I had a heat pump installed in my house in July 2022, and a year later a Vaillant gateway, which provides monitoring and logging services via a ‘myVaillant’ smartphone app. Since I now have a year’s worth of data, this note reviews what has been logged to see how the system is performing.
The heat pump is a 10 kW Vaillant Arotherm plus (pictured above), serving my house of approximately 90 square meters internal area. I have previously posted about the building’s energy efficiency and additional first floor insulation added. This pump’s capacity exceeds the current requirements of the house, but we are planning an extension in the near future. In terms of the operation of the heating and hot water system, space heating is set to run continuously with a fixed temperature of 19 degrees C is set in the downstairs hallway (a central point in the house), and hot water is set to run continuously at 50 degrees C. In the first year of running the system, I experimented with running heating and hot water for set periods during each day of the week, but I found that the heat pump had to work hard to bring the system back up to temperature, particularly with the heating during the winter where the not only is the system heating the water but also the fabric of the building. I was advised that continuous operation was more efficient and since the house is almost always occupied, this matched our use.
The myVaillant app usefully provides a button to download all logged data per year in a set of CSV files. I have summarised the year period from July 2023 to July 2024 in the following charts. Having read this review by Michael de Podesta of the myVaillant app and the quality of the logged data, I am aware that there are accuracy/quantisation issues. The review measured a ~10% error in electricity consumed and a ~20% error in the heat energy produced, as well as pointing out many logged values are quantised to kWh units, which I also observe in the data. Unfortunately I don’t have a point of comparison with with no other means of measuring the metrics. I hold out some hope that Vaillant have improved the reported measurements in the last year, but even with these errors the data is still useful for a high-level review of the system.
Diving into the data, the first chart below shows the consumption of electrical energy for heating (red) and hot water (blue).
Most obviously, electricity consumption during the winter months increases significantly when the heat pump is using approximately 10 kWh per day. There are two spikes in November and January where temperatures went below zero, and at these times consumption went up close to 30 kWh per day. As expected, electricity consumption for hot water is more consistent throughout the year with a lesser increase through the winter months. The short periods where both drop to zero are due either to a holiday and the system being in ‘absence’ mode or a technical issue. Regarding the latter, these have only been due to a loss of pressure in the system, requiring a manual top up from the mains. The total electrical energy for heating for the year is 1.35 MWh and 1.04 MWh for hot water.
The next chart is the heat energy generated, which is tightly correlated with the consumption graph.
Combining the previous two charts by calculating the ratio between generated and consumed, gives the coefficient of performance (COP). The average COP in this period is 3.3 (the red line). The Vaillant app reports 3.4, which may be taking into account the quiescent periods, which broadly means the system is functioning well. However, caution is again advised by Michael de Podesta because of the inaccuracies in the Vaillant data, but noting that he found that the Vaillant data underestimated COP.
The following chart shows the temperature of the contents of the hot water tank. As expected, this stays constant, with a few exceptions: when I changed the temperature from 45 to 50 degrees C in October 2023; when the system has been off or out of order; and when every week on a Monday the immersion heater kicks in to perform a Legionella purge (which curiously it has stopped since March 2024, something I need to investigate).
Finally, we have a plot of internal (red) vs external (blue) temperature in degrees C. This clearly shows that the 19 degrees C target was maintained throughout the year, notwithstanding the periods of absence/downtime and on particular hot days when the temperature rose above the target. Given how quickly our summers are changing in terms of heat waves, having a system that can also perform cooling would be a big benefit. I think this chart represents well the benefit of having a heat pump, and a home with a continuous temperature throughout the year.
For the same period I obtained the electricity use and cost data from Octopus using their excellent API via the Octograph tool, visualised below on a Grafana dashboard. According to the Vaillant data, the heat pump used 2.39 MWh of energy, which is only 32% of the total electricity use. I am suspicious that this is incorrect, even factoring in a 10% underestimate from the Vaillant system. I would expect the heat pump to be using more like 50-60% of total electricity on average, given that other electricity use is cooking and appliances etc. Otherwise the usage profile matches between the two data sets.
References and further reading
- Protons for Breakfast, articles about heat pumps is a fantastic set of articles by physicist Michael de Podesta.
- Energy Stats provides pricing data for various Octopus Energy tariffs.
- Guy Lipman’s Octopus Energy resources is a collection of notes focusing on using the Octopus API to access energy data.
- Octopus Energy API is the landing page for using their API.
- Octograph (Github) A Python tool for extracting Octopus Energy meter readings to InfluxDB and Grafana.