Wednesday, 27 November 2013

How much do you really save by insulation or a new boiler?

The latest report form the National Energy Efficiency Data-Framework (NEED) reports on the impact of energy saving measures (insulation or new boilers) on household energy bills [1]. This year they report on individual measures and combinations of measures and they also have some interesting data on how savings persist over the years. This data is compiled by comparing before and after bills of properties which have had a measure installed (these days usually through the Green Deal and ECO, formerly by CERT) with before and after bills of comparable properties which have not had anything done to them.

The overall mean savings are given in the table below. The gas bill data has already been weather corrected, so these savings take into account that some winters are colder than others.

Measure in 2010Mean Saving (Expected)%* Median Saving (Expected)%*
Cavity wall insulation7.8 8.9
Solid wall insulation (only 790 in sample)16.114.7
Loft insulation1.7 2.2
New boiler9.2 10.7
Cavity wall insulation and loft insulation9.4 (9.5)11.6 (11.1)
Cavity wall insulation and new boiler15.7 (16.3)19.6 (18.6)
Loft insulation and new boiler13.1(10.7)14.3 (12.7)
Cavity wall insulation, loft insulation and new boiler19.0 (18.3)24.1 (19.8)

*In practice the actual savings you can expect depend very much on the shape of your house and the level of insulation and draught proofing you have elsewhere. For example, loft insulation has a much bigger impact on a bungalow because it has more roof compared to walls. In the NEED data the median saving for cavity wall insulation was 9% overall, but it was 11.5% for detached homes and 6.9% for mid-terrace. Using my home heat loss model (which considers heat loss only, so overestimates savings on bills which include hot water and cooking too) cavity wall insulation gives you typical savings of 6% for a mid-terrace house compared to 16% for a semi-detached house, assuming 100mm loft insulation and good draught management to 'cosy' level). For solid wall insulation my tool gives savings of 11% and 24% in the same mid-terrace and semi-detached cases - this is after making a correction to the U-values for solid walls which I hope to explain about in another post.

The NEED data has an interesting anomaly in that combinations of insulation and new boiler often save more than you would expect. You would expect the savings to be less than the same of the individual measures* but actually you often get more. The authors do not have an explanation for this. I suspect it is because when a combination of measures is installed there may be a general overhaul, perhaps including draught stripping and a review of heating controls which isn't recorded.

Another odd point is that the mean saving is less than the median for all measures except for solid wall insulation (SWI). This means that most interventions are pretty good though there are some bad ones pulling down the mean. For SWI the mean is higher than the median which means most installations are not brilliant but there are some giving excellent savings which drag the mean up. Now granted there were only 790 SWI cases compared to 16000 cavity wall insulation cases, but I suspect the skew in this data is significant. Consider, when exterior SWI is fitted it is tricky to get the details right around windows to avoid thermal bridging. Also in most cases it will stop at the damp course level but it is also possible to insulate below ground level to trap heat underneath the house. Getting these extra details right is costly and probably not specified in the majority of cases. In comparison cavity wall insulation is not so complicated and it is often pretty obvious if it is not done well. For example, if the installers don't drill enough holes, leaving parts of the wall unfilled, the missing holes are easy to spot. Obviously this is pure conjecture on my part and I don't mean to accuse the majority of SWI installers of doing a shoddy job!

Another significant finding in the data is that even homes which have not had energy efficiency measures recorded have reduced energy consumption, and this reduction often dwarfs the savings due to the interventions. For example, considering homes with cavity walls installed in 2005, the matching homes without intervention reduced energy consumption by more than 25% between 2004 and 2011. This can't be due to new boilers and insulation. Even if many homes have had measures installed which are not recorded in NEED they would have to be incredibly effective to make such a big difference. Most of it must be due to the householders managing their heating better. Perhaps they have lowered the thermostat and adjusted the timing, changed their habits to open windows less often, remember to close their curtains in the evening and not to let them hang in front of the radiator ...

The difference between the cavity wall insulation cases and their matched homes without installation narrows somewhat from 8.4% in 2005 just after installation to 4.1% less in 2011. This could be because the insulation becomes less effective over time, though it should last a great deal longer than that, or it could be that the people who have insulated then try less hard to reduce their heating by other means. The data for boiler installations in 2005 is similar, from an initial 11.9% saving down to 8.6% after six years. Again, this could be due to the boilers becoming less efficient over time or the residents being less careful than others with higher bills.

This all goes to show how incredibly difficult it is to validate energy saving models against reality. There are just so many variables that affect our bills not all of which can ever be known and accounted for.

* Consider a boiler which give you 20% savings. Your bill might go down from 1000 kWh to 800 kWh. If you put in insulation which saves you 10% then the effect of your boiler now reduces your bill from 900 kWh to 720 kWh. 20% on the boiler + 10% from the insulation has given you 28%.

[1] National Energy Efficiency Data-Framework (NEED) report: Summary of analysis 2013 Part 2 (

No comments:

Post a Comment