I installed an air-to-air heat pump for a friend and lived in their house for 10 days during winter. I wanted to understand its cold weather performance, limitations, and optimal use cases. Here are my findings.
First Impressions
Sound
The indoor unit varies in noise level with demand. At low load, which was most of the time, it was very quiet though not completely silent. Under higher load, like warming up a cold room, the noise was similar to a dishwasher – much quieter than a washing machine. The sound wasn’t intrusive and was easy to live with.
One issue with the indoor unit is that the fan only has five discrete speeds. This makes software control easier, but it impacts the user experience. In auto mode, it would be better if the fan speed changed in smaller increments rather than noticeable jumps between speeds.
Air Movement
Because the unit moves air around the room, I noticed drafts until the space reached the target temperature. The feeling of a draft seemed more related to temperature differences than just moving air – once the room and surfaces reached an even temperature, the air circulation wasn’t noticeable.
Operating Mode
The unit works best in automatic fan speed and direction mode. In this mode, it adjusts the fan speed to match heating demand and angles the vanes towards the floor. This mixes cold air near the floor effectively and prevents temperature layering in the room.
Finding this auto-mode wasn’t straightforward – I had to hunt through the manual to understand what it did and how to use it properly.
Efficiency Testing
My testing conditions weren’t ideal for precise measurements. The house had just been renovated – a new staircase was installed in the main living area where the unit went in, and a partition wall was removed to combine a small bedroom with the main living space. This meant heat could now travel upstairs, and the room was significantly larger than before.
To run some rough tests, I blocked off the staircase with duvets and mattresses. At -1 degree outside, the unit used about 500-700W to heat this large room, even with some heat escaping up the imperfectly sealed staircase to an uninsulated attic. Previously, it took 1.5 to 2kW to heat the smaller, pre-renovation space to a lower temperature. While this suggests improved efficiency, I’m hesitant to quote exact figures.
The unit could modulate down to about 130W when heating demand was low. I’ve seen some units that can go even lower, down to around 80W.
How Many Cassettes Are Needed To Heat A House?
When I installed the single indoor cassette, my goal was to heat the main living area, but I hoped it could heat the whole downstairs. What I found was that heat doesn’t travel particularly well down corridors. I’ve attempted to show this with the colour contours below.
I considered installing a small extractor fan/duct above the door to Bedroom 1 to help air circulate, but we decided against this. While it’s possible to use ducted warm air to reduce the number of cassettes required, this house really needs at least two units, which is what I had planned for. The number of cassettes required for a house is highly dependent on how it is used – in this case, only Bedroom 1 and the main living area are in constant use.
Refrigerant Line Considerations
A critical but often overlooked aspect of air-to-air heat pump efficiency is the refrigerant line installation. In my setup, there was a 5m distance between the outdoor and indoor units. I initially used standard industry insulation (about 22mm thick) on the refrigerant pipes. After some rough calculations, I discovered I was likely losing several hundred watts of heat just through the outdoor pipes. Adding extra fibreglass insulation noticeably reduced the unit’s power demand.
This highlights a significant issue: standard insulation thickness may not be sufficient, especially for longer pipe runs. The problem becomes more severe in cold weather when the temperature difference between the hot refrigerant and outside air is greater, and the unit needs to work harder to maintain indoor temperature.
The issue is particularly important for heat pumps compared to air conditioning because heating mode typically runs with higher refrigerant temperatures than cooling, creating larger temperature differentials with the outside air. The system also tends to run for longer periods in heating mode.
Understanding Heat Transfer Methods and Their Design Implications
Fundamentals of Heat Transfer
Air-to-air and water-based heat pumps transfer heat in fundamentally different ways, which creates distinct design challenges for each system.
In a water-based system, warm water flows through radiators and gradually loses temperature as it transfers heat to the room. While there’s some infrared radiation involved, most heat transfer happens through air naturally circulating over the radiator’s surface. This simple convection process is well understood and easy to calculate.
Air-to-air systems work differently. Instead of gradually losing temperature, they rely on phase change – the refrigerant actually switches from gas to liquid inside your house. When the refrigerant reaches the indoor unit, it’s a hot pressurised gas. As this gas condenses into liquid, it releases a large amount of energy into the room. This phase change releases far more energy than you’d get from just cooling down a fluid.
This same phase change process happens in air-to-water heat pumps too, but it’s contained within the outdoor unit where it heats your system water. Adding this extra step might seem inefficient, but it brings some important advantages.
Heat Exchanger Design Challenges
The heat transfer mechanics in air-to-air systems create some interesting design challenges. While their compact heat exchangers can achieve impressive heat transfer rates through forced convection and dense fin arrangements, this comes with trade-offs.
The performance of these heat exchangers is much more dynamic than radiators. Their efficiency depends on maintaining optimal conditions across several key variables:
- Air flow rates through the fin stack
- Refrigerant saturation temperature and superheat
- Air inlet temperature and humidity
This creates a more complex optimisation problem than radiator-based systems, where the main variable is just water temperature.
In water systems, installers can use straightforward calculations to size radiators for efficient operation at lower water temperatures. These calculations don’t exist for air-to-air systems, which leads to some hidden complexities in sizing indoor units.
When an air-to-air system needs to output high heat (like on cold days), it has to run hotter refrigerant through the indoor unit at higher flow rates, with faster fan speeds. Running hotter refrigerant means lower efficiency because the compressor has to work harder to generate those higher temperatures.
This creates an efficiency paradox: units with smaller indoor heat exchangers relative to their outdoor unit power will be less efficient. To maintain high efficiency, you need more indoor units so the refrigerant can run at more moderate temperatures – but this drives up installation costs.
Real World Example
The Mitsubishi Electric MSZ-LNxxVG2 series illustrates these challenges well. All their indoor units share the same dimensions and weights, suggesting identical heat exchanger sizes. To get higher power outputs from these same-sized units, they must either:
- Run the refrigerant hotter
- Increase refrigerant flow rates
- Run the fan faster (increasing noise)
Some of their outdoor units are actually just limited versions of larger ones – their 2.5kW unit is essentially a capped 3.5kW unit. Their efficiency figures show larger units having lower COPs, which makes sense given these compromises.
https://library.mitsubishielectric.co.uk/pdf/download_full/4164
Professional air-con installers likely have some knowledge that I am not privy too, but I suspect they are just picking units that are big enough for the room and use experience to choose the units rather than mathematical rigour. The experience in this case is obviously very useful. However, it would be useful to know more, and to be able to size units based on more numerical information rather than kW capacity & aesthetics.
Defrost Behaviour and System Sizing
Through testing, I found that defrost cycles weren’t just dependent on outside temperature – how hard the unit was running made a big difference. At 6 degrees or below, running the unit in ‘powerful’ mode would cause it to frost up quite quickly. However, at normal speeds and modes, it took much longer to build up frost, even at the same outside temperature.
I was purposely messing around with the unit to see what would happen under different conditions, the advantage of doing this with an air-to-air system vs air to water was that I could instantly see if the unit would freeze up more, or if it would run better under certain conditions. My playing made me realise something important about system sizing. When a unit is pushed hard relative to its heat exchanger capacity, it needs to run at more extreme refrigerant conditions, which leads to:
- More frequent defrost cycles
- Lower efficiency
- Potentially more noise
- Less comfortable air delivery temperatures
Final Thoughts
After this testing, my view on air-to-air units has evolved. I believe they can be very efficient when properly oversized – but determining the right amount of oversizing is currently impossible. Manufacturers don’t provide crucial information about:
- Heat exchanger surface areas
- Refrigerant temperatures at different outputs
- Performance curves at partial loads
- Detailed defrost behaviour
This lack of technical data makes it hard to optimize these systems the way we can with traditional water-based heating. You’re essentially designing blind, relying on rules of thumb rather than engineering calculations.
There is the question of whether the systems need to be ‘great’ or just good enough. I believe we should strive for greatness, as by having more efficiency heating systems it will mean that the electricity grid will need less storage and generating capacity. I do think these systems have the potential to be great, but not with the information that is currently out there..
I think they are a great option if installing in a house that requires cooling where there is no wet heating system currently. However, I’d be tempted to have an air-to-water system over an air-to-air system if there is already a wet heating system installed in the house.
Additionally, I think these systems provide an excellent way to somewhat cheaply supplement an undersized air to water system or improve efficiency on cold days. The advantage of air-to-water systems comes from having a relatively larger heat exchanger area and emitter area through multiple radiators. This means the system can run less hard and improve its efficiency and heat loss characteristics. This becomes especially valuable if space for radiators is limited and the house needs cooling in the summer.