Air conditioner vs hydronic heating

The living area in Cameron’s new rammed earth extension is passively heated to a large extent, but extra heating comes either from hydronic heating in the slab or the air conditioner above the dining table. They’ve been able to compare the two for effectiveness and efficiency over the last three winters.
Heat pumps can provide air conditioning or hydronic heating. But which is more efficient, and which more comfortable? Cameron Munro tells us what he learned by trying both in his super-insulated home.

In November 2016 we moved back in to our weatherboard home in Melbourne after having renovated it using Passive House principles. This involves very high levels of insulation, careful attention to airtightness and the use of mechanical ventilation and heat recovery (more background on our journey is provided in Renew 141 and 144). In this article I reflect on our decision to install a heat pump hydronic heating system and whether, in retrospect, it was needed—or whether our reverse-cycle air conditioner could have been our main source of heating, instead.

We chose to install a small Tivok air-source heat pump that heats a 420L tank that is used for both hot water and our small hydronic heating system. The hydronic heating system is connected to two 700W heated towel rails (one in each bathroom) and a heated concrete slab of roughly 20m2. We also have a 6kW Daikin Cora reverse-cycle air conditioner (another type of air-source heat pump) in the living room.

Our original thinking was that we’d rely on the hydronic heating in winter and use the air conditioner for summer cooling. However, since we moved into our home almost three years ago, we’ve been playing with these two units to find an efficient and comfortable way to manage heating. In this article I review what we’ve learnt so far and what we might do differently if we were to do it all again.

Fabric first and then heat pumps

Our first objective was to minimise our need for heating and cooling by doing everything possible to maximise the performance of the building fabric. This ‘fabric first’ approach is central to the Passive House methodology.

Of course, ‘passive’ doesn’t necessarily mean the complete elimination of active heating and cooling, particularly in our case: our home has a narrow frontage to the north and heritage protection which meant we had limited opportunity to improve our passive solar orientation, so some heating and cooling would be needed.

Even so, as shown in Figure 1, we have reduced our space heating and cooling loads by around 90% compared to ‘typical’ Victorian homes. Our largest energy usage comes from appliances (29%), followed by the electric car (26%), heating and cooling (24%) and domestic hot water (16%).

Figure 1. Annual energy consumption for Cameron’s home versus a standard home. For typical Victorian homes space conditioning makes up around half of energy use. For Cameron’s, it’s about 24%.

Evolving thermal comfort

Over the first winter we relied almost solely on the hydronic system to provide our active heating. However, the thermal lag associated with the slab heating in the living room didn’t correspond with our desire to quickly inject a bit of heat into the home on cool mornings. And, for most of the day, even in winter, we simply don’t need the slab heating—if we used it, it’d be too hot during the middle of the day and we’d be wasting energy.

It is important to consider this in the context of our home, which being so well insulated and airtight, doesn’t experience rapid fluctuations in temperature. Nor does the temperature fluctuate far—on the coldest mornings it will still be 17–18°C indoors without any overnight heating. By comparison, a poorly insulated home would be more likely to need to leave the slab heating on constantly.

Note that we aren’t heating a thick structural slab—it’s a 100mm concrete screed sitting on top of 80mm of foam insulation which in turn then sits on the structural slab. By sitting the exposed slab on the foam we insulated the heating from the ground and reduced the amount of concrete that would need to be heated.

However, even with this substantially reduced amount of concrete, the time lags still don’t match well with our requirements.
I crudely estimate the time required to heat the slab from cold (say 18
°C) to a typical slab heating temperature of 27°C would be just under three hours and require around 10.5kWh of energy.

So, instead of using the slab in the living area, from the second winter on we’ve only used the reverse-cycle air conditioner. This provides the quick injection of heat that we want. We’ve combined this with running the towel rails in the two bathrooms on the hydronic circuit (the slab and towel rails are on separate circuits controlled with simple manual valves).

Doing this appears to be about the optimum for both efficiency and comfort for our situation; the reverse-cycle air conditioner provides the rapid injection of heating into the kitchen and living area during the morning and the heated towel rails provide the extra couple of degrees of heating in the bathrooms to make them more comfortable in the middle of winter, as well as the added luxury of warm, dry towels.

Comparing heating efficiency

Efficiency can be measured objectively, so I can reach a definitive conclusion: to get a unit of heating into the home, the reverse-cycle air conditioner is about 30% more efficient than the hydronic system.

This isn’t a surprising conclusion: the air conditioner operates at COPs above four, while the Tivok hydronic heat pump operates at COPs of around three at typical heating temperatures of around 10°C outside and return water temperatures of 50°C. COP stands for ‘coefficient of performance’ and is a measure of efficiency: a COP of four means that for every kilowatt-hour of electricity in, you get 4kWh of heat out. (By comparison, the effective COP of resistance electric heating will always be one and for gas hydronic or gas ducted heating systems will be somewhere around 0.7.)

The hydronic system also has heat losses in the heat exchanger at the water tank and in the pipework that runs from the tank to the towel rails and slab. When the hydronic system was first installed it was very poorly insulated—the PEX plastic pipework was wrapped in a standard 3mm foam insulation resulting in a loss of around 16W/m at 53°C water temperature. The insulation was also cut back at junctions and floor penetrations, leading to what I estimated to be an energy loss of 800W just in this exposed pipework.

Because we’ve only got a maximum of about 3.4kW of radiators, I estimated the pipework accounted for 1300W of heat loss under the floor, or around a third of all the heat energy put into the water at the heat pump. To rectify this, I bought two additional 13mm thick layers of flexible pipe insulation, which I slit lengthwise, wrapped around the installed pipework under the floor and glued at the slit, staggering the slits to ensure the continuity of the insulation. This was a horrendously unpleasant and laborious activity but should have reduced the pipe heat loss to around 6W/m—and the total line loss to around 300W, giving an efficiency of closer to 90%. A thermal image of the result is shown in Figure 2.

The lesson here is clear—if using any form of ducted or piped heating ensure the ducts or pipes are well insulated and that the installer reinstates any insulation removed as part of the installation. However, even once the piping losses are minimised, the air conditioner is still about 30% more efficient at putting heat into the home than the hydronic system.

Figure 2. Surface temperatures of standard hydronic piping insulation (left) and with one additional 13 mm layer of Armaflex (right), showing a 10 °C decrease in the exposure surface temperature (albeit for a larger surface area).

Heat distribution

Another consideration is the distribution of heat throughout the home. Our house is fairly modest in floor area (143m2) but long and narrow. The air conditioner is at the rear of the home, so distributing this heat to the front relies on our centralised mechanical ventilation and heat recovery system.

Flow rates through the mechanical ventilation system are low, amounting to less than 30% of the air volume in the home per hour. As such, it can take an hour or two to perceive any change in temperature at the front of the home attributable to the air conditioning.

In practice we don’t find this to be a major issue, as the temperature variation is never more than a couple of degrees and on sunny winter mornings the front (north) aspect will rapidly warm from the sun. In retrospect though we would at least have wired in for a second small reverse-cycle air conditioner in the front to provide more immediate heating across the home.

Reverse-cycle air conditioner comfort

A common concern about using reverse-cycle air conditioners for heating is that they don’t produce a heat which feels as pleasant as the silent radiant heat from hydronic systems. My previous experience in poorly insulated homes supported this—it felt like ducted gas or air conditioners used for heating were constantly having to blast hot air into the rooms to keep those poorly insulated buildings warm, and the blast from the air movement created a sense of chill, even though the air itself was warm.

But the way in which our air conditioner works is very different to these experiences. It is massively oversized for our heating requirements because of the excellent thermal and airtight building envelope and, as a result, as illustrated in Figure 3, the unit runs at full capacity only when first turned on and quickly reduces to around 400–500W (or about 40% of maximum power).

We operate the unit on its lowest fan setting, and it sits directly above the dining table and doesn’t feel uncomfortable. We attribute this to the fact it’s not having to work hard to keep the home warm, that all the internal surfaces of the home are very close to the indoor air temperature (so the radiant chill one sometimes feels in poorly insulated homes isn’t present), and to the airtightness eliminating cold draughts.

Figure 3. Typical air conditioner power consumption curve in Cameron’s home.

Cost comparison

The air conditioner cost well under $3000 installed, whereas the combined hot water and hydronics system cost around $18,000. The cheaper alternative would have been to rely solely upon the air conditioner for heating and use a heat pump for hot water, (costing around $3000), so the equivalent comparison is $6000 vs $18,000. There can be no question that what we have is a very expensive heating system, and one that is less efficient than it could have been.

Balancing solar with hydronic loads

When I was looking for a hot water and hydronic heat pump supplier, I was attracted by the option of coupling these loads into the one unit for obvious reasons. However, doing so has repercussions for system efficiency and integration with solar PV.

During the shoulder and summer months when the hydronics system is off, I set the hot water heat pump to switch on at 11am. The 1.8kW heat pump heats the 420L tank in just over an hour which then is more than enough for our hot water needs through to the next morning. In this way we are effectively using the hot water tank as a solar battery. Indeed, it’s a very high capacity battery at low cost—420L of water heated from 15°C to 56°C is equivalent to 20kWh of energy.

In winter, though, using the hydronic system for heating means we need to run the hot water heat pump from 6am to 9am, and then again from 5pm to 7pm. If we instead kept to the summer running period, there would be inadequate heat remaining in the tank in the mornings to meet both our hydronic and hot water needs. The hydronic circuit quickly depletes the tank of heat, so we’ve found it’s not practical to run the hydronics without having the heat pump on. This, in combination with the way in which we use our electric car (it being at work during the day, returning around 5.30pm), means our consumption and generation patterns on winter weekdays are very poorly aligned (Figure 4).

While a modest battery would partially rectify this imbalance, in my view it is more a reflection of the compromises that come from coupling the hot water and hydronics loads. If they were separate systems we could continue to run the hot water heat pump during the middle of the day in winter. A separate hydronic system could then run in the early morning for a couple of hours without the heat pump also running (and so with the water temperature decaying) and then heat during the middle of the day. It might still be necessary to run the hydronic heat pump in the evening, however, and of course there are the complications and capital costs of operating two distinct heat pumps.

Figure 4. Hourly electricity consumption by circuit and solar generation for a typical winter weekday—note how the consumption generally occurs outside the solar generation.

While we are net energy producers across the year thanks to our 6kW solar system this does not mean we can operate independently of the grid. Our peak load is mid-winter, when our daily consumption can be up to 25kWh and daily solar generation varies from around 5kWh to 15kWh.

Our daily consumption by circuit varies across the year as shown in Figure 5; as a rule of thumb the car consumes 5kWh to 8kWh, and the appliances in total around 1kWh consistently across the year. Lighting demand is somewhat higher in winter, but it’s still fairly minor—up to 0.5kWh in winter and less than 0.3kWh in summer. The main seasonal variations are from the hot water and hydronic heat pump, which consumes 1kWh to 2.5kWh in summer (when it’s only providing hot water), increasing to an average of 4kWh in winter, and sometimes up to 9kWh. Similarly, the air conditioner is rarely used in summer and during winter uses on average 4kWh and up to 6kWh.

Figure 5. Daily consumption by month of year and circuit, plus daily generation for solar. Shaded areas represent the area within which 90% of values fall.


Since we’ve moved in, I’ve spent much time reflecting upon the relative merits of the reverse-cycle air conditioner and hydronic system for our heating needs. What I can say with confidence is that the hydronic system is complex, expensive and inefficient compared to the reverse-cycle air conditioner. But it also provides some gentle heating of our bathrooms and nice dry towels.

So would I do it this way again? Probably not given the cost and complexity. To achieve equivalent levels of comfort to the hydronics we’d probably look at some combination of a small reverse-cycle air conditioner in the front of the home and electric resistive towel rails or a far infrared heater in the bathrooms. However, neither of the latter two appeal to me completely as they don’t take advantage of the leverage that can be gained from using heat pumps.

What I can also say with absolute confidence is that if we were to build again we would use the Passive House approach. Ensuring a super-insulated and airtight building fabric represents by far the most environmentally friendly, healthy and comfortable investment we can make in our homes.

Cameron Munro
Cameron is an aerospace engineer with an interest in energy efficiency and building thermal performance. You can find a recent video about his house’s thermal performance here.

support our work

Renew is a not-for-profit organisation dedicated to arming you with the unbiased advice and information you need to make your home and community more sustainable. If you enjoy our content, please consider supporting us by becoming a Renew member or buying our publications. Your support is critical to helping us achieve our goals.
support renew
Further reading
Tradies and the transition

Tradies and the transition

Do we need as many tradies for electrification as many think? Not if we are innovative, writes Alan Pears.

Read more
Double glazing on the (very) cheap

Double glazing on the (very) cheap

Do we need as many tradies for electrification as many think? Not if we are innovative, writes Alan Pears.

Read more
Ditching the shower fan

Ditching the shower fan

By fitting a lid on the shower, exhaust fans are not needed when showering. John Rogers describes this simple retrofit, using both a commercial product and a great looking DIY version.

Read more