Thursday, 30 July 2015

The Rebound Effect in Home Heating (review)

The Rebound Effect in Home Heating: 
A guide for policymakers and practitioners
by Ray Galvin
The rebound effect is of enormous importance because it affects energy savings – it explains why when we try to save energy by improving efficiency we don’t generally do as well as we expect.  Plus this book is about the rebound effect in home heating which is of special interest to me. If you insulate your home, or update your heating system, you will no doubt have been given an estimate of how much you will save and this helps you decide to go ahead or not. However, if this estimate has not taken the rebound effect into account then it may be overly optimistic. This book explores why the shortfalls occur, how much we can achieve in practice and the policy implications of this. It also exposes a woeful lack of data and the need for more research.

The reasons for the rebound effect in heating are partly to do with technology but also to do with people and difficulties with the human-technology interface. Galvin explores these at some length. Technology related issues include things like insulation not being fitted correctly so leaving gaps, or installing a too-large boiler that ends up running inefficiently at part capacity.  People reasons include heating more rooms than before, or even being at home more often than before because it is cosier and nicer. In between there can be issues with controlling the new system and making it do what you want it do to. Even if you don’t touch the control systems the upgraded home will probably be warmer than before simply because it loses heat less quickly. Also new technology such as heat pumps and underfloor heating may mean that it doesn’t warm up so quickly and you end up leaving the heating on when you aren’t at home.  Effects that mean your house is warmer than before are often lumped together as comfort taking, but they are not always due to a positive decision to be warmer – they may be just the way the new system works.

Galvin also explains the prebound effect, which he practically invented, or at least first described. This is closely related in that a good part of the rebound effect is because we aren’t as warm as we want to be to start with. Many people in older homes heat only a small part of them or just to a low temperature simply because they can’t afford more heat. In fact Galvin omits to mention another reason from my own experience – when we moved into our current house the existing heating system was simply incapable of providing the comfort level we wanted. Even if it was on all the time parts of the house were still cold. 

Surprisingly to me, it seems that if you plot average actual heating energy consumption against the expected requirement (based on a German standard of 19C all the time) then you get a consistent power law curve over quite a large range of energy efficiencies, from about 100 -400 kWh/year/m2. The equation is of the form
Consumption = k x expected energy (1-rebound effect)

Based purely on this, the rebound effect comes for Germany comes to 36% - i.e for every 1% improvement in energy efficiency, 0.36% is used to get more warmth and the actual energy savings are 0.64%.  The figure is not the same however in other countries. For example in France the average is 50%. It would be very interesting to explore the reasons why this might be the case. I suspect it has something to do with different heating technologies and housing designs as well as cultural norms and economic issues.

However, there may also be issues to do with how the data is collected. It is in practice quite difficult to get reliable data on either efficiency or consumption. On the consumption side, most homes use gas or electricity for heating and also cooking, water heating, appliances and so on – and it is difficult to get use figures that are adjusted to separate out the heating component from the other things. On the efficiency side, those figures depend on modelling example homes to estimate their heat loss, but in practice all models are simplifications and often inaccurate. 

One particularly surprising graph showing the rebound effect for each of the 27 EU countries is derived from time series data of energy use and energy efficiency. The idea is that you can correlate energy use with energy efficiency improvements over time across a national housing stock. However, this apparently shows countries such as Spain and Greece having negative rebound– so energy demand is reducing faster than efficiency is improving, while other countries such as Lithuania and Malta have rebound greater than one meaning energy demand is increasing even though efficiency is improving. Clearly there are other factors involved, such as decreasing or increasing wealth, but as Galvin points out it is the overall pattern that policymakers need to take into account when they decide how to tackle their energy use targets. 

Also, efficiency is normally calculated from the nominal heat demand which depends on a nominal heating pattern – in Germany for example this is to 19C and with 0.7 air changes per hour. Galvin relies mainly on Germany for his data but also includes examples from France, Belgium, the UK and other European countries and they all have different heating standards. Fortunately, a feature of the mathematics behind the rebound effect is that you can still get a robust figure even though the energy use models vary. So for example, if the German heating standard is warmer than the UK standard and needs 15% more energy, this does not affect the calculation of the exponent in the consumption equation, only the constant k. 

However, that constant is very important when other measures are used such as the energy performance gap (EPG). The EPG describes how much extra energy is used compared to what should be required for standard heating services. The EPG is –ve when the actual energy use is less than expected, and +ve when the home uses more. In all countries that Galvin has data for, the EPG is strongly –ve for inefficient homes but often +ve for very efficient ones. The –ve EPGs are partly explained by fuel poverty and Galvin has a whole chapter on this. Also he has a chapter on very efficient homes and why the EPG is +ve at that end. But in my view he doesn’t nearly cover the issues. This is an area where we just don’t have enough data.

Although the book’s title suggests it is about homes he does also have a short chapter on non-residential buildings. These are less important in that they consume less energy overall  - only about a third as much as residential buildings in Europe and also, they are much harder to study because there are many different categories. He shows us data from three small datasets including office buildings in Germany and supermarkets in Hong Kong. The rebound effect in office buildings is apparently considerably lower than in homes, perhaps because retrofits are typically less ambitious and better targetted. However, the EPG shows a similar pattern in that it is +ve for supposedly very efficient buildings and –ve for less efficient ones. Despite the different use patterns, Galvin tells us that there are lessons to be learnt from the commercial sector that can be applied to the residential sector. For example, taking living patterns into account when planning a retrofit, and choosing an appropriate balance between giving the residents control over their building versus more automation, especially for households with weak management skills.

This book is a must-read for anyone who needs or wants to understand the reasons why home retrofit energy savings are so often disappointing. It is generally very readable too, at least as long as you take the mathematics at face value and don’t try to understand the proofs and derivations in the appendix. (I think this would be much improved with some illustrative graphs.) Galvin gives a comprehensive overview of the issue as we understand it today and I found it very thought provoking. He concludes for Germany that although it may be economically feasible to double energy efficiency, as required by current building regulations, the rebound effect means that in practice the reduction in energy use would only be about 38%. This means that we need to focus hard on renewable heating fuels to achieve anything like our target reduction of 80% of carbon emissions in this sector.

The Rebound Effect in Home Heating: A guide for policymakers and practitioners
by Ray Galvin
Published on 26 June 2015
BRI book series, Earthscan/Routledge
ISBN 978 1 138 78834 3

Ray Galvin is a colleague and a friend of mine but I only review books I can be honest about.


  1. Nice review Nicola. Thank you! I just purchased the book since it looks like it has a lot of relevant information. This thing about Spain and Greece vs Lithuania and Malta... I can only say "yikes" and wonder what's up - what aren't we taking into account! Does the book give any insights into that? What are the socio-economic differences that would give Greece and Malta for instance opposite rebound responses?!

    1. The book suggests the main factors are relative wealth, as Malta is growing and people are becoming more wealthy, whereas in places like Greece the opposite it true. Fuel poverty is a big factor.

      In the UK domestic gas consumption has reduced by 30% between 2005 and 2013 (see I don't believe that was all new boilers and insulation, but there is also a lot people can do to save energy by behaviour change, especially when motivated by rising energy prices.

  2. This comment has been removed by a blog administrator.

  3. This comment has been removed by a blog administrator.