How to over-estimate heat loss in pre-2000 terraced houses!

I’ve updated the blog on the heat loss calculation I did for my own house. Worthwile reading I hope for anyone with a terraced property with solid stone walls built before the year 2000, and interested in getting a heat pump installed. Feedback most welcome!

https://trystanlea.org.uk/roombyroomheatloss

Quick summary

  • Default CIBSE/MCS assumptions for pre-2000 air change rates, unheated neighbours, design outside temperature and stone wall U-values result in a calculated heat loss of 7.5 kW for our house! This would typically result in a 8.5 kW or 10kW heat pump being installed. We are seeing a lot of Eco4 installations in our area (North Wales) installing this size of heat pump as a result.

  • In reality the highest cold weather heat demand that I recorded, on the 1st of December 2023, was 2.9 kW over 24 hours with an outside temperature of -1°C and an average living room temperature of 20°C. I extrapolate this and other measurements to indicate a measured heat loss of ~3.3 kW at a design outside temperature of -1.4K (± 0.6 kW). Our 5kW Ecodan with a real world maximum output of 4.2-4.5 kW during heavy defrost conditions has been plenty sufficient for our requirements.

  • With better input assumptions for air change rates, stone wall U-values, heated neighbours, and design outside temperature (based on local weather data) it’s possible to reproduce the measured heat loss in the heat loss calculation. Some additional over-sizing to cater for unheated neighbours and strong wind storms may be reasonable but it’s worth thinking it through carefully.

  • There’s also a section on choosing design flow temperatures and why I think a design flow temperature of 35-40C is required for a SCOP of 4 or above if the system is sized based on an accurate heat loss.

6 Likes

I’ve also updated my blog on the original mostly single panel radiator system that we had in our house before I replaced it all with new pipework and K2 radiators throughout. In hindsight the original radiators may have worked ok due to the lower actual building heat loss https://trystanlea.org.uk/original-heating-system. The performance would have been lower but perhaps difficult to justify the cost of upgrading it before it’s lifetime replacement point. I’m not totally convinced that my calculation for the expected SCOP of that original system running at higher flow temperatures is correct, it would be interesting to collect more data on systems with real world design flow temperatures around 50C!