Air Source Heat Pumps vs. gas boilers

First the bad news for heat pumps. The Energy Saving Trust heat pump survey in 2009 found that many users were not impressed at all. The follow up in 2013 improved the results but the final average system COPs of 2.45 (air source) and 2.82 (ground source) were still way below the headline figures quoted for these machines a lot of which are claiming over 4 these days. So maybe heat pumps are intrinsically good but tricky to install?

Despite all that, what is really good about heat pumps is that they can deliver more energy than they consume in electricity.

The power multiplier
The power multiplier

So a small one would be just like this diagram; working on the power of an electric kettle but delivering the power of 3 to your hot tank – a COP (Coefficient Of Performance) of 3 then. By contrast your immersion heater delivers and also consumes the power of an electric kettle so it has a COP of 1.

Heat pumps are all sold with an industry standardised COP. This is misleading to say the least and the reason why optimism is defeated by experience. Far from being a fixed figure the COP actually swings widely depending on outside air temperature and temperature delivered in the home. The COP plots here show how a kick is engineered to give a good headline figure; that kink in the graph is exactly at the publication point.

A sneaky kink
A sneaky kink

You might buy a machine with a quoted COP of say 3.75 but while making domestic hot water on a cold night it will be working at less than 2. There are benign swings however and given a sunny winter day with some warm air to chew on an ASHP can see COPs almost up to 5. You can see this on the blue line on the graph. Delivering 35c water to the underfloor heating the COP goes over 5 as the outside air goes over 10c. Note that the pale blue line (delivery temperature 50c for radiators) still only goes to around 3 so for most of the time the average COP will only be near 2.5. So, heat pump with radiators – think carefully.

Gas per costs about a third of electrical power so after adjusting for efficiency a gas boiler is similar to a heat pump with a COP of 3. Many people in the survey would be comparing their new heat pump to a gas boiler; a formidable opponent when running on cheap gas. A gas boiler is much more powerful than most heat pumps and delivers at usefully high temperatures so a heat pump must have an overall COP of over 3 to justify a hefty purchase price.
Perhaps the performance could be lifted further?

The next bit is a bit dull – you might want to skip on to conclusions below.

To winkle out some ideas we’ll take daily temperature data for January in Guildford ( and relate that to a COP matrix made from the published data from a modern ASHP (inverter drive scroll, r410a, delivering to under floor heating at 35 degrees).
We will be looking to lift the COP by running the ASHP at the warmest ambient temperatures possible.
A look at a January temperature trace shows:
There is usually a 5 degree swing between the mean night time temperatures and the daytime mean.
Night time temperatures are flatter and longer than the sharper daytime peak at 1-2pm.
The morning transition from lows to highs is halfway there by 10am.
Temperature rises coincide with sunrise, not surprisingly.

Relating the above to the COP matrix:
Running a 7hr shift from 10am gives an average COP of 3.86 – much better than gas.
The equivalent night time shift only gives a COP of 2.92 – but almost as good as gas.
If the pump has to make hotter water for radiators these day/night figures drop to 2.7 and 2.11 and for 55 degree hot water making 2.3 and 1.85– gas beats this hands down.
Storing daytime running means that delivery temperatures probably need to be around 50 degrees leading to an average COP of under 3 although bigger storage tanks improve this.
ASHPs can be smaller if they run continuously day and night on an average COP of 3.4 – still 13% better than gas.
Direct electrical heating is often used to boost hot water making (COP = 1) and this can lower the average COP. If we can avoid this practice and run predominantly in the daytime it should theoretically be possible to get a COP of 3.35 (7hrs day, 2hrs night, 2hrs hot water).

Transmission: Put 100W/square metre through your floors and your feet will be uncomfortably hot so somewhere near half that will be a good yardstick for calculating the power you need.

A small ASHP can run a bit more efficiently than a gas boiler in a modest well insulated house. Fan-coil units in bedrooms and underfloor heating elsewhere are essential. The heat pump should run in daylight except maybe for a boost before dawn to guarantee morning showers and take the chill off the floors.

Of course, with PV panel prices falling relative to electricity prices, the time is coming when your heat pump will run for free while the sun shines. At the moment it looks like we are firmly in no brainer territory and it is certainly worth checking now to see how the sums stack up.

A tip to make your heat pump installation cheaper. Use the PV panels and an energy diverter (e.g. Eddi) to heat your existing immersion heater. then you won’t need the special tank to accompany the heat pump which will just do central heating. Your installation might get a lot nearer to the £5,000 grant.

And while we are at it, don’t forget to consider a mini-split heat pump air conditioner. It’s an air to air unit and relatively so cheap it’s hard not to go and get one right away. I did.

Like what you’ve read? Find more essential reading on my e-book

‘Dream House – Down To The Details’ here:-

More on this topic in LIST OF POSTS


Ukraine battle Drone?

Quadcopter/puck ready for launching

Quadcopter launcher

Quadcopter launcher

This idea was first published here in 2015. A pity there has been no interest. Remember that long Russian convoy menacing Kviv? What if a few hundred drones had blown the tyres off those vehicles and any supply vehicles?
The Original Twist concept looks like a fat Frisbee or puck, at least it does for the first few hundred metres of flight with not an arm or propeller in sight.
Transportation: The propellers and arms are all tucked safely out of harm’s way, folded into the base of the puck. The folded puck can then be handled roughly and easily stacked up in racks.
Launching: Here lies the real Original Twist. As you will see in the launcher description below, the robust pucks can be flicked out like clay pigeons at the astonishing rate of 600 a minute. There is no battery power used for taking off, getting under way and reaching height which adds to their range. Neither is there a warning howl as hundreds of drones start up, just a line of discs silently darting across the sky to a location away from the launch site and therefore no giveaway of the origin. At the end of the launch trajectory the propeller arms flip out and the journey towards the enemy continues using onboard GPS. The propellers sit at a slight angle to the body so that when in flight the puck is perfectly aligned to the airflow. Reduced drag with some lift from the domed top allows a good range of at least 12 miles which allows one launch vehicle to cover an area of over 400 square miles.
Attack: The quadcopter/puck bodywork is made of moulded plastic explosive so they are very much like intelligent flying bombs. They can, for example, fly to a given location and using infra-red cameras locate human sized heat signatures for immediate targeting; no sniper will be safe from being blown out of his hiding place. Pucks can communicate with each other and with the host computer using a meshed network where signals are passed down the line. Other pucks can be programmed to cluster into a much larger bomb before simultaneously exploding. Pre programmed pucks will be immune to radio jamming and execute their missions come what may.
Tactical use: The near silent deployment can have a myriad of other uses. Surveillance from on high or with sound and vision lying on the ground, simulated radio transmission sources to attract enemy fire, diversionary attacks from various directions and overhead flares to light up the enemy. They can even provide ‘Shoot me’ targeting information to overhead drones and airborne weaponry.
What could be more demoralising to the enemy than intermittent attacks from the air at no significant cost or risk to the other side? The biggest fear will be a ‘Cluster puck’ attack where several successive droves are launched and are resting on the ground nearby in preparation for a massive orchestrated attack.
A set of pucks with an affinity for vehicle wheels could be launched and without any difficulty a whole convoy of vehicles could be stopped in its tracks, or what’s left of them!
A major comfort to troops in hostile territory will be escorted manoeuvres. Protection from ambush is provided by pucks flying apparently random recce patrols but in fact checking ahead for any heat signatures.
Drone warfare will cause a change to military clothing; a wetted and cool outer shell being necessary to avoid being ‘seen’ by a drone’s camera. Even so, once terrain has been optically scanned any changes in position will highlight potential targets. If a drone flies over your hiding place it is likely that you will be attacked by the next one if you move. Computers and drones together make awesome weapons and there is no doubt that ground based warfare is about to enter a new era.

The puck launcher

Imagine a Landrover carrying a few thousand pucks in racks and on the roof a 6 foot diameter Catherine wheel spinning at 600rpm (car engine tickover) and flicking out up to 5 pucks a second at a launch speed of 120mph. The pucks are introduced into the calm middle of the wheel from an overhead magazine and then nudged sideways into the 3 radial arms where they queue to be released onto the extended launch ramps where they accelerate out to the open edge and away. They gain spin from the friction side of the launch ramps which initially swing out under centrifugal force to make the diameter bigger. The curved channels enable more pucks to be in the queue and also ease the centrifugal force at the end of the curve where the release catch is situated.
The loading magazine is itself fed by conveyor that has passed through the arming station. Here a fresh battery, much like a small puck itself, is fitted into the middle of the puck while the mission computer installs targeting instructions via blue tooth.
No other system will deploy quadcopters this fast and it may even be necessary to slow it down.
Pucks not on suicide missions will return to land on a wide conveyor belt on top of the launch vehicle and from there mechanically re-folded, de-batteried and added to the stack heading for the launcher. In this way it will be possible to have hundreds of drones permanently out on various tasks. Separating the charging function makes long term storage very much easier and every puck receives a freshly charged battery just before taking off. Once deployed returning pucks can be re-batteried and launched back into the fray on a continuous basis.
With theoretical launch rates of a mind boggling 600 a minute there would never be time for a person to decide on the mix of missions being fed into the pucks just before they are launched. To do this efficiently the mission computer receives more generalised commands from several operations directors and these are collated and then automatically programmed onto the pucks.

Of course the puck concept is perfect for aerial launching too – no need for the launcher, just eject boxes of them to glide under low power for great distances. You would never see or hear the plane that launched a mass of pucks at you and it would be 100 miles away when the pucks arrive.

So to anyone who is not smelling the coffee yet here’s the question: Would you commit troops to a ground action where they can’t move, can’t hide, where attacks come relentlessly from every direction, where vehicles aren’t viable and all these things are cheaply applied by an enemy who, barring one launch vehicle, is immune to retaliation?