterior electronic circuit is cut out to save energy.

  4. I wish I had read this prior to purchasing my ASHP. My September use increased from 73 KWH to 163 KWH for the month.The fitters said it was my fridge. A meter has been fitted to the indoor and outdoor units which confirms my figures. It is turned off because the increased pressure caused my radiators to leak All replaced then the valves went. Still turned off as the feed pipes are leaking. Waste of space

  5. Thanks for sharing this - as usual the headline figures aren't always what they seem. I was looking for data on how an inverter compressor ASHP will absorb power whilst on and if it can be varied to the lowest quoted value or if it jumps straight up to max power to start (I have PV so would aim to only use when there is spare capacity). Anyway, your observations are interesting as this puts a fly in the ointment as it's effectively a phantom load doing nothing useful! Thank you :)

    1. Unfortunately, almost all HPs will tend go full power until the setpoint is reached. It is a great shame that the speed isnt controllable. Mitsubishi have an 'eco' setting for DHW mode. this simply allows it to ramp up slowly, but it doesnt stay on minimum. I only know a few heat pump controllers. The Ecoforest GSHP allows you to limit the input power, but this is inside the 'service setup' so you cannot vary it in response to available PV. No doubt manufacturers will come up with PV compatible systems at some point in the future. When they do, I would like to know.

    2. Thanks, appreciate the reply! I was intending to go down the split AC route which would cover cooling and heating (when needed) - the cooling side shouldn't be a problem with solar as it tends to be sunny when it's hot enough to need AC! I'm wondering whether to bite the bullet with a cheap split AC and take some measurements (I already use emonCMS for my power monitoring) and see if the specs for min>max apply depending on what the situation is!

    3. i wonder if you could get into some clever electronics. All or most AC have IR controller, so you can set the set-point using IR (you can read what is being sent if you are clever enough) you might be able to keep the unit on a minimum input speed by automatically changing the setpoint (so it tracks minimum speed). It might be problemtic, but a nice thing for someone with the time and the aptitude to try it.