How much energy does it take to keep a swimming pool warm?

Heat pumps are brilliant at heating swimming pools. The relatively cool temperatures required (pool water is typically 28-31°C) and the presence of steady, year-round demand suit heat pumps perfectly. But how do we select the right size of heat pump to deliver high performance and good value-for-money??

In a retrofit context, especially if the gas-heated system is being retained as backup, it ought to be easy – you can replace like-with-like or choose a heat pump that’s a bit smaller and run it for longer. But what if the pool water heating is just one part of a bigger system? What if there’s no heat metering, no retained design information and no prospect of modelling heat loss from the swimming pool basin and water surface*??

Natural Power’s Renewable Heat team found itself in just such a situation when working at a well-used community leisure centre. Our solution was to use temporary heat meters to directly measure the load on two LTHW-to-pool water heat exchangers. However, site-specific circumstances conspired to produce multiple challenges for what should normally be a straightforward process.?

Unable to consider the intrusive work that would be required to fit in-line heat meters – which would anyway have been disproportionately costly – we used clamp-on ultrasonic heat meters. These meters require a certain length of straight pipe upstream and downstream of the location you want to measure (the fluid needs to be flowing neatly rather than swirling around). At our site, the ‘hot’ side of the system (LTHW) just didn’t feature any suitable locations. On the pool water side, there were some reasonable locations – but the plastic pipe material was far from ideal either for flow measurement, or for getting a good reading of the temperatures of the inlet and outlet pipes. We did our best to clamp the temperature sensors firmly to the outside of the pipe with plenty of thermal coupling gel and lots of insulative padding to keep the sensors as warm as possible. At the end of the day, our calculations needed a temperature difference rather than an absolute temperature, so we hoped this approach would be OK.?

Two years later, the heat pump is up and running with a heat meter measuring its output - so now we can look back and see how good our measurements and the demand estimates we derived from them were. The two profiles aren’t directly comparable due to the buffer vessel that now moderates fluctuations in heat demand (to help the heat pump to run better), and the fact that the recent heat meter data is from the month of October when ground temperatures around the pool basin should be above average. However, there is still a satisfying similarity between the two curves. In the end, we had overestimated the heat load by something between 0 and 20%, depending on the impact of various other factors that had changed between the two measurement periods. This is not a bad outcome – indeed, the heat pump performs at its most efficient when it’s running at between 50% and 75% of its full output. This, along with the sensitivity analyses that we performed at the feasibility stage, mean that the eventual energy and carbon savings are within the range that stakeholders were expecting.???

Taking the sometimes-uncomfortable step of comparing past estimates with recent reality gives us the opportunity to improve our understanding and our processes in pursuit of providing an even better service for our clients.?

Meanwhile, back at the leisure centre, swimmers continue to enjoy their local pool. With the gently humming heat pumps tucked away around the side of the building, and the pool water the same comfortable temperature it always was, not many of them will notice that anything has changed…?

*The physics is surprisingly complicated, and without information about the ground around the basin modelling is badly unreliable.?